Abstracts from the 37th Annual Meeting of the Society of General Internal Medicine, 2014, San Diego, CA, USA.

Abstracts from the 37th Annual Meeting of the Society of General Internal Medicine

SCIENTIFIC ABSTRACTS

“A ONE-STOP SHOP:” PERCEIVED BENEFITS OF DIABETES GROUP VISITS IN THE SAFETY NET CLINIC SETTING Arshiya A. Baig1; Amanda Benitez1; Amanda Campbell2; Cynthia T. Schaefer3,2; Loretta J. Heuer4,2; Michael T. Quinn1; Deborah L. Burnet1; Marshall Chin1. 1University of Chicago, Chicago, IL; 2 MidWest Clinicians’ Network, Lansing, MI; 3University of Evansville, Evansville, IN; 4North Dakota State University, Fargo, ND. (Tracking ID #1934849) BACKGROUND: Diabetes group visits, shared appointments in which patients with diabetes receive self-management education in a group setting and have a medical visit, are an innovative and promising way to deliver diabetes care. The group visit model may be especially promising in safety net clinics where many patients with diabetes in underserved settings receive their care. However, little is known about safety net providers’ perceptions of diabetes group visits in the community health center setting. METHODS: The research team conducted site visits at community health centers across the Midwest to assess their experiences in providing diabetes group visits. Site visits were conducted at centers that had implemented diabetes group visits in the past or currently had them. Two members of the research team interviewed health center personnel at each site regarding their diabetes group visit program. Health center personnel included chief executive officers, medical and nursing directors, care coordinators, physicians, nurses, physician assistants, diabetes educators, mental health professionals, dietitians, medical assistants, pharmacists, and other recommended staff members who were familiar with the diabetes group visits at the site. Interviewees were asked about the benefits of having diabetes group visits at their site. All interviews were audio-recorded and transcribed. Systematic qualitative analysis techniques were used to identify the range and consistency of opinions and experiences across interviewees. RESULTS: The research team visited five health center sites across four states in the Midwest and conducted a total of 26 interviews with health center personnel. One site was rural, three were urban, and one was suburban. Health center personnel noted many benefits of group visits to the health center, providers, and patients. The benefits of group visits to the health center included an alignment of the group visit model with the mission of becoming a patient centered medical home, the ability to bill for group visits, and an efficient way to improve guideline-driven care for patients. Provider benefits from group visits included boost to provider morale, opportunities to collaborate with multidisciplinary

colleagues, and having more time to focus on other medical concerns during patient’s routine follow-up visits. Patients derived many benefits, including receiving education and medical care in a single appointment, e.g. “the one-stop shop” or “best bang for your buck,” the opportunity to obtain social support and enhanced motivation through peers, and the potential to improve their clinical outcomes by attending group visits. CONCLUSIONS: Diabetes group visits can offer many unique benefits to safety net clinics by providing patient-centered care, boosting provider morale, increasing multidisciplinary collaboration, and offering patients the convenience of combining an educational session with a primary care appointment. Further studies need to assess best practices in implementing group visits in health centers and assess their impact on patient outcomes.

“A PLACE AT THE TABLE:” EVALUATION OF COMMUNITY MEMBERS’ EXPERIENCES AND EXPECTATIONS FOR ACADEMIC-COMMUNITY PARTNERSHIPS IN HIV/AIDS RESEARCH Stella Safo1; Chinazo Cunningham3,2; Alice Beckman2; Joanna L. Starrels3,2. 1Montefiore Medical Center, Bronx, NY; 2Albert Einstein College of Medicine, Bronx, NY; 3Montefiore Medical Center, Bronx, NY. (Tracking ID #1937992) BACKGROUND: The foundation of community based participatory research (CBPR) is collaboration between academic researchers and members of a given community. Community advisory boards (CABs) are one mechanism through which academic-community partnerships are formed, but current research about CAB members’ opinions on barriers to collaborations with academics is limited. This qualitative study examined CAB members’ expectations and experiences in working with academic researchers in the field of HIV/AIDS. METHODS: We conducted 10 semi-structured one-on-one interviews with individuals serving on a CAB for HIV-related research at an urban academic medical center. Participating CAB members were leaders of HIV/AIDS community organizations in Bronx, NY and had at least 5 years of experience working in the field of HIV/AIDS. Interview questions focused on participants’ current and previous experiences with academic research and researchers, trust in these relationships, and best practices for collaboration. Interviews were professionally transcribed and data was analyzed using a grounded theory approach. Transcript data were coded by two independent researchers using NVivo 10 software and analyzed in an iterative process to identify emergent themes. RESULTS: CAB members described positive aspects of inclusion on a CAB, including improved access to information about current HIV topics and the opportunity to help shape HIV research in their communities. However, CAB members also described negative previous interactions with researchers, and a lack of trust in researchers themselves or in the process of conducting research with academic institutions. A major reason for distrust was that power was perceived as unequal,

ABSTRACTS often resulting in miscommunications that led to lasting negative impressions. For example, several members recounted researchers approaching subjects directly, without informing the home organization about the intended study focus. In addition, CAB members felt that expectations for collaboration were not clearly established. CAB members also shared suggestions for improving the collaborative process, including outlining expectations from the outset of the partnership and holding CAB meetings in the community rather than at the academic center. CONCLUSIONS: We identified complex interpersonal issues in relationships between CAB members and academic partners, particularly around trust and powersharing in conducting HIV/AIDS research. Our findings suggest that academiccommunity collaborations could be enhanced by improving transparency, explicitly addressing issues of distrust from previous negative experiences, and ensuring equitable power-sharing among researchers and community members. Further research should be conducted to determine the best ways to engage and maximize the impact of CABs in the field of HIV/AIDS.

“ARE YOU AFRAID OF PNEUMONIA?” PATIENTS’ FEEDBACK ON THE DEVELOPMENT OF AN EDUCATIONAL VIDEO ABOUT PNEUMOCOCCAL VACCINATION Kenzie A. Cameron; Tiffany Brown; Shira N. Goldman. Northwestern University, Chicago, IL. (Tracking ID #1926320) BACKGROUND: Invasive pneumococcal disease causes a mortality burden among people age 65 and older. Consequently, the Advisory Committee on Immunization Practices recommends pneumococcal vaccination (PnVx) for these adults. However, national rates of PnVx for those age 65 and older are just over 60 %. Further, racial and ethnic disparities in vaccination remain: 2011 estimates reported 66.5 % vaccination among whites age 65 and older, as compared to only 47.6 % among blacks. Informed by our previous research on perceptions toward PnVx among unvaccinated adults, we developed a 2 min patient education video targeted at unvaccinated 65 and 66 year olds to encourage uptake of PnVx. Our objective was to discuss patient attitudes toward vaccination and beliefs about pneumonia, and obtain feedback regarding suggested revisions for the video. METHODS: At an academic general internal medicine clinic, we queried the electronic health record to identify 65 and 66 year old black and white patients who had no record of receipt or refusal of PnVx. In early 2013, patients were recruited to participate in race-concordant focus groups. Prior to the start of each group, participants completed a brief demographic survey, which included questions regarding past experience with pneumonia and PnVx. During the audio-recorded focus group, we asked participants to describe pneumonia, tell us what they knew about PnVx, and specify what they would want included in educational messages about the vaccine. RESULTS: Four white and two black patient focus groups were conducted, with a total of 26 participants. Mean age was 65.9 (3.1), 81 % were female, 88 % had attended college, and 39 % reported receipt of PnVx. Based on our previous research, the video included a strong focus on individual susceptibility to pneumonia, as well as personal testimonial from a black man about his late wife who did not think she needed PnVx. Black patients in particular noted that the testimonial was a persuasive part of the video. Participant response to the video was positive, with some individuals indicating a newly found motivation to consider vaccination (e.g., “Even though I refused the pneumonia vaccine already, next time I see my provider, I’m going to tell her I want to get shot,” and “I’m going to go get my shot now. I didn’t realize how potent and prevalent it is.”). However, some participants felt that the video focused too heavily on severity and fear (e.g., “I find it dire. Very uncomfortable watching it.”). The video included an appeal to consistency and the need for vaccination across the lifespan (“Just as you got vaccinated as a child, you need to receive important lifesaving vaccines as an adult”), which was met with favorable responses from participants. Several participants highlighted the importance of discussing PnVx with their providers and indicated that viewing this video prior to an appointment could provide an opportunity for such discussion (e.g., “I might talk to my doctor. I don’t know if she could persuade me to get it. Maybe she could explain it to me differently”). CONCLUSIONS: Feedback from the focus group participants led to revisions to the patient education video, specifically with regard to reducing the number of times the severity of pneumonia was stressed. The personal narrative was retained as numerous individuals spoke of its persuasiveness, as was a focus on personal susceptibility to pneumonia and the need for vaccination across the lifespan. The revised video is under 2 min long, has been integrated into our electronic health record system, and is being tested at the point-of-care to assess its effect on receipt of PnVx among patients eligible for vaccination.

JGIM

“BEYOND OUR CONTROL”: HOW ORGANIZATIONAL CONTEXT IMPACTS PERFORMANCE MEASUREMENT Molly Harrod1; Jane Forman1; Claire Robinson1; Adam Tremblay2,3; Leo Greenstone2,3; Eve A. Kerr1,3. 1VA Ann Arbor Health Care System, Ann Arbor, MI; 2VA Ann Arbor Health Care System, Ann Arbor, MI; 3University of Michigan, Ann Arbor, MI. (Tracking ID #1937968) BACKGROUND: The Patient Centered Medical Home (PCMH) model requires a shift from a physician centric to a team-based approach to care, as well as improvement in process measures that reflect PCMH goals. As a result, as primary care (PC) providers are learning to work in teams, they also must learn how to change their clinic processes to meet PCMH-specific performance measures. The Veterans Health Administration (VHA) transitioned over 900 PC sites to a PCMH model beginning in 2010 and has implemented national metrics of PCMH success that include same day access and continuity with the patient’s usual provider. One large VHA health system implemented a coaching model to help newly formed PC teams redesign their delivery processes to improve access and continuity measures. This study examines how these PC teams discussed and responded to these measures during their coaching sessions. METHODS: Nine of 20 PC teams were longitudinally observed during coaching sessions (25+ hours) that entailed discussion of both performance measures and processes to improve measure results. Conversations and interactions were recorded via hand-written field notes. We analyzed data using a grounded approach. RESULTS: As PC teams reviewed their access and continuity measures, two predominant, and often overlapping, themes emerged: lack of control and lack of consistency across the organization. PC teams felt they did not have control over processes both within and outside PC. For example, a centralized call center with variable understandings of PC redesign continued to refer patients to the Emergency Department even though PC providers had open appointments within their schedules, thus impacting continuity. A lack of consistency across the organization as a whole was apparent given that many of the other departments did not change their patient care processes to align with PCMH goals. For example, inpatient physicians followed a rule that patients discharged from the inpatient setting needed “follow-up with primary care physician in 7–10 days”. Because this 7–10 day mandate meant fewer open slots in a PC provider’s clinic, it decreased same day access. Thus, organizational processes were often working against one another resulting measures that were not reflective of all the changes. CONCLUSIONS: Improving even straightforward performance measures, like same-day access, requires an understanding of the entire practice context. In our study, measurement results that fell short of goals reflected the organizational inconsistency of processes that impacted individual measures. It is important to develop processes designed to meet PACT goals across the organization rather than within each team or department. As more PC practices become part of Accountable Care Organizations and are assessed using these and other performance measures, a better understanding of what is being measured, how these measures reflect patientcentered care, unintended consequences of measures and other organizational actions, and how to align organizational performance goals is needed so that organizational performance reflects true quality of care and patient-centeredness, and not just performance on individual metrics.

“CODE R”: INTRODUCTION OF A HOSPITAL-WIDE PEER REVIEW PROCESS TO ASSESS RESIDENTS’ POTENTIAL IMPACT ON PATIENT SAFETY AND QUALITY Daniel I. Kim1,3; Lawrence Loo2; Huy Au1,3; Ramiz Fargo1,3; Roger C. Garrison1,3; Gary Thompson1,3; Minho Yu2,3. 1University of California, Riverside, Riverside, CA; 2Loma Linda University Medical Center, Loma Linda, CA; 3Riverside County Regional Medical Center, Moreno Valley, CA. (Tracking ID #1928321) BACKGROUND: The hospital physician peer review process is a valuable method of reviewing physician competency and safety, but it is rarely used to evaluate residents, who frequently work under indirect supervision. In 2008, our institution introduced “Code R” to our peer review process to highlight the potential resident involvement in cases undergoing peer review. To our knowledge, there are no published studies that utilize a hospital-wide peer review process to evaluate and track residents in training. Our study sought to characterize resident errors using the ACGME’s core competencies to determine whether interns are more prone to errors than senior residents and to see if there were patterns of resident errors by time of day or month of year. METHODS: The study is a retrospective review of all peer review cases recorded from 2008 to 2011 by the Quality Management department at Riverside County Regional Medical Center, an academic safety net institution. The institutional peer

JGIM

ABSTRACTS

review process spans all the departments within the hospital and reviews the cases involving patient deaths, unexpected adverse outcomes, incident reports, quality measure fallouts and random physician quality reviews. Cases designated a Code R were identified and the following information was collected using a standardized form: peer review level assigned, area of deficiency, resident post-graduate year, originating department, month, and time of day. Care issues were characterized into the five core competencies of patient care, medical knowledge, professionalism, communication/interpersonal skills and systems-based practice. The chi-square and Fisher’s exact tests were used to make comparisons and an a priori statistical level of significance of p value ≤ 0.01 was used to adjust for multiple comparisons. RESULTS: Between 2008 and 2011, 8052 individual cases were peer reviewed by the medical staff and 258 (or 3.2 %) had “Code R” assigned. Of the total charts initially identified, nine cases were determined to not be resident related, nine charts were unavailable and two cases had incorrect medical record numbers. Two hundred thirty-eight charts qualified for final data analysis. The peer reviewed designations and prevalence in each category were Level 1 (Meets standard)- 13 (5.5 %), Level 2 (Meets standard - opportunity for improvement) −196 (82.4 %), Level 3 (Deviation from standard) −26 (10.9 %), and Level 4 (Unacceptable care) −3 (1.3 %). Compared with the non-Code R, or attending physician cases, there were significantly more errors in all the competencies (5.5 % vs. 89.3 % Level 1 assignments for resident vs. attending physicians). In the Code R cases, 60 % (143/238) of the charts had documentation issues. One-third (79/238) of the charts had deficiencies in patient care issues, which include clinical judgment, decision-making and lack of follow-up. The next most common errors were in the competencies of Systems-based Practice and Interpersonal Communication, 9.2 % and 8.0 % respectively. We did not find any significant differences in resident errors by time of day or calendar year. However, interns (PGY-1) were more prone to errors than senior residents (PGY≥2) (p=0.002). CONCLUSIONS: The introduction of “Code R” helped identify cases in which resident involvement potentially impacted patient care and safety. Our analysis of Code R cases suggest specific areas for focused resident supervision and education are needed in the areas of documentation and patient care. This simple peer review process can be readily adopted by other institutions to help improve resident education, facilitate attending supervision, and potentially improve patient safety. Use of this instrument can be used to track the individual resident and attending performance that has not been previously described.

for favors, and because the consultation is stored in the electronic medical record and is conveniently linked to information about related tests and treatments. As one PCP explained, “I like that it is more formal and that it gets into the patient’s chart and it is recorded and I can refer to it later.” Nonetheless, the occasional use of curbside consultations persisted even among PCPs who routinely use eConsults, and was not seen as incompatible with eConsult use. The perception that an eConsult renders the specialist part of a patient’s care team, while simultaneously obviating the need for an office visit, was also reported as a source of increased patient and PCP satisfaction. 2) Providing better care: The use of eConsults was associated with increased clarity about specialists’ and PCPs’ roles, more transparent and efficient care coordination procedures, and PCPs’ enhanced participation in determining a patient’s follow-up plan. “I just feel like I’m providing better primary care”, said one PCP. Some participants reported that patient management using an eConsult is more timeconsuming, and entails greater responsibility on the part of the PCP, than a traditional consultation, but most felt that this was an acceptable tradeoff for the satisfying learning opportunities and increased sense of mastery accompanying the incorporation of eConsults in their workflow. 3) Uneven adoption patterns: A minority of participants was less enthusiastic about eConsults and the frequency of use varied widely among individual providers. Some PCPs described a reluctance to initiate an eConsult because of difficulty with technical aspects of the program, frustration with slower-than-promised responses from consultants, or concern about how and when to communicate with patients regarding a consultant’s recommendations. Others reported “pushback” from patients who wanted in-person advice from a specialist. Most participants reported weighing both clinical and non-clinical factors when deciding when to request an eConsult. “That is why I choose the right patients…”, said a PCP about the need to balance clinical appropriateness and patient preference when deciding between an eConsult and a conventional referral. CONCLUSIONS: Our results suggest that eConsults present an effective strategy for improving PCPs’ ability to obtain specialist expertise, and that they increase PCPs’ engagement and satisfaction with their clinical work. However, eConsults do not uniformly replace informal “curbside” consultation practices, and decisions about their use are influenced by non-clinical factors such as convenience, usability of the electronic interface, and patient preference. eConsult program designers should elicit continual feedback from providers in order to increase the ease of use and appeal of this mode of consultation.

“HAVE YOU THOUGHT ABOUT SENDING THAT AS AN ECONSULT?”: PRIMARY CARE PROVIDERS’ EXPERIENCES WITH ELECTRONIC CONSULTATIONS AT AN ACADEMIC MEDICAL CENTER Sara Ackerman1; Gina Intinarelli4; Nathaniel Gleason2; Michael Wang3; Sereina Catschegn2; Delthia McKinney4; Ralph Gonzales2. 1University of California, San Francisco, San Francisco, CA; 2University of California, San Francisco, San Francisco, CA; 3 University of California, San Francisco, San Francisco, CA; 4University of California, San Francisco, San Francisco, CA. (Tracking ID #1936109)

“IN THE MILITARY, YOUR BODY AND YOUR LIFE AREN’T YOUR OWN:” UNIQUE FACTORS INFLUENCING HEALTH BEHAVIOR CHANGE IN OVERWEIGHT AND OBESE VETERANS Melanie Jay1,3; Katrina F. Mateo1,2; Molly Horne2,1; Allison Squires4,2; Adina Kalet2; Scott Sherman1,3. 1VA NY Harbor Healthcare System, New York, NY; 2NYU School of Medicine, New York, NY; 3NYU School of Medicine, New York, NY; 4NYU College of Nursing, New York, NY. (Tracking ID #1920967)

BACKGROUND: Electronic consultations (eConsults) are a recent innovation enabling primary care clinicians to request advice from specialists, via a shared electronic health record (EHR), for lower-complexity clinical problems that do not require an in-person evaluation of the patient by the specialist. Prior studies suggest that eConsults can improve the time-to-access to specialty care and reduce specialty care office visit utilization. Using in-depth interviews, we sought to elucidate primary care providers’ (PCPs) experiences with a recently adopted eConsult program at an academic medical center. METHODS: We recruited a purposive sample of PCPs from an academic medical center with a recently implemented EHR-based program in which specialists and PCPs receive reimbursement for completed eConsults. Semi-structured interviews were conducted to explore the benefits and limitations of eConsults and their impact on clinical work - including information transfer and clinician-patient and PCPspecialist interactions. Interviews were audiotaped, professionally transcribed, and checked for accuracy. Each interview transcript was reviewed by at least two authors in addition to the interviewer. The constant comparative method was used by the team to collaboratively and iteratively develop themes related to the study objectives. RESULTS: Nine interviews have been completed to date, with participants representing a range of primary care professions and levels of experience, including faculty physicians, medical residents and nurse practitioners. A total of 20 interviews will be completed by March 2014. We identified three key thematic domains related to eConsult benefits, limitations, and influence on clinical work. 1) Access to expert advice: Most participants reported that eConsults were preferable to the use of informal, or “curbside”, consultations because they offer timely and standardized access to specialist advice without requiring PCPs to call on social networks or ask

BACKGROUND: Obesity affects 36 % of the Veteran population. The Veterans Affairs (VA) Healthcare System offers an intensive weight management program called MOVE!, but only approximately 10 % of eligible patients attend even one session. Since veterans see their PCPs 3.6 times per year, developing primary carebased weight management interventions may improve MOVE! attendance. To inform intervention development, we conducted a single-center qualitative study to explore unique VA patient 1) attitudes, barriers, and facilitators to healthy behavior change; 2) uses and understandings of goal-setting; and 3) weight management-related experiences with health care providers, technology, and MOVE!. METHODS: We sent recruitment letters to overweight and obese patients at the New York Harbor VA. Eligibility criteria included Veterans aged 18–75, having a BMI ≥ 30 or ≥ 25 with at least one co-morbidity, and self-reported ability to read and participate in focus groups. We organized patients into focus groups of 6–11 participants, grouped by gender and MOVE! attendance. The interview guide was informed by the Theory of Planned Behavior (TPB). Sessions were audio-recorded, contextually supplemented with field notes, and transcribed by a professional company. Initial codes were determined based on session recollection and field notes. Using an iterative coding approach, two coders separately reviewed segmented transcripts, modified the codebook as new codes emerged, and met to negotiate codes. We utilized NVivo software to assist with coding and subsequent thematic analysis. RESULTS: We screened 161 patients for eligibility, 77 were scheduled to attend, and 54 attended one of six focus groups (two female and four male). Participants were predominantly male (63 %), African-American (46 %) or White/Caucasian (32 %), college-educated or higher (74 %), and reported having attended MOVE! (61 %). We found that although we attempted to stratify focus groups by MOVE! attendance, the VA electronic health record did not accurately capture this information, and some

ABSTRACTS study participants were confused or unaware of their attendance. Thematic analyses revealed how military service informed participants’ attitudes, social norms, and perceived behavioral control with regards to weight management. Participants described how the structured environment and strict standards of the military motivated them to maintain weight control while enlisted, but that lack of autonomy was a barrier to learning how to manage their health when they left service. The military also influenced their perceived need for more personalized and tailored health counseling. Physical activity was largely considered the predominant method to lose weight, and participants felt the MOVE! program lacked a physical activity-related component. While participants acknowledged the potential usefulness of technology to help identify and achieve weight management goals, they expressed the need to have a knowledgeable person (i.e. health coach, PCP, dietician, etc.) provide guidance and support. CONCLUSIONS: This study revealed how military service may influence veterans’ intentions to lose weight and their weight management-related experiences. This will guide the development of a computer-assisted intervention to treat obesity within primary care and improve access and quality of weight management services. The implications from this study may be limited as it was conducted at a single site within a specific and a mostly self-selected patient population.

“NONE OF THEM APPLY TO ME:” A USABILITY STUDY OF THE VA’S MOVE!23 ONLINE WEIGHT MANAGEMENT SOFTWARE IN LATINA WOMEN Hector R. Perez4; Michael W. Nick5; Katrina F. Mateo1,2; Scott Sherman1,3; Adina Kalet2; Melanie Jay1,3. 1VA NY Harbor Healthcare System, New York, NY; 2NYU School of Medicine, New York, NY; 3NYU School of Medicine, New York, NY; 4NYU School of Medicine, New York, NY; 5NYU School of Medicine, New York, NY. (Tracking ID #1920971) BACKGROUND: The obesity epidemic disproportionately affects Latina women, but there are few interventions in this population. The MOVE!23 software program is a publicly available, English-only, online questionnaire used by the Veteran’s Affairs (VA) Administration to facilitate weight management. It evaluates patients’ current behaviors and stage of change and provides computer-generated, individualized advice. Completion of MOVE!23 during an intensive weight management program is associated with greater 6-month weight loss outcomes. Thus, adapting this tool for use in Latina women may be beneficial. However, the MOVE!23 is not specifically tailored for use in non-VA populations and its usability in this population is uncertain. The International Organization for Standardization (ISO) defines usability as “the extent to which a product can be used by specified users to achieve specified goals with effectiveness, efficiency, and satisfaction in a specified context of use.” In order to determine the utility of tailoring the MOVE!23 for Latina women and translating it into Spanish, we conducted usability testing of this software. METHODS: We recruited English-speaking, adult Latina women. A researcher observed each participant individually as they used the MOVE!23 . Participants were asked to verbalize their thoughts and reactions as part of a “Think-Aloud” protocol. A semi-structured interview followed that elicited reflections on this experience and how it might influence lifestyle behavior change. We utilized ScreenFlow (Version 4.0.3) to capture on-screen interactions, and each session was digitally recorded and professionally transcribed. A coding guide was derived from the ISO definition of usability and focused on positive and negative interactions between four domains (Tool, Task, User, Context). Two researchers reviewed and coded the transcripts independently using the guide and secondary codes were derived inductively. The researchers met to review the codes and resolve discrepancies. NVivo software was used to facilitate data analysis. RESULTS: Six female Latina participants completed the study (mean age 39, SD= 17); all had lived in the US for more than 20 years. Analysis indicated general satisfaction with tool navigation and the content, sequencing and segmentation of materials. Participants wanted individualized patient handouts to be more accessible and easily available on paper and suggested that images and icons might enhance visual appeal. Significant tool-user interactions included problems with navigation, difficulty understanding questions and answer choices, and sensitivity to the personal nature of the topics. User-context interactions included lack of experience with technology and requests for more support during completion of the questionnaire. In addition, subjects described personal experiences or beliefs while interacting with the materials. The extent to which questions or advice approximated their own experiences influenced the ease with which they were understood. CONCLUSIONS: This study supports the utility of translating and tailoring the MOVE!23 for use in Latina women and identified several areas for improvement. Findings suggest that task execution may be enhanced when users are able to compare and reconcile the online materials with personal experiences and beliefs,

JGIM

highlighting the need to take these into consideration when tailoring software. Further research will aim to tailor and pilot the MOVE!23 software in populations of Latina women.

“SOMEDAY IT WILL BE THE NORM”: PHYSICIAN PERCEPTIONS OF THE CLINICAL UTILITY OF WHOLE GENOME SEQUENCING Jason L. Vassy1,2; Kurt D. Christensen2; Melody J. Slashinski3; Denise Lautenbach2; Jill Oliver Robinson3; Jennifer A. Blumenthal-Barby3; Lindsay Feuerman3; Robert C. Green2; Amy L. McGuire3. 1VA Boston Healthcare System, Boston, MA; 2Brigham and Women’s Hospital, Boston, MA; 3Baylor College of Medicine, Houston, TX. (Tracking ID #1934488) BACKGROUND: There is excitement that medicine is entering a genomic era, in which knowledge of one’s genetic makeup can guide medical decision-making. Whole genome sequencing (WGS) can identify 3–5 million genetic variants for an individual. However, the clinical significance of most of these variants is currently unknown, and the specific clinical contexts in which WGS might add value to current standards of care have not been identified. The uptake of WGS into medical practice will largely depend on whether and how physicians perceive it to improve patient care. METHODS: We have enrolled primary care physicians (PCPs) and cardiologists to participate with their patients in the MedSeq Project: a randomized trial of WGS in the clinical care of healthy adult primary care patients and cardiomyopathy patients. We used mixed methods to describe the clinical utility that physicians perceive WGS to have. In a baseline survey on enrollment, we asked physicians to rate on a scale of 1–10 (“Not at all useful” to “Extremely useful”) how useful they thought WGS would be for “managing [their] patients’ health” at two times: now and in the future. We categorized responses as not useful (1–5) or useful (6–10) for analyses. We then conducted in-depth semi-structured interviews to explore whether and in what circumstances physicians might perceive WGS to have clinical utility. Transcribed interviews were consensus-coded with thematic content analysis to identify emergent themes. RESULTS: Of 17 physicians (mean age 52 years), nine were women, and five were of non-white race. Nine were PCPs, and 11 reported no genetics training “beyond the typical medical school curriculum.” Only 4 (24 %) said that WGS had utility now, but 15 (88 %) predicted it would have future utility for their patients’ care. We identified three major themes influencing physician perceptions of utility. 1) Inevitability: Physicians expected WGS to be widely integrated into medical practice in the near future. When asked his opinion about the routine use of WGS in clinical care, one responded, “That’s like saying, ‘What do I think about getting cholesterol tests on people?’ I mean, it’s a reality, and I think it’s going to happen.” 2) Uncertainty: Physicians perceived WGS results to have too little analytic and clinical validity at present to be useful for patient care. One commented, “It’s like doing a CT scan and finding a nodule and not knowing what it means. And the next thing you know, you have a biopsy, and then you have a complication from the biopsy.” 3) Contexts of utility: In particular, physicians anticipated that WGS may 1 day be useful for nudging patient behavior change, tailoring drug therapy (pharmacogenomics), and efficiently targeting preventive health maintenance. One said, “If it turns out that we don’t have to do colonoscopies every 10 years on 100 % of the people, maybe that will save a lot of money. If we know who we really should be doing PSA’s or mammograms on, instead of doing them on everybody, we can save a lot of money, stop creating a lot of harm and disability.” Physicians often used analogies to current medical testing to describe their perceptions of WGS. CONCLUSIONS: Physicians are skeptical about the current clinical utility of WGS for general patient care but anticipate a near future when the clinical integration of WGS will be widespread. Evidence that WGS improves health outcomes will likely be necessary to persuade physicians of its utility in patient care.

“WE TORTURE THEM BEFORE THEY DIE, EVEN THOUGH WE KNOW THAT THEY ARE GOING TO DIE:” THE INFLUENCE OF HOSPITAL POLICIES AND CULTURE ON ETHICAL DNR DECISION MAKING IN THE US AND UK Elizabeth Dzeng1,2; Michael P. Kelly2; Sydney M. Dy1; Thomas J. Smith1; Martin Roland2; Stephen Barclay2; David M. Levine1. 1Johns Hopkins School of Medicine, Baltimore, MD; 2University of Cambridge, Cambridge, United Kingdom. (Tracking ID #1939234) BACKGROUND: Physicians face ethical challenges when there are conflicts between autonomy and beneficence, which is defined as acting in the best interest

JGIM

ABSTRACTS

of the patient. Shared decision-making, where the patient and doctor form a partnership to decide on treatment plan, has emerged as a model on the spectrum in between medical paternalism and patient autonomy. In the United States, there are significant variations in hospital policies regarding Do Not Resuscitate (DNR) decision-making. In the United Kingdom, it is standard practice for doctors to make unilateral decisions to institute a DNR order in situations where survival is unlikely. The aim of this study is to explore how doctors in the US and UK balance autonomy and beneficence in decisions regarding DNR status at the end of life and how their ethical values and clinical approach are shaped by their institution’s culture and legal frameworks. METHODS: Semi-structured in-depth interviews were conducted with 58 internal medicine doctors across four sites (New York, Baltimore, Seattle and Cambridge, England), who were routinely involved in DNR conversations with end of life patients. Participants were purposively sampled by stage of training and medical subspecialty to provide a wide range of perspectives and contribute to understanding emerging patterns and themes. Interviews lasted on average 60 min and were audiotaped and transcribed. Transcripts were analyzed and double coded using thematic analysis with an interpretive approach. RESULTS: 13–16 doctors participated at each site. Approximately equal numbers of senior faculty, junior faculty, fellows and residents were interviewed. Experienced doctors at all sites generally felt comfortable engaging in shared decision-making and when clinically appropriate, making more paternalistic recommendations against resuscitation. However, there was greater variation amongst residents and fellows that was site specific. At each hospital, there appeared to be a dominant mode of decision-making that impacted trainees’ perspectives on DNR decision-making. Although hospital policies primarily dictate whether patient preferences should be honored, trainees at hospitals with policies that prioritize patient autonomy often interpreted them to mean that they should not provide clinical recommendations. They felt compelled to offer choice even if the chance of successful resuscitation was negligible. In contrast, trainees at hospitals that had policies that encouraged paternalism or shared decision-making, which included the hospital in the UK and one in the US, felt more comfortable expressing clinical judgment against resuscitation when appropriate. Trainees at hospitals that prioritized autonomy more frequently perceived conflict between honoring autonomy and acting with beneficence and recalled more conflicts with patients regarding preferred decision. CONCLUSIONS: Institutional norms and policies influence how doctors develop their professional attitudes and practices regarding DNR decisionmaking. Doctors at the American hospital that encouraged shared decisionmaking, had attitudes and beliefs about DNR decision-making that were more similar to British doctors than doctors at the other American sites. Hospitals characterized by higher levels of shared decision-making and paternalism provided an environment where trainees felt more comfortable withholding resuscitation when success was unlikely and in doing so, experienced fewer patient conflicts.

“WHAT DOES A GOOD LIFESTYLE MEAN TO YOU?” A QUALITATIVE ANALYSIS OF FIRST YEAR MEDICAL STUDENTS’ VIEWS Kimberly Clinite1; Alex Garnett1; Shalini Reddy1; Kent DeZee2; Stephanie Kazantsev1; Steven Durning2; Jennifer R. Kogan3. 1University of Chicago, Pritzker School of Medicine, Chicago, IL; 2Uniformed Services University, Bethseda, MD; 3 Perelman School of Medicine at the University of Pennsylvania, Philadelphia, PA. (Tracking ID #1938451) BACKGROUND: Medical students consider a specialty’s lifestyle when selecting a career; meeting the physician workforce shortage requires an understanding of how students define a “good” lifestyle. Understanding the lifestyle expectations of entering students may help medical educators guide students towards careers that will fulfill their ambitions and serve the health needs of our society. The purpose of this study was to determine how first year medical students define a good lifestyle. METHODS: First year medical students at 11 schools were surveyed at matriculation in 2012. The survey included the open-ended question, “When someone says ‘That specialty has a good lifestyle,’ what does that mean to you?” Two investigators performed a directed content analysis of responses. Descriptive statistics of the themes were calculated. RESULTS: 60 % (1020/1704) of students completed the survey; 934 answered the open-ended question. Responses encompassed four major themes: schedule control (66 % of students), off-time (54.5 % of students), financial aspects (35 % of students), and work-life (21 % of students). Schedule control was described with multiple

subthemes (reasonable hours, flexible hours, predictable hours, and limited call). Nine student responses were categorized as disparaging. CONCLUSIONS: This study has helped establish how matriculating medical students define lifestyle, a known factor in career selection. Matriculating medical students’ definition of lifestyle is largely defined by schedule control, though students’ concept of schedule control varies. Salary associated with a good lifestyle ranged from $175,000–$400,000. Understanding early perceptions of career decisionmaking may provide a context for medical educators to begin conversations with students about career choice. Additionally, specialties experiencing physician shortages should consider how their specialty will evolve to meet the needs of society while addressing the concerns of trainees. Policies that hope to successfully address the medical workforce shortages must account for students’ desire to have control of their work schedule.

“WHEN YOU ARE OLD, HAVE CANCER AND A STORM IS APPROACHING” -THE EFFECTS OF HURRICANE SANDY ON CANCER PATIENTS AND PROPOSALS ON POTENTIAL INTERVENTIONS Misako Nagasaka1; Hassan Alsabbak1; Zaid Aljuboori2; Koji Sasaki3; Alfred Burger1; Benjamin Levy1. 1Beth Israel Medical Center, New York, NY; 2Bronx Lebanon Hospital, New York, NY; 3MD Anderson Cancer Center, Houston, TX. (Tracking ID #1939110) BACKGROUND: The estimated loss to the nation’s economy from Hurricane Sandy has climbed to as much as $50 billion, making it one of the nation’s most costly disasters. While data on the economic impact of natural disasters are often reported in a timely manner, very little data exists on the impact of natural disasters to health care, especially in cancer care. METHODS: A retrospective chart review was performed to examine the impact of Hurricane Sandy and the subsequent blackout in southern Manhattan on emergency room (ER) visits of patients with a diagnosis of cancer. The records of cancer patients who had visited ER during the week of Hurricane Sandy (10/ 29/2012 to 11/4/2012) were reviewed. For control, ER visits of cancer patients of the same dates 1 year prior (10/29/2011 to 11/4/2011) were reviewed and compared. All data were extracted from the ER and/or inpatient electronic medical records. Collected data included age group, sex, cancer type, chief complaint, ER diagnosis and admission status. The diagnosis of cancer was extracted from the “History of Present Illness” or “Past Medical History” and included solid as well as hematologic malignancies. For patients who were admitted, additional data including metastatic disease, comorbidities, length of stay and mortality were extracted. RESULTS: During the week of Sandy, 144 ER visits of cancer patients were identified. One hundred eighteen were identified the prior year. In patients over age 70, 84 visits were observed during Sandy, while only 41 visited the previous year (p value: 0.000143). The chief complaint of “shortness of breath (SOB) or asthma” was significant, with 21 ER visits in 2012 and only eight in 2011 (p value: 0.045). Out of the 21 visits of SOB, 10 required continuation of their current therapy (8 requested home oxygen therapy and two requested dialysis), while these requests were unobserved the prior year. During the Sandy week, 77 ER visits resulted in admission, while 64 were admitted the year prior (p value: 0.902). Out of 77, seven expired and from 64, four expired (p value: 0.759). The ER admitting diagnoses of the seven who expired during Sandy included a septic patient with fungating head and neck cancer, a brain cancer patient with respiratory failure, a lung cancer patient with pleural effusion, a sarcoma patient with pneumonia, an ovarian cancer patient with anemia, a melanoma patient with acute coronary syndrome and a lung cancer patient with SOB requesting home oxygen therapy. Corresponding data from the year prior included a lung cancer patient with pancytopenia, a CLL patient with pneumonia, a septic patient with breast cancer and a lung cancer patient with suicidal ideation. CONCLUSIONS: Although there were no significant increase in the total ER visits of cancer patients during Sandy, there were significantly more ER visits during Sandy in cancer patients over age 70, suggesting increased vulnerability to this cohort. To our knowledge, this is one of the first few studies to evaluate cancer care related to natural disasters. In retrospect, the admission and death of a lung cancer patient with SOB requesting continuation of home oxygen therapy may have been the one potentially preventable. Cancer patients requiring additional medical assistance such as home oxygen therapy and dialysis may benefit from a natural disaster back-up plan. House calls, early and prioritized evacuation of these patient groups may be some options to minimize the effects of an approaching natural disaster.

ABSTRACTS Top 10 Chief Complaints Chief Complaints

2011

2012

p value

SOB or asthma Chest pain or discomfort Abd pain Fall Altered mental status Weakness Pain in extremeties Requesting dialysis Nausea and vomiting Intoxication

8 7 10 3 4 4 9 0 3 4

21 11 10 10 8 7 7 6 6 6

0.049 0.6 0.8 0.1 0.5 0.7 0.43 0.034 0.51 1

30 DAY READMISSION RATE FOR PATIENTS DISCHARGED WITH CHRONIC OBSTRUCTIVE PULMONARY DISEASE (COPD): ANALYSIS OF 1,858,618 ADMISSIONS. WHICH PATIENTS ARE MOST AT RISK? Devin B. Malik; Sourabh Aggarwal. Western Michigan University School of Medicine, Kalamazoo, MI. (Tracking ID #1939043) BACKGROUND: COPD is a major cause of morbidity and is associated with a high 30-day re-admission rate (RR) and economic burden on health care. This study was done to determine demographic parameters associated with high 30-day RR secondary to COPD METHODS: Nationwide Inpatient Sample data was used to extract data of patients discharged with DM with complications for years 2009–2011 using clinical classification software (CCS). NIS represents 20 % of all hospital data in US. All the patients who were discharged with primary diagnosis of COPD and readmitted within 30 days were identified and categorized based on admitting diagnosis. Patients were classified as readmissions secondary to COPD as a primary cause, readmissions with COPD as a secondary cause and non-COPD associated readmissions. Statistical analysis was done using SPSS. RESULTS: We identified a total of 1,858,618 admissions for COPD nationwide during the study period with total 30-day RR of 21.06 % and RR of 7.22 % secondary to COPD as a primary cause (Table 1). Age group (45–64), males, patients under Medicaid, patients in low median income for zip code and in metropolitan areas had higher 30-day re-admissions secondary to COPD as a primary cause (P1 year from 1/2009 to 7/2012 in Aetna’s national HMO (2.1 million), 278,975 persons aged 18 to 64 years were selected if they filled ≥2 Rxes for non-injectable Schedule II or III OAs. From claims and enrollment files for these subjects, a database was developed including demographics, encounters, ICD-9-CM codes, and filled Rxes. For this analysis, we selected persons aged 45–64, no cancer diagnosis (except basal cell) within 6 months of an OA Rx, and no OAs for opioid dependence. Follow-up after the first OA Rx was divided into 6-month intervals until 1/2012 (excluding incomplete intervals). The total morphine equivalent dose (TMED) was calculated from all filled OAs Rxes for each 6month interval and categorized for analysis as none and approximate quartiles: NO risk (0 days); LOW risk (1–190 TMED); MEDIUM risk (191–450 TMED); HIGH risk (451–1830 TMED); and VERY HIGH risk (>1830 TMED). Days covered by filled Rxes for other drugs used for CNCP—benzodiazepines, zolpidem, and antidepressants—were analyzed per 6-month interval. From ICD9 codes, we created indicators for five categories of CNCP, mental health conditions, and substance abuse. Repeated measures logistic regression was used for hospitalization (Y/N) and repeated measures Poisson regression was used for inpatient days. All analyses adjust for patient demographic, clinical, and treatment factors. No significant interaction was observed for OA risk and age; therefore, we only report results for OA risk. RESULTS: The cohort totaled 109,292 persons with a mean age of 53.7 years (SD =5.5), 54.7 % female, and residence in South (46.9 %), West (16.8 %), Midwest (6 %) and Northeast (30.3 %). In the cohort’s 253,531 six-month follow-up intervals, the proportion hospitalized was 6 % and median hospital days was 4 (IQR=3–6). For hospitalization, the adjusted odds ratio [95 % CI] for OA risk categories vs. NO risk were: VERY HIGH: 1.45 [1.38, 1.54]; HIGH: 1.15 [1.08, 1.21]; MEDIUM: 0.98 [0.93, 1.04]; LOW: 1.04 [0.98, 1.10]. For inpatient days, the incident rate ratio [95 % CI] for OA risk vs. NO risk were: VERY HIGH: 1.47 [1.34, 1.62]; HIGH: 1.10 [1.01, 1.20]; MEDIUM: 0.87 [0.80, 0.95]; and LOW: 0.95 [0.87, 1.03]. CONCLUSIONS: High or very high total dose of filled OA Rxes in a baseline 6month interval is associated with an increased risk of subsequent hospitalization and more inpatient days independently of age, pain-related comorbidities, and other prescribed medications commonly used for pain. Reasons for these hospitalizations need to be examined to inform interventions to address this high use of costly inpatient care.

A MIXED METHOD STUDY OF THE QUALITY OF WRITTEN FEEDBACK BY ATTENDINGS OF INTERNAL MEDICINE RESIDENTS Cynthia Kay1; Jeffrey L. Jackson2,1; Wilkins Jackson3. 1Medical College of Wisconsin, Milwaukee, WI; 2Zablocki VAMC, Milwaukee, WI; 3University of Wisconsin-Milwaukee, Milwaukee, WI. (Tracking ID #1920218) BACKGROUND: Residents uniformly criticize the quality of feedback they receive. The purpose of this study was to evaluate the quality of feedback in an internal medicine residency program. METHODS: Subjects for this retrospective analysis were Medical College of Wisconsin (MCW) internal medicine residents who completed residency training between 2004 and 2012. Over the 8 years, there were 6603 evaluations by 334 attendings. These evaluations followed ACGME guidelines and assessed residents in six domains and included a comment box. From these 6603 written comments, we randomly selected 500 and coded them using grounded theory methods. An initial coding tree was developed by two raters (CK, JLJ). Once the tree was saturated, 25 % of utterances were double coded with good inter-rater reliability (Spearman: 0.94). Utterances were further classified as high, moderate or low quality, also with good inter-rater reliability (Quadratic Kappa: 0.83). Focus groups were also conducted with medicine faculty and housestaff. First written comments were placed on notecards and the groups sorted the cards into high, moderate and low quality feedback. As they did this, they discussed the characteristics that merited this rating. The last few minutes of each focus group was spent reflecting and developing criteria for high, moderate and low quality feedback. RESULTS: The majority of the attending’s 500 written comments were rated as moderate quality (65 %); 22 % were rated as high and 13 % low. The most common type of feedback (29 %) was nonspecific comments, often about the resident’s personality, 16 % about patient care, 14 % interpersonal communication, 7 % medical knowledge, 6 % professionalism, 4 % each commented on practice based learning or systems based practice. The majority of written feedback comments (88 %) were positive, 8 % were negative and 4 % were neutral; three ABIM domains were more likely to include negative comments (medical knowledge: OR: 3.5, 95 % CI: 2.2–5.6; Practice Based Learning: OR: 2.5, 95 % ci: 1.3–4.8, Systems Based Practice: OR: 4.6, 95 % CI: 2.5–8.3). Nonspecific comments and comments on resident’s attitude or personality were much less likely to include negative comments than the other domains (nonspecific: OR: 0.22, 95 % CI: 0.13–0.39, attitude/personality OR: 0.53, 95 % CI: 0.34–0.82). Focus groups uniformly sorted quality into low, moderate and high quality feedback based on three criteria: specificity, quantitative ratings, actionable, Low quality feedback was nonspecific, moderate feedback had at least some specificity and high quality feedback included specificity plus either some aspect of quantification or was actionable. Despite the uniformity of the criteria espoused by focus groups, the inter-rater reliability between the focus groups was low, (quad kappa: 0.24). CONCLUSIONS: Written feedback suffers from lack of balance, lack of specificity and lack of actionability. Much of it is focused on personality characteristics of the resident. While the focus groups developed nearly identical criteria for quality, agreement between the groups was low because the groups disagreed as to what constituted specific and actionable feedback. Residents are justified in expressing chagrin at the quality of feedback they are receiving.

A MIXED-METHODS EVALUATION OF A VA PATIENT-CENTERED MEDICAL HOME FOR SPECIALTY CARE Michael Ho2; Catherine Battaglia1; Leah M. Haverhals2; Katherine M. Fagan2; Jackie G. Szarka3; Rebecca Alexander3; George Sayre3; Christian Helfrich3; Susan Kirsh5; David Au3; David Aron6; Julie Lowery4. 1 Eastern Colorado Health Care System, Denver, CO; 2Eastern Colorado Health Care System, Denver, CO; 3Puget Sound Health Care System, Seattle, WA; 4Ann Arbor Health Care System, Ann Arbor, MI; 5Case Western Reserve University, Cleveland, OH; 6Louis Stokes VA Medical Center, Cleveland, OH. (Tracking ID #1938016) BACKGROUND: The Veterans Health Administration (VA) implemented a pilot program called Specialty Care Neighborhood (SCN) to align with the VA’s Patient Centered Medical Home to deliver comprehensive patient-centered care. SCN is a team-based specialty care model, in which teams of specialists engage in proactive and long term collaboration with primary care in order to provide coordinated, optimal care for Veterans with the greatest specialty care needs. SCNs were implemented at 25 VAs in 20 specialties including Heart Failure, Diabetes, Cancer, Pain, Orthopedic Surgery, and HIV. We completed an evaluation of the SCN initiative for the VA Office of Specialty Care (OSC). METHODS: Findings are presented here from semi-structured interviews and staff surveys. A goal of the evaluation was to identify factors that differentiated four high

ABSTRACTS from two low performing sites as determined by OSC from regular site reports on implementation progress. The evaluation consisted of 27 semi-structured interviews with Primary Care Providers, specialists, and key stakeholders (administrators, chiefs of staff, and facility directors) at the participating sites. Additionally, SCN staff from the six sites returned 39 surveys about their experience with SCN. Interview and survey data were coded using the Consolidated Framework for Implementation Research. RESULTS: SCNs were viewed positively. Participants from all sites described improved professional relationships and communication as well as improved quality and quantity of specialty services. Interviewees felt improved relationships led to increased job satisfaction and engagement. Participants expressed frustration as to lack of clarity regarding how to implement SCN, and despite this barrier, pulled together to creatively implement SCNs best adapted to meet local needs. In structured ratings of implementation facilitators and barriers, High performing sites, in contrast to low performing sites, engaged leadership, utilized a thoughtful planning process that allowed for ample implementation time, and used systems redesign techniques to streamline processes. Additionally, there were factors from interview data that facilitated implementation at both high and low sites: 1) positive knowledge and beliefs about the program; 2) obtaining patient and provider feedback; 3) recognizing patient needs; 4) active champions; and 5) ability to adapt the program easily. Common barriers included 1) problems with hiring staff, 2) lack of commitment to long-term staff funding, 3) lack of administrative support, 4) insufficient materials for facilitating implementation, and 5) coordination and communication challenges among different providers and services. Interview data and staff survey data were largely congruent. However, survey participants reported high leadership engagement despite interview data suggesting this was less positive among low performing sites. CONCLUSIONS: The findings were used to identify areas on which to focus attention for improving dissemination of the program. Recommendations were based on factors that distinguished high from low performing sites and included recommendations to overcome barriers: 1) Provide opportunities for sharing best practices across sites; 2) Encourage innovation through site-based flexibility and individualization while limiting micromanaging; 3) Facilitate access to systems redesign expertise; 4) Minimize complexity by utilizing programs in place; and 5) Begin planning process early and involve all key stakeholders.

A MODEL TO PREDICT RISK OF CERVICAL ABNORMALITIES IN PRIMARY CARE Michael B. Rothberg; Bo Hu; Sarah Schramm; Kelly Nottingham; Glen Taksler; Xian W. Jin; Laura Lipold; Andrea L. Sikon. Cleveland Clinic, Cleveland, OH. (Tracking ID #1937479) BACKGROUND: Cervical cancer screening tests are highly sensitive but have low specificity, resulting in high false positive rates among low-risk women. Although there are known risk factors for cervical cancer, there are no validated tools for assessing risk in primary care. Using data from electronic health records, we derived and validated a risk prediction model for cervical intraepithelial neoplasia grade 2 or higher (CIN2+) lesions. METHODS: Our dataset included information on all women aged ≥30 years who underwent cervical cancer screening with cytology and human papilloma virus (HPV) testing at 33 physician practices associated with the Cleveland Clinic between 2004 and 2010. Patient risk factors at the time of screening (age, race, marital status, insurance type, smoking history, median income based on zip code and physician practice) were determined by reviewing the electronic health record. Pathology results were obtained from the Cleveland Clinic’s pathology database (for patients who underwent colposcopy with biopsy) or by imputation (for patients with abnormal cytology who did not undergo colposcopy). Imputation was based on observed associations between patient risk factors, HPV and cytology results and CIN2+ status among women who underwent colposcopy. Using CIN2+ as our outcome, we then derived a hierarchical generalized linear model with a logit link that included all patient risk factors available before screening (i.e. not HPV or cytology results) as independent variables. We then stratified patients into tertiles of low, medium and high predicted risk of CIN2+. We then compared the actual rate of CIN2+ (either observed or imputed) in the low-risk tertile to an implied threshold value, which was based on combinations of HPV and cytology results for which colposcopy is typically not recommended. The threshold was 0.20 %, which is equivalent to the rate of CIN2+ among patients with low risk ASCUS and negative HPV. The final model was validated using 100 bootstrap iterations for the entire population and noting the proportion of low risk patients who had CIN2+. RESULTS: The dataset included 99,388 women. Of these, 574 had pathology findings of CIN2+ and 175 had CIN2+ assigned by imputation (overall rate of CIN2+=0.75 %). The final multivariable model had a c-statistic of 0.73. All factors

JGIM

but race were associated with CIN2+: older age (OR 0.69 per 5 years, 95 % CI 0.66, 0.73), Medicare vs. private insurance (OR 0.51, 95 % CI 0.34, 0.75) and higher income (OR 0.97 per $5 K, CI 0.94, 0.99) were negatively associated, while being single vs. married (OR 1.50, 95 % CI 1.24, 1.82) and smoking (OR 1.66, 95 % CI 1.35, 2.04) were positively associated. Tertiles of predicted risk were 0.20 %, 0.53 %, and 1.51 %, with observed rates of CIN2+ over the same tertiles of 0.19 %, 0.52 %, and 1.53 %. Therefore, before cervical cancer screening, women in the lowest risk tertile had observed CIN2+ risk slightly below that of women with ASCUS and HPV negative (0.20 %). In 100 bootstrap iterations, the observed risk in the lowest tertile ranged from 0.13 to 0.25 % (IQR 0.18 % to 0.22 %). CONCLUSIONS: A multivariable model based on data derived from the electronic health record was able to risk stratify women across an 8-fold gradient of risk for cervical abnormalities and to identify a large group of women who are at very low risk for cervical cancer and may not require screening. After further validation, use of a similar model could enable more targeted cervical cancer screening.

A NATIONAL SURVEY OF PATIENT CENTERED MEDICAL HOME INITIATIVES Samuel T. Edwards1,2; Asaf Bitton3,4; Johan S. Hong4; Bruce E. Landon4,5. 1Boston VA Healthcare System, Boston, MA; 2Harvard Medical School, Boston, MA; 3Brigham and Women’s Hospital, Boston, MA; 4Harvard Medical School, Boston, MA; 5Beth Israel Deaconess Medical Center, Boston, MA. (Tracking ID #1932105) BACKGROUND: The reorganization and improvement of primary care through the implementation of the patient-centered medical home (PCMH) is central to healthcare delivery system reform efforts. PCMH initiatives, often organized by health plans, states or multi-stakeholder groups, seek to create the appropriate context to promote transformation of participating primary care practices into medical homes by building learning collaboratives, providing practice coaching, and reforming physician payment to support medical home services. Although there has been expansion of PCMH activity across the country, the current national scope of PCMH initiatives remains undefined. To address this gap, we conducted a nationwide survey of PCMH initiatives. We describe their size, scope, organization and payment methodology and assess their approach to practice transformation and program evaluation. METHODS: We performed a cross-sectional survey of all PCMH initiatives nationally that included payment reform as an operational feature. We identified initiatives for the survey from existing databases supplemented by formal literature review, internet searches, and by contacting known experts, state Medicaid directors, and representatives from all major national health plans. We included initiatives that were active as of February 1st, 2013 but excluded programs directed solely toward specific medical conditions or serious mental illness. The survey included descriptive questions including initiative start and end dates, numbers of participating practices, physicians and patients, and identity of key stakeholders. We asked about the selection process for practices to enter the initiative, the use of formal PCMH recognition, and payment methods. Finally, we asked about how initiatives facilitated practice transformation and plans for program evaluation. The survey was fielded from February 1st 2013–October 31st 2013 via web or telephone. Respondents were physicians, senior administrators, or senior health policy experts from each initiative. We present descriptive statistics to compare initiatives sponsored by commercial health plans, state Medicaid programs, and multiple payers. RESULTS: Of 119 identified initiatives, we received 114 responses (96 % response rate). Forty-one programs were small commercial single payer initiatives, 29 were large single payer initiatives, 23 were Medicaid only and 21 were multi-payer. In total, the programs included 63,011 providers caring for 20,389,327 patients nationally. Small commercial programs had a median of 3,896 patients and typically included a small number of practices (median = 4). Large commercial single payer initiatives included a median of 160,000 patients, often covering most or all practices in a specific region or an entire state (median 105 practices). Medicaid initiatives were typically large statewide programs (median 224,040 patients) run directly by state Medicaid agencies. Multi-payer initiatives included a median of six health plans (IQR 4–7) with a median of 187,343 patients, and 68 practices. These initiatives included commercial health plans, Medicare and Medicaid. Formal PCMH recognition was required by 69 % of initiatives. Nearly half of initiatives use practice consultants (46 %) or learning collaboratives (48.7 %), but this was more common in multi-payer initiatives (90.5 % and 95.2 %, respectively). Nearly all initiatives had a formal evaluation plan (92 %). A combination of fee for service payments (FFS) for office visits, a per-member-per month care management fee (PMPM), and pay for performance bonuses was the most common payment structure (55 % of programs). Median PMPM was $4.90 (range $0.25–$60.00), but was higher in multi-payer initiatives (median $7.00). Since 2009, fewer initiatives are time-limited pilots, shared

JGIM

ABSTRACTS

savings programs are more commonly used, and more initiatives are performing formal evaluations (table). CONCLUSIONS: In this national survey of PCMH initiatives, we demonstrate there has been a large expansion of PCMH activity over the last several years, with 114 initiatives representing over 60,000 providers and nearly 21 million patients in 2013. Initiatives are on average larger, feature more robust transformation efforts, and have higher care management fees. This expansion of PCMH initiatives suggests that the PCMH will continue to play an important role in national delivery reform efforts. Comparison of PCMH initiatives in 2009 and 2013

Total patients Time limited initiatives (%) Median patients per initiative Number multi-payer initiatives Median patients in multi-payer Use Shared Savings PMPM (range) Use Learning Collaboratives and Consultants Performing Program Evaluation

2009 (N=26)

2013 (N=114)

4,971,070 62 % 30,000 9 39,000 0% $0.50–$9.00 15 % 40 %

20,970,277 20 % 40,986 22 146,636 44 % $0.25–$60.00 37 % 92 %

A PROFILE OF HOSPITALIZED SMOKERS AND THE CARE THEY RECEIVE Ellie Grossman1; Alissa R. Link2; Erin Rogers2,3; Binhuan Wang2; Scott Sherman2,3. 1NYU School of Medicine, New York, NY; 2NYU School of Medicine, New York, NY; 3VA New York Harbor Healthcare System, New York, NY. (Tracking ID #1939163) BACKGROUND: Identification and treatment of patients who smoke is a priority treatment area for hospitals. Hospitalization can be a ‘teachable moment’ for patients contemplating behavior change, and they may be newly motivated by health consequences. Evidence suggests that appropriate treatment should begin during an inpatient stay and continue for at least four weeks post-discharge. To optimize success, treatment should include both counseling and medications. Patients with mental illness tend to smoke at higher rates than those without, and tend to die earlier of cardiovascular disease, but there is a paucity of evidence about the state of treatment offered to these patients. We aimed to describe the care offered to hospitalized smokers at two urban safety-net hospitals. METHODS: As part of a randomized controlled trial of smoking cessation interventions focused on the post-discharge time period, we assessed the care for smokers hospitalized at two large public hospitals in New York City. Adult patients were eligible for study inclusion if they: smoked at least one puff of a cigarette in the past 30 days; spoke English, Spanish, or Mandarin; were not incarcerated or in police custody; were not pregnant or breastfeeding, and had a U.S. phone number. At the time of discharge, study staff reviewed participants’ electronic medical records (EMR) for information about type of inpatient unit, history of medical or psychiatric illnesses, and tobacco treatment received during hospitalization and on discharge. RESULTS: From July 2011 to December 2013, we enrolled 1486 study participants. 511 (35 %) were on psychiatric inpatient units and 934 (65 %) were on medical/ surgical units (n=41 missing). Mean age was 47.9 years, 79 % were male. Of these patients, 17 % had EMR documentation of coronary artery disease, 12 % chronic obstructive pulmonary disease, 14 % asthma, 5 % cerebrovascular disease, 4 % congestive heart failure, 55 % significant alcohol or drug use history, and 10 % schizophrenia or schizoaffective disorder. We found EMR documentation that smoking status was assessed at admission in 1444 (97 %) and bedside counseling was provided in 1294 (87 %). Of those that received counseling, the EMR indicated that 1015 received it from nurses and 145 received it from physicians (for the rest the personnel was not specified). For 647 patients (44 %), physicians ordered smoking cessation pharmacotherapy during hospitalization (36 % NRT patch, 3 % NRT gum, 2 % NRT lozenge, 1 % other). On discharge, 274 patients (18 %) were prescribed cessation medication. Patients on inpatient psychiatric units were more likely to be counseled by their physician regarding smoking cessation (odds ratio (OR) 1.93, 95 % CI 1.20–3.09) and were more likely to be prescribed cessation medication in hospital (OR 3.21, 95 % CI 2.51–4.11) than were patients on non-psychiatric units. There was no association between type of inpatient service (psychiatric vs. non-psychiatric) and likelihood of prescription of smoking cessation medication at discharge. CONCLUSIONS: At urban safety-net hospitals, patients who smoke have high rates of mental illness and substance use disorders. Although assessment and counseling are provided at relatively high rates, there is much room for improvement in prescribing smoking cessation pharmacotherapy in the hospital and on discharge.

A QUALITATIVE ANALYSIS OF HOSPITAL LEADERS OPINIONS ABOUT PUBLICLY REPORTED MEASURES OF HEALTH CARE QUALITY Sarah L. Goff1,2; Tara Lagu2,1; Amy S. Shatz2; Penelope S. Pekow2; Peter K. Lindenauer2,1. 1 Baystate Medical Center/Tufts University School of Medicine, Springfield, MA; 2 Baystate Medical Center, Springfield, MA. (Tracking ID #1937659) BACKGROUND: Public reports of hospitals’ performance on quality measures are intended to both encourage hospital leaders to stimulate improvement activities and to enable consumers to make more informed decisions about where to seek care. How effective public reporting is at fostering quality improvement may be influenced by whether and to what extent hospital leaders find the measures valid and meaningful. We sought to qualitatively describe hospital leaders’ views regarding quality measures reported on the Centers for Medicare and Medicaid Services’ Hospital Compare website. METHODS: We mailed or e-mailed a 21-item questionnaire to the chief executive officer and the senior executive responsible for quality at a stratified random sample of U.S. hospitals. Stratification was based upon performance on selected quality measures to achieve a balance of high, average and low performers. While the quantitative portion of the study assessed attitudes regarding publicly reported quality measures, opinions regarding quality measures were elicited using an open-ended prompt at the end of the survey, “Please share your additional thoughts about publicly reported measures of healthcare quality, including strengths or weaknesses of current measures, ideas for new measures, etc.” We qualitatively analyzed responses to this prompt, identifying the presence and frequency of pertinent themes in an iterative process. RESULTS: Leaders (n=480) from 380 of the 630 hospitals surveyed (60 %) completed the questionnaire. Of those who completed the questionnaire, 146 leaders from 131 hospitals responded to the open-ended prompt. The majority of respondents to the open-ended prompt were chief executive officers (27 %), chief quality officers or vice presidents of quality (17 %) or chief medical officers (19 %). Half of the responses to the prompt came from leaders at urban hospitals and 68 % were from leaders at non-teaching hospitals. Most respondents were from hospitals that had fewer than 200 beds (41 %), with 23 % from hospitals with more than 400 beds. Forty-one percent of the respondents’ hospitals had performance scores higher than the national average and 27 % scored lower than the national average. Major themes identified included: 1) Positive views of the value of quality measures (12.5 %, e.g., measuring quality can stimulate change and enhance patient engagement); 2) Concerns regarding the validity and utility quality measures (63 %, e.g., measures fail to measure what is important, inadequate risk-adjustment); 3) Challenges to improving quality scores (e.g., inability to impact factors affecting care outside of the hospital, such as patient adherence to prescribed medication); and 4) Recommendations for improving measures (e.g., compare similar hospitals to each other). Many of the positive views expressed were general support for the concept of measuring quality: “Oversight and comparison with peers/standards is necessary and beneficial.” Criticisms of quality measures included concerns about whether doctors can actually influence change in the entities measured: “…the scores for which the doctors are held accountable are subject to profound influence by others besides the doctor to whom the score is attributed.” Challenges described included lack of resources: “Resources are stretched to the limit with meaningful use and data reporting!” Recommendations for improvement in quality measures included an emphasis on the continuous learning cycle, suggesting new targets for quality improvement should be identified as older measures reach acceptable levels of compliance: “…care measures should evolve as compliance becomes ingrained thus continuing the improvement cycle.” CONCLUSIONS: Among hospital leaders who responded to an open-ended prompt eliciting their views about quality measures some expressed support for the concept of measuring quality, while the majority offered criticisms concerning the validity and utility of current quality measures or described challenges to improving quality. As current quality measures are reassessed and new measures developed, the concerns of those responsible for implementing system changes to address quality should be taken into account.

A QUALITATIVE ANALYSIS OF PATIENT FACTORS ASSOCIATED WITH HOSPITAL READMISSION Colleen M. Rafferty; Tyler Sechrist; Heather Stuckey; Nicole Swallow; Cynthia H. Chuang. Penn State Hershey Medical Center, Hershey, PA. (Tracking ID #1936483) BACKGROUND: Given the focus on hospital readmissions as a quality indicator, health care organizations nation-wide are interested in finding ways to reduce hospital readmissions. Existing readmission prediction models have relied mainly on clinical

ABSTRACTS and administrative data. A weakness in these methods is that they are not informed by the patient perspective. Exploring patient experiences with readmission may broaden our understanding of what leads to readmission, thereby facilitating a more patientfocused approach to identifying high-risk patients, and enabling institutions to allocate resources to the highest risk patients and to the most cost-effective and efficacious interventions. We conducted qualitative interviews with readmitted patients to identify more precisely the patient-centered factors associated with readmission. METHODS: We conducted semi-structured interviews with a convenience sample of hospitalized patients readmitted to the Internal Medicine service within 30 days of discharge to home from a previous hospitalization at our institution, a 500-bed tertiary care academic medical center. Patients were eligible if they received their primary medical care within our medical system and were readmitted from June–August 2011 (n=19). Open-ended interview questions addressed the patients’ perspectives on the medical care they received during the previous admission, access to outpatient care, the patient’s knowledge of his/her medical conditions and treatments, ability to provide self- care, and social support available to them. We also encouraged the patient to discuss from his/her perspective what could have been done differently to help prevent the readmission. The contents of the interviews were transcribed and a thematic analysis was conducted. We also conducted a medical chart review to determine if the patient had a scheduled follow-up appointment that they attended. RESULTS: We found that the majority of patients had a limited understanding of their medical conditions and their medications. Many patients relied heavily on an advocate (family member or friend) to assist them in their disease management. The vast majority of patients had identified advocates, but we found that not all advocates were effective in providing the needed assistance. Also, most patients did not view their health as under their control, but identified an external locus of control (i.e. the doctors, the disease process, God). All patients identified a primary care physician, and all had been scheduled for follow up appointments within 5–24 day after discharge, but 9 of 19 patients were readmitted prior to their scheduled appointment (8/19 readmitted within 7 day, 10/19 within 10 day). The majority of patients did not feel their readmission was preventable, but attributed it to the disease process, or an issue that was not viewed by the patient as avoidable or predictable. CONCLUSIONS: These findings suggest potential patient-centered factors that may be used to help identify patients at high risk for readmission: insufficient understanding of medical condition or medications, lack of an effective advocate, a belief that one’s health is not under the one’s own control, and lack of timely access to outpatient follow up care. Early assessment of these risk factors may help target appropriate interventions to high risk patients to help prevent hospital readmission.

JGIM

guide system assessments with a series of “regular” and “expert” users of these systems. Outputs from these evaluations were recorded from each system using screen shots, video and audio transcripts of user and designer comments. Recordings were transcribed, analyzed, and used to inform the content of onsite visits to each medical center to obtain additional system information from medical leadership, IT experts, users, and pharmacists. Inter- and intra-system characteristics were analyzed descriptively for patterns, themes, and variations. RESULTS: We identified a variety of ways drug names are displayed in CPOE systems that could potentially contribute to medication errors. Recurrent issues and themes that emerged included: (a) Brand versus generic name: Presentation of brand and generic names varied both across and within systems depending on task functionality. In nine of the ten CPOEs assessed, for example, one brand name search (Plavix) yielded results displaying the drug in eight different ways. Some systems used all caps, some only initial caps, some used both. Some were inconsistent in the drug name searched and drug name returned. One system displayed two columns which listed both the brand and generic names, others included only one, and others presented one or the other parenthetically, based on the search term. One system allowed ordering by generic name only. If one searched by brand name, the medication did not appear; we observed end user frustrations resulting from inability to search for and find desired drugs searched non-generically. Systems also used varied methods to display brand and generic options to steer prescribing towards preferred products (typically the generic). (b) Combination products: In most systems, searching for a combination drug often yielded the brand name with the ingredients listed parenthetically. We rarely saw systems provide explanations of ingredient strengths. In some systems, searching the brand name yielded only the brand name, without its generic ingredients. (c) Drug-name modifiers: Across systems, drug modifiers seemed to have some consistency in both representation and display. Drug modifiers appeared primarily at the end of the drug name, in all caps, and were abbreviated (e.g. XR, DS, etc.). They were rarely truncated. (d) Extraneous additions to drug name fields: Across systems, we observed several instances of nonsystematically populating (jerry rigging) name fields with additional information ranging from drug indication to which facility the drug could be ordered from. (e) Font, text size, capitalization: These all varied widely with no standardized approaches seen both across and often within systems. CONCLUSIONS: There is little consistency in the ways that leading CPOE systems search for and present drug names. This lack of standardization was seen both across systems and often within a single system or institution’s CPOE systems. Brand vs. generic, combination products and drug name modifiers were particularly inconsistent, each presenting the potential for user frustration, confusion, and erroneous orders.

A QUALITATIVE ASSESSMENT OF CPOE AND THE EFFECTS OF VARIATION IN DRUG NAME DISPLAY Alexandra Robertson1; Arbor J. Quist1; Thu-Trang Thach1; Lynn A. Volk2; Adam Wright1,3; Shobha Phansalkar1,4; Sarah P. Slight1,5; David W. Bates1,3; Gordon D. Schiff1,3. 1Brigham and Women’s Hospital, Boston, MA; 2Partners HealthCare Systems, Inc., Wellesley, MA; 3Harvard Medical School, Boston, MA; 4Wolters Kluwer Health, Indianapolis, IN; 5The University of Durham, Durham, United Kingdom. (Tracking ID #1938546)

A QUALITATIVE STUDY EXPLORING THE VULNERABILITIES OF COMPUTERIZED PHYSICIAN ORDER ENTRY SYSTEMS Sarah P. Slight1,2; Tewodros Eguale1,3; Mary Amato1,4; Andrew C. Seger4; Diana L. Whitney5; David W. Bates1,6; Gordon D. Schiff1,6. 1Brigham and Women’s Hospital, Boston, MA; 2 Durham University, Stockton on Tees, United Kingdom; 3McGill University, Montreal, QC, Canada; 4MCPHS University, Boston, MA; 5Baylor College of Medicine, Houston, TX; 6Harvard Medical School, Boston, MA. (Tracking ID #1935926)

BACKGROUND: Previous work suggests that prescribing errors account for the highest proportion of medication errors (Lisby, 2010) and that one in every four reported in the U.S. can be linked to drug name confusion (Lambert, 1999). Computerized provider order entry (CPOE) has been widely proposed, implemented, and shown to reduce errors associated with traditional, handwritten, medication prescribing. However, CPOE has not fully lived up to its predicted benefits and many attribute these shortfalls to suboptimal design of systems that are “neither interoperable nor easy to use” (Kellerman, 2013, Health Affairs). Given recent calls for CPOE performance testing post-implementation (Thompson, 2010 & Bates, 2013) and the proportion of medication errors which occur during the prescribing phase, we undertook a study to assess drug name design and display issues in CPOE including: look-alike-sound-alike pairs, similar name adjacency errors, font, and visual display to better understand their potential to contribute to medication errors. METHODS: As part of an FDA Task Order, this qualitative study assessed 10 leading vendor and home-grown CPOE systems in six major U.S. academic medical centers via remote walk-throughs and on-site visits. To standardize assessment, the team developed and iteratively refined a CPOE Assessment Tool (CAT) relating to 18 elements defined by the FDA to systematically test the potential for drug name/ identification errors across various stages of the medication ordering process. This tool incorporated specific medication ordering, review, and deletion scenarios used to

BACKGROUND: Computerized Physician Order Entry (CPOE) systems can prevent medication errors in both inpatient and outpatient settings. Depending on how they are designed, however, they can fail to optimally prevent various prescribing errors or introduce new errors. The Institute of Medicine report Health IT and Patient Safety: Building Safer Systems for Better Care recommended that specific examples of potentially unsafe processes and risk-enhancing interfaces be identified and shared amongst the health IT community. This study aims to test the vulnerabilities of a wide range of CPOE systems to different types of medication errors, and to develop a more comprehensive understanding of how CPOE human factors design could be improved. METHODS: As part of a National Patient Safety Foundation-funded project, we examined a range of leading vendor and home grown CPOE systems (e.g., Cerner; Epic; Medi-tech; LMR; BICS; GE Centricity) in diverse organizations in United States and Canada. Typical users at each of 16 sites were asked to enter 13 different orders on test patients based on scenarios of previously reported CPOE errors. Users were encouraged to use both usual practice and, where-needed, workarounds to enter the erroneous orders, as well as reflect on their overall knowledge and experience of using their system. A research pharmacist and research assistant independently observed test users enter each order and rated the ease or difficulty of these entries using standardized operational definitions.

JGIM

ABSTRACTS

RESULTS: Ease of entry of erroneous orders and the generation of alert warnings in different CPOE systems was highly variable and appeared to depend on a number of factors including how the order information was entered (i.e., in a structured or unstructured way); whether a specific alert functionality (e.g., duplicate-drug checking) was operational in the system; and which drugs or drug combinations were included in the clinical decision support algorithms. Test users found the wording of many of the alert warnings confusing and expressed frustrations with the way irrelevant warnings appeared on the same screen as those more relevant to the current order. The timing of alert warnings also differed across CPOE systems, with dangerous drug-drug interaction warnings displayed, for example, only after both Imdur® (isosorbide mononitrate) and Revatio® (sildenafil) had been entered and the order signed off in two CPOE systems. Alert warnings also varied in their level of severity in different systems even within the same institution (e.g., test user was presented with a hard stop alert warning in the inpatient system after entering Lovenox® (enoxaparin sodium injection) 40 mg subcutaneous daily followed by a second order for Lovenox® 100 mg subcutaneous twice daily, whereas the equivalent outpatient system produced an alert for the same order that was easily overridden). Test users commented on how there may be situations where they needed to prescribe the same drug twice, requiring workarounds such as entering the brand name of the drug for the morning dose and the generic name for the evening dose, to avoid getting duplicate drug alert warnings. CONCLUSIONS: We found a high degree of variability of vulnerability in ordering and alerting between different CPOE systems. Detailed qualitative analysis of both the observed CPOE functionality and users think-aloud comments provided rich insights into the ways both systems and users were susceptible to ordering errors. System developers and users need to be aware of these vulnerabilities and build in protection strategies to re-engineer their systems to ensure safer prescribing.

A RANDOMIZED TRIAL OF A COMMUNITY HEALTH WORKER LED INTERVENTION TO IMPROVE DIABETES INTERMEDIATE OUTCOMES AMONG LATINOS PATIENTS WITH POORLY CONTROLLED DIABETES Olveen Carrasquillo; Yisel Alonzo; Cynthia Lebron; Natalie Ferras; Ernesto ReyesArrechea; Hua Li; Sonjia Kenya. University of Miami, Miami, FL. (Tracking ID #1936786) BACKGROUND: Diabetes disproportionately impacts Latinos. Community Health Workers (CHWSs) are one approach that may improve health outcomes in this population. However, evidence from rigorous randomized studies is limited. METHODS: The Miami Health Heart Initiative examined the impact of a comprehensive community health worker (CHW) intervention on diabetes intermediate outcomes of systolic blood pressure (SBP), lipids (LDL) and Hemoglobin A1C (HbA1c). We recruited 300 Latino patients with poorly controlled diabetes (HbA1c>=8.0 %) ages 35–65 from the primary care clinics of Miami-Dade county’s public hospital. Subjects were randomized to usual health care or a comprehensive structured CHW 1 year intervention consisting of home visits, phone calls and group education sessions. The intervention included patient navigation as well as assistance with social and non-medical needs. An RA blinded to group assignment conducted initial and follow-up evaluations at 12 months. We used linear mixed models to statistically test for the impact of our intervention on outcomes using intention to treat analyses. RESULTS: The mean age of our patients was 56 +7 years, 55 % were female and mean BMI was 31.6 +7.4 kg/m2. Over half (55 %) were sedentary (IPAQ) and median daily fruit and vegetable was two (BRFSS). Cubans made up 29 % of our sample with no other Hispanic ethnic subgroup representing over 15 % of the sample; 47 % of respondents scored the lowest possible acculturation score (Marin-Marin). Intervention patients received a median of five home CHW visits and 25 phone calls; 84 % of intervention group participants had at least 12 CHW contacts over the course of the year. Participation at group session was more skewed with 52 % of participants not attending any sessions. Of those that attended at least one session, average attendance was four sessions per participant. At 1 year the retention rate was 72 % and similar in both arms. Data on outcomes by group is shown in Table 1. CONCLUSIONS: In a heterogeneous Latino population with poorly controlled diabetes, a rigorous CHW intervention resulted in statistically and clinically meaningful changes in HbAIC but not in blood pressure or lipids. The impact of the intervention on the latter two outcomes may have been limited by the fact nearly half of patients were at the SBP and LDL targets prior to the intervention.

Figure 1: Baseline and One Year Outcomes in MHHI

HbA1C (%)*

SBP (mm/Hg)

LDL (mg/dl)

Baseline One Year Change Baseline One Year Change Baseline One Year Change

Usual Care

CHW

9.03 9.25 +0.21 135 134 −1 108 111 +3

9.19 8.81 −0.38 131 128 −3 100 99 −1

*p=0.01 for HbA1C NS for SBP and LDL

A REDUCTION IN TELEMETRY ORDER DURATION DOES NOT CHANGE TELEMETRY ORDERING FREQUENCY AND REDUCES UTILIZATION Joel Boggan1,2; Ann Marie Navar-Boggan4; Ryan D. Schulteis3,2; David Simel6,5. 1Durham Veterans Affairs Medical Center, Durham, NC; 2Duke University Health System, Durham, NC; 3Durham Veterans Affairs Medical Center, Durham, NC; 4Duke University Health System, Durham, NC; 5Duke University Health System, Durham, NC; 6Durham Veterans Affairs Medical Center, Durham, NC. (Tracking ID #1940176) BACKGROUND: American Heart Association guidelines for appropriate use of telemetry have not decreased inappropriate use of in-hospital cardiac monitoring. We evaluated how changing the default order set for telemetry affected overall telemetry utilization. METHODS: The computerized ordering system at the Durham VA Medical Center was changed so that the maximum number of hours of cardiac monitoring was reduced by 33 %, from 72 to 48 h, before requiring a renewal order. No other intervention or general education about telemetry appropriateness was done. We compared the frequency of orders/week for telemetry and duration of cardiac monitoring for 12 weeks before through 6 weeks after this intervention. To assess the safety of the intervention, we measured the number of rapid response team (RRT) or code events occurring at any time during the hospital stay of general medical or general surgical patients. We compared event rates on telemetry before and after the intervention. RESULTS: Over a 4-month period, there were 1051 telemetry orders placed for general medical or surgical patients before the intervention and 450 after. The average number of orders/week did not differ significantly before vs after the reduction in telemetry order duration (88 +/− 10 vs. 75 +/− 22 orders/week, p= 0.11). The mean duration of telemetry was reduced by 32 %, which was the same percent reduction required by the change in policy (mean reduction in telemetry time decreased from 64 +/− 3.1 h/patient to 44 +/− 1.2 h/patient, p< 0.01). The rate of code events and rapid response team events in hospitalized general medical or surgical patients decreased from 2.5 events/week to 2.2 events/week. Only seven RRT/code events occurred in patients on telemetry prior to the intervention, while one RRT/code event occurred in a patient on telemetry after the intervention (p= 0.28). CONCLUSIONS: A simple change in the default maximum number of hours a provider can prescribe cardiac monitoring before requiring a renewal order led to a dramatic decrease in telemetry utilization. Although the number of orders for patients to receive cardiac monitoring was unchanged, the duration of cardiac monitoring dropped substantially. Even with less monitoring, we found no increase in the rate of adverse events. Similar reductions at other medical centers may reduce telemetry utilization, technologist and nurse staffing requirements, and cost while preserving patient safety.

A RESIDENT INITIATIVE TO INCREASE USE OF NALTREXONE FOR ALCOHOL USE DISORDER AT A SAFETY NET HOSPITAL Maria E. Otto1; Jeffrey K. Hom1; Shilpa M. Shah1; Natalie C. Young1; Weston S. Fisher2; Edgar Pierluissi1. 1University of California San Francisco, San Francisco, CA; 2University of California San Francisco, San Francisco, CA. (Tracking ID #1939058) BACKGROUND: Alcohol use disorder (AUD) is extremely common at San Francisco General Hospital (SFGH), with nearly one in four inpatients on the Medicine service affected. These patients frequently present to the Emergency Department (ED) with alcohol-related falls, seizures, and symptoms of

ABSTRACTS withdrawal. Their 30-day readmission rate is 16 %, above the 13 % for the remainder of the hospital. Hospitalization provides a window of opportunity to help this vulnerable population, as patients are often motivated to make changes following an alcohol-related illness but are frequently without primary care doctors. Despite the effectiveness of naltrexone, an inexpensive, FDA-approved medication for treatment of AUD, it has been underutilized. A pilot intervention in 2011 at SFGH increased prescriptions of naltrexone to eligible patients at discharge, but was not sustained. METHODS: A resident-led, quality improvement (QI) project was implemented in February 2013 to 1) sustainably increase the prescribing of naltrexone to eligible inpatients with AUD upon discharge and 2) to measure the effect of the intervention on ED visits and readmissions within 30 days. The intervention 1) identified inpatients with AUD via the electronic handoff system, 2) educated providers about naltrexone, 3) prompted providers to prescribe naltrexone to eligible patients at discharge planning rounds, and 4) utilized principles of QI to continuously evaluate and improve the intervention. Data were collected via retrospective chart review to identify patients with AUD and determine eligibility for naltrexone. Among patients with AUD, number of ED visits and readmissions within 30-days of discharge were also obtained. RESULTS: Among 118 patients with AUD identified between March 1 and April 15, 2013, 43 % (n=51) were eligible for naltrexone. Following the initial intervention, naltrexone offering increased from 2 to 58 % of eligible patients. This effect was sustained through continuous cycles of improvement, with 67 % of patients offered it most recently. Compared to patients not prescribed naltrexone (n=31), patients prescribed the medication (n=20) had 30-day readmission rates and hospital visits that were 74 % and 88 % lower, respectively. CONCLUSIONS: These results are promising and suggest that an intervention to systematically identify and offer treatment to patients with AUD is important in caring for our patient population and decreasing hospital utilization. Additionally, this QI initiative demonstrates that a deep-rooted problem can be addressed with the commitment of an interdisciplinary team, which can include a rotating group of trainees. By assessing the effect of this intervention on hospital utilization, the project shows that measuring true outcomes, rather than just processes, can help build a business case for sustainability. Next steps include 1) expanding the intervention to other hospital services, 2) ensuring the safe transition of patients on naltrexone from the hospital setting to their outpatient providers, and 3) collaborating with hospital leadership to improve systems to identify and offer treatment to all patients with AUD.

A ROLLING, INTEGRATED EVIDENCED-BASED MEDICINE CURRICULUM’S IMPACT ON EVIDENCE-BASED PRACTICE SKILLS, KNOWLEDGE, BEHAVIOR AND ATTITUDES ON INTERNAL MEDICINE STUDENTS AND RESIDENTS AT THE WASHINGTON, DC VETERANS AFFAIRS MEDICAL CENTER Matthew Tuck; Samantha Mcintosh; David Maron. Department of Veterans Affairs Medical Center, Washington, DC. (Tracking ID #1941446) BACKGROUND: Teaching evidence-based medicine (EBM) skills to students and residents has become an imperative for medical education. The AAMC made recommendations to incorporate problem-based clinical skills in medical schools’ curricula. Similarly, the Next Accreditation System of the ACGME requires that residency programs train and evaluate residents in practice-based learning and improvement. Developing effective techniques for integrating these practices across the years of medical education remains a challenge for medical schools and residency programs. METHODS: We developed a rolling EBM curriculum for medical students and residents rotating through the Medical Service of the Washington, DC Veterans Affairs Medical Center (VAMC) that provides repeated opportunities for students and residents to learn, practice, and evaluate their skills within the context of real-time patient care. The curriculum is integrated into morning report over the course of a month-long rotation. Using a patient case on the wards, learners are guided through the process of formulating a PICO question, searching the literature to find an article that answers the question, appraising the validity of the article and applying the literature to the case. The aim of the study was to evaluate the curriculum on learners’ self-reported patient care practices, attitudes and knowledge using validated rubrics. We hypothesized that exposure to the curriculum would improve learners’ EBM-related patient care practices, attitudes and knowledge. Using a before-and-after design, consented learners were given questionnaires. We used validated rubrics on attitudes and behavior as well as a rubric to assess EBM knowledge (the so-called Berlin

JGIM

Questionnaire) at the beginning and end of their rotation at the VAMC. Incentives were provided for completion of the questionnaires. Means, standard deviations, percentages and univariate paired t-tests were calculated for pre- and post-exposure scores. The research was funded by a Medical Education Research Grant awarded by George Washington University. RESULTS: 231 residents and students were exposed to the curriculum. Sixtythree learners completed the pre-curriculum questionnaire and 21 completed both the pre- and post-curriculum questionnaires. Up to 74 % of respondents had never received formal education on EBM prior to their rotation at the VAMC. Mean scores on attitudinal and belief questions improved after exposure to the curriculum, but to a non-significant level. Similarly, knowledge of EBM, as assessed by the Berlin Questionnaire, improved to a non-significant level (pre-curriculum mean score 4.5, post-curriculum mean score 5.3, p =0.10). A regression was conducted to determine if the number of days between the pre and post-curriculum assessment predicted improved performance on the Berlin Questionnaire and was non-significant (p= 0.13). CONCLUSIONS: We successfully implemented a rolling, integrated EBM curriculum for internal medicine students and residents. Attitudinal, behavioral and knowledge scores improved to a non-significant level after the month-long exposure to the curriculum. The lack of significant findings may be in part due to small sample size.

A SALARY EQUITY STUDY: DIFFERENCES BY GENDER AND RACE, DEPARTMENT OF MEDICINE, UNIVERSITY OF MINNESOTA Anne Joseph; Wesley Miller; Jeremiah Menk; Stewart McMullan. University of Minnesota, Minneapolis, MN. (Tracking ID #1938954) BACKGROUND: Salary inequities for female faculty have been documented in many academic disciplines. METHODS: The University of Minnesota Department of Medicine (DOM) conducted cross-sectional studies in 2011, 2012, and 2013 with the objective of examining gender equity issues related to salary (n = 165, n = 196, and n = 213 respectively). A pre-specified multivariable linear regression was used to adjust the salary comparison between men and women and Whites and nonWhites. The analysis included the following variables: gender, race, academic rank (and time in rank), leadership roles, degree(s), number of ABIM certifications, time from highest degree to current position, percent clinical effort, and subspecialty. RESULTS: In 2011, 8 % of women were full Professors compared to 35 % of men. Leadership roles were held by 7 % of women compared to 19 % of men; similar differences in career characteristics were observed across all 3 years. The unadjusted mean salary for women was $69,717–$76,009 lower compared to men depending on the year. In 2011, 2012, and 2013 the adjusted mean comparisons for women vs. men were $-10,513, −$9,377, and −$16,132 respectively. For non-Whites vs. Whites the comparisons were −$9,018, −$741, and −$13,002. R-squared values for the models were between 0.757 and 0.767. As predicted, the model confirmed that the following were associated with increased salary: a PhD or Masters in addition to MD degree, number of ABIM certifications, rank, time variables, and holding a leadership position. For clinicians, procedural subspecialties (cardiology, gastroenterology, pulmonary, renal) conferred higher salaries than hematology-oncology-transplant, endocrinology, rheumatology, infectious disease, general internal medicine (in that order). CONCLUSIONS: Gender differences persisted over the 3-year period despite preferential salary adjustments for women. Professional accomplishment characteristics do not fully explain faculty salary differences by gender and race. In response, the DOM has formed a Salary Equity Committee to systematically address these differences abiding by principles of transparency, simplicity, value of clinical, educational and research missions, validated benchmarks of productivity, equal access to leadership positions and supplemental income, and sustainability. Adjusted Mean Comparisons

R-squared Sex Women vs. men Race Non-White vs. White

2011 0.767

2012 0.757

2013 0.759

−$10,513

−$9,377

−$16,132

−$9,108

−$741

−$13,002

JGIM

ABSTRACTS

A SIMPLE SCREENING METHOD TO DETECT MALIGNANCY USING RED CELL DISTRIBUTION WIDTH (RDW) Hiroaki Wakakuri; Toshihiko Ohara; Makoto Suzaki; Sonoko Kirinoki; Naoko Onodera; Takashi Araki; Hideya Hyodo; Makoto Kawai; Masahiro Yasutake. Nippon medical school, Bunkyokusendagi, Japan. (Tracking ID #1928358) BACKGROUND: RDW is rapidly calculated parameter using an automatic cell counter. Recent studies showed that RDW is strongly associated with prognosis in disorders such as chronic heart failure, idiopathic pulmonary fibrosis, sepsis, various cancer, and chronic kidney disease. However, it is still unclear that the relationship of RDW and diagnosis of malignancy. The aim of this study is to evaluate the usefulness of RDW to screening malignancy at the first visit outpatient clinic. METHODS: Fifty one consecutive patients diagnosed as malignancy at ER and general internal medicine of our hospital were reviewed in the period between September 2012 and September 2013. Fifty one patients, matching age and sex, with common diseases, including hypertension and dyslipidemia were assigned to control group. The patients complicated severe anemia (Hb99 %. Of the 49 “false positives,” according to the chart, 10 visits were for other ARIs, seven visits were to other practices, seven the patient had been coughing for more than 3 weeks, five the patient did not have cough, five the patient had chronic lung disease, and 15 the patient had a visit in the prior 30 days. Of the 53 “false negatives”—for which the clinician diagnosed the patient with acute bronchitis in their note—52 did not have an acute bronchitis diagnosis code (most common diagnosis codes were cough, hypertension, upper respiratory infection, and viral infection) and one had a diagnosis code of asthma that was not mentioned in the chart. The antibiotic prescribing rate for true positives was 22 %, for false positives was 4 %, and false negatives was 30 %. CONCLUSIONS: Compared to chart review, the acute bronchitis performance measure has a sensitivity and positive predictive value of less than 50 %. The performance measure omitted visits with antibiotic prescribing and falsely included visits without antibiotic prescribing, suggesting gaming.

ADAPTIVE RESERVE AT COMMUNITY HEALTH CENTERS: THE CANCER PREVENTION AND CONTROL RESEARCH NETWORK (CPCRN) MULTI-STATE SURVEY Shin-Ping Tu1,2; Alan Kuniyuki2; Vicki M. Young3; Maria Fernandez4; Rebecca S. Williams5; Amanda Kimura2. 1Virginia Common-

wealth University, Richmond, VA; 2University of Washington, Seattle, WA; 3South Carolina Primary Healthcare Association, Columbia, SC; 4University of Texas at Houston, Houston, TX; 5UNC Chapel Hill, Chapel Hill, NC. (Tracking ID #1936531) BACKGROUND: Adaptive Reserve comprises intangible elements of a practice’s capacity (i.e., human relationships, teamwork, a learning culture) that provides flexibility, resilience, and enhance positive change. Our objective is to determine the Adaptive Reserve at Community Health Centers (CHCs) in seven states (California, Colorado, Georgia, Missouri, South Carolina, Texas, and Washington). METHODS: We selected the Patient Centered Medical Home (PCMH) and the Consolidated Framework for Implementation Research to guide this research. Our survey included questions on colorectal cancer screening PCMH best practices and items from the Clinician Staff Questionnaire to examine the practice Adaptive Reserve (PAR) of participating CHC clinics. PAR scores were scaled from 0.00 to 1.00, with higher scores representing greater Adaptive Reserve. In collaboration with Primary Care Associations and CHCs, we recruited a convenience sample of 76 CHC clinics and invited their clinical staff (providers, nurses, and medical assistants) to participate in our CPCRN Community Health Center Survey. RESULTS: Survey respondents (N=246) are mostly female (82 %) and non-Hispanic (61 %). Forty-two percent reported their clinic had 7–8 colorectal cancer (CRC) screening PCMH best practices (maximum 8). Preliminary PAR scores in the states ranged from 0.61 to 0.7. In a study on PCMH implementation, 36 practices expected to have significant capability for change had a mean baseline PAR score of 0.69. Our survey indicates higher levels of PAR to be positively associated with CRC screening best practices in the clinics (adjusted OR=3.62; 95 % CI 1.31, 10.02) and how frequently staff performed the best practices (p=0.005). CONCLUSIONS: Evidence from primary care transformation indicates that practices with strong Adaptive Reserve are more successful at incorporating change. With the Affordable Care Act, understanding the Adaptive Reserve of CHC clinics will be timely for the many changes needed to successfully expand primary care at CHCs.

ADVANCE DIRECTIVE INTERVENTION FOR CHRONICALLY HOMELESS PEOPLE: ACCEPTABILITY, FACTORS ASSOCIATED WITH COMPLETION, AND TREATMENT PREFERENCES Alexander Leung1; Dhruv Nayyar1; Manisha Sachdeva1; John Song2; Stephen Hwang1. 1St. Michael, Toronto, ON, Canada; 2University of Minnesota, Minneapolis, MN. (Tracking ID #1933803) BACKGROUND: End-of-life care planning is relevant for homeless individuals because they experience high morbidity and mortality. A previous study has shown that homeless persons are more likely to complete advance directives with a one-onone counselor-guided intervention than if the advance directive is self-completed. However, the acceptability of a counselor-guided intervention for the completion of advance directives among chronically homeless individuals has not been investigated. The aim of this study was to determine the acceptability of a counselor-guided intervention for the completion of advance directives among chronically homeless individuals. Secondary objectives were to identify characteristics associated with willingness to complete an advance directive and to characterize treatment preferences in this population. METHODS: In a single-group intervention study, homeless men residing at a shelter in Toronto, Ontario were recruited from April to June 2013. Recruitment took place among programs that served chronically homeless individuals who were residents in a managed alcohol harm reduction program, a medical respite program that treats individuals with acute or uncontrolled chronic illnesses, and a long-term program for individuals who have been homeless for extended periods of time. Two hundred five homeless men (89.1 % of those approached) participated in the study. Participants completed a survey which obtained information on potential predictors of advance directive completion, including socio-demographic characteristics, health status, health care use, and attitudes towards end-of-life care and death. After completing this survey, all participants were offered an opportunity to complete an advance directive with a trained counselor. The primary outcome was the proportion of participants who completed an advance directive. The characteristics associated with completion of an advance directive were examined. RESULTS: The duration of homelessness was 12 months or longer in 72.8 % of participants. A total of 103 participants (50.2 %) completed an advance directive. Socio-demographic characteristics, health status, and health care use were not associated with completion of an advance directive. Participants were significantly more likely to complete an advance directive if they reported thinking about death on a daily basis, believed that thinking about their friends and family was important, or reported knowing their wishes for end-of-life care but not having told anyone about

ABSTRACTS these wishes. Of 103 individuals who completed an advance directive, the proportion expressing a preference to receive cardiopulmonary resuscitation in the event of a cardiorespiratory arrest was 94.1 %, 81.3 %, 64.7 %, and 36.3 % if there was a chance of returning to their current state of health or hypothetical health states of moderate dementia, severe dementia, and permanent coma, respectively. CONCLUSIONS: As the homeless population continues to age, health care professionals will encounter an increase in chronically homeless persons facing complex end-of-life care needs. A single-encounter, one-on-one counselor-guided intervention can achieve a high rate of advance directive completion among chronically homeless individuals. Health care providers should initiate advance care planning discussions with their patients who are homeless. Future research aims to ascertain whether the completion of an advance directive has an effect on subsequent care in homeless individuals.

ADVANCED DIRECTIVE USE IN INTERNAL MEDICINE RESIDENCY CLINICS: A RETROSPECTIVE REVIEW Briana N. Ketterer; Corinne Self; Guadalupe Martinez; Jennifer Gabbard; Macy Whitley; Ellyn Lee. University of Arizona, Tucson, AZ. (Tracking ID #1923316) BACKGROUND: Advanced directives (AD) allow patients to voice preferences about life sustaining therapies and end-of-life (EOL) care. ADs improve satisfaction, reduce stress, anxiety and depression at EOL. However, this information needs to be discussed and documented to be effective. Resident physicians play a key role as one of the first providers to encounter patients thus, making them an integral hub in the health delivery system. We sought to evaluate Internal Medicine (IM) Residents’ practice habits and assess participation in EOL/AD planning, and documentation of those preferences. This information will serve to address potential weaknesses in the Residency curricula and areas for improvement. METHODS: Retrospective chart review of patients seen in IM Residency clinics at The University of Arizona Medical Center (UAMC) from July 2011–2013. Inclusion criteria: ≥65 years of age with at least one chronic, debilitating diagnosis associated with morbidity or mortality (dementia, AIDS, cancer, neurodegenerative disease, chronic obstructive pulmonary disease, chronic kidney disease, cerebral vascular accident or transient ischemic attack, liver cirrhosis, coronary artery disease or myocardial infarction, post-transplant) documented in the electronic medical record, AllScripts. Documentation was reviewed and ADs were recorded along with demographics (age, gender), number and type of diagnoses. RESULTS: Reviewed 1075 charts, 165 met inclusion criteria (M 47 %, F 53 %; average age, 75; average number of problems, 7) with most frequent diagnoses of: cardiovascular disease (52 %), cancer (34 %), CKD (17 %). AD documented in 17 of 165 charts (10 %) which included: AD discussion (2), no specific qualifiers (1), Medical Power Of Attorney (8), DNR/DNI (9), Living Will (4), Hospice (1), and documentation for first responders (4). CONCLUSIONS: IM Residents at UAMC clinics documented ADs in a mere 10 % of seemingly tenuous patients ≥65 years of age. This represents a cohort who would benefit from AD conversations to preserve their wishes at EOL. Although residents have some training in EOL care and AD, the curricula appears insufficient as the rate of AD documentation is very low and residents feel uncomfortable with those discussions as unpublished data from our group similarly suggests. One way to improve upon this is to implement an educational intervention to enhance training, improve AD discussions, and address all AD components with patients. Improvement in this area is crucial as it would not only fulfill residency competency requirements but also lead to more complete ADs allowing patients greater autonomy in their EOL care.

AGING WITH HIV: SYMPTOM BURDEN AND IMPACT ON QUALITY OF LIFE Meredith Greene1,2; Christine Ritchie1; Jeff Martin4; Joy A. Madamba3; Monica Mattes5; Yinghui Miao2; Steven Deeks3; Victor Valcour1,6. 1University of California San Francisco, San Francisco, CA; 2San Francisco VA Medical Center, San Francisco, CA; 3University of California San Francisco, San Francisco, CA; 4 University of California San Francisco, San Francisco, CA; 5University of Central Florida College of Medicine, Orlando, FL; 6University of California San Franscico, San Francisco, CA. (Tracking ID #1932204) BACKGROUND: As HIV infection has shifted from a life-threatening disease to a chronic illness, the HIV-infected population is aging and often facing multiple comorbid conditions. Symptom assessment is an important clinical tool in the management of patients with complex chronic disease, yet limited data exists about symptom burden in older HIV-infected adults especially in the modern treatment era.

JGIM

The primary aim of this study was to describe the symptom burden (frequency and severity of symptoms) among HIV-infected adults over age 50 on combination antiretroviral therapy. We also examined factors associated with symptom burden in this population and the association of symptom burden on health related quality of life. METHODS: We conducted a cross-sectional survey in 2012–2013 of HIV-infected adults over age 50 that had an undetectable viral load (HIV RNA) on antiretroviral therapy and were enrolled in a clinic based cohort at a county medical center in San Francisco. Sixteen symptoms were measured by self-report in a two part question based on the 12 item HIV symptom index:1) yes/no for presence of each symptom in the past week and 2) degree of severity of each symptom (not a problem, mild, moderate, severe, very severe) as judged by symptom frequency, intensity, and duration. Symptom severity was analyzed by examining the number of symptoms rated as moderate, severe, or very severe. Health related quality of life was assessed by a single-item self-report of health as excellent, very good, good, fair or poor. Demographics, HIV related factors, co-morbidities and medications were measured by self-report and verified by chart review. Descriptive statistics were used for data analysis. RESULTS: 142 participants were enrolled, of which 94 % were male, 63 % were Caucasian, with a median age of 57 (range 50–74). The median CD4 count was 576 (IQR 393–715) and the median length of HIV infection was 22 (IQR 18–25) years. Participants had a median of 4 (IQR 3–6) co-morbid conditions, most commonly hyperlipidemia (61.3 %) and hypertension (52.8 %) and were taking a median of 9 (IQR 5–12) non-antiretroviral medications. Sixty-nine percent of participants reported two or more symptoms, with a median of 3 (IQR 1–5) symptoms reported. Fatigue (n=72, 51 %), “feeling sad or depressed” (n=60, 42 %), “pain, tingling, loss of feeling in feet or hands” (n=54, 38 %) and insomnia (n=53, 37 %) were the most commonly reported symptoms. Fifty-five percent of participants reported at least one moderate to very severe symptom with a median of 1 (IQR 0–2) moderate to very severe symptom reported. Fatigue (n=31, 22 %), “feeling sad” (n=31, 22 %) and insomnia (n=25, 18 %) were the most frequent severe symptoms. The total number of co-morbid conditions was associated with both the total number of symptoms reported (rs=0.21, p=0.01) and symptom severity (rs=0.27, p1.0 might be the most reliable marker for this condition. Various techniques, including Mercaptoacetyltriglycine (MAG3) renogram, may aid in establishing the diagnosis of urinothorax by demonstrating leakage of the tracer from the urinary tract into the pleural space. While urinothorax may resolve with relief of urinary obstruction, iatrogenic/traumatic cases associated with a direct tract to the pleura may persist or recur, as in our case. As urological procedures become more common, urinothorax will likely increase in incidence. As our case demonstrates, urinothorax can often go undiagnosed. Timely diagnosis requires a high index of clinical suspicion and should be considered whenever a pleural effusion occurs in the setting of urinary tract obstruction or recent urological intervention. The creatinine level of pleural fluid, a biochemical parameter not routinely measured, should be determined. Furthermore, earlier thoracentesis is more likely to reveal the characteristic parameters of urinothorax.

RECURRENT PNEUMONIA OR PULMONARYALVEOLAR PROTEINOSIS: A CONCEALED DIAGNOSIS Fayruz Araji1; Omar Mousa2; Daniela Moran3. 1 Universidad Francisco Marroquin, Guatemala, Guatemala; 2SUNY Upstate Medical University, Syracuse, NY; 3Lung and Asthma Clinic, Houston, TX. (Tracking ID #1934266) LEARNING OBJECTIVE 1: Diagnose Pulmonary Alveolar Proteinosis when the presentation is challenging with different mimickers of the disease. LEARNING OBJECTIVE 2: Recognize that outcomes are worrisome for disease progression with high rates of recurrence and mortality.

JGIM

CASE: We hereby present a case of a 49-year old Middle Eastern male patient who presented with worsening exertional dyspnea and productive cough with thick secretions for 3 months. In the preceding 6 months, he was evaluated multiple times for abnormal chest X-rays and clubbing and he was treated for recurrent pneumonia with oral antibiotics and glucocorticoids. He noticed weight loss of 20 lb but denied other symptoms. He had a history of recurrent sinusitis with no lung disease or a family history of similar illnesses. He worked as an aircraft maintenance technician and reported exposure to fiberglass and metal dust. He had a 25 pack-year smoking history with no recent travel. Physical examination revealed tachycardia (110 bpm) and hypoxemia (O2 saturation 92 % on 6 L O2). He had clubbing, bilateral diffuse crackles and the exam was otherwise unremarkable. He had bilateral interstitial alveolar infiltrates on the chest X-ray and the CT scan showed diffuse ground glass opacities bilaterally. Bronchoscopy was unremarkable and a single bronchoalveolar lavage culture was positive for nontuberculous mycobacterium. His symptoms worsened despite inpatient care and close monitoring. He required a video-assisted thoracoscopic surgery and anatomic pathology showed filling of the alveolar spaces with eosinophilic acellular finely granular material that stained positive with PAS, consistent with Pulmonary Alveolar Proteinosis. He received high volume bilateral lung lavage, oxygen therapy, GM-CSF 250 mcg SQ daily and anti-mycobacterial treatment with rifampin and ethambutol. The hospital course was complicated by nonsustained ventricular tachycardia. He improved slowly after a prolonged hospital stay. DISCUSSION: Pulmonary alveolar proteinosis (PAP) is a very rare disease of the lungs with an estimated annual incidence and prevalence as high as 0.49 and 6.2 cases per million, respectively. The natural history of PAP has been studied over the past 6 decades through less than 1000-reported cases. The diagnosis is easily missed and delayed due to the mimickers of its various presentations, including disorders with similar radiographic manifestations. With only 8 % of patients experiencing clinically significant spontaneous improvement, delay in the management increases the mortality risk. Healthcare providers should have a high index of suspicion for PAP in patients with recurrent pneumonia. Raising awareness about the stepwise plan of treatment with whole lung lavage, inhaled GM-CSF then rituximab improves the clinical outcomes and the pulmonary function. Early intervention with such therapeutic modalities remains the standard of care until further immunologic mechanisms are discovered.

RECURRENT RASH WITH FEVER Sadie Trammell Velasquez. 1University of TX Health Science Center at San Antonio, San Antonio, TX; 2South Texas Veterans Healthcare System, San Antonio, TX. (Tracking ID #1936647) LEARNING OBJECTIVE 1: Identify the differential diagnosis of fever and rash in an adult. LEARNING OBJECTIVE 2: Recognize the clinical presentation of Adult-Onset Still’s Disease and understand the treatment and clinical course of Adult-Onset Still’s Disease. CASE: A 20 year-old woman presented with 3 days of fever, rash and arthralgias. Arthralgias were of elbows, ankles and back. She describes her rash as non-painful and non-pruritic that has progressed to cover her arms, palms and legs. She has also had associated myalgias and a sore throat. Her temperature was 39.4οC; heart rate 145 bpm; blood pressure 127/76 mmHg. She was tachycardic with regular rhythm. Skin exam revealed erythematous, somewhat blanching thin plaques with slight edema scattered on her bilateral arms, bilateral and medial aspects of legs, lateral thighs/hips and lower chin/neck and spared her torso; bilateral palms had macular lesions, slightly raised papules (PICTURE). The remainder of her exam was normal. Her WBC was 8 with 31 % bands; hemoglobin was 13 and platelet count of 253,000. Her liver function tests were normal. Her LDH was elevated at 550 IU/L and ferritin elevated to 1411 ng/mL. Her RPR was negative in addition to ANA, HIV antibody, HIV PCR, respiratory viral PCR, monospot heterophile antibody, CMV and throat culture. Her complements were normal. The patient was admitted and started on high dose ibuprofen with continued fevers yet resolution of her rash. On hospital day four she was initiated on prednisone 20 mg daily with resolution of her fevers. She followed up with Rheumatology 3 weeks after discharge. DISCUSSION: Fever and rash are common findings that the general internist (hospitalist) encounters. A stepwise approach to determining the etiology of this constellation of symptoms should be taken and careful attention should be taken to assure that a treatable cause is not overlooked. The differential diagnosis with erythematous, slightly edematous rash with fever includes reactive, viral exanthema, autoimmune/connective tissue disease, Adult Onset Still’s disease (ASD). Viral etiologies are to be considered given her sore throat, arthralgias and other nonspecific complaints. In our patient with fever and rash, the associated arthralgias clued the diagnosis of ASD. Currently there is no specific test or combination of tests that can be used to establish the diagnosis of ASD. The Yamaguchi criteria have the highest

JGIM

ABSTRACTS

sensitivity in patients with a definite diagnosis of ASD. This criteria requires five features for diagnosis, two being the major criteria. ASD has been associated with markedly elevated serum ferritin concentrations in as much as seventy percent of patients.

RECURRENT VS. LINGERING LEGIONELLA? A CASE HIGHLIGHTS RISK FACTORS, DIAGNOSTIC CHALLENGES AND KNOWLEDGE GAPS IN THE AFTERMATH OF HURRICANE SANDY Madeline R. Sterling. New York Presbyterian - Weill Cornell, New York, NY. (Tracking ID #1927313) LEARNING OBJECTIVE 1: Review clinical features of Legionella Pneumonia and understand how a clinical test (Urine Antigen) is applied and interpreted. LEARNING OBJECTIVE 2: Demonstrate how the incidence of Legionnaire’s disease is affected by environmental exposures and patient characteristics. CASE: A 73 year-old man with CLL (stage 3, on immunosuppression) and hospitalization 1 year prior for Legionella pneumonia, presents from home to our hospital with 3 days of intermittent fevers and 1 week of blood-tinged sputum. He reports a maximum oral temperature of 103 F, associated with a productive cough. He notes dyspnea and hemoptysis of blood-tinged sputum for the last week. He denies night sweats, weight loss, lymphadenopathy, rhinorrhea, headache, neck stiffness and chest pain. He is retired and is able to walk 20 city blocks without dyspnea at baseline. He denies recent travel or Tuberculosis exposure. The patient’s apartment was victim to flooding from Hurricane Sandy; his apartment sustained extensive water damage. He did not relocate due to finances. He has not been to a primary care provider since his last hospitalization. On presentation he was afebrile and breathing comfortably (O2sat 98 %) on 2 L NC. He was alert and oriented, speaking in full sentences, without blood in his oropharynx. His lung exam was notable for decreased breath sounds on the right, without dullness to percussion. CT was remarkable for diffuse bilateral groundglass opacities, right greater than left, suspicious for diffuse alveolar hemorrhage. Blood cultures, PCR for atypical pneumonias and fungal cultures were negative. Urine Legionella antigen was positive and he was treated with Levaquin for 7 days. The patient was hemodynamically stable and improving; a diagnostic bronchoscopy was deferred. DISCUSSION: Legionella, which was first discovered in a 1976 Philadelphia outbreak linked to water supply and air conditioning units, is a common cause of community-acquired and nosocomial Pneumonias. When the clinical situation appropriate, a urine antigen test is used to detect L. pneumophila. The test, which is relatively cheap and rapid, has a sensitivity of 70 % and specificity of 99 %. Although easy to diagnose, clearance is difficult to pinpoint, particularly in immunocompromised hosts such as this patient. Following treatment, it is not recommended to check urine antigen in the outpatient setting, since false positive tests are likely in the short term. In fact, immunocompromised patients can shed the antigen upwards of 300 days post-treatment. Not only is this patient at risk due to his lymphoma treatment(s), but he remained in his flood-zone apartment during (and after) the largest Atlantic hurricane on record, Sandy. Although Legionella has an incubation of 2 to14 days after exposure, several hard-hit regions took several months to be restored and cleaned. It is unclear how flooding and stagnant water affected his clinical state and susceptibility for this bacterium. Further research ought to examine the association between hurricane-affected areas and disease incidence. If Legionella incidence is higher in flood zones, then patients with clinical risk factors or prior infections ought to be screened more vigilantly and counseled regarding early evacuation by their primary care providers.

RED, HOT, SWOLLEN, AND TENDER: A CASE OF CELLULITIS? Aderonke Salau-Okeleji2,1; Margaret C. Lo2,1. 1Univeristy of Florida, Gainesville, FL; 2 University of Florida, Gainesville, FL. (Tracking ID #1939038) LEARNING OBJECTIVE 1: Distinguish between the clinical, diagnostic, and therapeutic features of refractory cellulitis vs. drug-induced myositis LEARNING OBJECTIVE 2: Recognize the adverse health effects of popular energy drinks, the associated risk factors in the development of these adverse effects, and the need for screening for abuse of energy drinks. CASE: A 30 year-old healthy, athletic Caucasian male was admitted by his PCP for refractory cellulitis after he failed to respond to a 7-day course of oral doxycycline for worsening right arm swelling with pain and redness. He denied any trauma or associated fevers, chills, right arm paresthesia or weakness. Vital signs on admission were all normal including temperature. Physical exam revealed a severely erythematous right upper extremity with increased warmth, dependent +2 pitting edema and taut skin extending from proximal biceps to distal elbow. Rest of the cardiopulmonary, skin, musculoskeletal and neurologic exam was unremarkable. Admission laboratory values including BMP, CBC, CRP, ESR, and

blood cultures were all normal. Upon hospitalization, the patient was started empirically on IV Vancomycin for presumed refractory cellulitis that had failed out patient management. However, when he failed to improve on hospital day 2, investigation for noninfectious causes was initiated in light of no other objective evidence of systemic infections. Right upper extremity Doppler ultrasound was negative for DVT. Further studies including CPK level came back severely elevated at 11640 U/L with corresponding elevated aldolase value of 30.8 U/L. Urine myoglobin and anti-Jo 1 antibody titers were negative; ESR/CRP levels remained normal. Subsequent MRI of the right arm showed nonspecific myositis and swelling of the biceps, brachialis, and pronator teres. Further history from the patient revealed that 1 week prior to presentation he and his friends had vacationed at South Beach for Spring Break week where he had “gone clubbing all night, every night” and had gone weight-lifting during the day. To stay awake, he would drink 3–4 cans/day of Red Bull and 2–3 shots/day of 5-hour Energy. With this history, antibiotics were discontinued and supportive measures including IV fluids, arm elevation, and ice were started. His right arm symptoms resolved and his CPK level normalized 2-weeks later at his PCP follow-up visit. DISCUSSION: Energy drinks such as Red Bull, Monster, and 5-hour Energy have become quite the craze in young adult culture with self-report surveys of consumption in 30–50 % of adolescents and young adults. Touted for positive effects on physical and mental endurance, growing evidence in the literature has now surfaced that energy drinks are not as harmless as advertised. Multiple studies have revealed adverse health effects such as obesity, seizures, cardiac anomalies (hypertensive urgency, tachyarrhythmias), strokes, rhabdomyolysis, mood disorders (mania, anxiety attacks), and even death. Our case adds to the literature by highlighting the potential for energy drinks to cause myositis. Such clinical presentations can become a diagnostic challenge by mimicking cellulitis. However, unlike cellulitis, clinicians must recognize that drug-induced myositis exhibits no systemic signs of infection (i.e. fevers, leukocytosis, elevated ESR/CRP), often fail to respond to antibiotics, and produces severe CPK and aldolase elevation. Risk factors for incurring the detrimental effects of energy drinks include mixture with alcohol and dehydration. Sleep deprivation and extreme exercise workouts can also be included as risk factors, as revealed in our case. Since there are no current diagnostic guidelines for energy drink abuse, a detailed and targeted social history for energy drink ingestion, either alone or with alcohol, is instrumental in the prompt recognition and management of their adverse effects. Screening by clinicians is particularly important for athletes, young adults with high-risk behaviors, certain health conditions (i.e. seizures, diabetes, obesity, hypertension, cardiac anomalies), and those with unexplained myalgias/myositis, behavioral/mood changes, anorexia, or sleep problems.

REDISCOVERING AN ANCIENT DIAGNOSIS Stephanie Behringer; Gianni Carrozzi. Montefiore Medical Center, Bronx, NY. (Tracking ID #1939336) LEARNING OBJECTIVE 1: Recognize visceral manifestations of tertiary syphilis in a patient with hepatic lesions CASE: A 32 year-old man with HIV (normal CD4 count, on HAART), high risk sexual behavior and polysubstance abuse presented with 4 weeks of vague abdominal pain. The pain was diffuse and dull in nature, intermittent and would last several days before resolving on its own. He denied fevers, chills or weight loss and was well appearing. Infectious workup was notable for a positive serum RPR test with a titer of 1:256, but was otherwise negative. His primary care doctor and the New York State Department of Health confirmed that he had been diagnosed with syphilis in the past. However, it was unclear whether he had been successfully treated, and he continued to engage in high-risk sexual behavior, exposing himself repeatedly to reinfections and had poor medical follow up. A CT scan of the abdomen revealed scattered illdefined hypodense lesions within the liver. DISCUSSION: Syphilis is caused by the spirochete Treponema pallidum and is virtually always sexually transmitted. About 6 % of patients with HIV infection have serologic evidence of syphilis. If left untreated, 25–40 % of patients will develop late phase syphilis, also known as tertiary syphilis. Tertiary syphilis can manifest itself anytime from 1 to 30 years after primary infection and has become a forgotten diagnosis in parts of the world where screening and antibiotic treatment have made advanced stages of this disease a rare finding. In the past, however, tertiary syphilis used to be a common diagnosis leading to visceral involvement with space occupying lesions known as syphilitic gummata (or syphilomata). These lesions are granulomas that contain a central necrosis and are surrounded by inflammatory tissue that can mimic metastatic disease or sarcoidosis. In some cases, they are found in the liver, but can develop in any organ. In cases of hepatic involvement, patients can present with abdominal discomfort, mild hepatitis, jaundice and fatigue. Depending on size and location, syphilitic gummata can cause portal hypertension, variceal bleeding, hepatic vein thrombosis, ascites and liver failure. They are not malignant in nature, but can lead to biliary obstruction. Antibiotic treatment usually results in complete resolution of the lesions. The diagnosis of tertiary syphilis with visceral involvement is confirmed by high serologic markers (serum RPR titer >1:32) and by response to treatment with penicillin. Biopsies of

ABSTRACTS gummata may be indistinguishable from other granulomatous diseases unless spirochetes are visualized within the tissue samples. Treatment of tertiary syphilis requires a prolonged antibiotic course with penicillin as spirochetes are slowly growing bacteria. Therapeutic levels are achieved with long-acting depot penicillin preparations. The current guidelines recommend Benzathine penicillin G 2.4 million units IM once a week for a total of 3 weeks.

REDUCING POLYPHARMACY IN THE ELDERLY-ROUND AND ROUND WE GO Celeste Newby; Alegra Venditto. Tulane University Health Sciences Center, New Orleans, LA. (Tracking ID #1924994) LEARNING OBJECTIVE 1: Discuss the challenge of managing polypharmacy in the elderly. LEARNING OBJECTIVE 2: Highlight physician barriers to reducing polypharmacy in the elderly and the need for greater evidence-based strategies to reduce polypharmacy CASE: A 69-year-old man with multiple medical problems presented to his primary care physician (PCP) for medication review after expressing confusion about his regimen. He also sees four subspecialty physicians who manage his chronic diseases. Review of his medication list via the electronic medical record revealed twenty total medications; all five of his physicians had prescribed medications. As requested, the patient presented to clinic with his home medications in two large duffel bags containing a total of 45 medication bottles. In addition to his currently prescribed regimen, the patient had two expired antibiotics, three duplicate medications, and medications for blood pressure and muscle pain that had been previously discontinued. After review, 25 medication bottles were given to pharmacy for destruction. A complete medication list, including essential and as-needed medications, was reviewed with patient. The electronic medical record was reconciled accordingly, with all extraneous medications eliminated. The patient returned to primary care clinic 1 month later after seeing several specialists. Four new medications had been added to his list. Again, the medication list was reviewed and edited with the patient as it had been the month prior. DISCUSSION: Polypharmacy is a well-known topic to most internists, and an important consideration in caring for any elderly patient. One in six hospital admissions (one in three for age >75) for older adults can be attributed to an adverse drug effect. Older patients often have multiple chronic diseases, and long medication lists consisting of both essential and as-needed prescriptions. The clinical picture is complicated further by multiple prescribers. Medication lists can easily grow or change when patients have new complaints or are involved in care transitions. Even with electronic medical records, keeping medication lists current and appropriate can be a daunting task. The PCP often feels a responsibility to reduce polypharmacy, as he or she is the main coordinator of a patient’s care. While studies have reported that the number of prescribing physicians is an independent risk factor for adverse drug reactions in the elderly, physicians report awkwardness and reluctance to discontinue a medication prescribed by another physician. Additional barriers include difficulty in convincing patients to discontinue long-term medications, and problems distinguishing between new complaints and medication side effects. While there is great interest in improving appropriate medication use in the elderly, how to best accomplish this goal is still unclear. A recent Cochrane review found that while the Beers criteria and Medication Appropriateness Index (MAI) appeared beneficial in reducing medication-related problems and inappropriate prescribing, it was not clear that these interventions translated into clinically significant improvements. More evidence-based strategies are needed to implement meaningful change in the area of polypharmacy to improve safety for older patients.

REFRACTORY HYPOGLYCEMIA IN A TYPE 1 DIABETIC PATIENT CLUE TO ADDISON’S DISEASE AND AUTOIMMUNE POLYENDOCRINE SYNDROME TYPE 2 Anene Ukaigwe; Adetokunbo Oluwasanjo. The Reading Health System, West Reading, PA. (Tracking ID #1936686) LEARNING OBJECTIVE 1: Recognize features of Autoimmune Polyendocrine Syndrome type 2 (APS-2). LEARNING OBJECTIVE 2: Manage and appropriately screen APS-2 patients. CASE: A 47-year-old Caucasian female presented for the third time to the emergency room with hypoglycemia and hypotension. For 3 months, she had recurrent hypotension, hypoglycemia, fatigue, orthostatic dizziness, weight loss and vomiting. She had a 16-year history of T1DM. Glycemic control was stable for 6 years on same dose of insulin. She had hypothyroidism for 5 years, also stable on levothyroxine. She denied recent travel, lifestyle or medication changes. Insulin dose was adjusted several times but hypoglycemic episodes persisted, occurring 2–3 times weekly. For her symptoms, she had seen a Cardiologist, Neurologist, Otorhinolaryngologist,

JGIM

Primary Care and Emergency Physicians. Work up thus far excluded Multiple Sclerosis, cobalamin deficiency, myasthenia gravis, hemochromatosis, vestibular, cardiac and rheumatologic disease. On examination she was alert and oriented, hypotensive; Blood Pressure = 87/53 mmHg and tachycardic; Heart rate = 124/min. Her skin was diffusely tan. The remaining exam was unrevealing. Labs showed Complete Blood Count, renal, liver and thyroid function tests were within normal range. Blood glucose was 70 mg/dl. Adrenal insufficiency was suspected based on hypoglycemia, hypotension, tachycardia and tan skin. This was confirmed with serum cortisol = 0.8mcg/ml (normal 8.7–22.4mcg/dl) and ACTH = 298 pg/ml (normal 10– 60 pg/ml). After intravenous hydrocortisone and fluids, symptoms resolved. Given hypothyroidism, adrenal insufficiency and T1DM further antibody tests confirmed APS-2. DISCUSSION: APS-2 is an autosomal dominant disease affecting HLA genes. APS2 is characterized by Addison’s disease, autoimmune thyroid disease and type I diabetes mellitus. Other associated disorders include Pernicious anemia, celiac disease, Hepatitis, Hypogonadism, Hypophysitis, vitiligo. Unlike APS-2, APS-1 affects only children. APS-2 occurs more in females and is often diagnosed between ages 30s to 40s. Although APS-2 is uncommon (1.4–4.5 per 100,000), its prevalence increases when subclinical cases are included (150 per 100,000) implying it is not as rare as believed. The clinical presentation depends on specific APS-2 component disorders and involved organs. However, conventional treatment for each component disorder is often sufficient. Specialist care is required only as needed. The importance of APS-2 is illustrated in our patient. More than one APS-2 disorder should raise vigilance for associated disorders given that clinical features of these disorders are nonspecific. In addition, the interval between development of one disorder and the next can be as long as 20 years, making screening necessary for early initiation of disease-specific management. Experts recommend that patients with two APS-2 disorders should be screened every 1–2 years until 50 years old. In addition to detailed history and physical examination, screening includes autoantibody testing for T1DM, thyroid disease, Addison’s disease, Celiac disease and autoimmune hepatitis. Vitamin B12, sex hormone and blood glucose levels including liver and thyroid function tests are also useful.

RELATED CEREBRAL HEMORRHAGE WITH AN INFECTED ABDOMINAL AORTIC ANEURYSM AND VERTEBRAL OSTEOMYELITIS? Suzue Hirano-Kuwata; Yumiko Kanzaki; Hirokazu Mizohata; Keishi Kuriyama; Kentaro Isoda; Kenichiro Hata; Yukimasa Ooi; Akira Ukimura. Osaka Medical College Hospital, Takatsuki, Osaka, Japan. (Tracking ID #1934970) LEARNING OBJECTIVE 1: Manage the complications of bacteremia. LEARNING OBJECTIVE 2: Recognize the causes of intracranial hemorrhage. CASE: A 73-year-old man with hypertension, coronary artery disease, and peripheral artery disease presented with severe lumbago and fever. Despite several different types of intravenous antibiotic treatments at another hospital, his symptoms remained over 2 weeks. He was transferred to our institution for further treatments. On admission, his body temperature was 38.0, his heart rate was 74 beats/min, and respiratory rate was 18 breaths/min. Laboratory studies showed a WBC count of 10160/μL (79 % neutrophils, 13 % lymphocytes), and a CRP concentration of 155 mg/L. He was diagnosed as vertebral osteomyelitis (disk-space infection) based on a hyperintense lesion of the spine on MRI. Streptococcus sanguinis was detected in one of three blood cultures. The transthoracic echocardiogram revealed no valvular vegetation, no other evidence of infective endocarditis, nor any regurgitant murmur was heard. His symptoms gradually improved with antibiotic therapy, but a CT scan showed a newly developed abdominal aortic aneurysm with periaortic soft tissue infiltration. The diameter of the abdominal aorta was 35 mm; it had been 20 mm on admission and we recognized it as an infectious aortic aneurysm. The aneurysm was adjacent to the vertebral osteomyelitis and despite sustained antibiotic treatment, the aneurysm enlarged. Eventually he had a graft implantation operation and was discharged 22 days after surgery. Two months after the aortic operation, he presented with left hemiplegia. A CT scan of the head showed a right putamenal hemorrhage, and he had an open craniotomy and an evacuation of the hematoma. He survived but was left with significant disability. DISCUSSION: We describe a case of vertebral osteomyelitis, an infected abdominal aneurysm, and cerebral hemorrhage, which occurred in a short period of time. One possible cause of all three phenomena is hematogenous spread of infection. Vertebral osteomyelitis most often results from hematogenous seeding. Infected aortic aneurysms are also related to hematogenous spread of infection. Co-existence of an infected aneurysm and vertebral osteomyelitis is occasionally reported, and is thought to be caused by contiguous involvement of the vessel wall from adjacent pyogenic vertebra. Our patient presented with cerebral hemorrhage only 4 months after the initial symptoms. The most common cause of cerebral hemorrhage is hypertension, but intracranial infectious aneurysm is one of the

JGIM

ABSTRACTS

causes. A minimum of 4 to 6 weeks of antimicrobial therapy is recommended for intracranial infectious aneurysm, but there remains a high risk for rupture even during appropriate medical treatment. It may be beneficial to perform cerebrovascular imaging on patients with a suspected infectious aneurysm. We are not sure whether this cerebral hemorrhage is relevant to an infected cerebral aneurysm, without previous examinations of cerebral arteries. This case involved vertebral osteomyelitis with an adjacent infected aneurysm and subsequent cerebral hemorrhage. Hematogenous spread of infection can explain the sequence of phenomena, though it is not proven. This case provides us an opportunity to consider further therapeutic plans that might lead to a better prognosis.

RESISTANT HYPOGLYCEMIA FROM GLYBURIDE-CIPROFLOXACIN INTERACTION Suzanne J. Supplee; Anthony A. Donato. Reading Health System, West Reading, PA. (Tracking ID #1939107) LEARNING OBJECTIVE 1: Recognize a drug interaction between glyburide and ciprofloxacin may lead to potentially fatal hypoglycemia. CASE: A 50 year-old male with a history of diabetes mellitus type 2 on long-term combination glyburide/metformin therapy who was recently treated with a 14 day course of ciprofloxacin for venous stasis ulcers presented to the ED after his sister noticed he was disoriented, mumbling, diaphoretic and unarousable, but exam was otherwise non-focal. EMS recorded a finger stick blood glucose at 33 mg/dl. The patient was given one ampule of D50 en route. Upon arrival to the ED, the patient required another 3 ampules of D50 to get serum glucose over 100 mg/dl. A 5 % dextrose in half normal saline solution at 150 ml/h was required to stabilize his blood sugars over the next 24 h. The patient denied any recent change in medications except for the ciprofloxacin course completed 10 days prior, and in fact had not taken the oral hypoglycemic in the 24 h leading up to admission. He denied medication misuse, exogenous insulin or alcohol use. He denied any previous episodes of hypoglycemia. Laboratory values included a normal TSH and WBC count. Abdominal CT was negative for pancreatic mass. Abnormal laboratory values at the time of hypoglycemia included a C-peptide of 8.1 ng/ml (normal: 1.1–4.4 ng/ml) and free Insulin of 59 μ IU/ml (normal: 1.4–14 μ IU/ml). DISCUSSION: Hypoglycemia is a common medical complication in patients receiving pharmacotherapy for diabetes that may result in permanent neurologic deficits and even death if unrecognized. With over 285 million people diagnosed with diabetes worldwide, many require oral hypoglycemic agents for disease management. A possible interaction between glyburide and ciprofloxacin has been reported, resulting in reduced drug clearance and a prolonged hypoglycemic effect of glyburide. Recognition of potential drug side effects is critical in the management of complex medical patients. Ciprofloxacin causing hypoglycemia in patients taking glyburide may be under-reported as the mechanism of action is poorly understood, therefore a causal relationship overlooked. A thorough medication history in hypoglycemia is essential to uncover potential adverse drug interactions. In patients on sulfonylurea therapy presenting with hypoglycemia of unknown etiology, ciprofloxacin should be considered as a potential cause and prescribed cautiously in patients currently on glyburide therapy.

RETIFORM PURPURA AS A CLUE TO DIAGNOSIS Sarah Apgar. Yale-New Haven Hospital, New Haven, CT. (Tracking ID #1938728) LEARNING OBJECTIVE 1: Recognize the clinical features of retiform purpura LEARNING OBJECTIVE 2: Understand the differential diagnosis of retiform pupura CASE: A 50 year-old asplenic man with a history of hypertension presented with 4 days of nausea, vomiting and diarrhea and 1 day of lower extremity burning pain and numbness. Two days prior to admission he noted red spots on his upper thighs which became larger blotches on his feet and ankles followed by bluish discoloration of his third toe. Upon admission, he was afebrile, tachycardic at 125 with a blood pressure of 123/82. Physical examination demonstrated an ill man with a nonblanching purple mottled rash on the legs and feet without neurologic deficit. Blood work revealed a creatinine of 4.8 mg/dl, white blood cell count of 40 with 36 % bands, platelet count of 67, normal coagulation studies and lactic acid of 3.3. He was admitted to the intensive care unit and started on broad spectrum antibiotics and intravenous fluids. Further investigation revealed a D-dimer of >10,000, fibrinogen of 959, negative anti-cardiolipin antibodies and ANA1.35 g/l) and histopathological features, including storiform fibrosis, obliterative phlebitis, and lymphocyte and IgG4+ plasma cell infiltration (IgG4+ plasma cells/IgG+ plasma cells ratio >50 % on a highly magnified slide); tissue eosinophilia may also be present. Glucocorticoids are typically the first line of therapy. They appear to be effective in the majority of patients but disease flares are common. Azathioprine, mycophenolate mofetil, and methotrexate are used frequently as glucocorticoid-sparing agents or remission-maintenance drugs after glucocorticoid induced remissions. For patients with recurrent or refractory disease, B-cell depletion with rituximab appears to be a useful approach. Additionally, there have been case reports and studies about the development of lower extremity DVTs leading to PEs in patients with retroperitoneal fibrosis due to involvement of iliac vessels and in those cases treatment was steroids in addition to anticoagulation.

RETROPERITONEAL HEMORRHAGE ARISING FROM A RENAL MASS: A CASE REPORT Jean Canham. University of North Dakota School of Medicine and Health Sciences, Bismarck, ND. (Tracking ID #1924635) LEARNING OBJECTIVE 1: Recognize the possibility of spontaneous hemorrhage as a complication of anticoagulation in the setting of renal cell carcinoma. CASE: A 91 year old female with a past medical history significant for stasis dermatitis, lower extremity edema, and a known left renal mass presented to the emergency department with complaints of a swollen and painful right leg of 1 day’s duration. The renal mass had been previously detected on a CT scan, but the patient and her family had declined any work-up or treatment for it because of her advanced age. Physical exam revealed 3+ edema and redness to the mid-thigh of the right lower extremity. Initial labs included a CBC demonstrating WBC of 11.6 and a microcytic anemia with a hemoglobin of 10.5. Doppler ultrasound of the right lower extremity revealed a large occlusive thrombus extending from the common femoral to the popliteal vein. She was admitted and started on therapeutic heparin and warfarin. The patient initially improved on anticoagulation therapy, but 36 h after starting heparin she began to experience left flank pain. A drop in her hemoglobin was noted. An abdominal CT was performed which revealed internal hemorrhage of the left renal mass, which appeared to represent likely renal cell carcinoma. A large amount of blood and hematoma were seen in the left retroperitoneum. An extensive tumor thrombus extending from left renal vein to the inferior vena cava was also seen. Heparin was immediately stopped and vitamin K was administered. Despite these measures, her hemoglobin continued to drop and she became hypotensive. Resuscitative efforts were initiated, but her hemodynamics failed to improve and the patient expired. DISCUSSION: Renal cell carcinoma is known to be highly vascular, posing bleeding complications in surgical settings. However, little is known about the risk of spontaneous hemorrhage in RCC when anticoagulation therapy is initiated. Anticoagulation brings with it the inherent risk of bleeding, but this case forces us to ask if patients with renal cell carcinoma are particularly vulnerable to catastrophic spontaneous hemorrhage when anticoagulated.

REVENGE OF VIRCHOW’S TRIAD: MAY-THURNER SYNDROME Megan Sisk; Paula Wichienkuer; Mark Gibson. PPMC, Portland, OR. (Tracking ID #1896509) LEARNING OBJECTIVE 1: May-Thurner syndrome should be considered in cases of DVT in young patients, unprovoked clot, or extensive proximal clot burden. LEARNING OBJECTIVE 2: Definitive treatment should be pursued if the patient is symptomatic from complications related to May-Thurner. CASE: A 59 year old man presented to the ED with acute onset of left lower extremity swelling. His leg was normal at 6 am and by noon he had significant swelling, pain, and was unable to ambulate. On physical exam he had 3+ edema to the thigh, cyanosis, and the dorsal pedis pulse could be found using Doppler only. Ultrasound showed extensive occlusive clot burden extending from the left common femoral, profunda femoris, and proximal femoral veins. Due to concern for limb ischemia, he was taken for catheter directed thrombolysis. CT imaging for preparation of vascular access showed occlusive clot and compression of the left common iliac vein consistent with May-Thurner Syndrome. After 24 h of thrombolysis, the catheter was removed and angioplasty with stent placement of the left common iliac vein was performed. By day 4, the patient was discharged with complete resolution of his symptoms. DISCUSSION: In 1851, Virchow first proposed the idea that a LLE clot could be caused by this compression of the vein by the artery. In 1957, May and Thurner confirmed this idea and the principles of Virchow’s Triad through autopsy studies where they noted effects of endothelial damage and turbulent flow. May-Thurner Syndrome leads to DVT of the left lower extremity due to the narrowed turbulent flow caused by the compression and repetitive

JGIM

trauma on the left common iliac vein. The vein traverses between the right common iliac artery where it can be compressed against the spine at L5. Additionally, the constant pulsatile force leads to fibrosis of the vein wall. May-Thurner Syndrome is being identified more often due to an increase in catheter directed endovascular treatments. While findings are often found incidentally during CT or ultrasound, venogram remains the gold standard. Treatment typically involves angioplasty with stent placement to support the vein from further compression. The exact length of anticoagulation and recommendations regarding prophylactic stenting remain controversial. This case demonstrates the importance of evaluation for additional causes in an unprovoked DVT. May-Thurner syndrome should be considered in cases of DVT in young patients, unprovoked clot, or extensive proximal clot burden. Definitive treatment should be pursued if the patient is symptomatic from complications related to May-Thurner.

REVERSIBLE ENCEPHALOPATHY IN AN ELDERLY PATIENT WITH ENDSTAGE RENAL DISEASE Naoki Misumida; Krupa Majmundar. Beth Israel Medical Center, New York, NY. (Tracking ID #1925564) LEARNING OBJECTIVE 1: Recognize cefepime-induced encephalopathy as a cause of altered mental status in patients with chronic kidney disease. LEARNING OBJECTIVE 2: Review the typical clinical course of cefepimeinduced encephalopathy. CASE: A 93 year-old female with a history of end-stage renal disease on hemodialysis presented from nursing home with a five-day history of shortness of breath and dry cough. Her blood pressure was 131/66 mmHg, heart rate was 65/min and SpO2 was 92 % on room air. She was afebrile. Her cognitive function was unchanged from her baseline and neurological examination was unremarkable. She had coarse crackles on right lung base. Laboratory data did not show leukocytosis. Chest X-ray showed right lower lobe infiltration. She was admitted for suspected healthcare-associated pneumonia and fluid overload, and treated with cefepime (2 g/ day), vancomycin (500 mg, once) and azithromycin (500 mg/day). Hemodialysis was performed on the following day and her dry weight was lowered. After hemodialysis, her respiratory condition improved and the infiltration on chest X-ray resolved. All of the antibiotics were discontinued based on her clinical improvement on the second day. On the third day, she became confused with temporo-spacial disorientation and expressive dysphasia. Extensive work-up including laboratory tests, head CT scan and brain MRI did not reveal the cause of acute mental status change. Metabolic encephalopathy and central nervous system infection were considered and excluded. She started to improve gradually on the fifth day and returned to her baseline level of mental status on the sixth day. Because of the temporal association of the symptom onset and drug administration, as well as complete improvement after drug discontinuation, a final diagnosis of cefepime-induced encephalopathy was made. Patient was discharged without any sequelae. DISCUSSION: Cefepime is a fourth-generation cephalosporin used to treat severe infection, which is mainly excreted by the kidney. Cefepime-induced encephalopathy has been sporadically reported worldwide over the last decade. The neurotoxic side effects are likely secondary to suppression of inhibitory neurotransmission mediated through gamma-aminobutyric acid-A receptor. A review of 42 cases of cefepimeinduced encephalopathy most commonly observed confusion with temporo-spatial disorientation (96 %), myoclonus (33 %) and seizures (13 %) in affected patients. Almost all affected patients have had renal failure and the dose of cefepime was relatively higher for the level of renal failure. In our case, the dosage of cefepime should have been adjusted more precisely according to the creatinine clearance even though it was stopped on the second day. The latency varied between 1 to 10 days. Symptoms usually regress within 2 to 7 days after stopping cefepime. In conclusion, physicians need to be aware of this potential adverse effect although it is uncommon. To avoid this complication, dose adjustment should be undertaken in patients with chronic kidney disease.

REVERSIBLE MYELONEUROPATHY IN A YOUNG WOMAN Ximena A. Levander1; Traci A. Takahashi3,1; Maxwell Ma2. 1University of Washington, Seattle, WA; 2University of Washington, Seattle, WA; 3VA Puget Sound Healthcare System, Seattle, WA. (Tracking ID #1928649) LEARNING OBJECTIVE 1: Diagnose B12 deficiency and distinguish it from other etiologies of myeloneuropathy. LEARNING OBJECTIVE 2: Recognize nitric oxide abuse as a potential cause of B12 deficiency. CASE: A 35-year-old woman presented to clinic with 1 week of bilateral upper and lower extremity numbness. The paraesthesias occurred suddenly and felt like her extremities “fell asleep but never woke up.” She also reported difficulty ambulating,

JGIM

ABSTRACTS

requiring a cane to walk even short distances. Symptoms had neither improved nor worsened since onset. Review of systems was notably negative for vision changes, headaches, new neck or back pain, and urinary or bowel incontinence. She denied any recent infections, sick contacts, travel, injury or immunizations. Medical history was significant for PTSD, depression, chronic pain, and cervical and lumbar surgeries. Her medications included morphine, etodolac, citalopram, lorazepam and a multi-vitamin. Family history was unremarkable. She had a 20-pack-year smoking history but denied any alcohol or illicit drug use. On physical exam she had normal vital signs except for a BMI of 30. Her neurological exam was notable for decreased pinprick sensation in a stocking and glove distribution bilaterally and hyper-reflexia in the lower extremities. Her gait exam demonstrated spasticity and sensory ataxia such that she was unable to ambulate unassisted. The neurology service admitted her and initial laboratory results were notable for normal electrolytes and a slight normocytic anemia. She had normal TSH and copper levels. Her RPR, treponemal antibody and ANA panel were negative. Her B12 level was 253 pg/mL (normal >211 pg/mL). Both her methylmalonic acid and homocysteine were elevated. MRI brain was unremarkable. MRI C and T spine demonstrated non-contiguous bilateral dorsal column lesions with high T2 signal that was non-enhancing with contrast. On further questioning, she revealed almost daily use of inhaled nitrous oxide for the past few weeks. She was started on cobalamin 1,000 mcg IM daily for 6 doses then switched to 1,000 mcg PO daily for 2 months. She was discharged to rehab and advised to stop using nitrous oxide. Within 1 month her ataxia resolved and she only had minor residual sensory deficits. In 2 months her B12 level was 1,092 pg/ml. DISCUSSION: Nitrous oxide is an inhaled anesthetic used in minor medical and dental procedures. Its recreational use started shortly after its discovery in the eighteenth century, when English chemist Humphry Davy demonstrated its psychogenic effects. Today, nitrous oxide continues to be an inhaled drug of abuse, obtained from medical settings or from whipped cream cartridges, commonly referred to as whippets. The toxicity of nitrous oxide is thought to be caused by irreversible binding of the cobalt ion in cobalamin, rendering it unable to function as a coenzyme with methionine synthase. The enzymatic reactions catalyzed by methionine synthase are critical for DNA synthesis and for regeneration of carbon donor groups. Primary care providers need to be aware of the medical complications of nitrous oxide abuse since patients may present to clinic with its sequelae and may not readily volunteer information regarding abuse on social history screening. The differential diagnosis for our patient’s initial presentation was extensive and most worrisome for anatomic injury given her spinal surgical history but also included autoimmune conditions, malignancies, nutritional deficiencies and infections. Her MRI findings were consistent with many conditions on this differential and a lumbar puncture with cerebrospinal fluid analysis was the next step in evaluation. However, the patient later revealed her substance abuse history deeming this diagnostic procedure unnecessary. A high degree of suspicion was required given she denied any substance abuse both in clinic and to the initial inpatient team. Case reports of nitrous oxide toxicity have been described in toxicology, neurology, emergency medicine and anesthesiology texts and highlight patients presenting with myeloneuropathy and/or megaloblastic anemia. Most patients reported prolonged nitrous oxide abuse and had true B12 deficiency, unlike our patient who had a borderline low normal B12 level. Because of this lab value and her well-nourished appearance, the inpatient team tested her methylmalonic acid and homocysteine levels, which came back elevated. Given her presenting symptoms, MRI findings and the lab results, as well as her response to high dose B12 repletion, nitrous oxide had caused an effective B12 deficiency. Fortunately, symptoms of nitrous oxide abuse related cobalamin deficiency often resolve upon stopping nitrous oxide use and aggressively repleting B12.

REVERSIBLE POSTERIOR LEUKOENCEPHALOPATHY SYNDROME: MRI SAYS IT ALL Charu Ramchandani; Waleed Quwatli; Chris Burke. Unity Health System, Rochester, NY. (Tracking ID #1939239) LEARNING OBJECTIVE 1: Recognizing the clinical and radiologic findings in reversible posterior leukoencephalopathy syndrome for the timely initiation of treatment. CASE: A 28 year old African American female presented to the hospital with a constant headache in the frontal region, 9/10 in severity not responding to over the counter pain medications. The pain was associated with fever, nausea, right sided neck pain, blurred vision and photophobia. She was seen by her primary care doctor a week ago who prescribed her amoxicillin for sore throat. Her past medical history was significant for viral meningitis in 2006 and atopic eczema. There was no significant family history. She denied any recent sick contacts or travel outside the country. On physical examination, vital signs were stable, oral examination revealed mild pharyngeal erythema, and pain on flexion of the neck. There was no associated photophobia, vesicular rash or any focal neurologic

deficits. Laboratory values were remarkable for white blood cell count 17000/cu mm; Cerebro-spinal fluid (CSF) showed - 8/cu mm of red blood cell, 15/cu mm WBC with 26 % segmented cells and no organisms on gram stain. CSF protein and glucose were 17.5 and 54.5 mg/dl respectively. Head Computed tomography (CT) scan with contrast was normal. Patient was seen by the infectious disease concerning for meningitis for which she was started on Vancomycin, Ceftriaxone and Acyclovir with a pending CSF culture and polymerase chain reaction (PCR) for Herpes simplex virus (HSV) 1 and 2. Three days after the admission, patient continued to have persistent headache, nausea, vomiting and her antibiotics were stopped because PCR for HSV and CSF culture came back to be negative. On the 7th day, she experienced two episodes of generalized tonic clonic seizure along with hypertensive urgency/persistently elevated blood pressure readings. Immediately, she was started on Keppra for seizures and Nicardipine drip for her blood pressure control and was taken for a repeat CT scan. Repeat CT head showed decreased attenuation within the right parieto-ocipital region which was suspicious for an acute infarct. Magnetic resonance imaging (MRI) showed numerous areas of patchy ill-defined high T2 and flair signal in bilateral supra- and infra-tentorial white matter with effacement of overlying sulci and fissures suggestive of reversible posterior leukoencephalopathy. Over the next few days, Nicardipine was stopped and she was started on oral Labetalol. After controlling for her high blood pressure, her headache and blurry vision resolved, and she was discharged home on a regimen of oral Labetalol. A repeat MRI month after the discharge showed resolution of the high T2 flair signal in bilateral supra- and infra-tentorial white matter along with continued improvement in her symptoms. DISCUSSION: Reversible posterior leukoencephalopathy syndrome (RPLS) is a clinic-radio-neurologic entity presenting as headache, visual symptoms, seizure with accompanying symmetrical white matter edema in posterior cerebral hemispheres on MRI brain. Hypertensive encephalopathy, eclampsia and use of cytotoxic or immunosuppressant drugs are the most common causes of RPLS. Etiology remains unclear, but some of the proposed mechanisms which could lead to RPLS are: (1) failure of auto-regulation of cerebral blood flow in severe hypertension leading to cerebral hyperperfusion and breakdown of blood brain barrier; (2) damage of vascular endothelium with cytotoxic agents leading to disruption of blood brain barrier and vasogenic edema. Primary involvement of the parieto-occipital region is not well understood. MRI is the best diagnostic modality to support the diagnoses of RPLS. Treatment consists of rapid lowering of blood pressure and discontinuation of any offending agent as mentioned earlier. In this case report, the patient was neither pregnant nor on any immunosuppressive drugs and the most likely underlying etiology/triggering event could be the rapid elevation of blood pressure on an unrecognized essential hypertension. Improvement of her symptoms with control of hypertension and resolution of MRI findings further supports the diagnosis of RPLS. RPLS can be easily confused with a stroke/stroke mimic based on CT scan findings and MRI should be obtained to help differentiate the diagnoses. Differentiating RPLS from stroke is very important for early management strategies in both the conditions. RPLS is a reversible condition if recognized early and treated promptly. It is essential for the internists to recognize this entity early to avoid long term complications in patients presenting in a similar fashion.

RIBBON-SHAPED STOOL IS NOT ALWAYS CANCER Maryam Sattari; Ravneet Bajwa; Robert S. Egerman. univeristy of Florida, Gainesville, FL. (Tracking ID #1935612) LEARNING OBJECTIVE 1: Recognize signs and symptoms of retroperitoneal fibrosis LEARNING OBJECTIVE 2: Become familiar with diagnosis, management, and treatment options for retroperitoneal fibrosis CASE: A 47-year-old male presented with 2 months of abdominal pain, constipation, fatigue, and 20-lbs. weight loss. He described “something pushing” on his colon and the urge to defecate. He had also noted a change in his stool caliber and described straining 4–6 times daily and difficulty passing “ribbon shaped” stools. Past medical history was remarkable for migraine headaches and use of ergot-based medications 10 years prior to presentation. Physical exam, including vital signs, was normal. Labs, including CBC and renal function, were also normal. Abdominal and pelvic CT scan showed periaortic soft tissue thickening surrounding the infrarenal abdominal aorta and proximal common iliac arteries, suggestive of retroperitoneal fibrosis (RF). Peripheral blood flow cytometry did not show atypical findings. Colonoscopy and EGD were normal. Patient was started on high dose steroids (1 mg/kg), reported complete symptom resolution within days, and had marked radiographic improvement in the extent of disease. After a month, daily prednisone dose was decreased to 10 mg. After remaining symptom-free and having stable imaging for 8 months, patient presented to an outside facility with recurrent symptoms, including “flat stools.” Outside CT reported aortitis. Addition of cellcept has resulted in improvement of his symptoms to date.

ABSTRACTS DISCUSSION: RF encompasses a range of diseases characterized by presence of fibro-inflammatory tissue in the retroperitoneum. Presenting signs and symptoms are non-specific and include fever, pain in abdomen/flank/back, anorexia, weight loss, renal failure, hypertension, and deep venous thrombosis. While the majority of RF cases are idiopathic, others are associated with recognized etiologies, such as malignancy, certain infections, drugs (e.g. ergotomines, hydralazine and beta blockers), abdominal aortic aneurysm, radiation therapy, surgery, or trauma. It has also been suggested that RF is mediated by the immune system. In fact, idiopathic RF may be part of a systemic fibrosing disease, such as immunoglobulin G4 (IgG4)related disease. We believe our patient’s RF is either IgG-related or associated with his previous use of Ergot compounds. Laboratory findings are also nonspecific and consist of anemia and elevated erythrocyte sedimentation rate and serum creatinine. CT and MRI are the best diagnostic modalities for RF. As illustrated by our case, classical radiographic findings include a confluent mass surrounding the infrarenal aorta and other retroperitoneal structures. Biopsy is not necessary for the diagnosis, but is appropriate if malignancy or infection is suspected, location of fibrosis is atypical, or there is inadequate response to initial treatment. Careful history and physical exam would reveal causes of secondary RF, such as malignancy or drugs. Suspected drugs should be discontinued. If appropriate, patients should undergo ageappropriate malignancy screening. Surgical treatment might be required for obstructive complications. Successful nonsurgical management of idiopathic RF with various immunosuppressive agents (e.g. corticosteroids, azathioprine) and hormonal medications (e.g. tamoxifen) have been reported. This case highlights the importance of maintaining a high index of suspicion for RF, even in patients without classical urinary symptoms. Prompt diagnosis and treatment not only provide symptom-relief, but can also preserve renal function, prevent other organ involvement, and reduce need for surgical intervention.

RIZATRIPTAN INDUCED RENAL INFARCT IN A MIDDLE AGED MALE WITH MIGRAINES Saurabh K. Bansal; Emani Vamsi. University of Illinois College of Medicine at Peoria, Peoria, IL. (Tracking ID #1938756) LEARNING OBJECTIVE 1: With this case we would like to emphasize to practicing physicians to be able to recognize renal infarction as one of the potential side effects of the Rizatriptan. LEARNING OBJECTIVE 2: With this case we would like to emphasize to practicing physicians about the vaso-occlusive and ischemic side effects of Triptan class of medications. CASE: 44 year old male presented with 1 week history of nausea, vomiting and vague diffuse abdominal pain. He denied fever, chills, dyspnea, diarrhea and abnormal weight gain/loss. He had history of migraines and was on Topamax for preventive therapy and Rizatriptan for as needed therapy for recurrent headaches. He took approximately two tablets of rizatriptan per week for last few weeks. He denied any personal or family history of autoimmune disorder or clotting disorders. Reported smoking marijuana rarely but no cocaine. On visual appearance he appeared dehydrated. Vital signs were normal, pulse rate 92 beats/min and blood pressure of 115/72 mm of Hg. Abdomen was soft with reproducible tenderness in the left CVA region. No murmur was heard during the cardiovascular examination. CT scan with contrast of abdomen showed hyperechoic wedge shaped shadow on the upper pole of the left kidney. This was highly suggestive of a renal infarct. Hemoglobin, creatinine, electrolytes, liver function tests and urinalysis were all within normal limits. Urine drug screen was only positive for cannabinoids. ESR, CRP elevated at 33 and 6.79 respectively. Prothrombin time and PTT minimally elevated at 16.9 s and 41 s respectively. ANA, RF and ANCA, Factor V leiden, lupus anticoagulant and homocysteine tests were all negative. Protein C&S levels were slightly lower than normal levels. EKG showed sinus rhythm and 2D Echocardiogram did not show any LV thrombus. Renal Doppler ultrasound did not show any renal artery stenosis. He was managed conservatively and improved slowly in the next few days. No further complications were noted. Repeat urinalysis and creatinine were normal. DISCUSSION: Rizatriptan/maxalt is fairly safe medication in terms of side effects. Myocardial infarction, stroke, GI ischemia or splenic infarction had been mentioned on the prescriber’s sheet. Renal infarction is not mentioned anywhere in the list of side effects. Renal Infarction is a rare disorder. Most common causes of renal infarction are - Thromboemboli or in-situ thrombosis. Patients with embolic phenomenon have atrial fibrillation, cardiac thrombus, endocarditis or aortic atheroma. Patients with non-embolic occlusion have renal artery dissection, fibromuscular dysplasia or hypercoagulable state. In our patient, hypercoagulable workup was negative by lab testing and negative renal duplex study ruled out renal artery stenosis or dissection. EKG and in hospital cardiac monitoring for 72 h showed normal sinus rhythm. Transthoracic Echocardiogram was negative for any cardiac thrombus or vegetations. We believe in this healthy adult middle aged male, renal

JGIM

infarct was a result of prior Rizatriptan use. Rizatriptan had been previously associated with coronary ischemia, gastrointestinal ischemia and splenic infarcts due to peripheral vasoconstrictive properties. So this is completely plausible to have vasoconstriction elsewhere resulting in ischemia, such as what happened, presumably, in a segmental branch of left renal artery in our patient. Renal infarct due to Rizatriptan had only been reported once before (see reference 1), describing two such cases1. Our patient was discharged home with advice not to use Rizatriptan again. No anticoagulation was started since we did not find any evidence of hypercoagulability or clotting disorder or this as an acute embolic/thrombotic event. References: 1) Renal infarction during the use of rizatriptan and zolmitriptan: two case reports. Fulton JA, Kahn J, Nelson LS, Hoffman RS. Clin Toxicol (Phila). 2006;44(2):177–80.

ROUX EN YAM I SO TIRED? Rachel Sandler; Catherine Jones. Tulane University Health Sciences Center, New Orleans, LA. (Tracking ID #1926241) LEARNING OBJECTIVE 1: Identify gastric bypass surgery as a risk factor for iron deficiency anemia LEARNING OBJECTIVE 2: Recognize the nutritional deficiencies associated with gastric bypass surgery CASE: A 43-year-old African American woman presents with complaints of fatigue and shortness of breath worsening over the past year. She also noted associated chest pressure when having long conversations and with exertion. Past medical history was notable for recurrent episodes of anemia requiring transfusion along with laparoscopic gastric bypass surgery that was converted to an open procedure over 10 years ago. Her only medication was a multivitamin. On physical exam, vital signs were normal. She had anterior rim pallor and prolonged capillary refill. She had a 2/6 systolic flow murmur and lungs were clear to auscultation. Complete blood counts showed hemoglobin of 4.4 g/dL and a mean corpuscular volume of 52.7 fl. Peripheral blood smear showed microcytes and polychromasia. Ferritin was 1.9 ng/mL. Hemoglobin electrophoresis showed no evidence of hemoglobinopathy. On chart review, 3 years ago iron studies revealed iron of 9 μg/dL, iron saturation 9 %, and total iron binding capacity 419 μg/dL. Transferrin was 322 mg/dL.. Ferritin at that time was 2.9 ng/mL. During her hospital course, she was transfused 2 units of packed red blood cells with appropriate rise in her hemoglobin and significant improvement in her symptoms. She was initiated on a trial of oral iron therapy prior to discharge. DISCUSSION: While gastric bypass surgery has become a common and effective approach to treating obesity, this surgery, particularly Roux-en-Y, is not without its long-term complications, including iron deficiency anemia. The Roux-en-Y gastric bypass works by both restricting the size of the stomach through the creation of a remnant pouch and malabsorption by shortening the small intestine. Positive effects of this procedure include significant weight loss, reduction in cholesterol and triglyceride levels, and reversal of type 2 diabetes. However, as in this patient, 6– 50 % of patients develop clinically significant iron deficiency anemia after Roux-enY surgery secondary to disruption of iron absorption in the duodenum and proximal jejunum, which are bypassed in the procedure. As gastric bypass procedures have become a more common method for treating obesity, a growing body of evidence has shown that nutritional deficiencies are common after surgery. Iron and vitamin B12 are particularly common micronutrient deficiencies, along with the fat-soluble vitamins (A, D, E, and K) and thiamine (vitamin B1). In addition to factors directly influencing iron absorption, chronic inflammation due to obesity and changes in iron hemostasis related to decreased hepcidin can also play a concomitant role in anemia in bariatric surgery patients. In some cases, oral iron supplementation is sufficient to replete iron stores, especially at higher doses, and with reducing agents like Vitamin C. Some cases do require IV iron for resolution. Recognition of iron deficiency in the setting of bariatric surgery is essential to recognize as a complication of this common obesity treatment.

RUPTURE OF SINUS OF VALSALVA ANEURYSM: A RARE BUT IMPORTANT CAUSE OF HYPOTENSION AND RENAL FAILURE Ahmed Salman; Patrick Quinlan; Anthony Donato. Reading Health System, West Reading, PA. (Tracking ID #1934613) LEARNING OBJECTIVE 1: In patients presenting with acute hemodynamic instability and new cardiac murmurs, cardiac etiology should be suspected . LEARNING OBJECTIVE 2: A bedside transthoracic echocardiogram is an effective approach to rule out rupture of sinus of valsalva aneurysm . CASE: Aneurysms of Sinus of Valsalvas (ASVs) are rare cardiac anomalies found at autopsy studies in 0.09 % of the population, as the result of congenital malformations, connective tissue disease (e.g. Ehler-Danlos, Marfan’s), infections or thoracic trauma. While typically asymptomatic, rupture of the aneurysm will result in profound hypotension and

JGIM

ABSTRACTS

renal failure. Fifty-seven year old man presented to emergency department with a 5 day history of non-radiating mild precordial chest pain associated with nausea, vomiting and decreased urine output. On examination he was afebrile with a blood pressure of 70/40 and a pulse rate of 95 beats per minute. Despite receiving 4 l of IV normal saline, his blood pressure continued to range between 80 and 100 systolic. Initial workup revealed elevated BUN and creatinine concentrations (190 mg/dl (Normal 5–26 mg/dl) and 17.6 mg/dl (Normal 0.5– 1.5 mg/dl) respectively). On physical exam the patient was noted to have a continuous 3/6 murmur on left sternal border with a friction rub. Given the refractory hypotension and new murmur, a transthoracic echocardiogram was urgently performed, which revealed a ruptured 1.4 cm aneurysm of the sinus of valsalva. Cardiothoracic surgery was consulted, who took the patient urgently to open repair. The patient required 10 days of hemodialysis support postoperatively but recovered kidney function (Creatinine 1.65 mg/dl) by time of discharge. DISCUSSION: Rupture of Sinus of Valsalva should be suspected in patients presenting with acute hemodynamic instability and new cardiac murmurs. Prompt diagnosis of this rare condition is required for appropriate surgical management.

SCLEROSING MESENTERITIS - A RARE DIAGNOSIS THAT MAY PRESENT AS ACUTE ON CHRONIC ABDOMINAL PAIN Anil Kumar Reddy Anumandla; Gaurav Kistangari; Omair Javed. Clevaland Clinic, Cleveland, OH. (Tracking ID #1939221) LEARNING OBJECTIVE 1: Learning the differential diagnosis and treatment of Sclerosing Mesenteritis CASE: A 66 year old man presented with complaints of worsening epigastric abdominal pain for 4 days and inability to tolerate oral intake due to severe pain. He reported similar but milder post-prandial epigastric discomfort with associated 15 lb weight loss over the past year. His past history was significant for lung cancer (treated 10 years ago) and autoimmune hemolytic anemia. At presentation his vitals signs were stable. Physical examination was benign aside from severe tenderness in epigastric region. Admission labs including CBC and BMP were normal except for mild leukocytosis. Lactate was normal. Computed tomography (CT) abdomen revealed a 10 cm non-heterogeneous fatty mass extending from the root of the mesentery suggestive of SM. There was no evidence of bowel obstruction so he was treated with medical management—a trial of immunosuppressive therapy with high dose prednisone 40 mg daily and colchicine 0.6 mg BID. His pain responded dramatically over the next 2 days after which he tolerated a regular diet without any discomfort. The need for biopsy was discussed extensively but given the significant improvement with medical therapy, it was felt that percutaneous biopsy might not be useful, as it would not change management. As such, the patient was discharged on prednisone and colchicine, with the option of doing laparoscopy for definitive diagnosis in the future if he fails to respond to medical treatment. Given the association of SM with malignancy and the patient’s history of lung cancer, Chest CT was done which showed no evidence of recurrence of his malignancy. DISCUSSION: Sclerosing mesenteritis (SM) is a rare idiopathic inflammatory and fibrotic condition affecting the mesentery. The treatment of SM is tailored according to the severity of the presentation. For patients who do not have significant symptoms, a conservative approach can be followed, as the disease course is often self-limited. For symptomatic patients without obstructive symptoms, a trial of medical therapy with immunosuppressants is recommended, as illustrated in this case. If the patient has bowel obstruction or is refractory to medical therapy, then surgery is an option. SM is associated with a number of other conditions including malignancies, autoimmune disorders, or abdominal surgery/trauma. As such, a thorough diagnostic evaluation should be performed to rule out any other underlying pathology. In this case, the patient’s history of lung cancer prompted a CT chest to ensure no recurrence of his cancer.

SEEING THINGS - AN UNUSUAL CAUSE OF VISUAL HALLUCINATIONS Payel J. Roy1; Anand Kartha2. 1Boston University Medical Center, Boston, MA; 2VA Boston Healthcare System, West Roxbury, MA. (Tracking ID #1938747) LEARNING OBJECTIVE 1: Assess the differential diagnosis of visual hallucinations CASE: Mr. L presented with a chief complaint of “I was hallucinating this morning.” He is a 64 year-old man with a history of cerebrovascular accident without residual deficit, coronary artery disease, hypertension, diabetes mellitus, and pulmonary embolism status post warfarin treatment. On admission, Mr. L noted increasing confusion, memory loss, and sudden onset of visual hallucinations (VH), including speaking to a nonexistent woman and wrestling a nonexistent man. Mr. L was aware of the hallucinations throughout the event and expressed concern about them. His family history was significant for Alzheimer’s dementia in his mother. Physical exam

revealed left abducens nerve palsy, right pronator drift, and slowed rapid-alternating movements. Mr. L was alert and pleasant, oriented only to name and place, had naming, abstraction, and perseverating difficulties, and 0/3 object recall at 5 min. Labs revealed a creatinine of 4.8 (baseline 1.1) and normal urinalysis. The rest of his labs, including CBC, CMP, urine toxicology, and hypercoagulable work-up, were normal. Electroencephalography was normal. MRI brain without contrast showed a subacute left hippocampal infarct. DISCUSSION: The differential diagnosis of Mr. L’s hallucinations included toxicmetabolic syndrome, possibly from acute kidney injury; seizures, given his confusion; dementia, given his family history; ophthalmologic causes, including macular degeneration, though usually only seen in patients with complete visual loss (Charles Bonnet syndrome); psychotic disorder, given the threatening nature of hallucinations; and peduncular hallucinosis, or VH arising from infarction of the midbrain. Mr. L’s acute kidney injury resolved with fluids but his hallucinations persisted and his electroencephalogram was normal, making toxic-metabolic syndrome and seizures less likely. Optometry evaluated patient; he had no visual deficits and his left abducens nerve palsy was deemed related to his diabetes mellitus. Patient had no prior history of psychiatric illness and the acute onset of symptoms makes psychotic disorder and dementia less likely. Infarction of the midbrain was ruled out by MRI, thus making hippocampal infarction the most likely cause. The hippocampus is supplied by the posterior cerebral artery and its branches. Hippocampal infarction can cause memory loss and confusion. Auditory hallucinations have been seen from a lesion restricted to the hippocampus, however VH from hippocampal infarction have not been expressly seen in the literature. Instead, patients with positive spontaneous visual phenomena, including VH, have been found to have associated hippocampal lesions in the setting of agitated delirium and hemianopia. The above presentation, work-up, and subsequent Neurology consultation affirmed hippocampal infarction as the cause of Mr. L’s VH. Visual hallucinations are not an uncommon presentation in general practice and can be a marker of serious medical illness. Internists should be able to assess for medical, neurologic, and psychiatric causes of these hallucinations. Key features leading to diagnosis in our patient included a thorough history of symptom acuity, risk factors for organic causes, a thorough neurologic exam, a broad differential diagnosis, and appropriate consultation and MR imaging use.

SEIZING THE DAY Rachel Sandler. Tulane University Health Sciences Center, New Orleans, LA. (Tracking ID #1925914) LEARNING OBJECTIVE 1: Identify key risk factors for intimate partner violence during history taking and physical examination LEARNING OBJECTIVE 2: Recognize the importance of screening for intimate partner violence in patients at risk CASE: A 55 year-old woman was admitted for new-onset seizures. She was noted to have nausea, vomiting, and blurry vision associated with a posterior occipital headache. She had an initial witnessed tonic-clonic seizure followed by multiple subsequent seizures while accompanied by EMS. She was also noted to have severe thoracic back pain. She had a history of multiple rib fractures. She did not smoke, but admitted to drinking 1 to 2 beers daily 6 times weekly. Physical exam was notable for elevated blood pressure at 165/89. She was also found to have diffuse ecchymoses on her arms and legs along with spinal tenderness along the thoracic spine. She had no focal neurologic deficits. Her laboratory studies revealed AST elevation 59 units/L with AST/ALT ratio of approximately 2:1. MRI brain showed no evidence of hemorrhage. MRI of the thoracic spine showed moderate compression fracture of T9 consistent with post-traumatic origin. During her hospital course, the patient’s seizures were treated with levetiracetam. She also underwent kyphoplasty of her compression fracture with improvement in her pain post-operatively. Prior to discharge, patient revealed that she had a long history of suffering from intimate partner violence including being hit and beaten by her husband for the past 30 years. DISCUSSION: Intimate partner violence includes threatened, attempted, or completed physical or sexual violence or emotional abuse by former or current intimate partners whether a spouse, ex-spouse, or dating partner. Such violence is experienced by 35.6 % of women and 28.5 % of men in the United States in their lifetime. Recognizing patients at risk is crucial to help prevent the morbidity and mortality associated with this condition. A variety of health conditions and risk behaviors have been associated with intimate partner violence including stroke, heart disease, arthritis, HIV risk factors, current smoking and heavy drinking. Conditions like irritable bowel syndrome, frequent headaches, chronic pain, depression, anxiety, and insomnia have all been found as consequences of intimate partner violence. Patients may also show physical evidence of trauma not consistent with the history provided. In this case, the patient had multiple risk factors including frequent alcohol use both by history and laboratory studies and history of multiple rib fractures as well as evidence of a traumatic spine fracture. In January 2013, the United States

ABSTRACTS Preventive Services Task Force released a grade B recommendation for screening women of childbearing age for domestic violence. Multiple screening tools including the HITS (Hurt, Insult, Threaten, Scream), HARK (Humiliation, Afraid, Rape, Kick), and STaT (Slapped, Threatened, and Throw) have been validated with high sensitivity and specificity for identifying patients at risk for intimate partner violence. Using these simple tools can help open up a larger conversation for patients about an incredibly personal topic. Opening up the conversation with patients will allow physicians and patients to recognize situations of abuse and move towards greater action towards prevention and providing treatment for victims of intimate partner violence.

SEROTONERGIC REACTION MANIFESTING AS DELAYED TOXICITY: R E S P O N S E TO C Y P R O H E P TA D I N E T H E R A P Y S a r a s c h a n d r a Vallabhajosyula1; Pranathi Sundaragiri1; Ariel M. Modrykamien1,2. 1AlegentCreighton University Medical Center, Omaha, NE; 2Alegent-Creighton University Medical Center, Omaha, NE. (Tracking ID #1939506) LEARNING OBJECTIVE 1: Understanding the clinical presentation and management of serotonin syndrome, highlighting the role for cyproheptadine therapy LEARNING OBJECTIVE 2: Identifying delayed manifestation of serotonin toxicity manifesting 3 months after medication alteration CASE: Introduction: Serotonin (5-HT) neurons participate in sleep-wakefulness cycles, mood, food and sexual behaviors, and thermoregulation via the central brainstem midline raphe nuclei and on vascular tone and gastrointestinal motility in the periphery1,2. Serotonin syndrome (SS) is a clinical triad of mental status changes, autonomic hyperactivity and neuromuscular abnormalities mediated via 5HT2A receptors1-3. First reported in 1980s, there were 26,733 cases and 93 deaths in 20022,3. Case Report: A 53-year-old Caucasian female with past history of diabetes, hypertension, osteoarthritis and depression presented with worsening dyspnea, tremors and loss of balance of one-day duration. Three months prior, the patient developed tremors after increase in sertraline dosage. Review of systems was positive for urinary retention, forgetfulness and cough. Home medications included promethazine, diphenhydramine, citalopram, bupropion, hydrocodone, morphine, metformin and sertraline. Vital signs at presentation—temperature 102.7 °F, heart-rate 103/min, blood pressure 212/183 mmHg, respiratory rate 20/min & oxygen saturation 80 % on ambient air. Examination demonstrated dry mucous membranes, rigid abdominal wall, tachycardia, confusion, disorientation and resting myoclonus. Laboratory studies revealed mild leukocytosis, hyponatremia and acute kidney injury (AKI) with creatinine 4.0 mg/dL (baseline 0.9 mg/dL). A mixed respiratory/metabolic acidosis (pH 7.27), with elevated serum lactate (2.4 mmol/L) and creatinine phosphokinase (CPK) (921 U/L), was noted. Thyroid and liver function testing was normal and imaging studies for chest, abdomen and retroperitoneal regions were unremarkable. Home medications of sertraline, promethazine, metformin and diphenhydramine were held. Patient did not respond to empirical physostigmine therapy for suspected central anticholinergic syndrome. Supportive measures including intravenous hydration, positive pressure airway support, bronchodilator and diazepam therapy were initiated. Patient’s labile blood pressure ranging 94 to 224 mmHg was controlled with esmolol as needed. In view of insignificant improvement, cyproheptadine therapy was initiated—loading 12 mg, then 2 mg for 8 doses (total 28 mg). Patient’s confusion, myoclonus and dyskinesia resolved with return of baseline renal function confirming a diagnosis of SS. She was transitioned to lower doses of citalopram and bupropion, with no recurrence of symptoms at sixmonth follow-up. DISCUSSION: SS is a dose-dependent continuum of clinical manifestations noted in all age groups 1–4, with onset usually immediate with ranges of 4–6 h after initial use, overdose or change in dosing with occasional reports of delayed onset (12 h) in slow release preparations2,3. Since it is a clinical diagnosis, multiple criteria in the past have been used, which were not very specific1,3,4. Currently, Hunter Serotonin Toxicity Criteria (sensitivity 84 %, specificity 97 %), which quotes clonus (spontaneous, ocular and inducible), agitation, diaphoresis, tremor and hyperreflexia as accurate predictors, is applied2,4. Occasionally, hyperreflexia and clonus are absent, which can be explained by the extreme state of muscle rigidity3,4. Bowel sounds, skin discoloration and rapidity of onset help distinguish from alternate diagnoses2. Non-specific laboratory findings include leukocytosis, high CPK and low bicarbonate with no correlation to serum serotonin levels. Management entails discontinuation of serotonergic agent and supportive therapy for prevention of respiratory failure, AKI and agitation with intravenous hydration, anxiolytics, chemical paralysis and mechanical ventilation 1–4. Direct-acting sympathomimetic agents and short acting anti-hypertensives are used to treat extremes of blood pressures2. Cyproheptadine, a histamine-1 receptor antagonist with non-specific action at 5HT1A and 5HT2A receptors, in a dose of 12–32 mg in a 24-hour period, has shown good response in management of SS. Other agents used include chlorpromazine, olanzapine,

JGIM

propranolol, bromocriptine and dantrolene, but are currently not recommended. However, further clinical trials are required to validate the use of cyproheptadine in SS2,3. References: 1. Birmes P et al. Serotonin syndrome: a brief review. CMAJ. 2003 May 27;168(11):1439–42. 2. Boyer EW, Shannon M. The serotonin syndrome. N Engl J Med. 2005 Mar 17;352(11):1112–20. 3. Isbister GK, Buckley NA. The pathophysiology of serotonin toxicity in animals and humans: implications for diagnosis and treatment. Clin Neuropharmacol. 2005 Sep-Oct;28(5):205–14. 4. Dunkley EJ et al. The Hunter Serotonin Toxicity Criteria: simple and accurate diagnostic decision rules for serotonin toxicity. QJM. 2003 Sep;96(9):635–42.

SEVERE HYPERBILIRUBINEMIA IN A HEALTHY 18-YEAR-OLD MALE Dennis D. Wang1,2; Chris Chu1,2; David McCants1,2. 1Baylor College of Medicine, Houston, TX; 2Ben Taub General Hospital, Houston, TX. (Tracking ID #1935177) LEARNING OBJECTIVE 1: Hyperbilirubinemia is a common symptom in newborns and mid-age patients with liver or biliary diseases, but is relatively uncommon in healthy 18-year-olds. It is important to take a complete and thorough family history in teenage patients who presents with acute onset jaundice because hereditary spherocytosis often precipitates bilirubin stones, and the patient may be easily be misdiagnosed with obstructive biliary disease instead. Obtaining a blood smears can also be helpful in making the diagnosis. LEARNING OBJECTIVE 2: In patients with a functioning liver presenting with hyperbilirubinemia secondary to hemolytic anemia, the hyperbilirubinemia can be predominantly direct rather than indirect. As in this case, the patient initially presented with a total bilirubin of 50.8 mg/dL and direct bilirubin of 38.5 mg/dL, but RUQ ultrasound and CT abdomen did not reveal any signs of biliary obstruction. Therefore, it is important to also work up any patient with direct-bilirubin predominant hyperbilirubinemia for causes that increases bilirubin production. CASE: An 18-year-old otherwise healthy Hispanic male presents with a 2-week history of RUQ and right flank pain and nausea, which were exacerbated by eating solid food. On the day before admission, he began noticing that his skin and the sclera of his eyes turning bright yellow, which prompted him to come into the emergency room. Over the same time frame, he also noticed that his urine had become darker than usual. The patient denies fevers, chills, diarrhea, constipation, hematochezia, melena, dysuria, unintentional weight loss, drug or heavy alcohol use, recent travels, or any sick contacts. He has no known history of biliary or liver disease. At the emergency room, the patient was found to have a total bilirubin level of 50.8 mg/dL (Ref: 0.2–1.0), direct bilirubin level of 38.5 (Ref: 2000 glucose, pH 5.0, no blood or protein. Formal UA confirmed >1000 glucose, pH 6.0, and no blood or protein. Pt had non fasting labs of Na 138 K 3.8 Cl 100 HCO3 24 BUN 17 Cr 0.77 Gluc 140 Ca 9.2. His A1c was 5.7 %. Repeat UA demonstrated >250 glucose, pH 6.0. STI screening was negative. The patient’s back pain improved with acetaminophen. DISCUSSION: The most common cause of glycosuria is diabetes mellitus where sodium-glucose co-transporters (SGLT2) in the proximal convoluted tubule, which reabsorb 90 % of filtered glucose, become saturated and glycosuria usually commences once blood glucose levels are >185 mg/dl. It is unlikely in this patient given his A1c of 5.7 %. Fanconi syndrome causes global proximal convoluted tubule dysfunction and can also lead to glycosuria. Causes include genetic abnormalities, heavy metal poisoning, medications, and amyloid. Patients typically have bicarbonate wasting, urine pH>7.5, and electrolyte abnormalities, none of which were present in this patient. Patients with nephrotic syndrome may have glycosuria but this patient did not have protein in his urine and lacked symptoms. Finally, familial renal glycosuria causes glycosuria via a defect in the SGLT2 transporter. It is very rare with a prevalence estimated to be 0.29 %. It is the most likely diagnosis in this patient as he repeatedly has large glycosuria in the absence of hyperglycemia or other renal dysfunction. The diagnosis can be confirmed via genetic testing, but it has a benign outlook so further testing is generally not indicated. Case reports indicate that some patients may have an increased risk of UTIs but otherwise no long term adverse effects. Further outcomes can be extrapolated from SGLT2 inhibitor drugs, such as canagliflozin, which are now being used to treat diabetes. Benefits include an A1c reduction ~0.5–0.7 %, weight loss via glycosuria ~1.8 kg, modest blood pressure reduction ~4.5 mmHg, and rare hypoglycemia. However, patients have an increased risk of genitourinary infections and the long-term cardiovascular and renal safety profiles are still unknown. This patient’s glycosuria likely protected him from becoming diabetic despite multiple risk factors, including obesity, Hispanic ethnicity, and family history.

SWOLLEN FEET AND PALPITATIONS AFTER DRAINING A PNEUMOCOCCAL EMPYEMA Ankita Tandon; Cynthia H. Chuang; Jed Gonzalo; Eileen Hennrikus. Penn State University Hershey Medical Center, Hershey, PA. (Tracking ID #1934365) LEARNING OBJECTIVE 1: Recognize that large volume thoracentesis can result in systemic effects of hypoproteinemia. CASE: Empyema is a life threatening complication of pneumococcal pneumonia requiring tube thoracostomy to drain the exudate. Common complications of chest tube placement include pneumothorax, tension pneumothorax, infection and organ perforation. The lesser recognized complication of systemic hypoproteinemia with resulting fluid shifts can result in significant complications. Fifty-three-year-old woman with no medical

JGIM

ABSTRACTS

history presented with fatigue, low-grade fevers, dyspnea and left-sided pleuritic chest pain. Physical exam revealed no breath sounds in the left lung, labs were significant for leukocytosis, hyponatremia, and a slightly low albumin level at 3.2 g/dL. Chest CT showed numerous large loculated left-sided fluid collections. Three simultaneous chest tubes were placed and streptokinase was administered, resulting in the drainage of 3 l of purulent fluid over the next 3 days. By day four, her leukocytosis, fever, dyspnea and chest pain had improved, but she developed new 2+ lower extremity edema and palpitations. Physical exam revealed a pericardial rub. Subsequent work-up revealed a markedly low pre-albumin level (10 mg/dL), an electrocardiogram showed atrial flutter with rate of 170 beats per minute, and echocardiogram revealed a moderate pericardial effusion without tamponade physiology. Presumed diagnosis was pericardial effusion and peripheral edema as a complication of protein loss associated with large volume thoracentesis. With protein supplements and diuretics, the effusion and lower extremity edema improved. After a week, the chest tubes were removed. She was discharged on anticoagulation and a beta- blocker for heart rate control. She was cardioverted a month later. At subsequent follow-up appointments, she has clinically improved and repeat chest x-rays are clear. DISCUSSION: We report a case of lower extremity edema and pericardial effusion resulting from protein loss and decreased oncotic pressure upon removal of 3 l of purulent fluid from the pleural space. Although, she was also hypoalbuminemic to begin with, so a part of these fluid shifts was from the disease process itself. While total protein loss is known to occur after tube thoracostomy, systemic manifestations resulting from decreased oncotic pressure is a less reported complication. Therefore, it is unclear if there are patients at particularly high risk who may benefit from a high protein diet prior to large volume empyemarelated thoracostomy. It is also unclear whether such patients need to be preemptively placed on cardiac monitors despite no cardiac history due to risk of developing arrhythmias with pericardial effusions. Either way, one must be vigilant of subsequent systemic side effects of a seemingly localized pulmonary process that was treated and improving. SYMPTOMATIC HYPOCALCEMIA, VITAMIN D DEFICIENCY AND ANEMIA Eunice Chuang. Mount Auburn Hospital, Cambridge, MA. (Tracking ID #1940011) LEARNING OBJECTIVE 1: To recognize that hypocacelmia, vitamin d deficiency and anemia are presenting symptoms of celiac disease. CASE: A 22-year-old female without past medical history presented with numbness and tingling in the extremities as well as cramps in her hands and calves for the past few weeks. She also noted that she has a history of fatigue for many years. Denies weight loss. Has had intermittent nonbloody diarrhea for many months. Physical exam showed stable vitals and remarkable for palor. Lab showed hemoglobin of 5.8 g/dl, MCV of 61.7, MCH 16.2 and calcium of 5.8 mg/dl, albumin of 3.2 mg/dl, PTH was elevated at 296 and vitamin D was low at 5 ng/ml. Iron studies showed undetectable iron and ferritin of 4. She was started on IV calcium supplementation with good response. Endocrinology and GI were consulted. Celiac studies were sent with positive endomysial antibody and gliadin IGA. Patient was discharged with instructions of gluten free diet and on calcium and calcitriol supplemtation and close follow up with primary care physician. DISCUSSION: Celiac disease is an immune-mediated inflammatory disorder of the small intestines induced by the prolamins of certain cereals causing loss of villous height, crypt hypertrophy leading to malabsorption. It is one of the most common lifelong disorders in Europe and the USA. In adults, presentation of celiac disease varies from asymptomatic to severely symptomatic, depending on the severity and the extent of mucosal involvement. Gastrointestinal symptoms are more common and include other conditions such as diarrhoea, weight loss, malaise, lethargy and abdominal pain. It is important to recognize vitamin D deficiency and hypocalcemia as presenting symptom of celiac disease. Hypocalcaemia is thought to occur due to the negative calcium balance, due to loss of villous surface area, with unabsorbed calcium binding to excess fatty acids in the intestinal lumen as a result of fat malabsorption. There is impairment of active intestinal calcium transport mechanisms because of depletion of calbindin from enterocytes. Decreased vitamin D also contributes to decreased calcium absorption. Hypocalcaemia stimulates PTH to correct for low calcium level. Secondary hyperparathyroidism occurs due to vitamin D and calcium malabsorption. It is important to recognize that hypocalcemia, vitamin D deficiency, elevated PTH in a patient with anemia can be signs of malabsorption, which can lead to diagnosis of celiac disease.

SYNCOPE DUE TO COLD WEATHER Sparsha Kukunoor; Poojita Shivamurthy. University of Connecticut, hartford, CT. (Tracking ID #1939777)

LEARNING OBJECTIVE 1: Identify the association between acute hemolytic anemia and syncope. LEARNING OBJECTIVE 2: Recognize the importance of finding cold agglutinins on peripheral smear. CASE: 81 year old female presented to the emergency room after witnessed syncope while working outside in the cold doing volunteer work for her church. Her syncopal event was not preceded by aura, fatigue and lasted a few seconds. On coming to the emergency room she was short of breath and fatigued. Further questioning revealed a similar episode 2 years ago, during which time she was diagnosed with cold agglutinin disease with titers of 1: 512. Laboratory investigations revealed hemoglobin of 6.9 g/dl from a baseline of 10.0 g/dl. Peripheral smear was found to have cold agglutinins. This was confirmed with direct anti-globulin test which showed positive poly-specific test and mono-specific test positive for complement protein C3d/C3b and negative for IgG. She was admitted and was given warmed packed red blood cells. Other laboratory data revealed monoclonal gammopathy, IgM kappa type with M component of 0.67. Bone marrow biopsy and CT of abdomen and pelvis showed no evidence of lymphoproliferative disease. She dramatically improved the next day and did not require any further transfusions. DISCUSSION: Cold agglutinins (CA) are antibodies that agglutinate erythrocytes at an optimum temperature of 0-4C. CA may be found in the serum of normal individual in the polyclonal form however in Cold Agglutinin Disease (CAD) they are monoclonal and in high titers (1). Usually effects female population in the 7th decade. CAD has an incidence of one in one million per year. A single institutional experience showed that in patients with 43,000 patients having a monoclonal gammopathy, less than 1 % had cold-reactive autoantibodies (2). The autoantibodies can be primary as in our case or secondary due to autoimmune disease, malignancy or infections. Clinical features include anemia, cold induced symptoms like acrocyanosis, Raynaud disease, livedo reticularis and sometimes cutaneous necrosis. Diagnosis is made by positive Coombs test to anti - C3, CA titers and thermal activity. Work up including a bone marrow aspiration should be done for secondary causes. Non-pharmacological management like avoidance of cold is the cornerstone of therapy. Studies as lacking and thermal amplitude of the CA antibody in many patients is such that physiologic peripheral cooling results in antibody binding sufficient to cause anemia. Pharmacological therapy has been unsuccessful until recently. Half of the patients respond to rituximab monotherapy. Fludarabinerituximab combination therapy is very effective, resulting in 75 % response rate, Complete remissions in about 20 %, Toxicity is a concern, and benefits should be carefully weighed against risks (1).

SYSTEMIC LUPUS ERYTHEMATOUS MIMICKING SEPSIS Ming Zhao1; Charu Ramchandani1; Peter M. Deane2; Carlos D. Palacio1. 1Unity Health System, Rochester, NY; 2Allergy-Asthma-Immunology of Rochester, PC, Rochester, NY. (Tracking ID #1939134) LEARNING OBJECTIVE 1: Recognize neuropsychiatric symptoms in systemic lupus erythematous (SLE). LEARNING OBJECTIVE 2: Distinguishing SLE flare from sepsis. CASE: Patient is a 38 year old female who was admitted to hospital due to worsening weakness, fatigue, shortness of breath and constant headache. Her past medical history was significant for systemic lupus erythematous (SLE) diagnosed 20 years ago. Patient stopped treatment for SLE 3 years ago. At the time of her admission, she was found to have systemic inflammatory response syndrome (SIRS), with WBC 14500, respiratory rate 36/min and heart rate 120/min. On examination patient was somnolent, breath sounds were decreased at the left lung base with scattered rhonchi and fine rales. Her labs were remarkable for white blood cell count (WBC) 14500, Hemoglobin 10 g/dl and serum creatinine 1.52 mg/dl. Shortly after admission, patient spiked high fever of 40° celsius along with worsening severe headache, increased confusion, irritability, memory impairment and lethargy. Given her severe headache, fever and SIRS, there was concern for sepsis secondary to meningitis/encephalitis. CT head and MRI/MRV did not show any abnormality. Cerebrospinal fluid (CSF) showed normal WBC, glucose and protein level with gram stain negative for organisms. She was empirically treated with Vancomycin, Ceftriaxone, Ampicillin and Acyclovir. However, given her history of SLE, there was also concern for SLE flare which could have led to her symptoms. Urine analysis was remarkable for 3+ proteinuria and numerous red cells. Chest x-ray showed cardiomegaly and small bilateral pleural effusions. Echocardiogram showed biatrial enlargement, right ventricular dilatation, severe tricuspid regurgitation, moderate mitral regurgitation, and 55 % left ventricular ejection fraction. We checked her serum C3 and C4 complement level, which were very low. We also checked her serum C-reactive protein (CRP), erythrocyte sedimentation rate (ESR) and ds-DNA, which all came back highly elevated. Rheumatology was consulted, following which she was started

ABSTRACTS on intravenous cyclophosphamide and methylprednisone. Antibiotics were stopped. Eventually her fever and headache resolved and her mental status improved markedly. Over the course of her stay her ESR came back to normal and CRP improved. Her renal biopsy showed stage 4 lupus nephritis. DISCUSSION: Neurologic and psychiatric symptoms are commonly seen in patients with SLE, either prior to the diagnosis or during the course of their illness. Neuropsychiatric systemic lupus erythematosus (NPSLE) is associated with a worse prognosis and more cumulative damage and represent an important cause of morbidity and mortality in SLE. Unfortunately, there are no diagnostic tests that can establish a specific diagnosis of NPSLE, and thus, the approach to patients with neuropsychiatric symptoms require establishment of the diagnosis of SLE and distinguish from other causes of nervous system abnormalities, such as medication effects, metabolic abnormalities and infection. It is always important to suspect infection in patients with SLE and central nervous system manifestations. Immune response in SLE is impaired due to defect in different components of humoral and cell mediated immune process, making patients prone to infection. Immunosuppressive action of the standard therapeutic regimen also increases the susceptibility to infection. However, sometimes, clinical manifestations of SLE can be indistinguishable from infection. Patients with highly active SLE may present with a picture mimicking infection or even sepsis. With severe inflammation and disease activity, such patient may present with high fever, tachycardia, hypercapnea and high WBC, just like our patient. In such circumstance, patients may benefit from immunosuppressive treatment such as corticosteroids, rather than antibiotics. Certain medications, such as Acyclovir, may even give patient a higher risk of worsening renal function and should be avoided as possible as we can. A typical workup for such circumstance may include a detailed history and physical exam to look for any clues of infection, serologic and imaging studies to look for the severity and activity of the disease as well as the possibility of underlying infection. A high ds-DNA titer may indicate highly active SLE rather than infection. Also complement levels are low in SLE flare versus high in infection. Other biological markers reported in several studies to identify infection in SLE patients are CRP, sedimentation rate, ferritin, procalcitonin, and delta neutrophil index. In conclusion, neuropsychiatric symptoms are commonly seen in patients with active SLE and should be distinguished from infection such as meningitis/encephalitis. Highly active SLE may also present with a picture mimicking infection or even sepsis. The diagnoses and treatment should always be based on a comprehensive evaluation.

TAKOTSUBO CARDIOMYOPATHY WITHOUT AN ACUTE STRESS TRIGGER Ranjan Pathak; Paras Karmacharya; Leena Jalota; Benjamin Lloyd. Reading Health System, West Reading, PA. (Tracking ID #1938289) LEARNING OBJECTIVE 1: Clinicians should consider Takotsubo cardiomyopathy in the differential diagnosis of patients presenting with chest pain, even without a history of recent stress. CASE: A 92 year old female with no significant cardiac history presented to the emergency room complaining of non-radiating pressure-like chest discomfort of 2 h, dyspnea, nausea and vomiting. She denied any history of recent stress. Her Serial EKGs revealed dynamic changes with inferior, anterolateral ischemic changes with ST elevation and T inversion. A diagnosis of inferior wall STEMI and acute congestive heart failure was made. Given her age, a decision was made to manage her conservatively. However, she continued to have chest discomfort with rising Troponin (peak 14.09), CK-MB levels and was ultimately taken for cardiac catheterization on the third day. Coronary angiography showed non-obstructive coronary artery disease. Left ventriculography revealed hyperdynamic basal segment with a normal apex. The mid segment was severely hypokinetic to akinetic consistent with atypical TC. She was medically managed with aspirin, statin, beta blocker, and ACEI with significant improvement in symptoms. DISCUSSION: Takotsubo cardiomyopathy (TC) is an acute segmental myocardial dysfunction of the left ventricle without significant lesion of the coronary artery. It presents with chest pain, electrocardiographic changes and elevated troponins frequently masquerading as acute coronary syndrome. Of the four variants, the most common type consists of apical ballooning and with hyperkinetic or normal basal region. TC is usually associated with acute or chronic emotional or physical stress. While the role of chronic stress in TC is unknown, stress as trigger may be absent in 15 % of cases. The mid ventricular variant is seen in about 17 % of the cases. The absence of the common apical left ventricular abnormality and lack of preceding stressor makes our case atypical. Clinicians should be aware of this unique syndrome, its variants and consider this in the differential diagnosis of patients presenting with chest pain, especially in post-menopausal women even without a history of recent stress.

JGIM

TB OR NOT TB: TUBERCULOUS ENTERITIS PRESENTING AS APPENDICITIS Jaimin Amin; Arthur Jeng; Stanley K. Dea. Olive View-UCLA Medical Center, Sylmar, CA. (Tracking ID #1939731) LEARNING OBJECTIVE 1: Recognize the clinical presentation of tuberculous enteritis and discuss the appropriate diagnostic and therapeutic management LEARNING OBJECTIVE 2: Differentiate between M. tuberculosis and M. bovis CASE: A 56 year-old Mexican man presented with 2 days of severe abdominal pain accompanied by fevers, weight loss, and non-bilious emesis. On examination, he was febrile and tachycardic with rebound tenderness of the epigastrium and periumbilical abdomen. Laboratory studies were significant for WBC 7.1 with 80 % neutrophil predominance, hemoglobin 8.4 with MCV 79, BUN 31 and Cr 1.1 and isolated alkaline phosphatemia 303. An abdominal CT scan demonstrated two large complicated cystic lesions with lymphadenopathy. The first lesion measured 5× 5 cm and was located in the right lower quadrant adjacent to the third portion of the duodenum with a dilated appendix concerning for perforated appendicitis. The second lesion measured 3×3 cm and was found within the body of the pancreas suggestive of pancreatic malignancy. Surgery was consulted for exploration and appendectomy but deferred. Endoscopic guided biopsy of the pancreatic cyst was unremarkable. CT guided drainage of the right abdominal cyst revealed acid-fast bacilli, which grew M. tuberculosis complex, identified as M. bovis. Initiation of rifampin, isoniazid, and ethambutol resulted in resolution of symptoms and normalization of imaging. DISCUSSION: This case emphasizes a lesser common clinical presentation of tuberculosis, that involving the GI tract. Although the initial evaluation was suggestive of appendicitis and/or pancreatic malignancy, further evaluation by CT guided drainage led to the demonstration of acid fast bacilli and the diagnosis of tuberculous enteritis from M. bovis. M. tuberculosis and M. bovis are related to the extent that it is not possible to distinguish these two mycobacteria by genetic amplification means (eg. PCR). However, they are distinguishable by biochemical means. They can both tuberculosis in humans, although their epidemiological reservoirs are different. M bovis is known to cause tuberculosis in cattle and may have evolved into M. tuberculosis. However, humans can be infected with this zoonotic pathogen by consuming unpasteurized dairy products from afflicted cattle. Because of the oral route of entry, M. bovis most commonly presents within the GI tract. Due to their ability to survive in mononuclear leukocytes and the density of lymphoid tissue within Peyer’s patches. M. bovis enteritis frequently mimics appendicitis or colon cancer, which are commonly the initial diagnoses. Suspicion for M. bovis, rather than M. tuberculosis, should arise when the isolate is resistant to pyrazinamide (PZA), given its universal genotypic resistance to this anti-mycobacterial drug. Treatment involves 6–9 months of standard anti-tuberculous medication, with the exception of PZA.

TENSION PNEUMOTHORAX IN OUT-PATIENT CLINIC Manisha Bhide. University of Colorado Denver, Aurora, CO. (Tracking ID #1923481) LEARNING OBJECTIVE 1: To recognize that certain causes of dyspnea need urgent triage. Relatively stable outward appearance may be deceptive in cases of chest pressure and dyspnea CASE: 39 year old male teacher, nonsmoker came to out-patient medical clinic with 7 days of worsening dyspnea, moderately severe chest pressure, dry cough and pleuritic chest pain. Past medical history was significant for hypertension, seasonal allergies and mild intermittent asthma. Family history was noncontributory. Exam revealed afebrile patient with heart rate of 96/min, BP 130/86 and respiratory rate of 18/min with increased work of breathing. Breath sounds were decreased on Rt side. Rt sided hyper resonance was present on percussion. EKG showed normal sinus rhythm with Rt axis deviation. Chest X-ray showed large right tension pneumothorax. Urgent Chest tube with water seal device and suction placement was done. Pneumothorax resolved over 4 days. CT Chest did not show any blebs or lung abnormality. DISCUSSION: Differential diagnosis of dyspnea and pleuritic pain in a young male includes pulmonary embolism, pleurisy, autoimmune diseases and pericarditis. Chest pain with dyspnea and respiratory abnormalities should stimulate us to put pulmonary pathology higher on the list of differential diagnosis. Hyper resonance on percussion can indicate COPD, but in a young male, nonsmoker it could be due to pneumothorax. Presence of air in the pleural space, with no obvious underlying lung pathology or HIV or trauma, is termed as spontaneous pneumothorax. The incidence ranges from 7.4 to 18/100,000 per year among men. The incidence is lower in women at 1.2/100,000. Smoking, connective tissue disease and family history are thought to increase the risk by sub-pleural bullae formation and degradation of elastic tissue. The American College of Chest Physicians has established guidelines for management of spontaneous pneumothorax. For large pneumothorax, hospitalization and placement

JGIM

ABSTRACTS

of chest tube with water seal device, with or without suction is recommended. Procedures to prevent recurrence, like pleurodesis are usually reserved only for second occurrence or for professionals like pilots and divers. Routine Chest CT is not recommended to find cause for first time pneumothorax, but CT may be done for management of secondary or complex pneumothorax. Shift of mediastinum along with unstable vitals like tachycardia, hypotension are indicative of tension pneumothorax. This case is important to remember because outwardly stable patients can conceal a tension pneumothorax which can lead to compression of the mediastinum and can have rapid clinical deterioration and death if unrecognized.

TERTIARY LGV DIAGNOSIS AFTER FISTULOTOMY: UNDERSCREENED, UNDER-DIAGNOSED, AND UNDER-TREATED Robert Spencer; David M. Levine; Richard E. Greene; Andrew A. Chang. New York University Medical Center, New York, NY. (Tracking ID #1939016) LEARNING OBJECTIVE 1: Formulate a differential diagnosis for rectal pain in MSM LEARNING OBJECTIVE 2: Recognize and diagnose lymphogranuloma venereum CASE: A 27 year-old gay man presented to the ED with 2 weeks of much-worsened 7/10 rectal pain with bloody purulent discharge, fevers, chills, tenesmus, and a week of a rapidly enlarging tender L-sided inguinal mass. Five years prior to presentation he was the victim of a sexual assault involving forced receptive rectal penetration by the assailant’s penis and a sharp object causing a rectal abscess requiring incision and drainage (I&D) and infection with HIV. Until last year, he noted only mild rectal tenderness without drainage. At that time (CD4 800 [40 %], VL 2 weeks) is also associated with higher risk. Hepatotoxicity is clearly linked to the clavulanic acid moiety, with a 5- to 9-fold increase for the combination versus amoxicillin alone, giving an estimated incidence rate of 9.91 cases of jaundice per 100,000 prescriptions of amoxicillin-clavulanate. The cause is often elusive because the symptoms typically arise after completion of the antibiotic regimen (average delay of 17 days) and therefore a thorough inquiry into drug history is key. Treatment includes withdrawal of the offending agent and symptom management with ursodeoxycolic acid (which may improve the cholestasis) and cholestyramine. Despite this some patients may progress to fulminant hepatic failure resulting in transplantation or death. Amoxicillin-clavulanate is a one of the most widely prescribed antibiotics worldwide. Reports of adverse reaction with cholestasis, abnormal liver function test were less than 1 %, however, while rare, the morbidity and mortality associated with drug induced hepatoxicity is significant. This case emphasizes the importance of the judicious use of antibiotics.

THE CASE OF MS. A. EARLY SUPPORT FOR A GRASS-ROOTS, TEAMBASED APPROACH TO AMBULATORY CARE IN A RESIDENT CLINIC Nora Segar; Manik Chhabra; Janet J. Ho; Megan S. Lemay; Faye Farber; Sarita Soares; Julie R. Rosenbaum; Stephen Holt; Tracy Rabin. Yale University, New Haven, CT. (Tracking ID #1939139) LEARNING OBJECTIVE 1: Identify the challenges of caring for complex patients in a traditional model of ambulatory resident education. LEARNING OBJECTIVE 2: Describe the implementation and potential benefits of a resident-organized, team-based outpatient care model. CASE: Ms. A is a 52-year-old, unemployed, high-school-educated, Puerto Rican mother of three, and the primary care giver for her grandson and frequently hospitalized mother. She has uncontrolled diabetes with average hemoglobin A1C of 11, hyperlipidemia, asthma, tobacco dependence, hidradenitis suppurativa with recurrent groin abscesses, uncontrolled depression with history of a prior suicide attempt, and chronic pain secondary to spinal stenosis and adhesive capsulitis, managed with opioids. Of note, she has previously violated the clinic’s controlled substance agreement by taking non-prescribed opiates for worsening pain. Lead author NS has followed Ms. A at our resident continuity clinic since starting training 2 ½years ago. During this time she has seen Ms. A 20 times and spoken with her over the phone on a weekly basis. In our established care model, NS cares for her with the support of a faculty preceptor. When NS is unavailable, other residents not familiar with Ms. A provide interim care. Our clinic does not currently employ social work staff, full-time mid-level providers, or a dedicated behavioral health team integral to the successful multidisciplinary care of patients like Ms. A. These resource limitations, in combination with complex resident schedules that often prioritize inpatient rotations, result in a fragmented clinical experience. For example, in addition to three hospitalizations and six emergency department visits, Ms. A has been seen 15 times by 11 distinct clinic providers other than NS in the past 2 ½years. One model for effective outpatient care and improved training environment is the Patient Centered Medical Home (PCMH), a team-based, multidisciplinary approach to primary care for complex patients. At our clinic, a core group of interested

ABSTRACTS residents have organized a pilot PCMH model called the Chase Ambulatory ICU (ChICU) designed to coordinate care for a subset of our highest risk patients and to improve resident satisfaction with the clinic experience. Enrolling Ms. A in the ChICU meant: 1) Establishment of mutual provider-patient goals, which, in her case, included compliance with the clinic’s controlled substance agreement, smoking cessation, and improved diabetes control; 2) Direct access to residents familiar with the patient through a direct-to-provider pager system; 3) Structured monthly team sign-out of patient updates and goals with a core group of known covering resident providers; 4) Completion of a home visit assessment. Two weeks post enrollment, while NS was away on a six-week international health elective, Ms. A was hospitalized for an asthma exacerbation under the care of a medicine resident who had received sign-out, supervised by the faculty preceptor. The same resident then saw her in post-hospital follow up. She has since met the goal of tighter diabetes control with improvement in her AIC from 11 to 8 through frequent structured visits to residents who are familiar with her care. She has been compliant with the controlled substance agreement as evidenced by multiple appropriate urine toxicology screens. Additionally, the home visit helped the care team to appreciate her biopsycho-social complexity as well as to identify further barriers to accomplishing our mutual goals. DISCUSSION: Patients with low health literacy, complex medical disease and low socioeconomic status may have difficulty navigating a fragmented and complex primary care system. Residency training models focused on inpatient training at the expense of outpatient continuity of care further complicate this fragmentation. Ambulatory ICU models designed to improve continuity of care may improve patient health outcomes for the under-served, medically complex patients served by residency clinics. Although these models traditionally rely heavily on incorporation of non-physician health care providers, we feel the above clinical vignette demonstrates the potential benefits of a structured, resident-driven intervention to improve the quality and continuity of care we provide for patients like Ms. A. Key teaching points: 1) Team-based care can improve outcomes for the complex patients cared for in residency clinics 2) Grassroots, resident-organized teams have the potential to improve outcomes even in the absence of interdisciplinary support and full-time clinic staff.

THE DEVIL IS IN THE (BLOOD SMEAR) DETAILS Naveen Nannapaneni; Adeeb Bulkhi; Marwan Al-Hajeili; Carter Bishop. Wayne State University/Detroit Medical Center, Detroit, MI. (Tracking ID #1923332) LEARNING OBJECTIVE 1: Recognize the utility of a peripheral smear. LEARNING OBJECTIVE 2: Diagnose thrombotic thrombocytopenic purpura. CASE: Thrombotic thrombocytopenic purpura (TTP), characterized by low levels of ADAMTS13, abnormal platelet and endothelial activation, microvascular fibrin deposition, and hemolysis, is a disease that carries a high mortality rate. The early recognition of, and emergent plasmapharesis for, TTP results in favorable outcomes. Delay in making the diagnosis can lead to unnecessary investigations, incorrect therapies, prolonged hospitalization and even mortality. A 53 year-old woman with lupus was brought to the ER with progressive confusion, generalized pain, weakness and fatigue going on for 3 days. She was unable to provide further history due to her confusion. On exam her vitals were stable and aside from being confused, had no other abnormality. CBC showed an anemia, 11.2 g/dL, and severe thrombocytopenia 11/μL. She had evidence of hemolysis with an elevated LDH 816 U/L and low haptoblogin 10 mg/dL. After being admitted she was transfused with platelets pending a hematology consultation. Hematology suspected a broad differential including immune thrombocytopenia (ITP), plaquenil-induced and hydrochlorothiazide-induced thrombocytopenia and recommended starting corticosteroids while holding the implicated medications. The patient continued to deteriorate without improvement in her platelet count despite receiving platelet transfusions and incurred further mental status decline. Neurology was consulted and suspecting a stroke, performed a CT followed by MRI/MRA of the head, which were all negative for an acute process. Additional evaluation by EEG revealed no epileptiform activity. A second opinion by another hematologist was requested, and after review of the peripheral smear revealed schistocytes, a diagnosis of TTP was made. Immediate daily plasmapharesis was performed and platelet transfusions were discontinued resulting in progressive clinical improvement. Further workup showed a decreased ADAMTS13 activity level, 35 %. She was discharged after plasma exchanges resulted in a return to her baseline mentation and stabilization of hemoglobin and platelet count. DISCUSSION: The classic presentation of TTP involves the pentad of fever, hemolytic anemia, thrombocytopenia, acute renal failure and neurologic changes. While all five findings are not necessary for diagnosis, whenever a patient presents with microangiopathic hemolytic anemia and thrombocytopenia, the diagnosis of TTP

JGIM

must be entertained. When included, interventions such as platelet transfusions, and corticosteroids alone, would not be considered as they cause further harm by additional platelet activation and confer no benefit, respectively. Review of a blood smear should be performed as the presence of numerous schistocytes provides rapid, inexpensive, and reliable evidence of TTP. Once diagnosed, the patient should undergo plasmapharesis. As in this case, if the individual problems in TTP are not assessed in a greater constellation there can be inappropriate investigations and therapies which can lead to an unfavorable outcome.

THE DILEMMA OF MANAGEMENT OF CARDIAC TAMPONADE IN THE PRESENCE OF SEVERE PULMONARY HYPERTENSION: OUR EXPERIENCE Vimalkumar Veerappan Kandasamy; Ajaykumar Kaja; Tammy O. Wichman. Creighton University, Omaha, NE. (Tracking ID #1939129) LEARNING OBJECTIVE 1: To recognize pericardial effusion as a common cause of worsening respiratory status in a patient with pulmonary hypertension. LEARNING OBJECTIVE 2: To understand the complications of pericardiocentesis in patients with severe pulmonary hypertension presenting with cardiac tamponade. CASE: A 33 year old Caucasian female with past medical history of pulmonary hypertension (PH) WHO group 1, functional class 3 at baseline, was admitted with complaints of worsening shortness of breath (SOB) and cough for the last 2 weeks. She had SOB at rest with increased oxygen requirements at the time of admission. She was on treprostinil continuous infusion, tadalafil, ambrisentan, digoxin and warfarin for PH. Jugular venous distention and positive Kussmaul sign were noted on examination. The patient went into hypoxic respiratory failure and subsequently intubated. Echocardiogram showed a large pericardial effusion with tamponade physiology with left atrial and ventricular collapse. The right ventricle (RV) was severely dilated and systolic function was severely reduced with RV systolic pressure of 161 mmHg. Pericardiocentesis was performed through a subcostal approach and 40 ml of hemorrhagic fluid aspirated. Pigtail catheter was left in the pericardial cavity and 50–100 ml of pericardial fluid was drained every 4 h. Patient also was started on antibiotics and vasopressors. A total of 2300 cm3 of fluid was aspirated from pericardial cavity with slow intermittent drainage over the next few days. Her clinical condition slowly improved over the week and vasopressors were weaned off. She was slowly diuresed and subsequently extubated and oxygen requirements came down. Although patient improved, her baseline oxygen requirement was higher than preadmission levels and she was discharged with 2 L oxygen at rest and 4 L oxygen with activity and at night. She was referred to transplant center for lung transplantation. DISCUSSION: Pulmonary arterial hypertension (PAH) in its advanced stages is complicated by right heart failure (RHF) and pericardial effusion. The optimal treatment of large or hemodynamically significant pericardial effusions in this group is not defined. The prevalence of pericardial effusion can be as high as 54 % and it is one of the predictors for mortality. It is hypothesized that a high PA pressure reduces lymphatic and venous return to a pressure-overloaded RA, leading to the pericardial effusion and eventual tamponade. The full picture of RV failure with low cardiac output and elevated RV filling pressures is typically seen in PH groups 1 and 4. When patients present with worsening RHF, echocardiogram to evaluate for pericardial effusion is an essential component of management. Multiple case reports have been published with various treatment strategies. In a case series of 6 patients by Hemnes et all, both pericardiocentesis and pericardial window placement were associated with significant postprocedural mortality. It was postulated that after removal of pericardial fluid, the already pressure- and volume-overloaded RV can enlarge further, resulting in further septal bowing and subsequently diminishing the LV size leading to worsening left heart filling, left HF, hypotension, and death. Our case report emphasizes the fact that removal of pericardial fluid in small quantities over longer period with adequate hemodynamic support could be a good treatment option for these patients, as it provides time for the right ventricle to adapt to the changing hemodynamics. More experience is needed before concluding this as a standard of care.

THE ELUSIVE DIAGNOSIS OF PULMONARY EMBOLISM; A CASE REPORT Mai Kaga; Leonard Amoruso; Peter Rattner. Beth Israel Medical Center, New York, NY. (Tracking ID #1897760) LEARNING OBJECTIVE 1: Recognize the limitations of using a single imaging modality for diagnosis of pulmonary embolism LEARNING OBJECTIVE 2: Diagnose pulmonary embolism using imaging modalities in addition to laboratory values and clinical decision making tools CASE: 35 year old male with history of heroin abuse presented with altered mental status after a witnessed heroin overdose. In the emergency department (ED), after

JGIM

ABSTRACTS

receiving intravenous naloxone, the patient’s mental status improved. The patient disclosed that in the days prior to admission, he had a cough productive of yellow sputum. He had also been experiencing non-radiating chest pain occurring at rest, described as pins and needles for 1 day. He denied fevers, chills, shortness of breath, dysuria, and muscle aches. The patient’s initial vitals were remarkable for mild tachycardia to 105; otherwise, he was afebrile and normotensive. Heart sounds were normal with no murmurs, lungs were clear, and there was no evidence of lower extremity edema or JVD. Initial laboratory values were remarkable for elevated creatinine to 2.31, elevated AST/ALT to 585/678 respectively, as well as an elevated troponin of 2.1 with CK/CKMB elevated to 7857/49.3. Initial EKG showed TWI in leads V2-V4 with no old EKG available for comparison. Initial CXR was consistent with a multifocal pneumonia. Bedside TTE, done in the ED, showed left ventricular dysfunction with preserved ejection fraction (EF), as well as right ventricular dilation with possible strain. Concern for a PE led to a ventilation/perfusion (V/Q) scan, the result of which was low to very-low probability for PE. The patient was started on treatment for community acquired pneumonia as well as aspirin, statin and heparin drip for possible NSTEMI. The patient was admitted to telemetry for monitoring. While on the telemetry unit, the patient’s troponin continued to rise, peaking at 9.8. The patient was subsequently transferred to the cardiac care unit. An official TTE was performed which showed normal left ventricular size and function with an EF of 65 %. However, there was right ventricular hypokinesis and dilation. The patient underwent a left heart catheterization, which showed no evidence of obstructive disease. During the course of stay, the patient’s creatinine improved and a CT Angiogram of the chest was performed, which showed a large segmental pulmonary embolism. The patient was continued on a heparin drip while being bridged to warfarin. He completed his course of antibiotics for community acquired pneumonia and was discharged on warfarin to be followed as an outpatient. DISCUSSION: The current gold standard for diagnosis of PE remains pulmonary angiography. In more recent years, studies, including PIOPED I/II, have shown both CT angiography as well as V/Q scan in combination with the use of clinic decision making tools, such as modified Well’s criteria, as both sensitive and specific. IThe initial modified Well’s score of this patient, 1.5, put the patient into a low risk category. The V/Q scan was very low probability for PE. When using the information put forth by the PIOPED and PIOPED II study, this patient still had a 4 % chance of having a PE, despite the negative V/Q scan results. It is critically important to not abandon the diagnosis of PE, despite having a near normal V/Q scan, in the setting of additional clinical information which may indicate the need for further imaging to diagnose PE.

THE FAçADE OF FAMILIARITY—A MISSED CASE OF MYXEDEMA COMA Laura Harrison; Leena Rahmat. Baystate Medical Center, Springfield, MA. (Tracking ID #1939256) LEARNING OBJECTIVE 1: Diagnose myxedema coma, and recognize it as a medical emergency. LEARNING OBJECTIVE 2: Recognize difficulties created by experiential thinking when diagnosing a rare condition in the geriatric patient. CASE: An 87 year-old male with a past medical history of advanced dementia, coronary artery disease, and hypertension was admitted from a long-term care (LTC) facility with unresponsiveness in the setting of subacute on chronic cognitive decline. On arrival to the hospital he was hypothermic with a temperature of 96.2 °F and bradycardic with a heart rate of 22. Physical examination was remarkable for lethargy, dry mucus membranes, and Glasgow coma scale (GCS) of 7. Labs were remarkable for serum sodium of 166 and serum creatinine of 3.5. Despite hydration and objective improvements in serum sodium level and renal function, clinically he remained unchanged. A thyroid stimulating hormone (TSH) level was checked on hospital day number 7, which returned at 219. The previous TSH level was normal. Intravenous thyroxine was started and the patient’s GCS increased from nine to 12. He was discharged to the LTC facility. Because of significant improvement in cognitive function, comfort goals were instituted and the patient expired 19 days after discharge. DISCUSSION: Myxedema coma as a diagnosis makes this case unique. As a rare medical emergency, this case provides beneficial teaching points. Such uniqueness creates an environment where experiential thinking perpetuates a poor clinical outcome. This geriatric patient’s multiple, and common, co-morbidities further illustrate the downfalls of experiential thinking when clinicians face unique diagnoses in a façade of familiarity.

THE FEBRILE TRAVELER TO INDIA: A CASE OF TYPHOID FEVER Lucinda B. Leung. UCLA Medical Center, Los Angeles, CA. (Tracking ID #1900811)

LEARNING OBJECTIVE 1: Recognize immigrants returning home to visit friends & relatives (VFRs) as a population vulnerable to typhoid fever and target them for immunization. LEARNING OBJECTIVE 2: Manage typhoid fever when antibiotic-resistant strains of salmonella typhi are involved. CASE: 41-year-old healthy male presents with fever and malaise, following travel to India one-month prior. The patient emigrated from India as a child, frequently visiting without routine vaccinations. One week after returning, he developed subjective fever, malaise, fatigue and non-bloody diarrhea. He denied abdominal pain, upper respiratory symptoms, cough, headache, palpitations, rash, and sick contacts. His diarrhea had improved but he continued to feel weak and febrile. On admission, he was febrile (Tm 40s), tachycardic (max HR 100 s), normotensive, and hypovolemic on exam. Labs showed mild leukocytosis, hyponatremia and transaminitis. Viral serologies, urine cultures, stool cultures, ova and parasite, hepatitis panel, malaria prep, dengue antibody, HIV, MTB quantiferon gold were negative. Blood cultures grew Salmonella typhii, resistant to cipro. His antibiotic was narrowed to ceftriaxone for a total 2-week course. Three weeks after antibiotic completion, his fevers returned and his blood cultures again grew Salmonella. Liver ultrasound did not reveal any abscess. Echocardiogram did not reveal any valvular vegetation. Tagged WBC scan did not reveal a metastatic source of relapse. He was treated with another 4-weeks of ceftriaxone, then transitioned to 2-additional weeks of azithromycin. He remains afebrile and well to date. DISCUSSION: Typhoid fever should be suspected in a febrile returning traveler, along with malaria, rickettsial diseases, hepatitis, leptospirosis and dengue. There are 200 to 300 cases of S. typhi annually in the United States. However, most cases occur in endemic regions, where there is overcrowding and poor sanitation (e.g., India, Southeast Asia). Typhoid fever is acquired by direct or indirect fecal-oral spread. Since they have less control over their diet, travelers who visit friends and relatives are at increased risk for enteric fever. Typhoid fever is caused by S.typhi, following an incubation period of 5 to 21 days. The serotype S. paratyphi also causes enteric fever. Humans are the only reservoir for Salmonella. Patients present with sustained fever, anorexia, malaise, and vague abdominal discomfort. Diarrhea often appears early in the course and resolves before fever. Physical exam may reveal pulse-temperature dissociation, hepatosplenomegaly or rose spots. Complications include intestinal perforation, encephalopathy, relapse after clinical cure, and asymptomatic chronic carriage. Laboratory findings are nonspecific, such leukocytosis, anemia or transaminitis. S. typhi may be isolated in blood, stool, urine, bone marrow and duodenal aspirate cultures. Typhoid fever occurs in unimmunized travelers and 30 % of immunized travelers. Multidrug-resistant Salmonella strains are usually found in travelers to the Indian subcontinent. Due to increasing resistance, there has been a shift away from chloramphenicol, ampicillin, and trimethoprim-sulfamethoxazole. For susceptible strains, fluoroquinolones appear to be most effective and used empirically most often. Ceftriaxone (and oral cefixime) and azithromycin are also used for empiric therapy.

THE GREAT MASQUERADER, A BYGONE ACQUAINTANCE OR A PRESENT-DAY FOE? Maria Fan. University of California, San Francisco, San Francisco, CA. (Tracking ID #1937836) LEARNING OBJECTIVE 1: Perform evidence-based lab and imaging assessment of cognitive impairment and dementia LEARNING OBJECTIVE 2: Interpret RPR titers to guide clinical suspicion for neurosyphilis CASE: An 82 year-old African American woman with a history of HTN, hypothyroidism, smoking, and chronic daily alcohol use was seen in clinic for follow-up of cognitive impairment, including difficulty recalling prior clinic conversations with her physician. The patient reported some general “forgetfulness” but declined to provide specifics. She also reported that her father had had syphilis so that she was treated for syphilis at age 9 with penicillin. Her examination was notable for Montreal Cognitive Assessment (MoCA) scores of 17/30 and 16/30 on two separate occasions with a PHQ-9 score of 3/27 suggesting minimal depressive symptoms. The remainder of her physical exam was unremarkable, without any focal neurologic deficits, rigidity, tremor, or gait abnormalities. Laboratory testing revealed normal TSH, B12, and thiamine levels. HIV was negative. RPR testing returned reactive with a titer of reactive undiluted, confirmed with reactive treponemal antibody. After discussion with Infectious Disease, a plan was made to treat the patient for late latent syphilis, with no plan for lumbar puncture to examine for neurosyphilis given the rarity of neurosyphilis in the era of antibiotics, low serum RPR titer, and patient preference. DISCUSSION: The estimated prevalence of dementia in the United States is 6.5 % for those aged 60 years or greater, with Alzheimer disease representing 60–80 % of cases and vascular dementia 10–20 %. The American Academy of Neurology (AAN) has published guidelines for the laboratory and imaging evaluation of cognitive impairment and

ABSTRACTS dementia, but what is done in clinical practice remains variable. The current AAN recommendation includes screening for B12 deficiency, hypothyroidism, and structural neuroimaging with either non-contrast CT or MRI. This is in contrast to the prior AAN 1994 practice guidelines that recommended examination of CBC, electrolytes, glucose, BUN/creatinine, folate, B12, thyroid function, and syphilis serology. In 2011, there were 18,576 cases of late and late latent syphilis in the United States (case rate of 6 per 100,000), with Los Angeles County, CA, and Cook County, IL, accounting for the most cases. Given the decline of neurosyphilis cases in the antibiotic era, the AAN now recommends not screening for syphilis unless the patient has a specific risk factor, evidence of prior syphilitic infection, or resides in a high prevalence area. Importantly, studies have found that a serum RPR titer of ≥1:32 highly predicts presence of neurosyphilis, defined as a positive CSF VDRL or CSF pleocytosis, regardless of HIV status or prior syphilis treatment history. A serum RPR titer of ≥1:32 has been suggested as a threshold to consider lumbar puncture to examine for neurosyphilis. In summary, this case reviews the laboratory and imaging evaluation of cognitive impairment and dementia in addition to illustrating the complexity of syphilis serology in considering neurosyphilis.

THE INCREDIBLE FUSS OVER ATYPICAL HUS Anita Bhagavath; David Chun; Jason Filopei; Gina Elhammady; Andrew Bohmart. Beth Israel Medical Center, New York, NY. (Tracking ID #1928693) LEARNING OBJECTIVE 1: Recognize the clinical features of atypical hemolytic uremic syndrome. LEARNING OBJECTIVE 2: Recognize the differences between typical HUS and atypical HUS. CASE: 60yo F with PMHx of Hepatitis C presented with shortness of breath for 2 days, associated with chest tightness and lightheadedness. Further questioning revealed she had 2 days of bloody stools, hematuria, and minor nose bleeds. Physical examination was remarkable for dried blood in the nares, pale conjunctiva, and a flow murmur. Rectal exam showed brown stool. Admission labs revealed acute on chronic kidney disease, new normocytic anemia, and mild thrombocytopenia (Creat 4.75 g/ dL, baseline 1.55, Hgb 7.8 g/dL, baseline 11.9, Plt count 130 K/uL, baseline 237, all 1 month prior). Stool sample was occult positive. Urinalysis was yellow, with large blood (>182 RBCs). Anemia work up showed schistocytes on peripheral smear, lactate dehydrogenase (LDH) 999, haptoglobin 16, and normal bilirubins. C3 was elevated (250 mg/dL) and C4 was low (3 mg/dL). The patient underwent emergent plasmapheresis and hemodialysis, with marked improvement in clinical status and laboratory abnormalities (Hgb 10.8, LDH 498, Plt 206, Creat 2.98). Stool cultures were negative and ADAMTS13 activity was normal, ruling out toxin-mediated hemolytic uremic syndrome and thrombotic thrombocytopenic purpura, respectively. Renal biopsy findings were consistent with HUS and a final diagnosis of atypical HUS (aHUS) was made. She was discharged with a plan to continue hemodialysis and begin outpatient ecluzimab therapy. DISCUSSION: HUS is characterized by the triad of hemolytic anemia, severe thrombocytopenia, and renal impairment. Commonly this is secondary to Shiga toxinproducing enteric infections and presents as bloody diarrhea. In contrast, aHUS is both toxin and diarrhea negative, and is secondary to dysregulation of the alternate complement pathway. Our case of aHUS was unique because it presented with bloody diarrhea, only mild thrombocytopenia, and elevated rather than decreased C3 levels. These atypical presentations can result in delayed diagnosis and worsened outcomes. Plasmapheresis is the first-line therapy for aHUS and early initiation leads to better outcomes. Response is monitored by normalization of platelets, hemoglobin and LDH. Recently, a better understanding of the pathophysiology behind aHUS has led to the emergence of ecluzimab, a humanized anti-C5 monoclonal antibody. This drug inhibits the critical steps in the alternative complement pathway and is currently being evaluated in clinical trials. This new treatment option is changing the therapeutic approach to aHUS, One should recognize that aHUS can present with or without bloody diarrhea and the hallmark laboratory data may not always follow the classic pattern. This is important as early intervention with plasmapheresis and more cutting edge therapy such as ecluzimab may be critical to improved outcomes.

THE INDELIBLE DIRECTIVE: A CASE SERIES REPORT OF PATIENTS WITH DNR TATTOOS Yevgeniy Mikityanskiy; Sagar Patil; Paul Leis; Jonah Feldman. Winthrop University Hospital, Mineola, NY. (Tracking ID #1937593) LEARNING OBJECTIVE 1: To help guide treatment decisions for patients with DNR tattoos who may not be able to communicate their wishes. CASE: Tattooing as a form of artistic expression remains as popular as ever, and some people are now choosing their ink based upon their attitudes towards health care. A 69-year-old widowed man with “Do Not Resuscitate” tattooed on his left

JGIM

forearm and past medical history of lung cancer presented to our emergency room with shortness of breath and fever. He was found to be in acute respiratory failure from pneumonia. Though he had difficulty speaking, he was able to communicate that he wished full resuscitation efforts to be made, including intubation, if needed. Mechanical ventilation as well as appropriate treatment was initiated, which led to a complete recovery. He explained that the tattoo represented his frustration with the health care system after losing his wife to cancer several years prior, but his attitude about health care changed after establishing a rewarding doctor-patient relationship with a new primary care physician. He was healthy when the tattoo was inked, and he never intended for it to be taken as an advance directive. DISCUSSION: To help guide treatment decisions for future patients who may not be able to communicate their wishes, we searched Pubmed and Google scholar for articles containing the phrase “do not resuscitate” and the word “tattoo.” Review of these articles revealed four previously published case reports, which formed a series of five cases when added to our own. Based on these reports, we identified three reasons why a DNR tattoo should not be actionable in clinical practice. 1) A tattoo inked years ago may not be an accurate reflection of a patient’s current wishes. 2) DNR tattoos are ambiguous and are often intended to make a public statement unrelated to decisions about resuscitation. 3) DNR tattoos are not legally binding. While the third reason reflects a technical concern, the first two give insight into the true feelings of patients who get this type of tattoo. In our case series, patients with tattoos had a change in preference regarding resuscitation status in 2/5 cases. The meaning behind the tattoo did not coincide with specific decisions about resuscitation in 3/5 cases. In total, we found that tattoos did not accurately reflect patient resuscitation preferences in 4/5 cases. From this first known case series of patients with DNR tattoos, we conclude that clinicians should strongly consider resuscitating patients unless they have appropriate documentation. Tattoos, however, should be used as a starting point for conversations regarding preferences for care. In our case, this discussion led to a narrative that enriched the relationship between the patient and his healthcare providers.

THE MYSTERIOUS GREEN URINE! Venu M. Ganipisetti; Pratyusha Bollimunta; Sathish kumar Krishnan; Malav P. Parikh; Harvey Friedman. presence saint francis, Evanston, IL. (Tracking ID #1923585) LEARNING OBJECTIVE 1: 1) Recognize that Propofol administration can be associated with green urine in patients. LEARNING OBJECTIVE 2: 2) Recognize that this finding is benign and does not cause harm to the patient. Discontinuing the medication will normalize the urine color in few hours and will establish the diagnosis. CASE: A 44-year-old African-American male with no known past medical history, presented to our Emergency department with blunt head trauma resulting in intracranial hemorrhage. Upon arrival, he was agitated, minimally responding to verbal stimuli, and was intubated emergently in the ER and later transferred to the ICU for close observation. He was started on a Propofol and Fentanyl infusions for sedation and analgesia. Sedation vacation was tried each day during the hospital course, however the patient became agitated, restless, and hypertensive upon weaning. He continued to require propofol and fentanyl at high doses. On day 10 of admission, we observed that his urine in the foley bag was green. At that time, his labs were normal except for elevated alkaline phosphatase 177 and serum albumin 2.9. His medications at this time were Lisinopril, Amlodipine, Unasyn (for facial injuries), and Dilantin. Cultures were negative for bacterial growth. Ultimately, Propofol was stopped and patient was switched to an Ativan drip. We monitored the urine color, which changed back to yellow within 6 h. No further work up was done. DISCUSSION: Propofol is a commonly used sedative in critical care settings. Green urine color with propofol is a benign and rarely reported finding. To date, only 13 cases are reported in the literature. It has been proposed that green color occurs due to the presence of metabolites in urine. Propofol is metabolized in the liver by glucoronide conjugation and excreted in urine predominantly as the 1-glucuronide, 4glucuronide, and 4-sulfate conjugates of 2,6-diisopropyl-1, 4 quinol which imparts green color to urine. Urine alkalinization increases the formation of these metabolites. A second proposed potential mechanism is entero-hepatic circulation failure due gallbladder hypomotility and extra-hepatic glucuronidation in the kidneys, which might be the cause of green urine in our patient. Other common causes for green urine include pseudomonas infection (Pyocyanin pigment), hartnup disease, indicanuria, ingestion of triamterene, amitriptyline, indomethacin, methocarbamol, promethazine, cimetidine, food coloring, listerine mouthwash, clorets mints, indigo dyes, and methylene blue. All these were excluded in our patient. The temporal relationship of normalization of urine color upon discontinuing propofol indicates it is the most likely cause of this finding. Presence of propofol metabolites in urine is benign and does not affect kidney function or cause harm to the patient. Although this finding is benign, it may cause unnecessary apprehension in Physicians and family members.

JGIM

ABSTRACTS

Notably, in an informal survey of over 30 Anesthesiologists, only one was aware that propofol could cause green urine. This case illustrates that an awareness of this benign finding is important and may prevent unnecessary testing. We were able to avoid all unnecessary further workup in this patient while providing reassurance to the patient’s family members.

THE OUTPATIENT MANAGEMENT OF KETOSIS-PRONE DIABETES AT BELLEVUE HOSPITAL Arnab K. Ghosh; Jamie Osman; Natalie K. Levy. NYU, New York, NY. (Tracking ID #1920760) LEARNING OBJECTIVE 1: 1) Recognize the epidemiology and classification of Flatbush/Ketosis-Prone Diabetes Mellitus and its common presentation as diabetic ketoacidosis. LEARNING OBJECTIVE 2: 2) Recognize the need to carefully monitor glucose levels in ketosis-prone diabetic patients, who are at risk for hypoglycemia to prevent hypoglycemia. CASE: A 53YO Mexican previously well male was admitted with a week-long history of polyuria, polydipsia, dry mouth, and a 30 lb weight loss over the past 2 weeks. He denied any personal or family history of diabetes mellitus. On examination he was clinically dehydrated. His labs were significant for K 5.5 mmol/L, HCO3 21 mmol/L, chloride 88 mmol/L glucose 1190 mg/dL, anion gap of 22, urinalysis showing ketones and glucose. Venous blood gas pH was 7.284. The patient was diagnosed with diabetic ketoacidosis and treated with intravenous fluids, potassium supplementation and insulin therapy. His symptoms resolved without incident. No precipitating factor was found. He was discharged home on glargine 10 units at night and metformin 1000 mg twice a day. The patient was seen in clinic. His glutamic acid decarboxylase (GAD) antibody and islet cell antibody were negative, fasting C-peptide was 2.23 ng/mL (N: 0.8–3.1 ng/mL) and HbA1c 15.8 %. He was only taking 6 units of glargine for unclear reasons. Based on his home blood glucose readings, his glargine dose was increased to 20 units nightly over these first two visits. For 2 subsequent visits over the next 5 weeks this insulin dose, plus metformin, led to controlled fasting blood glucose. The patient was lost to follow up for the next 2 months. When he returned to clinic 4 months after his initial discharge, he had ceased taking his insulin and metformin, owing to symptoms of hypoglycemia. Repeat HbA1c was 7.1 %. Because his fingerstick glucose remained elevated, the patient was restarted on metformin. DISCUSSION: This case represents a case of a patient with Ketosis-Prone Diabetes (KPD) initially presenting with DKA. In the literature, KPD has been noted in the Hispanic and African-American populations, the diagnosis occurring in the fourth or fifth decade of life. Most patients diagnosed with KPD tend to be obese, and 80 % have a family history of type 2 diabetes mellitus. To place a KPD patient into the AB classification system, they need to be tested for both autoantibodies and beta cell reserve. Our patient was GAD and Islet Ab negative. His fasting C-Peptide level was normal, placing him in the A-B+ category, the most common category for new diabetics presenting in DKA. Patients with a new diagnosis of DM in DKA have a varied post discharge course. Some will need lifelong insulin. Others will have recovery of beta cell function and be able to come off of their insulin. A classification of A-B+ is associated with the highest likelihood of achieving insulin independence. Smiley et al. recommend KPD patients be followed every 2 weeks for the first 2 months for insulin adjustment. If found to have fasting blood glucose values of 39 °C. After 2.5 weeks, our differential diagnosis included military tuberculosis (his son was being treated for latent TB), an unclear autoimmune process, sarcoidosis, and hemophagocytosis. Anti-TB treatment with moxifloxacin, ethambutol, and rifampin was started but didn’t show improvement after 4 days. Then the anti-TB medications were stopped and IV Solumedrol was started to treat a possible autoimmune process. Once steroids were started his blood counts improved and he began to feel well enough for discharge to inpatient rehab. At his follow up appointment in Rheumatology a month later, his ANA was 1:320 and his repeat ds-DNA was positive. Despite feeling 85 % of his normal self, his protein/creatinine ratio was uptrending, raising concern for developing lupus nephritis. DISCUSSION: This case demonstrates the protean manifestations of SLE in a young, otherwise healthy male. SLE is a chronic inflammatory disease of unknown etiology that typically presents with a constellation of constitutional, hematologic and serologic findings and can affect almost any organ. Almost all patients experience fatigue, fever and arthralgias. A significant portion also have hematologic findings, including anemia, leukopenia, thrombocytopenia, and pancytopenia. A formal diagnosis of SLE requires the presence of 4 out of 11 criteria: malar rash, discoid rash, photosensitivity, oral ulcers, arthritis, serositis; renal, neurologic and hematologic disorders; and elevated ANA and one other serologic marker (anti-dsDNA, anti-Sm, APLA). Our patient has 3 criteria: pancytopenia, elevated dsDNA, and elevated ANA. His developing lupus nephritis would make it 4 criteria and a definite diagnosis. Approximately half of patients who present with a positive ANA and fewer than 4 criteria eventually accumulate enough criteria for a diagnosis of lupus. It is interesting to note that although dsDNA only has a 75 % sensitivity for SLE, it is a risk factor for developing lupus nephritis. Other risk factors include being young, male, black, and having positive anti-Smith and lupus anticoagulant. Serologies may be equivocal and can

JGIM

fluctuate throughout the disease course, so it is important to interpret them in the context of the clinical picture, and to recheck them if clinical suspicion of SLE remains high. SLE can be difficult to diagnose due to its variable presentation. It’s important to remember that although 84 % of patients with SLE are female, there are 25,000 males with this disease in the US, so keep it on the differential for patients presenting with systemic, nonspecific symptoms. Next, SLE is resistant to type 1 reasoning due to its unpredictable pattern. It is therefore important to apply type II clinical reasoning and to systematically develop a differential. Finally, cases like this require the collaboration of multiple teams who need to practice good communication and documentation. The coordination of care falls to the primary team to ensure sound (and safe) clinical decision making based on synthesized data and opinions.

THE WRONG SIDE OF THE TRACKS Uchechukwu Obih; Beth Susi; Thomas Montgomery. Carolinas Medical Center, Charlotte, NC. (Tracking ID #1926364) LEARNING OBJECTIVE 1: Discuss the importance of obtaining thorough social history. LEARNING OBJECTIVE 2: Identify a noninfectious cause of thrombotic thrombocytopenic purpura (TTP) CASE: A 21 year old man presented with weakness and lethargy of 1 week’s duration. He complained of nausea, vomiting, diffuse abdominal achiness, chills, and headaches. His past medical history was notable for chronic back pain secondary to a horsing accident as a child. Exam showed a heart rate of 100, diffuse abdominal tenderness, and splenomegaly. His sclera, face, chest, back, and arms were jaundiced, with petechiae on his upper back, and a large tattoo on his left forearm and upper back. No focal neurologic deficits were present. Laboratory was notable for white blood cell count of 7.4 g/dL with 20 % bandemia, hemoglobin of 5.5 g/dL, MCV 114, platelets of 13 K/uL. Other pertinent labs included potassium of 2.8 mmol/L, creatinine of 1.59 mg/dL, albumin 2.6 g/dL, total protein 5.1 g/dL, total bilirubin 4.0 mg/dL, alkaline phosphatase of 54 IU/L, ALT 241 IU/L, AST 137 IU/L, and CPK 78 IU/L. Lactate dehydrogenase was elevated at 1,627, haptoglobin was low at 2 and peripheral smear showed 1+ schistocytes. Urine drug screen was negative for Tylenol, salicylates, ethanol, barbituates, marijuana, or opioids. Hepatits C was positive with viral load of 51,000 copies. HIV was negative. He was admitted to the intensive care unit to begin plasma exchange and steroids for concern of thrombotic thrombocytopenic purpura (TTP). His thrombocytopenia and anemia had slow resolution and plateaued by cycle ten of exchange. As his jaundice and petechiae improved, track marks were noted in the left antecubital fossa. The patient revealed that his brother presented to another hospital with similar symptoms and was also diagnosed with TTP. He admitted to injecting extended-release oxymorphone with his brother 1 week prior to the start of his symptoms. Plasma exchange and steroids were discontinued with improvement in anemia and thrombocytopenia. DISCUSSION: Obtaining a thorough social history, including exposure to medications and illicit drugs is important in diagnosing thrombotic thrombocytopenic purpura (TTP). TTP is defined by the pentad of microangiopathic hemolytic anemia determined by the presence of schistocytes, elevated LDH and low haptoglobin, thrombocytopenia, fever, neurologic symptoms, and renal dysfunction. Etiology has been associated with ADAMSTS-13 activity deficiency and with Shiga-like toxin from diarrheal illness caused by enterohemorrhagic Escherichia coli O157:H7. A number of medications, such as cisplatin, bleomycin, and mitomycin have been implicated in the development of TTP in some patients. TTP induced by intravenous injection of extended release oxymorphone is a rare cause, of which there have been a few reported cases to the Centers for Disease Control. Treatment for drug induced TTP has not been established. While plasma exchange has become the mainstay of treatment for TTP secondary to ADAMTS13 deficiency or E. coli, evidence of use in drug induced TTP is not robust and is anectodal. Supportive care is often the treatment of choice.

THERE’S AN UNCOMMON FUNGUS AMONG US Prathyusha Savjani. Tulane University Health Sciences Center, New Orleans, LA. (Tracking ID #1926142) LEARNING OBJECTIVE 1: Understand the importance of obtaining tissue diagnosis in patients with cerebral infections while on immunosuppressive agents to accurately identify the disease. LEARNING OBJECTIVE 2: Recognize how the differential diagnosis for infection changes based on the immunosuppressive agent being used CASE: A 62 year-old African American man with a history of cadaveric kidney transplant presented with sudden onset of right arm weakness. MRI of the brain was obtained due to concern for stroke. On imaging, it was discovered that he had

JGIM

ABSTRACTS

multiple peripherally enhancing lesions scattered throughout the cerebral parenchyma. The differential for such findings included metastatic disease, primary malignancy, infectious etiologies, and demyelinating processes. Cerebrospinal fluid analysis was normal as was serum testing for infectious etiologies of cerebral abscesses. He then underwent craniotomy with biopsy with preliminary fungal cultures showing light mold. Pathologic evaluation of the brain sample demonstrated acute granulomatous cerebritis containing pigmented hyphal forms consistent with dematiaceous fungi. Final fungal culture results yielded Fonsecaea pedrosoi. Investigation to identify a primary source of entrance for this organism included thorough investigation of the skin and sinuses. The only skin lesion noted was a chronic non-healing ulcer on his back that was previously biopsied and determined to be radiation dermatitis as a result of fluoroscopic procedures. Dermatology was consulted and did not believe that this skin finding was consistent with a fungal infection. The sinuses were patent with no thickening of any mucosal tissue on both CT and MRI of the head and thus sinus disease was ruled out. Treatment options were discussed including Amphotericin B, Itraconazole, Voriconazole, as well as surgical debridement of the abscesses. The patient chose not to undergo any further invasive procedures, and treatment was initiated with Amphotericin B and Voriconazole and the transitioned to Itraconazole at discharge. He was followed-up in the outpatient setting by the infectious disease team with a repeat MRI of the brain several weeks after treatment. It showed stable lesions that had neither grown nor shrunk in size. The patient’s symptoms remained stable with no improvement but also with no deterioration of neurologic deficits. DISCUSSION: Chromoblastomycosis of the central nervous system by F.pedrosoi is a very rare disease throughout the world and even more rare in the United States. This fungal organism is generally found in rotten wood and soil in humid regions. Most reported incidences of this infection involve skin and subcutaneous infection from direct inoculation, which can occur in both immunocompetent and immunocompromised hosts. Most of the previous cases come from Asia, South America and Africa where the environment is friendlier to this organism. Cases of brain abscesses with this organism have been reported in patients with an altered immune system and most had a primary cutaneous site of inoculation. The mechanism by which this organism reaches the central nervous system is not known. This patient was on Tacrolimus, which by inhibiting T-cell activation, likely predisposed this patient to developing this fungal infection. One should build a differential diagnosis based not only on imaging, previous cases, and patient’s exposures, but also on the immunologic effects of the immunosuppressant being used. T cells are the cornerstones of immunologic response to fungal infections, and suppression of T cell function by medications such as Tacrolimus may predispose to fungal infections consisting of very rare organisms.

THESE DEVICES CAN BREAK SKIN!!! Samian Sulaiman; Muhammed Sherid; Salih Samo; Shahriar Dadkhah; Nadia El Hangouche; Muhammad Shahreyar; Addis Asfaw; Ahmet A. Oktay; Ana Inashvili; Husein Husein; Preethi Subramanian. St. Francis Hospital. Evanston, evanston, IL. (Tracking ID #1938782) LEARNING OBJECTIVE 1: Despite improvements in the design and implantation techniques of permanent pacemakers, infection and skin erosion of the pacemaker pocket remain a serious problem. Continuous or recurrent pain at the pocket site should raise suspicion of infection. It is crucial to identify early signs of pocket infection before these devices breaks skin. Regular pacemaker check-up helps avoid late complications of pacemakers CASE: − A 77 y/o Male who presented with redness and discomfort over his pacemaker pocket in the left upper chest. The pacemaker was implanted 2 years ago. The current complaint started 4 weeks ago. Despite outpatient oral antibiotics, his erythema progressively worsened and developed into a superficial ulcer with draining pus. Patient was admitted to our hospital to receive IV antibiotics. - On physical exam, Tenderness and erythema were noticed at pacemaker site with a superficial small ulcer. Patient was afebrile and rest of vital signs were normal except for irregular heart beats. - His past medical history was significant for atrial fibrillation, Diabetes mellitus, hypothyroidism and Hyperlipidemia. His medications included Coumadin, levothyroxine, atorvastatin, Cardizem, metoprolol and metformin - Patient was started on IV vancomycin and Zosyn. Blood and wound cultures were neg. Transthoracic Echocardiography did not show any vegetation. CT chest (see Figure 3) showed Skin and soft tissue thickening superficial to the pacemaker device but no abscess was present. His erythema and drainage resolved and he was discharged home on oral augmentin on sixth day of admission. Ten days after discharge, he presented again with a skin break at pacemaker’s pocket. Patient reported episodes of local irritation to which he responded by rubbing his skin. On physical exam, part of pacemaker hardware was exposed (see Fig 1). Cardiothoracic surgeon replaced the pacemaker and extracted intra-cardiac leads. Pacemaker lead tip culture grew Staph Epidermis susceptible to multiple antibiotics including vancomycin. Patient received extended course of antibiotics for 4 weeks.

DISCUSSION: The overall incidence of pacemaker-related infections is 1.9 per 1000 device-years. Several factors are associated with an increased risk of Permanent pacemaker infections, including diabetes mellitus, malignancy, operator inexperience, advanced age, corticosteroid use, anticoagulation, recent device manipulation, chronic renal failure, and bacteremia from a distant focus of infection. Early post-implantation infections are most commonly caused by Staphylococcus aureus and late infections most commonly by Staphylococcus Epidermis. There is a general consensus that once there is deep pacemaker pocket or lead infection, removal of the whole pacemaker system followed by an extended course of appropriate antibiotics results in the best prospect for long term eardiaction of infection. If pacemaker erodes the overlying skin, it is considered contaminated and current opinion favors replacement of the generator and leads. The mortality of persistent infection when infected leads are not removed can be as high as 66 %.

THIS IS MORE THAN A SIMPLE AKI! Hani Snounou. Saint Francis Hospital, Evanston, IL. (Tracking ID #1934404) LEARNING OBJECTIVE 1: To recognize the reciprocal relationship between the uric acid level and acute kidney injury. CASE: A 62 year old male presented with melenic stool, generalized weakness, low Urine output for 1 week duration. He is a reformed alcoholic with underlying cirrhosis but he admits returning to drink 2 weeks before the admission due to social stress. Otherwise he has a past medical history of Diabetes mellitus, Hypertension, diastolic CHF and COPD. Home medications were metformin, glimepiride, lisinopril, Lasix, advair and albuterol PRN. Physical examination at the time of admission was remarkable for obesity with BMI of 44, jaundice and mild tenderness in the right upper quadrant and epigastric area, the rectal exam with melenic stool. His initial labs showed creatinine of 7.2 mg/dL (with a historical creatinine of 0.9 1 year ago), BUN of 98 mg/dL, NA of 130 mmol/L, chloride of 84 mmol/L, hemoglobin of 12.1 g/dL, PLT of 129 K/mL, normal INR/PT, AST of 311 IU/L, ALT of 194 IU/L, total bilirubin of 4.9 mg/dl with direct bilirubin of 3 mg/dL, alkaline phosphatase of 163 IU/L. Hospital Course: The initial AKI workup showed a CK of 424 IU/L, Uric acid of 23.2 mg/dL, FENA of 1.24 %, urine PH of 5 with +30 protein without casts seen; CT abdomen without contrast showed fatty liver, normal kidneys texture without hydronephrosis or ascites. The patient underwent EGD and colonoscopy and no active bleeding was found. He was placed on half saline with 75 Meq of sodium bicarbonate at 100 cm3/h and was given one dose of rasburicase 5 Mg. His creatinine and the Urine output improved during the admission as well as the uric acid level as following: Creatinine trend: 7.2———>6.6————>5———>4.0 Uric acid trend: 23.2——>22——>20———>16 He was discharged and asked to follow up with his PCP when he had a repeat blood workup after 2 weeks showing normalization of the creatinine to 0.9 mg/dL and a uric acid of 7.8 mg/dL. DISCUSSION: Acute uric acid nephropathy or UAN is a well known syndrome most frequently encountered in the course of rapid cell turnover such as leukemias and lymphomas, and the added acceleration of cell lysis that occurs with chemotherapy and radiation. But this pathology can be seen even with no malignancy history as in our case. The uric acid elevation can be the reason or the consequence of the AKI but a level above 12 is rarely to be due to the AKI by itself. High tubular fluid flow diuresis achieved by hydration is the primary mechanism of both protection and treating acute uric acid nephropathy. Urine alkalinization is controversial since it may sometimes accelerate phosphate precipitation in the kidneys and thereby worsen the renal failure, but this technique still commonly used. Using rasburicase helps to bring the uric acid level faster and can be considered as treatment option and a good outcome can be seen even with a single dose as in our case. Finally hemodialysis therapy may be required both to correct azotemia and to reduce the body burden of uric acid.

THROW ME A BONE: A CASE OF LANGERHANS CELL HISTIOCYTOSIS PRESENTING AS RIB PAIN IN A 35-YEAR-OLD WOMAN Julia P. Brockway1; Elena Katz2; Sudha Dubey1. 1Beth Israel Medical Center, New York, NY; 2Beth Israel Medical Center, New York, NY. (Tracking ID #1936915) LEARNING OBJECTIVE 1: Recognize Langerhans cell histiocytosis as an uncommon etiology of bone pain and lytic bone lesions in adults. LEARNING OBJECTIVE 2: Learn how to diagnose Langerhans cell histiocytosis and assess for multi-organ involvement. CASE: A 35-year-old African American woman presented with intermittent right breast pain for 2 months. Three months prior she had presented to an outside clinic with a right breast lump. Ultrasound of the breast at that time showed two breast masses and a suspicious enhancement at the right fifth rib. The breast masses proved to be fibroadenomas on core biopsy. She was then referred for CT scan of the chest,

ABSTRACTS after which she transferred her care to our institution. She now had right breast pain and a throbbing rib pain inferior to the breast. She also noted an unintentional 7 lb weight loss over 2 months. She denied fevers, night sweats, shortness of breath, or cough. Her right breast was tender to palpation in the lower inner quadrant and had a palpable, non-tender mass at 10 o’clock. There was no cervical, supraclavicular, or axillary lymphadenopathy. Lung exam was unremarkable. CBC showed mild neutropenia with an ANC of 1400/uL, hemoglobin of 12.5 g/dL, and platelets of 180 K/uL. Liver and renal function tests were normal, and calcium level was 8.7 mg/ dL. Serum and urine protein electrophoresis showed no abnormal bands. CT scan of the chest revealed lytic lesions in the fifth and sixth ribs with an associated expansile soft tissue mass at the fifth rib. Biopsy of the rib mass was performed and pathology revealed eosinophilic granuloma, with no tumor seen. Immunohistochemistry stained positive for CD1a, S100, and CD207, consistent with a diagnosis of Langerhans cell histiocytosis (LCH). PET CT scan showed additional suspicious foci at the left hip and left maxilla, and interval development of numerous thin-walled cysts in the midto upper- lungs suggestive of pulmonary eosinophilic granulomatous disease. Bone marrow biopsy showed normocellular marrow. The patient was diagnosed with multisystem LCH without “risk organ” involvement. She was referred to hematology/ oncology to explore treatment options, and her care was ultimately transferred to an outside cancer center to participate in a clinical trial. DISCUSSION: LCH is an uncommon cause of lytic bone lesions in adults. The estimated incidence is one to two cases per million adults; however, the true incidence is not known, and LCH is thought to be under diagnosed. The differential diagnosis for lytic skeletal lesions, depending on a patient’s age, includes primary bone or soft tissue tumor, multiple myeloma, and metastatic disease. Definitive diagnosis can therefore only be made by biopsy of a suspicious lesion. LCH is a disorder of histiocytes that originates from the myeloid dendritic cell, a precursor of the Langerhans cells found in skin and mucosa. Pathophysiology is ambiguous, though evidence suggests a neoplastic process, a reactive process, or a combination of both. Histiocytes infiltrate various organs, including bones, skin, lymph nodes, the liver, the spleen, the lungs, and the central nervous system. While the most common presentation is pain due to lytic bone lesions, patients may also present with dyspnea due to lung involvement, rash, gingival hypertrophy, weight loss, fever, lymphadenopathy, splenomegaly, or polydipsia, polyuria, ataxia, or memory problems due to CNS involvement. About half of all patients diagnosed with LCH present without systemic symptoms and with involvement of only one organ, such as the lungs in primary pulmonary LCH. The rest are considered to have multi-system LCH, involving two or more organ systems with or without involvement of the “risk organs” (bone marrow, spleen, and liver) which portend a worse prognosis. Classification of a patient’s disease is critical, as it determines the specific type of treatment needed. All patients should be evaluated after initial diagnosis with a thorough physical exam, CBC, renal and liver function studies, skeletal survey, chest x-ray, and PET CT scan. Patients with multi-organ involvement or cytopenias must undergo bone marrow biopsy. Treatment of LCH depends on patient age and whether it is single or multi-system disease with or without “risk organ” involvement. For adults with single site disease, management ranges from watchful waiting to local methods such as curettage, topical steroids, or radiation therapy. For patients with multisystem disease, data is limited, and patients are encouraged to participate in clinical trials. The mainstay of treatment is chemotherapy. Prospective trials have shown that the majority of these patients respond to initial treatment; however, patients must be closely followed for disease recurrence after completing therapy. Further research is needed in order to better understand LCH, its pathophysiology, and its treatment. It is important for clinicians to be aware of this rare but potentially treatable disease and the various ways that it can present.

THYROID-INDUCED ATRIAL FIBRILLATION Farah Kaiksow; Philip J. Putnam. Tulane University Health Sciences Center, New Orleans, LA. (Tracking ID #1925927) LEARNING OBJECTIVE 1: Diagnose thyroid disease as a cause of new-onset atrial fibrillation. LEARNING OBJECTIVE 2: Recognize that thyroid disease workup can be limited by amiodarone administration. CASE: A 65-year-old man presented with sudden-onset shortness of breath for 1 day. He was at home watching television when the dyspnea started. After a few minutes he began to have blurry vision and lightheadedness. The patient denied any associated chest pain or tightness. Past medical history was significant for hypertension and Type 2 diabetes, both well controlled. Review of systems was positive for dyspnea on exertion over the preceding 2 weeks that was associated with symptoms of a mild upper respiratory tract infection including rhinorrhea and cough. The patient’s heart rate was 136 and electrocardiogram revealed atrial fibrillation with rapid ventricular

JGIM

response. Vital signs were otherwise within normal limits. The patient had mild crackles at both lung bases and an otherwise normal exam. A diagnosis of mild heart failure secondary to new arrhythmia was made and amiodarone was started to control the heart rate and rhythm. Upon further workup, blood tests revealed an undetectable thyroid stimulating hormone level and an elevated free T4. The patient had previously been euthyroid. The diagnosis was then amended to include hyperthyroid as the underlying cause of the atrial fibrillation. While viral thyroiditis was thought to be the most likely source of the hyperthyroid disease, further workup with radioactive iodine uptake scan was not able to be performed because of the initial administration of amiodarone. DISCUSSION: Atrial fibrillation is the most common arrhythmia diagnosed in the United States and thus a condition in which general internists should be well versed. Approximately 1 % of cases of this arrhythmia can be linked to hyperthyroidism, and about 1.7 % of patients with new hyperthyroid disease have simultaneous atrial fibrillation. Although hyperthyroid is much more common in women than in men, hyperthyroid-induced atrial fibrillation is more than twice as common in men than women. When a patient presents with new-onset atrial fibrillation, the minimum workup that should be performed includes a 12-lead electrocardiogram, chest x-ray, transthoracic echocardiogram, and thyroid functions tests. The goal of this workup is to rule out treatable causes of atrial fibrillation, such as hypertension, heart failure, myocardial ischemia, pulmonary disease, and thyrotoxicosis. In hemodynamically stable patients, there is no need to urgently control the abnormal rhythm. First line therapy is rate control with beta-blockers or calcium channel blockers. In cases where hypotension or heart failure is a significant concern, amiodarone is a reasonable alternative. However, if there is any concern for possible thyroid disease as the etiology of the arrhythmia, amiodarone should be avoided as it can both exacerbate thyroid disease and invalidate results of a radioactive iodine uptake scan. In this case, given the patient’s stable vital signs, absent history or exam findings of significant heart failure, and recent upper respiratory infection symptoms, it would have been reasonable to postpone amiodarone administration until further lab results were available.

THYROTOXIC HYPOKALEMIC PERIODIC PARALYSIS Saifullah Nasir1,2; Jeffrey T. Bates2,1. 1Baylor College of Medicine, Houston, TX; 2Houston VAMC, Houston, TX. (Tracking ID #1939642) LEARNING OBJECTIVE 1: Recognize Thyrotoxic Hypokalemic Periodic Paralysis (THPP) as a potential cause of weakness or paralysis. LEARNING OBJECTIVE 2: Manage the hypokalemia and hyperthyroidism of Thyrotoxic Hypokalemic Periodic Paralysis (THPP). CASE: A 36-year-old Vietnamese female without prior medical history presented with acute onset of several hours of bilateral lower extremity weakness. Over the previous 3 months, she had noticed a 10–15 lb unintentional weight loss, an intermittent throat pain which resolved with pain medications, and an occasional tremor of her hands. There was no trauma, recent illness, vision problems, or dizziness. Physical examination revealed tachycardia to 110, non-tender thyromegaly with a right thyroid lobe larger than the left, and an enlarged isthmus. There were no palpable thyroid nodules, no lymphadenopathy, and no thyroid bruit. There was no lid lag and no periorbital edema or chemosis. The patient’s lower extremity weakness was localized to her proximal thighs, but she was able to stand with some support. Laboratory data showed a potassium of 2.3 mmol/L, TSH of 100/high power field and dysmorphic red blood cells but no proteinuria. Two weeks later, though, she develops bilateral ciliary injection and photophobia; slit lamp exam reveals anterior chamber uveitis. Labs show BUN 26 mg/dL and Cr 2 mg/dL; anti nuclear antibody is 1:320 and IgA 497 mg/dL (normal 90–400 mg/dL). Urinalysis reveals protein 3200 mg/day and β2-MG 1347/μL (normal 0–250/μL). Kidney biopsy shows 1 cellular crescent, increased mesangial cell proliferation, and deposition of IgA and C3 in the glomeruli, consistent with IgAN. The interstitium, however, shows edematous changes and infiltration of granular leukocytes. Given her diagnosis of TINU syndrome complicated by IgAN, she is given methylprednisolone ×3 days with subsequent oral prednisolone ×2 months with clinical improvement. DISCUSSION: TINU syndrome, with onset usually in women 120 mg MED) doses before and after implementation of the strategy. FINDINGS TO DATE (IT IS NOT SUFFICIENT TO STATE â€œFINDINGS WILL BE DISCUSSEDâ€□): From 2010 to 2013, three or more training sessions on chronic pain management were held each year for residents. The multidisciplinary committee provided recommendations on 99 cases, an average of 2.4 per month. The number of patients receiving opioids for chronic pain fell 39 % from 591 to 358; the percentage on extremely high doses fell from 10 % (n=60) to 2 % (n=7), and the percentage on high doses fell from 35 % (n=197) to 19 % (n=69). KEY LESSONS FOR DISSEMINATION (WHAT CAN OTHERS TAKE AWAY FOR IMPLEMENTATION TO THEIR PRACTICE OR COMMUNITY?): We created a clinic-wide strategy to promote safer chronic opioid prescribing in a large academic safety-net primary care clinic. Through increased education, communication, and decision support around pain management, we reduced the total number of patients on chronic opioids and the number of patients on high-dose opioids.

A METHODOLOGY USING INTERNET BASED RESOURCES AND WEBSITES TO IDENTIFY COMMUNITY RESOURCES AND SOCIOECONOMIC FACTORS OF PATIENTS ATTRIBUTED TO A GENERAL INTERNAL MEDICINE PRACTICE LIVING WITHIN A SPECIFIC METRO ZIP CODE Scott Joy1,2; Jordan T. Wehe2; Gavriel Roda2; Edward A. Soto2; Cara Campbell2; Aaron B. Galbraith2. 1The Colorado Health Foundation, Denver, CO; 2 University of Colorado-Denver, Denver, CO. (Tracking ID #1936054) STATEMENT OF PROBLEM OR QUESTION (ONE SENTENCE): Providing referrals to community resources is a required element for PCMH certification, but community resources are widely dispersed within a given geographical region covered by a general internal medicine practice, and resources that catalog and illustrate these community services are not readily availble to patients and general internal medicine practices. OBJECTIVES OF PROGRAM/INTERVENTION (NO MORE THAN THREE OBJECTIVES): 1. Identify within the metro Denver area the most common zip code address of patients who are attributed to a particular primary care practice. 2. Explore internet based resources/websites that are available to identify physical community resources (recreational centers, churches, schools and parks) and their location within this zip code. 3. Develop a map of these community resources that can be used by the practice and given to patients who live within the particular zip code. DESCRIPTION OF PROGRAM/INTERVENTION, INCLUDING ORGANIZATIONAL CONTEXT (E.G. INPATIENT VS. OUTPATIENT, PRACTICE OR COMMUNITY CHARACTERISTICS): A query of existing billing data was performed to identify the most common address zip code of patients receiving care within a general internal medicine practice. It was determined that the metro Denver zip code 80205 was the most common address zip code for patients within the practice. A team of 4 undergraduate students in the BA/BS-MD program at the University of Colorado Denver was tasked with searching for internet based resources/websites that would provide detailed data on community resources available in this particular zip code. A comprehensive internet search located the following website (www.city-data.com/zips/80205.html) that allowed for detailed evaluation of physical community resources (churches, recreational centers, schools and parks) within zip code 80205. The website www.maps.google.com was used as an additional resource. The Colorado specific website www.piton.org was identified as a tool to identify relevant demographic information within each neighborhood in the specific zip code, including total population, average household income, % persons in poverty, and crime rate per 1,000 persons. MEASURES OF SUCCESS (DISCUSS QUALITATIVE AND/OR QUANTITATIVE METRICS WHICH WILL BE USED TO EVALUATE PROGRAM/ INTERVENTION): Determine the total number and addresses of physical community resources (churches, schools, recreational centers, parks) available within the targeted zip code. Create a map of the targeted zip code that includes locations of the physical community resources within the targeted zip code. Identify key socioeconomic factors for people living in each of the neighborhoods within the target zip code. FINDINGS TO DATE (IT IS NOT SUFFICIENT TO STATE â€œFINDINGS WILL BE DISCUSSEDâ€□): Our team’s analysis found that zip code 80205 has the following physical community resources: 9 Churches 9 Parks 3 Recreational Centers

JGIM

ABSTRACTS

1 Denver Police Community Resource Center 18 Schools Locations of the churches, parks, recreational centers and schools were placed on an electronic map of zip code 80205 found on the city-data.com website for reference. This document can be printed and given to patients to educate them on community resources available in their neighborhood. Zip code 80205 was found to have 5 distinct neighborhoods with the following socioeconomic demographics. Range of total population in each neighborhood: 3,106 to 12,710, average 5,927 Range of average household income: $35,519 to $44,847, average $40,313 (Metro Denver average is $55,129) Range of % of persons in poverty: 14.29 % to 31.5 %, average 26.1 % (Metro Denver average is 14.3 %) Range of crime rate/1,000 persons: 53.77 to 125.02, average 71.53 (Metro Denver average is 60.19) KEY LESSONS FOR DISSEMINATION (WHAT CAN OTHERS TAKE AWAY FOR IMPLEMENTATION TO THEIR PRACTICE OR COMMUNITY?): The existing internet based websites www.city-data.com and www.maps.google.com can be used by any general internal medicine practice within the United States to identify physical community resources for their patients. A map of resources within the zip code can be created from this data and provided to the practice and to the patients living within the zip code to allow all to understand the community resources available. A general internal medicine practice can also use demographic and socioeconomic factors to target clinical and social resources more effectively. Next steps will be to use the detail provided in this search to reach out and collaborate with community leaders to determine best ways to meet the needs of shared citizens living in these neighborhoods to maximize health outcomes.

A NURSE VISIT PROTOCOL COMBINED WITH POPULATION MANAGEMENT REPORTS IMPROVES QUALITY METRICS IN HYPERTENSION Jason Higdon; Monica Agrawal; Jennifer Zreloff. Emory University School of Medicine, Atlanta, GA. (Tracking ID #1937459) STATEMENT OF PROBLEM OR QUESTION (ONE SENTENCE): As the focus of healthcare shifts towards providing high quality care for chronic medical conditions, we must develop novel ways to provide better care, meet quality metrics, and manage ever-increasing demands in primary care. OBJECTIVES OF PROGRAM/INTERVENTION (NO MORE THAN THREE OBJECTIVES): In a university-based and NCQA recognized Patient-Centered Medical Home (PCMH), we sought to improve the percentage of patients with hypertension who achieved a goal blood pressure of less than 140/90. DESCRIPTION OF PROGRAM/INTERVENTION, INCLUDING ORGANIZATIONAL CONTEXT (E.G. INPATIENT VS. OUTPATIENT, PRACTICE OR COMMUNITY CHARACTERISTICS): Our PCMH pilot practice consists mostly of patients on an employer-sponsored health plan. Using a team-based approach we focused on staff education and empowerment, population management reports, and documentation in our electronic medical record to deliver protocol-driven care to our patients with hypertension. 1. An in-service was done with the entire team including all providers, nurses, and medical assistants to reinforce proper measurement of blood pressure including timing, position, and cuff size. Competencies were observed and scored by the team physicians to ensure buy-in and comfort with a nurse visit protocol. 2. A nurse visit protocol was written for hypertension follow-up visits. This included a standard nursing note for our electronic medical record which includes the patient’s medication list, an assessment of medication compliance/ side effects, current and most recent blood pressure readings, and follow-up of most recent self-management goals. The protocol includes parameters to identify potentially dangerous vital signs. For patients whose blood pressure is not at goal, the nurse automatically makes a 1 month follow-up visit for reassessment. In our patient population this type of nursing visit is offered at no cost to the patient. 3. We developed a monthly hypertension population management report that lists all patients with a diagnosis of hypertension whose last blood pressure anywhere in our system (excluding inpatient and ER visits) was above 140/90. It delineates if their most recent blood pressure reading in our clinic was at goal or not and notes if they have an upcoming appointment. We use this to generate health reminder letters to our patients which are largely delivered through our patient portal. MEASURES OF SUCCESS (DISCUSS QUALITATIVE AND/OR QUANTITATIVE METRICS WHICH WILL BE USED TO EVALUATE PROGRAM/ INTERVENTION): Each month, we track the percentage of all patients in our clinic who have hypertension whose last reading was

Abstracts of the 28th Annual Symposium of The Protein Society, July 27-30, 2014, San Diego, CA, USA.

Abstracts of the 35th Annual Meeting of the American Brachytherapy Society, April 3-5, 2014, San Diego, CA.

Abstracts of the SIR 2014 Annual Scientific Meeting Program Convergence, 22–27 March 2014, San Diego, CA.

Abstracts from the USCAP 103rd Annual Meeting, March 1-7, 2014, San Diego, California.

Abstracts from the 2016 Society of General Internal Medicine Annual Meeting.

Antibody Engineering & Therapeutics, the annual meeting of The Antibody Society December 7-10, 2015, San Diego, CA, USA.

Abstracts of the USCAP 103rd Annual Meeting, March 1-7, 2014, San Diego, California.

Abstracts of the American Academy for Cerebral Palsy and Developmental Medicine 68th Annual Meeting, 10-13 September, 2014, San Diego, California, USA.

Abstracts of the 37th Annual Meeting of the American Society of Primatologists, September 12-15, 2014, Decatur, GA.

Society for Maternal Fetal Medicine 2015 Annual Pregnancy Meeting supplemental episode: A Guide to San Diego.

Abstracts of the British Society of Gastroenterology Annual General Meeting. June 16-19, 2014.

Abstracts from the 1st Annual Meeting of the Scottish Society of Cytomics (SCC) 2014.

Neurosciences-2013: 43(rd) Annual meeting of the Society for Neurosciences (SfN), San Diego.

Abstracts of papers for the Seventeenth Annual Meeting of the Society of Toxicology, San Francisco, California.

Highlights from the 58th meeting of the American Society of Haematology, 1-6 December 2016, San Diego, USA.

Experimental Biology 2014. 26-30 April 2014, San Diego Convention Center, San Diego, CA.

Abstracts of the 56th Annual Meeting of the American Headache Society, June 26-29, 2014, Los Angeles, CA.

Abstracts of the 2014 International Society Developmental Psychobiology Meeting, November 12-14, 2014, Washington, DC, USA.

Best of the 2013 AUA Annual Meeting Part II: More Highlights From the 2013 American Urological Association Annual Meeting, May 4-8, 2013, San Diego, CA.

Report from the 37th san antonio breast cancer symposium, 9-13th december 2014, Texas, USA.

Abstracts of the Annual Meeting of the Danish Pathology Society, 13-15 March 2014, Herlev, Denmark.

Abstracts of the 37th Postgraduate Course and 51st Annual Meeting of the European Society of Paediatric Radiology, June 2-6, 2014, Amsterdam, The Netherlands.

Annual meeting of the World Federation of Neurology--Research Group on Neuroepidemiology. San Diego, California, May 4, 1992. Abstracts.

Abstracts of the 19th Annual Congress of the Belgian Society of Internal Medicine, December 12-13, 2014, Brussels, Belgium.