Other Settings

Expanded HIV Testing and Linkage to Care: Conventional vs. Point-of-Care Testing and Assignment of Patient Notification and Linkage to Care to an HIV Care Program Sara Bares, MDa Rebecca Eavou, MAa Clara Bertozzi-Villa, BSa Michelle Taylor, MSWa Heather Hyland, BSa Rachel McFadden, RNa Sachin Shah, MDa Mai T. Pho, MD, MPHa James Walter, MDa Sameer Badlani, MDa John Schneider, MD, MPHa Nik Prachand, MPHb Nanette Benbow, MASb David Pitrak, MDa

ABSTRACT Objective. The University of Chicago Medicine (UCM) led the Expanded Testing and Linkage to Care (X-TLC) program for disproportionately affected populations on the South Side of Chicago. The X-TLC program aimed to expand routine HIV testing to high-prevalence communities with disproportionately affected populations (i.e., minority men and women, men who have sex with men, and intravenous drug users) according to CDC guidelines at multiple clinical sites. Methods. The X-TLC program used standard blood-based laboratory testing vs. point-of-care rapid testing or rapid laboratory testing with point-of-care results notification. Site coordinators and the linkage-to-care coordinator at UCM oversaw testing, test notification, and linkage to care. Results. From February 1, 2011, through December 31, 2013, the X-TLC program completed 75,345 HIV tests on 67,153 unique patients. Of the total tests, 48,044 (63.8%) were performed on patients who self-identified as African American and 6,606 (8.8%) were performed on patients who self-identified as Hispanic. Of the 67,153 patients tested, 395 (0.6%) tested positive and 176 (0.3%) were previously unaware of their HIV-positive status. Seroprevalence was even higher for EDs, where 127 of 12,957 patients tested positive for HIV (1.0% seroprevalence), than for other patient care sites, including for new diagnoses, where 50 of 12,957 patients tested positive for HIV (0.4% seroprevalence). Of the 176 newly diagnosed patients, 166 of 173 (96.0%) patients who were still alive when testing was complete received their test results, and 148 of the 166 patients who were eligible for care (89.0%) were linked to care. Patients linked to X-TLC physicians did well with respect to the continuum of care: 77 of 123 (62.6%) patients achieved HIV viral load of 200 copies/milliliter. Conclusion. Lead organizations such as UCM were able to assist and oversee HIV screening and linkage to care for HIV patients diagnosed at community sites. HIV screening and linkage to care can be accomplished by incorporating standard testing for HIV into routine medical care.

University of Chicago, Department of Medicine, Chicago, IL

a

Chicago Department of Public Health, Chicago, IL

b

Address correspondence to: David Pitrak, MD, University of Chicago, Department of Medicine, 5841 S. Maryland Ave., MC 5065, Chicago, IL 60637; tel. 773-702-9078; fax 773-702-9078; e-mail . ©2016 Association of Schools and Programs of Public Health

Public Health Reports  /  2016 Supplement 1 / Volume 131

 107

108    Other Settings

Until recently, the incidence of new human immunodeficiency virus (HIV) infections in the United States had remained constant since the early 1990s, with 56,000 people becoming infected each year and the number of people living with HIV/acquired immunodeficiency syndrome (AIDS) continuing to increase.1 To avoid further challenges with treating more people with HIV and higher health-care costs, the White House Office of National AIDS Policy developed the National HIV/AIDS Strategy.2 This strategy provided a roadmap for the Centers for Disease Control and Prevention’s (CDC’s) Expanded HIV Testing for Disproportionately Affected Populations program and the Chicago, Illinois, Enhanced Comprehensive HIV Prevention Plan.3 Since implementation, Chicago has seen a decline in new HIV infections and new AIDS diagnoses.4 Recent progress in expanded routine testing, the dramatic effect of antiretroviral therapy on viral load, HIV treatment as prevention, pre-exposure prophylaxis, and other prevention measures make the goal of reducing or even eliminating new HIV infections a real possibility.5–7 While the city has made progress, not everyone has benefited. African American and Hispanic people have disproportionately higher rates of new HIV infection and AIDS diagnoses than those who self-identify as white. While new HIV infections in Chicago are declining, new infections among young men who have sex with men (MSM) have increased by 5% since 2007.4 Preventing HIV transmission in this group and other high-risk groups requires outreach using social and sexual network methodology, as well as new behavioral and biomedical prevention interventions. HIV screening in health-care settings, however, remains important for identifying individuals unaware of their HIV infection and needing linkage to care and antiretroviral therapy.8–11 Nationally, our health-care system does a poor job of engaging HIV-infected individuals at all points of the care continuum.12 In Chicago, too many people are still unaware that they are infected with HIV (e.g., an estimated 16% of people living with HIV/AIDS are undiagnosed), and only 61% of HIVinfected people are in care.4 Despite the failures of targeted testing, some authorities still believe targeted testing might have advantages over universal screening. Haukoos and colleagues developed a prediction tool, the Denver HIV risk score, which was derived from patients seen in a public sexually transmitted disease clinic in Denver from January 1, 1996, through December 31, 2008.13 This score was validated in a separate urban ED population in Cincinnati, Ohio, from January 1, 1998, through June 30, 2010, and did predict the probability of HIV

infection on subsequent testing. Lyons and colleagues, however, compared fully implemented targeted testing with universal screening in a prospective, randomized study in the urban ED in Cincinnati from January 2008 through December 2010.14 Targeted testing did not lead to a higher positivity rate or reduction in tests performed, and screening actually identified more cases. Uncertainty regarding best practices might also explain why few hospitals report that they conduct routine HIV screening, with failure to screen being common across all types of hospitals.15 Routine HIV screening has been conducted in varying clinical venues using different testing methodologies (standard HIV testing vs. rapid point-of-care testing), different processes for obtaining patient consent (opt-in vs. optout), and different strategies for informing patients of their results and linking HIV-infected patients to care. In general, little evidence or data suggest that one universal screening strategy that incorporates a particular testing methodology, consenting process, or approach to test notification is superior to another with respect to uptake, linkage-to-care rates, clinical outcomes, or harms from screening, according to the systematic review conducted to update the U.S. Preventive Services Task Force recommendations.16 Furthermore, although the sustainability of disparate models is unknown, running parallel programs with dedicated point-of-care HIV testers and/or patient navigators for linkage to care most likely costs more than incorporating HIV testing and linkage to care into routine medical care. The Chicago Department of Public Health (CDPH) sponsors and oversees the Expanded Testing and Linkage to Care (X-TLC) for Disproportionately Affected Populations on the South Side of Chicago initiative at the University of Chicago Medicine (UCM) as part of the overall HIV prevention program for the city. This program is part of a CDC initiative that began in 2007 and has sponsored 25 health departments, including CDPH. CDPH-sponsored prevention sites conducted more than 68,000 HIV tests in Chicago in 2013; 70.0% of tests were conducted in clinical settings, and 63.0% of these tests were performed by the X-TLC program (Personal communication, Nanette Benbow, Chicago Deputy Commissioner, HIV/STI Services, August 2014). Our program evaluated the number of HIV-infected individuals, the number notified of their test results and linked to care, and performance at other points in the HIV care continuum. We describe unique aspects of our program, including standard laboratory testing vs. rapid testing, test follow-up after clinical encounter vs. point-of-care testing, assigned responsibility for test notification and linkage to care, and the use of trained

Public Health Reports  /  2016 Supplement 1 / Volume 131

Expanded HIV Testing and Linkage to Care   109

social workers who are part of an HIV care program to assure linkage to care vs. a parallel process of dedicated testers or patient navigators. METHODS Purpose The X-TLC program aimed to expand routine HIV testing to high-prevalence communities according to public health guidelines. As the lead organization, UCM provided oversight and support to other community sites to assist with the development and implementation of policies and procedures that would promote routine screening as outlined in CDC’s 2006 recommendations. CDPH required all sites to provide care to the following populations that were disproportionately affected by HIV infection: (1) African American and Hispanic men and women and (2) MSM and injection drug users, regardless of race/ethnicity. CDPH’s goal was to screen 97,000 individuals: 27,000 in the first year and 35,000 per year for the next two years. These goals were set forth by the funding agency, CDPH, in its request for applications and were in accordance with CDC’s program for expanding testing for disproportionately affected populations. Patient population Chicago’s South Side neighborhoods primarily comprise minority communities of color and shoulder a substantial portion of the HIV burden within the city. The overall HIV prevalence rate in Chicago is 799.6 people per 100,000 population (people with HIV through April 1, 2013, using 2010 U.S. Census data),4 which is almost three times the national rate of 268.6 cases of HIV infection per 100,000 population (CDC HIV surveillance data through 2009).16 Eleven of the 19 communities with the highest prevalence of people living with HIV (1,372–2,507 per 100,000 population) are within the X-TLC catchment area (supplemental data available upon request). While the payer mix varies widely across X-TLC sites, many sites act as primary health-care providers for uninsured and underinsured individuals in their communities (Table 1). Site selection As part of the Urban Health Initiative at UCM, the South Side Health Collaborative comprises multiple providers that form a coordinated health-care network. All sites except Planned Parenthood of Illinois were recruited through this collaborative. The number of unique patients seen by each organization varied widely, and testing targets were selected based on patient census data (Table 1).

Institutional policy changes When the X-TLC program was proposed in November 2010, UCM adopted a new hospital policy promoting routine opt-out testing and eliminating written consent. In addition, ultimate responsibility for oversight of HIV screening, test notification, and linkage to care was assigned to the HIV Care Program at UCM. Partnering sites and UCM worked collaboratively to create policies and procedures modeled after UCM that would facilitate routine screening within their organizations. Institutional culture, provider comfort, and laboratory procedures were considered in the development of HIV testing policies at each site, and some sites, including the lead site, ultimately used an opt-in approach to assure that patients were aware they were being tested and could ask questions. Consent procedures Consent for HIV testing was obtained through opt-in oral consent at UCM and most other participating sites. Staff members at UCM and partnering sites preferred opt-in language, although policies allowed for the use of opt-out testing. Because Illinois statute requires evidence of oral consent for HIV testing in the patient’s medical record, UCM and partnering sites included this record of oral consent in the patients’ electronic medical records (EMRs). Additionally, several sites preferred to have brief pretest counseling with patients. UCM provided a patient education sheet discussing the rationale for screening, risk factors for HIV acquisition, and significance of the test result, whether it was positive or negative, for locations that did not provide oral pretest counseling. The sheet also provided a telephone number for the X-TLC program for further information and listed telephone numbers for testing information from the State of Illinois and CDC. Testing methods The test technologies used included standard thirdgeneration blood-based enzyme immunoassay and confirmatory Western blot or fourth-generation HIV testing with Multispot confirmation, depending on the site. Only Planned Parenthood used rapid pointof-care testing. Information technology (IT) interventions IT developments at UCM provided several solutions to common barriers to HIV screening. An electronic alert was implemented within the UCM EMR that fired when there was an encounter with an eligible patient who was aged 13–64 years and did not have a prior HIV antibody test result in the EMR. Physicians could order HIV antibody testing from the prompt or could

Public Health Reports  /  2016 Supplement 1 / Volume 131

110    Other Settings

Table 1. Characteristics of participating X-TLCa sites (clinics, family planning clinics, emergency departments, and inpatient hospitals) on the South Side of Chicago, at time of site initiation, 2011–2013 X-TLC sites (number of sites)

Annual visits and unique patients

Site type

Payer mix

Major testing venues

HIV provider

ACCESS Community Health Network (7)

CHC

48,738 visits 16,472 patients

30% Medicaid 4% uninsured 66% private

Clinics

Yes

Beloved Community Family Wellness Center (1)

CHC

8,239 visits 3,149 patients

54% Medicaid 24% uninsured 22% private

Clinics

No

Chicago Family Health Center (5)

CHC

114,601 visits 28,179 patients

61% Medicaid 28% uninsured 11% private

Clinics

No

Community Health (2)

CHC

27,500 visits 9,000 patients

100% uninsured

Clinics

No

Friend Family Health Center (5)

CHC

28,000 visitsb

80% Medicaid 9% uninsured 11% private

Clinics

Yes

Holy Cross Hospital (1)

Community hospital

10,000 admissions 42,000 visits 39,000 ED visitsb

40% Medicaid 4% uninsured 56% private

ED

No

Mercy Medical Center (1)

Community hospital

12,000 visits 3,000 patients

50% Medicaid 5% uninsured 45% private

ED, inpatient hospital, clinics

Yes

Planned Parenthood of Illinois (3)

CHC

20,032 visits 10,156 patients

Unknown

Family planning clinics

No

TCA Health, Inc. (1)

CHC

17,758 visitsb

60% Medicaid 10% uninsured 30% private

Clinics

No

Academic medical center

22,797 admissions 384,550 clinic visits 74,359 ED visitsb

25% Medicaidc 3% uninsured 72% private

ED, inpatient hospital, clinics

Yes

University of Chicago Medicine (1)

a The X-TLC program aimed to expand routine HIV testing to high-prevalence communities with disproportionately affected populations (minority men and women, men who have sex with men, and intravenous drug users) according to Centers for Disease Control and Prevention guidelines at multiple clinical sites in Chicago. As the lead organization, University of Chicago Medicine provided oversight and support to other community sites. b

The total number of patients was not known.

b

Payer mix for the Infectious Diseases/HIV Care Clinic

X-TLC 5 Expanded HIV Testing and Linkage to Care HIV 5 human immunodeficiency virus CHC 5 community health center ED 5 emergency department

exit the alert if a patient declined or had not been offered testing. Across most sites, electronic orders also included HIV testing for different patient care settings. The electronic order required documentation that the patient gave oral consent to testing, eliminating the need for other notation in the EMR. Our alert did not have any hard stops, and alerts could be ignored. As such, we were unable to collect data on the number of tests offered, the number of patients who refused, and the number of patients previously tested elsewhere.

The variability of EMRs across sites made it difficult to implement solutions for all locations and to capture data on patients who were offered but declined testing. Educational efforts The X-TLC initiative was promoted to institutional leadership, providers, and staff members at all sites through a series of in-service meetings at multiple locations. Program presenters reviewed the 2006 CDC guidelines for HIV screening, discussed the rationale

Public Health Reports  /  2016 Supplement 1 / Volume 131

Expanded HIV Testing and Linkage to Care   111

for universal HIV screening, presented local data on HIV testing at UCM, and discussed our local screening strategy. They also emphasized the failure of risk-based screening to identify a substantial proportion of people with HIV early in the course of their disease. Furthermore, presenters described the HIV test characteristics that made ideal screening tools and evaluated HIV testing according to accepted criteria recommended for any medical screening: (1) the condition is important, (2) the natural history is understood, (3) there are effective treatments, (4) early treatment is beneficial, and (5) the cost of testing is substantially lower compared with possible expenditures on medical care if diagnosis is delayed.17 Presenters also discussed the cost-effectiveness of screening when HIV prevalence is .0.1%18 and how screening benefits public health, because people who know their status are less likely to engage in high-risk behaviors that could transmit HIV to others. Testing review, test notification, and linkage to care Each week, the UCM linkage-to-care coordinator received an automated report of all patients tested for HIV. The coordinator reviewed the list to discuss active linkage to care for those with positive test results and supplemental or repeat testing for patients with indeterminate results. Coordinators at other sites regularly reviewed their own test results and linked patients to care. Although ordering providers were encouraged to discuss results with patients, the linkage-to-care coordinator at UCM notified HIV-positive patients if the site personnel were uncomfortable with test notification or unable to make contact. Patients without working telephone numbers were sent a generic letter requesting a return telephone call; those who did not respond were reported to CDPH for case finding. Mercy Medical Center, UCM, ACCESS Community Health Network, and Friend Family Health Center provided on-site HIV specialty care and were the primary care sites for patients tested through the initiative. The UCM linkage-to-care coordinator provided support for linkage to care, regardless of testing location, and made recommendations based on patient insurance and citizenship status. A patient was considered linked to care if he or she attended at least one medical care appointment with an HIV care provider. Site initiation and technical support An initiation visit was conducted at each site to review contracting requirements, discuss goals, negotiate scopes, assess technical support needs, and schedule site trainings. Technical support included assistance with building reports from the EMR, drafting testing

policies, and arranging training sessions for both clinical and nonclinical staff members with assistance from the Midwest AIDS Training and Education Center. Data collection and statistical analysis All sites used an EMR, and data abstraction occurred throughout the course of the program. It included the test date, site location, department, test result, patient demographics and (for people with reactive results only) patient name, prior visit date, previous HIV test result, department, CD41 count, viral load, medical appointment date, risk factor, and medical appointment attendance. Partnering sites were responsible for providing these data and tracking linkage-to-care and retention outcomes for patients who were linked to care sites outside the X-TLC care network. Identification of new vs. existing positive diagnoses was based on previous testing history within the agency (when possible) or self-reported patient testing history. Collected data were housed and updated in REDCap.19 All data were analyzed using Stata®.20 One-way analysis of variance (ANOVA) was used to examine starting CD41 counts of patients newly diagnosed through each program year. One-way analyses of covariance (ANCOVA) were used to identify the effect of other covariates on the CD41 counts and the year of diagnosis. The analysis of early deaths (i.e., death within one year after a new diagnosis of HIV infection) was analyzed by Fischer’s exact test. We considered p,0.05 to be statistically significant. RESULTS The X-TLC program was conducted with UCM and nine community partner sites from February 1, 2011, through December 31, 2013. Upon completion of the X-TLC initiative, 75,345 HIV tests were performed (Table 2). Some individuals had repeat testing, but the number of unique patients tested was 67,153 (data not shown). The number of tests represented only 77.7% of the planned number of tests (75,345/97,000 tests were anticipated), but testing increased each year (11,839 tests in 2011, 27,265 tests in 2012, and 36,241 tests in 2013) without any incremental increase in funding. The majority of patients tested met program goals for testing disproportionately affected members of minority groups, although universal screening was the goal at all sites, regardless of population demographics. Of the 75,345 HIV tests, 48,044 (63.8%) were performed on patients who self-identified as African American, 11,833 (15.7%) on patients who self-identified as white, 2,343 (3.1%) on patients who reported other races, 6,606 (8.8%) on those who self-identified as

Public Health Reports  /  2016 Supplement 1 / Volume 131

112    Other Settings

Hispanic or Latino, and 13,131 (17.4%) on those whose race or ethnicity was not known. The proportion of tests performed on patients who self-identified as African American was even higher in EDs (11,054 of 12,957, 85.3%) than in other clinical settings. Women accounted for 54,560 (72.4%) tests. Screening the 67,153 unique patients identified 395 HIV-infected patients for an overall seroprevalence of 0.6%. However, 127 of 12,957 patients tested in EDs were positive, for a seroprevalence of 1.0%. Of the 395

patients testing positive for HIV, 207 (52.0%) had been diagnosed previously, 176 (45.0%) were confirmed as new diagnoses, and 12 (3.0%) had an unknown prior status. The seroprevalence of undiagnosed cases of HIV was 0.3% (176/67,153), a substantial proportion and higher than the 0.1% cutoff for the seroprevalence accepted for cost-effective HIV screening. This seroprevalence did not vary substantially by site type: 62 of 28,574 (0.2%) HIV-positive tests at community health centers, 36 of 16,333 (0.2%) at community hospitals,

Table 2. X-TLC programa HIV test numbers and positivity rates by various sites, Chicago, Illinois, February 2011 through December 2013 Number of HIV tests (percent positive) Site Total All tests, n All confirmed positives Newly diagnosed patients Previously diagnosed patients CHC and FQHC  ACCESS    All tests, n    All confirmed positive tests    Newly diagnosed patients    Previously diagnosed patients   Beloved Community Family Wellness Center    All tests, n    All confirmed positive tests    Newly diagnosed patients    Previously diagnosed patients   Chicago Family Health Center    All tests, n    All confirmed positive tests    Newly diagnosed patients    Previously diagnosed patients   Community Health    All tests, n    All confirmed positive tests    Newly diagnosed patients    Previously diagnosed patients   Friend Family Health Center    All tests, n    All confirmed positive tests    Newly diagnosed patients    Previously diagnosed patients   Planned Parenthood    All tests, n    All confirmed positive tests    Newly diagnosed patients    Previously diagnosed patients   TCA Health, Inc.    All tests, n    All confirmed positive tests    Newly diagnosed patients    Previously diagnosed patients

2011

2012

2013

Overall HIV seropositivity (percent)

11,839 87 (0.7) 32 (0.3) 43 (0.4)

27,265 134 (0.5) 66 (0.2) 68 (0.2)

36,241 174 (0.5) 76 (0.2) 96 (0.3)

75,345 395 (0.5) 176 (0.2) 207 (0.3)

0 0 (0.0) 0 (0.0) 0 (0.0)

4,895 46 (0.9) 22 (0.4) 24 (0.5)

5,329 40 (0.8) 18 (0.3) 22 (0.4)

10,224 86 (0.8) 40 (0.4) 46 (0.5)

0 0 (0.0) 0 (0.0) 0 (0.0)

336 0 (0.0) 0 (0.0) 0 (0.0)

684 1 (0.1) 0 (0.0) 1 (0.1)

1,020 1 (0.1) 0 (0.0) 1 (0.1)

0 0 (0.0) 0 (0.0) 0 (0.0)

2,460 2 (0.1) 2 (0.1) 0 (0.0)

4,568 5 (0.1) 5 (0.1) 0 (0.0)

7,028 7 (0.1) 7 (0.1) 0 (0.0)

0 0 (0.0) 0 (0.0) 0 (0.0)

252 1 (0.4) 19 (0.4) 0 (0.0)

1,241 3 (0.2) 3 (0.2) 0 (0.0)

1,492 4 (0.3) 4 (0.3) 0 (0.0)

0 0 (0.0) 0 (0.0) 0 (0.0)

1,307 6 (0.5) 5 (0.4) 1 (0.1)

3,826 7 (0.2) 6 (0.2) 1 (0.0)

5,133 13 (0.3) 11 (0.2) 0 (0.0)

0 0 (0.0) 0 (0.0) 0 (0.0)

1,098 0 (0.0) 0 (0.0) 0 (0.0)

1,929 0 (0.0) 0 (0.0) 0 (0.0)

3,027 0 (0.0) 0 (0.0) 0 (0.0)

0 0 (0.0) 0 (0.0) 0 (0.0)

295 0 (0.0) 0 (0.0) 0 (0.0)

354 1 (0.3) 0 (0.0) 1 (0.3)

649 1 (0.2) 0 (0.0) 1 (0.2)

continued on p. 113 Public Health Reports  /  2016 Supplement 1 / Volume 131

Expanded HIV Testing and Linkage to Care   113

Table 2 (continued). X-TLC programa HIV test numbers and positivity rates by various sites, Chicago, Illinois, February 2011 through December 2013

2012

2013

Overall HIV seropositivity (percent)

831 6 (0.7)b 0 0

1,350 11 (0.8) 5 (0.4) 6 (0.4)

1,175 24 (2.0) 8 (0.7) 16 (1.4)

3,356 41 (1.2) 13 (0.4) 22 (0.7)

3,996 27 (0.7)b 4 (0.1) 17 (0.4)

4,326 22 (0.5) 10 (0.2) 12 (0.3)

4,655 26 (0.6) 9 (0.2) 17 (0.4)

12,977 75 (0.6) 23 (0.2) 46 (0.4)

7,012 54 (0.8) 28 (0.4) 26 (0.4)

10,946 46 (0.4) 21 (0.2) 25 (0.2)

12,480 67 (0.5) 29 (0.2) 38 (0.3)

30,438 167 (0.6) 78 (0.3) 89 (0.3)

Number of HIV tests (percent positive) Site Community hospital   Holy Cross Hospital    All tests, n    All confirmed positive tests    Newly diagnosed patients    Previously diagnosed patients   Mercy Medical Center    All tests, n    All confirmed positive tests    Newly diagnosed patients    Previously diagnosed patients Academic medical center   University of Chicago Medicine    All tests, n    All confirmed positive tests    Newly diagnosed patients    Previously diagnosed patients

2011

a The X-TLC program aimed to expand routine HIV testing to high-prevalence communities with disproportionately affected populations (minority men and women, men who have sex with men, and intravenous drug users) according to Centers for Disease Control and Prevention guidelines at multiple clinical sites in Chicago. As the lead organization, University of Chicago Medicine provided oversight and support to other community sites. b

Twelve patients were not classified as newly or previously diagnosed due to missing data.

X-TLC  Expanded HIV Testing and Linkage to Care HIV  human immunodeficiency virus CHC  community health center FQHC  federally qualified health center NA  not available or not applicable

and 78 of 30,438 (0.3%) at academic medical centers (calculated from data in Table 2). The seroprevalence for undiagnosed cases was higher at 63 of 12,957 (0.5%) patients tested in EDs compared with other health-care settings. The overall seroprevalence decreased, from 87 of 11,839 (0.7%) in 2011 to 134 of 27,265 (0.5%) in 2012 and 174 of 36,241 (0.5%) in 2013 (calculated from raw data in Table 2). The seroprevalence of previously undiagnosed HIV-infected individuals also decreased, from 32 of 11,839 (0.3%) in 2011 to 66 of 27,265 (0.2%) in 2012 and 76 of 36,241 (0.2%) in 2013 (calculated from raw data in Table 2). Of the 176 patients newly diagnosed with HIV infection, 142 (80.7%) self-identified as African American, 15 (8.5%) as white, 19 (10.8%) as either other (1.7%) or unknown (9.1%), and 11 (6.3%) as Hispanic (Table 3). A total of 119 (67.6%) of new HIV diagnoses were men, 55 (31.3%) were women, and two (1.1%) were transgender. This proportion of women was higher than that reported for Chicago overall, where women accounted for 19.0% of HIV diagnoses in 2011 even though more than twice as many women

were tested than men.4 Risk factors for HIV acquisition were reported for 133 of 176 (75.6%) newly diagnosed patients. Of the 133 patients reporting mode of acquisition, 77 (58.6%) reported heterosexual exposure, 54 (40.6%) reported MSM, and nine (6.8%) reported intravenous drug use. The risk factors for acquisition in the X-TLC program were also quite different from the demographics of new HIV diagnoses in Chicago in 2011, with 694 of 1,008 (68.9%) newly diagnosed people acquiring HIV through MSM and 215 of 1,008 (21.3%) through heterosexual contact.4 HIV screening in health-care settings identified patients who were not targeted for outreach HIV screening and prevention programs. The exception was screening in EDs, where MSM was the most common risk factor identified by 29 of 60 (48.4%) patients reporting risk factors. Of the 176 patients who were newly diagnosed, three patients died before testing was completed and could not be notified, and 166 of the remaining 173 patients (96.0%) received their test results. Fifty of the 166 patients (30.1%) were notified by the linkage-to-care coordinator by phone, which was the most common

Public Health Reports  /  2016 Supplement 1 / Volume 131

Public Health Reports  /  2016 Supplement 1 / Volume 131

(80.7) (8.5) (1.7) (9.1)

47 37 22 70

(26.7) (21.0) (12.5) (39.8)

11 (6.3) 151 (85.8) 14 (8.0)

142 15 3 16

(75.3) (11.6) (2.4) (10.6)

21 68 22 96

(10.1) (32.9) (10.6) (43.6)

11 (5.3) 161 (77.7) 35 (16.9)

156 24 5 22

42 (40–44) 43 (17–78)

148 (71.5) 56 (27.1) 3 (1.4)

119 (67.6) 55 (31.3) 2 (1.1) 33 (31–35) 29 (16–77)

207 (100.0)

176 (100.0)

Number of previously diagnosed (percent)b

(69.4) (11.3) (3.2) (16.1)

6 4 1 51

(9.7) (6.5) (1.6) (82.3)

7 (11.3) 49 (79.0) 6 (9.7)

43 7 2 10

28 (26–31) 26 (16–59)

36 (58.1) 25 (40.3) 1 (1.6)

62 (100.0)

Number of newly diagnosed (percent)b

(4.06) (28.0) (2.0) (24.0)

0 1 0 49

(0.0) (2.0) (0.0) (98.0)

6 (12.0) 40 (80.0) 4 (8.0)

23 14 1 12

36 (33–40) 32 (19–78)

41 (82.0) 8 (16.0) 1 (2.0)

50 (100.0)

Number of previously diagnosed (percent)b

CHC and FQHC

(80.6) (5.6) (2.8) (11.1)

11 3 3 19

(30.6) (8.3) (8.3) (52.8)

2 (5.6) 29 (80.6) 5 (13.9)

29 2 1 4

34 (29–39) 29 (16–72)

26 (72.2) 10 (27.8) 0 (0.0)

36 (100.0)

Number of newly diagnosed (percent)b

(92.6) (4.4) (0.0) (2.9)

3 12 6 47

(4.4) (17.6) (8.8) (69.1)

0 (0.0) 52 (76.5) 16 (23.5)

63 3 0 2

47 (44–50) 50 (19–77)

42 (61.8) 26 (38.2) 0 (0.0)

68 (100.0)

Number of previously diagnosed (percent)b

Community hospitals

(89.7) (7.7) (0.0) (2.6)

30 30 18 0

(38.5) (38.5) (23.1) (0.0)

2 (2.6) 73 (93.6) 3 (3.8)

70 6 0 2

37 (33–40) 34 (16–77)

57 (73,1) 20 (25.6) 1 (1.3)

78 (100.0)

Number of newly diagnosed (percent)b

(78.7) (7.8) (4.5) (9.0)

18 55 16 0

(20.2) (61.8) (18.0) (0.0)

5 (5.6) 69 (77.5) 15 (16.9)

70 7 4 8

42 (39–45) 42 (17–72)

65 (73.0) 22 (24.7) 2 (2.2)

89 (100.0)

Number of previously diagnosed (percent)b

Academic medical center

Percentages do not always sum to 100 due to rounding.

Includes Asian, American Indian, Pacific Islander, and mixed-race individuals

FQHC 5 federally qualified health center

CHC 5 community health center

X-TLC 5 Expanded HIV Testing and Linkage to Care

HIV 5 human immunodeficiency virus

c

b

a The X-TLC program aimed to expand routine HIV testing to high-prevalence communities with disproportionately affected populations (minority men and women, men who have sex with men, and intravenous drug users) according to Centers for Disease Control and Prevention guidelines at multiple clinical sites in Chicago. As the lead organization, University of Chicago Medicine provided oversight and support to other community sites.

Total Sex  Male  Female  Transgender Age (in years)   Mean (range)  Median Race   African American  White  Otherc  Unknown Ethnicity   Hispanic or Latino   Non-Hispanic or Latino  Unknown Insurance  Uninsured  Public  Private  Unknown

Characteristic

Number of newly diagnosed (percent)b

All site types

Table 3. Characteristics of HIV-infected patients identified through the X-LTC program,a by site type, Chicago, Illinois, January 2011 through December 2013

Expanded HIV Testing and Linkage to Care   115

mode of notification for patients who tested positive in the ED. The remaining seven patients who did not receive their test results were tested in the ED and did not have accurate contact information in the EMR. Despite attempting to contact them by phone and through registered letters, we were unable to notify them of their diagnosis; however, we did report them to CDPH for ongoing attempts at case finding. Of 166 patients eligible for linkage to care, 148 (89.0%) newly diagnosed patients were linked to care within a median of 16 days (range: 0–855 days) (Table 5). The linkage-to-care rate (88.0%) was almost identical for patients newly diagnosed in the ED. Patients had the final decision on where to go for HIV care, but most (123 of 148 patients, 83%) were linked to a provider in our program at one of the four sites that offered comprehensive HIV care. Of the 123 patients linked to care with our providers, 93 (75.6%) initiated highly active antiretroviral therapy and 62.6% achieved viral suppression (i.e., viral load 200 copies/milliliter [mL]) vs. 14.0% for new patients diagnosed through other CDPH programs.21 We did not have access to patient information for the 25 (17.0%) patients who were linked to 12 other sites, as our network of HIV physicians was not providing care for these patients and did not have access to clinical information after linkage. Although the X-TLC program’s primary goal was to identify patients unaware of their HIV infection, identifying patients with a prior diagnosis is another opportunity to capture patients who are out of care and relink them (Table 4). Of the 207 patients previously diagnosed with HIV infection, 92 (44.0%) were confirmed to be in care and nine (4.0%) had moved or died. Of the remaining 106 (52.0%) patients not in care, 78 (73.0%) were relinked to care. Of the 70 previously diagnosed patients who were subsequently linked to care with a provider in our program, 50 (71.0%) began antiretroviral therapy and 35 (50.0%) had a suppressed viral load (calculated from data in Table 4). We had baseline CD4 counts for 122 of the 176 newly diagnosed patients. Baseline CD4 counts at UCM have historically been 200 cells/cubic millimeter (mm3); during X-TLC, baseline CD4 counts at UCM increased from 186 (95% CI 104, 268) cells/ mm3 in 2011 to 316 (95% CI 222, 410) cells/mm3 in 2012 to 494 (95% CI 384, 604) cells/mm3 in 2013 (p0.09, ANOVA). Across all X-TLC sites, the baseline CD4 counts increased from 238 (95% CI 158, 318) cells/mm3 in 2011 to 324 (95% CI 249, 399) cells/mm3 in 2012 to 437 (95% CI 357, 517) cells/mm3 in 2013 (p0.007, ANOVA) (Table 5). A smaller proportion of patients were diagnosed

at CD4 counts 200 cells/mm3 during the three years of the program. We looked at other variables that might have affected the baseline CD4 count, including age, race, sex, risk factor for HIV acquisition, and type of health-care site. Women had higher baseline CD4 counts than men (mean  451329 cells/mm3 vs. 311235 cells/mm3, respectively, p0.03 by ANOVA), and patients diagnosed at community health centers had higher CD4 counts (mean  460249 cells/mm3) than patients diagnosed in the hospital setting (301205 cells/mm3 for community hospitals and 31028 cells/mm3 for the academic medical center, p0.02 by ANOVA). When we adjusted for these covariates, however, the year of diagnosis was still substantially associated with the baseline CD4 count (p0.02, ANCOVA), indicating earlier diagnosis of HIV infection throughout our program (Table 5). The diagnosis of fewer patients with advanced disease (i.e., CD4 count 200 cells/mm3; 11 in 2011, 14 in 2012, and 10 in 2013) (Table 5) was associated with a decrease in early deaths within one year of a new HIV diagnosis. Deaths occurred within one year of diagnosis for four of 32 patients with a new diagnosis in 2011, four of 66 patients with a new diagnosis in 2012, and only one of 78 patients with a new diagnosis in 2013. The differences in the proportion of early deaths between years of the X-TLC program was statistically significant by Fisher’s exact test (p0.04). DISCUSSION We described several unique aspects of the X-TLC initiative that increased the capacity for screening and linkage to care. First, the use of standard blood-based laboratory testing vs. point-of-care testing eliminated the need for a parallel process of care and made testing part of routine health care. Phlebotomy was available for all patients as part of routine medical care, so there was no limit to the number of tests that could be performed. Another unique aspect of the X-TLC program was the assignment of responsibility for oversight of HIV testing to the HIV Care Program at UCM. Previously, it was expected that the ordering physician was responsible for notification of test results. Making providers who have limited experience in conducting HIV pre- or posttest counseling or test result notification responsible for these activities is often a barrier to routine testing. Many physicians have little experience with providing results to patients who test positive, and they might be unable to discuss treatment options and linkage to care. Weekly electronic reports of HIV testing events helped our linkage-to-care coordinators actively follow up with individuals who tested positive.

Public Health Reports  /  2016 Supplement 1 / Volume 131

Public Health Reports  /  2016 Supplement 1 / Volume 131

(0.0) (90.6) (86.2) (96.6) (92.9) (69.2) (83.3) (66.7) (61.1)

0/32 29/32 25/29 28/29 26/28 18/26 15/18 12/18 11/18

5/8 4/5 3/5 3/5

(62.5) (80.0) (60.0) (60.0)

21/43 (48.8) 0/43 (0.0) NA 16/43 (37.2) 8/16 (50.0)

Number of previously diagnosed (n43) (percent)

44/56 36/44 32/44 32/44

0/66 66/66 63/66 61/66 56/61 (78.6) (81.8) (72.7) (72.7)

(0.0) (100.0) (95.5) (92.4) (91.8)

Number of newly diagnosed (n66) (percent)

32/34 20/32 11/32 14/32

(94.1) (62.5) (34.7) (43.8)

23/68 (33.8) NA NA 43/68 (63.2) 34/43 (79.1)

Number of previously diagnosed (n68) (percent)

2012

61/66 42/61 38/61 34/61

0/78 78/78 78/78 77/78 66/77

(92.4) (68.9) (62.3) (55.7)

(0.0) (100.0) (100.0) (98.8) (85.7)

Number of newly diagnosed (n78) (percent)

33/36 26/33 17/33 18/33

(91.7) (78.8) (51.5) (54.5)

48/96 (50.0) NA NA 48/96 (50.0) 36/48 (75.0)

Number of previously diagnosed (n96) (percent)

2013

Three patients died before testing was completed.

NA  not available or not applicable

LTC  linkage to care

X-TLC  Expanded HIV Testing and Linkage to Care

HIV  human immunodeficiency virus

c

Seven patients had moved or died before they attended a scheduled outpatient visit.

b

a The X-TLC program aimed to expand routine HIV testing to high-prevalence communities with disproportionately affected populations (minority men and women, men who have sex with men, and intravenous drug users) according to Centers for Disease Control and Prevention guidelines at multiple clinical sites in Chicago. As the lead organization, University of Chicago Medicine provided oversight and support to other community sites.

Notification and LTC outcomes   Already in care   Eligible for notificationb   Notified of result   Eligible for LTCc   Linked to care Retention and care outcomes   Linked to X-TLC network   Prescribed highly active antiretroviral therapyc   Retained in carec   Viral suppression (viral load 200 cells    per cubic millimeter)b

Linkage to care and outcomes

Number of newly diagnosed (n32) (percent)

2011

Table 4. Notification and linkage-to-care outcomes for patients testing HIV positive through the X-TLC program,a Chicago, Illinois, 2011–2013

Expanded HIV Testing and Linkage to Care   117

Table 5. Baseline CD4+ counts for patients newly diagnosed HIV positive through the X-TLC program,a by year, Chicago, Illinois, 2011–2013 Starting CD41 countsb Number of new HIV diagnoses Number of people with starting CD41 counts available:b   ,100 cells/mm3   100–199 cells/mm3   $200 cells/mm3 Mean CD41 count in cells/mm3 (95% CI) Median CD41 count in cells/mm3 CD41 count range in cells/mm3

2011

2012

2013

32 24 8 3 13 238 (158, 318) 232 4–768

66 47 11 3 33 324 (249, 399) 309 2–1,238

78 51 7 3 41 437 (357, 517) 403 3–1,147

a The X-TLC program aimed to expand routine HIV testing to high-prevalence communities with disproportionately affected populations (minority men and women, men who have sex with men, and intravenous drug users) according to Centers for Disease Control and Prevention guidelines at multiple clinical sites in Chicago. As the lead organization, University of Chicago Medicine provided oversight and support to other community sites. b

Baseline CD41 counts were unavailable for patients linked outside the X-TLC network.

HIV 5 human immunodeficiency virus X-TLC 5 Expanded HIV Testing and Linkage to Care mm3 5 cubic millimeter CI 5 confidence interval

­ oordinators initially contacted the ordering physicians C to see if they would notify the patients or if they wanted our group to do so. At the time of test notification, our coordinator was available for counseling and was able to immediately schedule a medical care visit with an experienced HIV care provider. With high patient volumes and high-risk patient populations, EDs are prime areas for screening. However, studies have revealed that missed opportunities for earlier diagnosis occur frequently in EDs.22–24 Uptake of screening in ED settings has been limited due to the time and financial burdens imposed by testing programs. Giving the responsibility of notification and linkage to care to an HIV care program reduces the burden for an already over-tasked ED. Our ED providers elected to have our linkage-to-care coordinators contact all HIV-infected patients. This approach previously led to a substantial increase in ED testing at UCM, and this earlier success with our ED has been repeated with the X-TLC program.25 Linking patients to care at the time of test notification worked extremely well, but it should be remembered that patients screened in this program were already engaged in care, with the possible exception of patients screened in the ED or in inpatient hospital units. It is clearly more difficult to link patients to care when they are tested in non-primary care settings.26 The use of social workers who are trained to assess needs and help overcome barriers to linkage to care very likely positively affected our results. It is also an advantage to have the same group responsible for

expanded HIV testing to also be involved in linkage to care and medical care. We have looked at the continuum of care cascade for patients with a new HIV diagnosis identified through the X-TLC program and linked to providers in our network, and have found that 63.0% have achieved a viral load ,200 copies/ mL. This increased proportion of HIV patients with viral suppression is reducing the community viral load and preventing new transmissions. Not only did our program have success in intermediate process outcomes (i.e., increased testing events and patients diagnosed with HIV and linked to care), but it also affected clinical outcomes. The earlier diagnosis, as evidenced by higher baseline CD41 counts in our program, is encouraging, as up until at least 2010, baseline CD41 counts in the United States had not changed since 1992.27 Earlier diagnosis is extremely important for improving outcomes for patients infected with HIV. Patients with a late diagnosis of HIV have a substantial risk of AIDS and/or death in the first year after diagnosis.28,29 Long-term survival is also improved with earlier diagnosis.30 Within our cohort, we have seen a reduction in early mortality within the first year after a new diagnosis when comparing years two and three with the first year of the program. It is possible that the lower baseline CD41 counts in the first year might represent more targeted testing early in the program and more screening in the latter part of the program. A change in physician behavior from targeted to routine screening is, however, exactly in line with the program’s purpose.

Public Health Reports  /  2016 Supplement 1 / Volume 131

118    Other Settings

Furthermore, despite testing a much higher number of patients over time, the absolute number, not just the proportion, of patients diagnosed with advanced disease also declined. The reduced proportion of patients diagnosed at an advanced stage of HIV infection indicates that there are fewer patients with advanced disease who remain undiagnosed at our participating health-care sites. The reduction in new diagnoses might indicate more targeted testing in the early phase of our program, or might be due to a reduction in the proportion of patients with undiagnosed HIV infection and/or a reduction in new transmissions—the goal of an effective screening program. We would hope that the number and proportion of previously undiagnosed patients would decrease over time. Although the identification of patients previously diagnosed with HIV infection was not the focus of our program, we believe that identifying patients who know they are HIV-infected but do not disclose their HIV positivity during the clinical encounter is another benefit of the universal testing approach. More than half (51.7%) of the patients testing positive in our program had a prior diagnosis. For whatever reason, they represented a group of patients that is very likely to be out of care. We were able to reconnect 72.9% of these patients to care. Identifying these individuals and linking them to care is extremely important, as 91.5% of new HIV transmissions are linked to people who are not in care, including 30.2% of transmissions from people who are undiagnosed and 61.3% of transmissions from people with known diagnosis who are not in care.31 Limitations The X-TLC initiative was subject to several limitations. First, our program was far from achieving universal screening for all eligible patients. We might, however, have tested at a level that effectively identified a substantial proportion of patients with HIV who were unaware of their status. A recent study suggests that health-care systems should see a substantial decline in undiagnosed HIV patients if they annually reach 20% of their eligible population.32 Interestingly, 59.0% of our newly diagnosed patients acquired HIV through heterosexual contact and 31.3% were women. Patients diagnosed through screening in health-care settings may not reflect the overall demographics of the HIV epidemic. Although our sites might care for more heterosexual men and women, it is also possible that heterosexual men and women were offered and/or accepted testing more frequently than MSM despite having a lower prevalence of HIV infection. The exception was ED screening, where MSM was the most com-

mon risk factor identified at 48.4%, but this percentage was still lower than the proportion of cases acquired through MSM for Chicago (69.0%). Health-care screening programs may not reach high-risk populations such as young African American MSM. On the other hand, heterosexual men and women are not necessarily targeted for screening in outreach programs and may only be diagnosed by testing in health-care settings. Changing provider behavior has been slow, even with champions promoting HIV in various clinical areas at UCM and at our community sites. Education has had limited impact. We have repeated our educational campaign many times for multiple groups at UCM and our other sites, with updates on our progress. To assess the impact of education, we have surveyed staff members in medicine, emergency medicine, pediatrics, and obstetrics and gynecology to assess knowledge, attitudes, and behaviors toward HIV screening.33 Our data indicate that despite increased knowledge of current recommendations and positive perceptions of screening (e.g., believing it is good for patients and for public health), actual testing practices did not improve substantially. Electronic alerts might be helpful, but alerts are frequently ignored. It would be helpful to have providerspecific data on how often testing is being offered, deferred, or refused by patients, and report individual performance as compared with that of peers and overall program goals. We reported our results to the Department of Medicine Safety Committee, highlighting the increased baseline CD41 counts for newly diagnosed HIV patients and the decrease in early deaths. This committee subsequently endorsed HIV screening as a quality performance measure for all physicians in the department. The designation as a quality performance measure could be the impetus that changes physician behavior to make HIV screening part of routine medical care. Although education did not change testing behaviors, it might have improved acceptance of new institutional requirements to test. One problem with universal screening is that risk factor assessment might be deferred, making it difficult to identify patients who need repeat testing and/ or a referral to designated prevention programs for preexposure prophylaxis or behavioral interventions. We also need to develop systems to assure that patients who test negative for HIV are notified of their result. Posttest counseling for negative results is an opportunity to educate patients about avoiding HIV, and this patient encounter might also be an opportunity to identify high-risk patients who could benefit from repeat testing and/or prevention programs. Finally, we developed our HIV screening efforts independent

Public Health Reports  /  2016 Supplement 1 / Volume 131

Expanded HIV Testing and Linkage to Care   119

of other health screenings. In the future, we need to take a more coordinated approach to implementing all recommended health screenings for our patients (i.e., develop health screening care bundles). CONCLUSION We described our program for expanded HIV testing using conventional laboratory testing instead of pointof-care testing across a network of hospitals and healthcare centers, with responsibility for test notification and linkage to care assigned to the lead organization. This program has expanded testing and successfully linked patients to medical care with experienced HIV providers. Clinical outcomes were also improved, with diagnosis at an earlier stage of disease and a reduction in early mortality. The Expanded Testing and Linkage to Care (X-TLC) group members, in addition to the authors, included Brian Bragg; Nancy Glick, MD, Kischa Hampton, Danielle Lazar, and Shirley Stephenson (ACCESS Community Health Network); Reneau Diallo, Ladonna Hall, and Laura Lucero, MD (Beloved Community Family Wellness Center); Kisti Catellano, MD, Lindsey Madgett, and Sara Schutz (Chicago Family Health Center); Rachel Bennett, Judith Haasis, Emily Hendel, and Laura Michalski (CommunityHealth); Sabrina Moore, Joyce Smith, MD, and Debra Sturgis-Hinton, MD (Friend and Family Health Center); Fred Giles and Michael “Casey” Kincaid (Holy Cross Hospital); Meshoun Dorsey, Art Moswin, MD, and Norma Rolfson (Mercy Medical Center); Jerrica Ferris, Haydee Ramos, and Kai Tao, MD (Planned Parenthood of Illinois); and Laverne Barnes, MD, and Veronica Clarke (TCA Health, Inc.). The authors and X-TLC group members thank those involved with or who supported this program, especially their patients. The authors also thank David Amarathithada and Randall Buffington, who were part of the X-TLC team with the Chicago Department of Public Health (CDPH) at the start of this program; the Midwest AIDS Training and Education Center for conducting training sessions for the sites; and the Gilead HIV on the Frontlines of Communities in the United States (FOCUS) program for its quality performance-improvement program for HIV screening. CDPH funded the X-TLC Program for Disproportionately Affected Populations on the South Side of Chicago at the ­University of Chicago Medicine as part of the overall HIV prevention program for the city. CDPH is funded by the Centers for Disease Control and Prevention as part of an initiative for expanded HIV testing for disproportionately affected populations. David Pitrak, MD, received grant support for expanded HIV testing from the Gilead HIV FOCUS program.

REFERENCES  1. Centers for Disease Control and Prevention (US). Estimates of new HIV infections in the United States. August 2008 [cited 2015 Sep 15]. Available from: http://www.cdc.gov/nchhstp/newsroom /docs/fact-sheet-on-hiv-estimates.pdf   2. The White House (US), Office of National AIDS Policy. National HIV/AIDS strategy for the United States, 2010 [cited 2015 Sep 15]. Available from: http://aids.gov/federal-resources/national-hiv-aidsstrategy/nhas.pdf

  3. Centers for Disease Control and Prevention (US). Expanded HIV testing for disproportionately affected populations. Monitoring and evaluation plan. Atlanta: CDC, National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention, Division of HIV/AIDS Prevention; 2010.  4. Chicago Department of Public Health. HIV/STI surveillance report, Chicago, 2013. Chicago: City of Chicago; 2013. Also available from: https://www.cityofchicago.org/content/dam/city/depts /cdph/infectious_disease/STI_HIV_AIDS/HIV_STISurveillance Report2013.pdf [cited 2015 Jul 30].   5. Cohen MS, Chen YQ, McCauley M, Gamble T, Hosseinipour MC, Kumarasamy N, et al. Prevention of HIV-1 infection with early antiretroviral therapy. N Engl J Med 2011;365:493-505.   6. Das M, Chu PL, Santos GM, Scheer S, Vittinghoff E, McFarland W, et al. Decreases in community viral load are accompanied by reductions in new HIV infections in San Francisco. PLoS One 2010;5:11068.   7. Grant RM, Lama JR, Anderson PL, McMahan V, Liu AY, Vargas L, et al. Preexposure chemoprophylaxis for HIV prevention in men who have sex with men. N Engl J Med 2010;363:2587-99.   8. Jenkins TC, Gardner EM, Thrun MW, Cohn DL, Burman WJ. Riskbased human immunodeficiency virus (HIV) testing fails to detect the majority of HIV-infected persons in medical care settings. Sex Transm Dis 2006;33:329-33.   9. Branson BM, Handsfield HH, Lampe MA, Janssen RS, Taylor AW, Lyss SB, et al. Revised recommendations for HIV testing of adults, adolescents, and pregnant women in health-care settings. MMWR Recomm Rep 2006;55(RR-14):1-17. 10. Chou R, Selph S, Dana T, Bougatsos C, Zakher B, Blazina I, et al. Screening for HIV: systematic review to update the 2005 U.S. Preventive Services Task Force recommendation. Ann Intern Med 2012;157:706-18. 11. Preventive Services Task Force (US). Final update summary: human immunodeficiency virus (HIV) infection: screening. Rockville (MD): USPSTF; 2015. Also available from: http://www.uspreventiveservices taskforce.org/page/document/update summaryfinal/humanimmunodeficiency-virus-hiv-infection-screening?ds=1&s=HIV [cited 2015 Jul 30]. 12. Gardner EM, McLees MP, Steiner JF, Del Rio C, Burman WJ. The spectrum of engagement in HIV care and its relevance to test-andtreat strategies for prevention of HIV infection. Clin Infect Dis 2011;52:793-800. 13. Haukoos JS, Lyons MS, Lindsell CJ, Hopkins E, Bender B, Rothman RE, et al. Derivation and validation of the Denver human immunodeficiency virus (HIV) risk score for targeted HIV screening. Am J Epidemiol 2012;175:838-46. 14. Lyons MS, Lindsell CJ, Ruffner AH, Wayne DB, Hart KW, Sperling MI, et al. Randomized comparison of universal and targeted HIV screening in the emergency department. J Acquir Immune Defic Syndr 2013;64:315-23. 15. Herrin J, Wesolowski LG, Heffelfinger JD, Bostick N, Hall HI, Ethridge SF, et al. HIV screening practices and hospital characteristics in the US, 2009–2010. Public Health Rep 2013;128:161-9. 16. Centers for Disease Control and Prevention (US). Diagnoses of HIV infection and AIDS in the United States and dependent areas, 2010. HIV Surveillance Report 2012;22:1-79. 17. Wilson JMG, Jungner G. Principles and practice of screening for disease. Geneva: World Health Organization; 1968. 18. Paltiel AD, Weinstein MC, Kimmel AD, Seage GR 3rd, Losina  E, Zhang H, et al. Expanded screening for HIV in the United States— an analysis of cost-effectiveness. N Engl J Med 2005;352:586-95. 19. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)—a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform 2009;42:377-81. 20. Stata Corp. Stata®: Release 12. College Station (TX): StataCorp; 2013. 21. Benbow N. Expanded HIV Testing and Linkage to Care (X-TLC) in healthcare settings on the south side of Chicago. Abstract presented at the 8th International Conference on HIV Treatment and Prevention Adherence; 2013 Jun 2–4; Miami, Florida. 22. Telzak EE, Grumm F, Coffey J, White DAE, Scribner AN, Wuan S, et al. Rapid HIV testing in emergency departments—three U.S. sites, January 2005–March 2006. MMWR Morb Mortal Wkly Rep 2007;56(24):597-601.

Public Health Reports  /  2016 Supplement 1 / Volume 131

120    Other Settings

23. Kelen GD, Hexter DA, Hansen KN, Humes R, Vigilance PH, Baskerville M, et al. Feasibility of an emergency department-based, risk-targeted voluntary HIV screening program. Ann Emerg Med 1996;27:687-92. 24. Rothman RE. Current Centers for Disease Control and Prevention guidelines for HIV counseling, testing, and referral: critical role of and a call to action for emergency physicians. Ann Emerg Med 2004;44:31-42. 25. Schrantz S, Bares S, Pouch S, Pitrak D. Retrospective analysis of HIV testing at a large, urban medical center: support for increased nontargeted testing. Poster presented at the Infectious Diseases Society of America 46th Annual Meeting; 2008 Oct 25–28; Washington, DC. 26. Gilman B, Hidalgo J, Thomas C, Au M, Hargreaves M. Linkages to care for newly diagnosed individuals who test HIV positive in nonprimary care settings. AIDS Patient Care STDS 2012;26:132-40. 27. Lesko CR, Cole SR, Zinski A, Poole C, Mugavero MJ. A systematic review and meta-regression of temporal trends in adult CD4(1) cell count at presentation to HIV care, 1992–2001. Clin Infect Dis 2013;57:1027-37. 28. Shouse RL, Kajese T, Hall HI, Valleroy LA. Late HIV testing—34 states, 1996–2005. MMWR Morb Mortal Wkly Rep 2009;58(24):661-5.

29. The Antiretroviral Therapy Cohort Collaboration. Life expectancy of individuals on combination antiretroviral therapy in high-income countries: a collaborative analysis of 14 cohort studies. Lancet 2008;372:293-9. 30. May M, Gompels M, Delpech V, Porter K, Post F, Johnson M, et al. Impact of late diagnosis and treatment on life expectancy in people with HIV-1: UK Collaborative HIV Cohort (UK CHIC) Study. BMJ 2011;343:6016. 31. Skarbinski J, Rosenberg E, Paz-Bailey G, Hall HI, Rose CE, Viall AH, et al. Human immunodeficiency virus transmission at each step of the care continuum in the United States. JAMA Intern Med 2015;175:588-96. 32. Casey E, Hamilton T, Schechter C. What proportion of patients need to be tested to realise test and treat goals? Trend analysis of patients screened for HIV within the New York City municipal health-care system. Lancet 2013;382 Suppl 2:[1 screen]. 33. Bares S, Steinbeck J, Bence L, Kordik A, Acree ME, Jih J, et al. Knowledge, attitudes, and ordering patterns for routine HIV screening among resident physicians at an urban medical center. J Int Assoc Provid AIDS Care 2014;[Epub ahead of print].

Public Health Reports  /  2016 Supplement 1 / Volume 131