Archives of Psychiatric Nursing xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Archives of Psychiatric Nursing journal homepage: www.elsevier.com/locate/apnu
Lessons Learned from Implementing a Pilot RCT of Transitional Care Model for Individuals with Serious Mental Illness Phyllis Solomon a,⁎, Nancy P. Hanrahan b, Matthew Hurford c, d, Marissa DeCesaris e, LaKeetra Josey e a
University of Pennsylvania School of Social Policy & Practice, Center for Mental Health Policy and Services Research University of Pennsylvania School of Nursing, Center for Health Outcomes and Policy Research Philadelphia Department of Behavioral Health and Intellectual Disability Service d University of Pennsylvania School of Medicine, Center for Mental Health Policy and Services Research e University of Pennsylvania School of Nursing, Center for Health Outcomes and Policy Research b c
a b s t r a c t We adapted an evidence-based transitional care model for older adults being released from acute care hospitals for patients with serious mental illness and medical co-morbidities being discharged from two psychiatric units of an acute care hospital (TCare) and evaluated implementation issues. An advisory group (AG) of community stakeholders assessed barriers and facilitators of a 90-day T-Care intervention delivered by a psychiatric nurse practitioner (NP) in the context of conducting a pilot randomized controlled trial. Minutes of AG and case narratives by NP of 20 intervention participants were content analyzed. Patients with immediate and pressing physical health problems were most receptive and actively utilized the service. Provider barriers consisted of communication and privacy issues making it difficult to contact patients in mental health facilities. In contrast, the NP was accepted and valued in the physical health arena. Psychosocial needs and relationship issues were demanding, and we recommend a team approach for TCare with the addition of a social worker, peer provider, and consulting psychiatrist for severely mentally ill patients being released from an acute physical health hospitalization. © 2014 Published by Elsevier Inc.
Adults with severe psychiatric disorders have excess medical morbidity and mortality as compared to the general population (Colton & Mandersheid, 2006; Piatt, Munetz, & Ritter, 2010). This psychiatric population dies on average 25 years sooner than others, often due to preventable medical illnesses (Colton & Mandersheid, 2006). Integrated medical and behavioral health care has been found to improve outcomes for this population (Druss, Rohrbaugh, Levinson, & Rosenheck, 2001; Lawrence & Kisely, 2010; Woltmann et al., 2012); yet their ability to access and interact with the medical care system is often compromised due to personal, provider, and system factors which result in inferior quality healthcare and consequently, poor health outcomes. For example, adults with severe psychiatric disorders often face motivational and cognitive challenges to access and communicate their concerns to health providers and to act as their own advocates in the complex medical arena. Furthermore, the mental health providers serving them do not have the medical All authors note no conflict of interest. Funding: Robert Wood Johnson Foundation Interdisciplinary Nursing Quality Research Initiative. ⁎ Corresponding Author: Phyllis Solomon, PhD, Professor, University of Pennsylvania School of Social Policy & Practice, Center for Mental Health Policy and Services Research. E-mail addresses: [email protected] (P. Solomon), [email protected] (N.P. Hanrahan), [email protected], [email protected] (M. Hurford), [email protected] (M. DeCesaris), [email protected] (L. Josey).
knowledge, time, or resources to deal with their physical health problems or to assist them in accessing these needed medical services. Analogously, medical providers are uncomfortable dealing with this population as they have inadequate knowledge of their psychiatric disorders and also have limited time and resources to deal with patients with complex co-morbidities compounded by social needs (Druss et al., 2010; Lawrence & Kisely, 2010; Lester, Tritter, & Sorohan, 2005). To further complicate the situation, system-level factors such as financing policies and privacy requirements create challenges to integrating care across these systems. Care coordination to assist with appropriate connections between systems such as hospitals and community health services is widely accepted as a cost-effective strategy to improve access and care quality (Griswold, Homish, Pastore, & Leonard, 2010; Griswold et al., 2005). The point of discharge from a psychiatric inpatient stay for those with severe psychiatric illnesses is an opportune time to intervene with those with medical co-morbidities to assist them in connecting with medical and behavioral health systems of care and educating them about managing both their mental and physical health conditions. A model that has been shown to be effective, particularly with elderly patients being released from an acute medical hospitalization, is the transitional care model (TCM) (Naylor, 2000; Naylor, Aiken, Kurtzman, Olds, & Hirschman, 2011; Naylor et al., 1999, 2013). This evidenced- based model seemed appropriate to adapt for adults with severe mental illness and a comorbid major medical problem being discharged from an acute
http://dx.doi.org/10.1016/j.apnu.2014.03.005 0883-9417/© 2014 Published by Elsevier Inc.
Please cite this article as: Solomon, P., et al., Lessons Learned from Implementing a Pilot RCT of Transitional Care Model for Individuals with Serious Mental Illness, Archives of Psychiatric Nursing (2014), http://dx.doi.org/10.1016/j.apnu.2014.03.005
2
P. Solomon et al. / Archives of Psychiatric Nursing xxx (2014) xxx–xxx
psychiatric unit by employing a psychiatric advanced practice nurse to deliver the intervention. We conducted a feasibility assessment of the implementation of the adaptation of TCM within the context of a pilot RCT naming it transitional care (TCare) (so as not to conflict with targeted case management also referred to as TCM in the mental health arena). Here we report on the results of the qualitative assessment of the implementation of TCare. The outcomes of the pilot RCT (to be reported in another manuscript) of rehospitalizations, emergency room use of both psychiatric and medical, and appointment keeping did not differ by conditions, and therefore, the need to assess the implementation of the TCare intervention was all the more pressing to understand. BACKGROUND LITERATURE The use of nurses to assist in the coordination of care of psychiatric patients from hospital discharge to community placement has long been considered an effective strategy. Public health nurses working with patients and their families in the beginning of deinstitutionalization in the 1960s were found to effectively help in maintaining patients to remain in their homes (Davis, Dinitz, & Pasamanick, 1972; Scarpitti, Albini, Baker, Dinitz, & Pasamanick, 1965). Forchuk, Reynolds, Sharkey, Martin, and Jensen (2007) developed a transitional discharge model (TDM) which included peer support and inpatient providers (including nurses) to help with the linkages to community services post-discharge for patients being released from a psychiatric hospitalization. They found that it took an average of 3 months for discharged patients to form a relationship with community providers and that the model did enhance social relationships and did reduce length of hospital stay, but did not improve functioning or quality of life or reduce hospitalization or emergency room use (Forchuk, Martin, Chan, & Jensen, 2005). However, when the model was tested employing an RCT in Scotland, the researchers did find TDM patients reported being less symptomatic, having higher levels of functioning, better quality of life, and had fewer hospital readmissions within 5 months of discharge (Reynolds et al., 2004). Recently, a variety of care navigator models to assist in connecting this psychiatric population with medical services have been developed. A randomized pilot study employed peer navigators to teach patients in situ self-advocacy skills within the healthcare system to engage more effectively with the medical system. Preliminary results indicated that those in the experimental intervention which utilized peer navigators changed their views about seeking care from primary health care providers away from emergency room use (Kelly et al., 2014). Another study connected with patients at the point of a psychiatric crisis using care navigators with support of peers to assist patients in linking with primary care and found that after 1 year, access to primary care was greater for the intervention condition than the control (Griswold et al., 2010). TCM AND PILOT STUDY INTERVENTION TCM was designed as a 90 day intervention employing an advanced practice nurse who does a comprehensive discharge plan for chronically ill older adult patients prior to their discharge from the hospital. The APN then sees the patient within 48 hours of discharge in their home with the family caregiver and provides education and support to them regarding managing the patient's illness. The APN engages in outreach, goes to primary care appointments with the patient and is available 7 days per week by phone during the intervention period (Naylor et al., 1999). The essential elements of TCM as delineated by Naylor et al. (2013) are: 1. coordination of care by an APN; 2. a plan developed prior to hospital discharge; 3. home visits by APN for approximately 90 days post-hospital discharge and available 7 days a week; 4. coordination with physicians in community, including accompanying patient on these visits; 5. inclusive focus on health needs of patient; 6. involvement of both patient and family in patient care through educating and supporting them; 7. early
detection and quick “response to health care risks and symptoms”; 8. patient, family caregiver, and providers function as a team; 9. collaboration of nurse and physician; and 10. information sharing with and among all members of the team. TCM has been found to effectively reduce readmissions and lower healthcare costs in three randomized controlled trials and a recent quasi-experimental design within routine care (Naylor, 2000; Naylor et al., 1999, 2011, 2013). Rose, Gerson, and Carbo (2007) conducted a pilot study of TCM for psychiatric patients discharged from hospital to family settings, i.e., study eligibility required that the patient had to reside with a family member. TCM was selected due to its effectiveness with highrisk populations in preventing rehospitalizations and reducing cost. The researchers felt that TCM had the potential to assist patients who are often discharged with symptomatic behavior, have “complex and unpredictable illness course”, are frequently medication non-adherent due to difficulty handling poly-pharmacy and therefore require support and education to prevent hospital readmission (Rose et al., 2007). This study included 10 patients and found that the intervention needed to address the concerns of caregivers in relating to non-adherence issues and potentially violent threats of the patient; assistance in connecting and engaging patients with structured daily activities; help in obtaining and developing strategies to improve adherence with prescribed medication, and to manage symptomatic behaviors. In conclusion, Rose and colleagues found that the advanced practice nurse needed to be “fluid and responsive” and consequently, taking on functions that could have been provided in partnership with lay persons or peers as had been done in the TDM model discussed above. This study showed potential for the model, but required further study. In our pilot study TCare included all of the elements of TCM, but the advanced practice nurse was specialized and had experience in psychiatric nursing as in the pilot by Rose and colleagues. The APN hired had both inpatient and outpatient psychiatric experiences as well as research coordination background, which made her an ideal candidate to deliver this intervention. Coordination in our study was not only with health care providers but also with psychiatric care. Furthermore, a psychiatrist was available for consultation. Also, caregivers were more broadly defined than in Rose's study as often psychiatric patients do not live with family members but in residential care facilities, boarding homes, and homeless shelters. Therefore, the APN worked with others besides families with whom the patient resided, such as residential staff. Education and support was offered to any caregivers in the patient's environment who could assist the patient with managing the illness and offer support for adherence to treatment. TCare was similar in some respects to targeted case management in that it offered assessment, planning, assistance in accessing medical and social services based on the patient's needs, including attending appointments with the patient and monitoring receipt of services. Targeted case management is offered by community mental health agencies to Medicaid eligible clients. However, case managers in the public mental health system often have limited training regarding medical issues and generally focus their efforts on mental health and social service needs of clients. Case managers may have a bachelor's degree or less in any discipline. Thus, their clinical training is minimal, if any. A master's degree, such as in social work, is often required for supervisors of case managers. The psychiatric APN was clinically trained in both medical and psychiatric assessment and treatment as well as had medication prescribing privileges. Thus, she was in a better position to ensuring that other physical and behavioral health providers had accurate clinical information and that patients understood their physical health management needs and she could offer education and support in regard to these issues. In addition, the APN met with patients in the hospital to engage with them prior to discharge, while the usual system is a referral to a mental health agency upon discharge from the hospital, where they are often assigned a case manager. The patients are thus left to their own devices to connect with community mental health services.
Please cite this article as: Solomon, P., et al., Lessons Learned from Implementing a Pilot RCT of Transitional Care Model for Individuals with Serious Mental Illness, Archives of Psychiatric Nursing (2014), http://dx.doi.org/10.1016/j.apnu.2014.03.005
P. Solomon et al. / Archives of Psychiatric Nursing xxx (2014) xxx–xxx
METHODS An advisory group (AG) of key stakeholders met throughout the study to translate the original TCM and to assess barriers and facilitators to TCare implementation. The AG met once a month for about 1 year and included key stakeholders: two mental health consumers; the director of behavioral health programs for the city of Philadelphia; an administrator from the public behavioral health managed care organization who was also a psychiatrist and the recruitment hospital staff including a primary care physician, a psychiatrist, and a social worker; a visiting nurse from the community; and one private insurance representative. University IRB approval was obtained, and all AG and RCT participants were consented. At monthly meetings, the research team provided examples to the AG of barriers and facilitators from the implementation of the TCare RCT. Between meetings issues that arose during the implementation were discussed by the research team for presentation and discussion at the next AG meeting. Minutes of these meetings were taken. Additionally, the psychiatric nurse practitioner who provided TCare wrote detailed case narratives of study participants in the experimental arm of the study. RCT participants were forty adult patients with serious mental illness, i.e., schizophrenia, bipolar, and major affective disorders and a major medical condition who were assigned to TCare or usual care. The 20 case narratives from the experimental arm along with the AG meeting minutes were content analyzed and used to generate the findings described below. These materials were read by three authors (PS, NH, & MH) and discussed to come up with the themes with input of the APN (LJ). Consensus regarding themes was reached by the three research team members, a social worker, a nurse and a psychiatrist, respectively. Thus a qualitative approach was employed. RESULTS Sociodemographics Mean age was 44 years, 55% male, and 45% African American. Forty percent had less than a high school education, were poor (mean income = $717 per month), and single (60%). At baseline, a quarter lived in emergency shelters, halfway homes or had no place to live. Most were unemployed (see Table 1). Complex Health and Behavioral Health Needs On average, participants had 1.6 psychiatric diagnoses and 3.3 comorbid medical diagnoses. Half of the participants had a diagnosis of major depression (50%), a third had a diagnosis of schizoaffective disorder and/or substance use (30% and 30%, respectively), and a quarter of participants had a diagnosis of schizophrenia (25%). The following medical problems were common: endocrine disorders (60% with hypothyroid and/or type II diabetes mellitus); hypertension (40%); respiratory conditions (30%), and hyperlipidemia (20%). Medication regimens were complex, and medication problems were common. Participants took an average of 6 different types of medication each day. Three participants misplaced their hospital prescriptions, and the TCare-NP replaced the prescription until an appointment with a psychiatrist could be made. One participant with schizophrenia did not fill his new prescription and instead used his pre-hospital antipsychotic medication. Within a week he began to relapse with auditory hallucinations and paranoia. The TCare-NP assessed the problem, corrected his dose of medication and his symptoms remitted. Another participant's insurance would not pay for his prescribed medication so the TCare-NP worked with the primary care provider to find a compatible substitute. A participant with schizophrenia became physically aggressive with a boarding
3
Table 1 Characteristics of the Intervention Group (n = 20). Characteristic Mean (SD) 44.1 ± 11.2 $717 ± $458 6.1 ± 3.03
Age Income (total money received previous month) Mean # of daily medications
Gender as observed Male Female Race Black/African American White Asian More than one race Other Hispanic/Latino Education Less than high school High school Post-high school technical training Some college College degree Some graduate study Marital status Single Married or with consistent partner Divorced Separated Widowed Current living situation Home, hotel, or apartment With parents or other family members With friends Emergency shelter Halfway house or board and care No home or regular place to live Employed No Yes ⁎Mental disorder diagnoses Major depression Bipolar disorder Schizoaffective disorder Schizophrenia Psychosis NOS Personality disorder Substance use ⁎Comorbid medical problems Endocrine (diabetes, hypothyroid) Hyperlipidemia Hypertension Cardiovascular Respiratory Gastric-GERD Seizure disorder M/S-arthritis Infectious disease (hepatitis B, C; HIV) Cancer Traumatic brain injury
n
%
11 9
55 45
9 6 0 2 2 1
45 30 5 10 10 5
8 6 0 5 0 1
40 30 0 25 0 5
12 3 2 2 1
60 15 10 10 5
12 3 0 2 1 2
60 15 0 10 5 10
15 3
75 15
10 2 6 5 2 2 6
50 10 30 25 10 10 30
12 4 8 4 6 2 3 3 1 2 1
60 20 40 20 30 10 15 15 5 10% 5%
Data source: Hospital medical record. ⁎ Note: Individuals may meet criteria for more than one diagnostic category thus percentages exceed 100%.
home staff member, and his housing was in jeopardy. Given that he was having difficulty sleeping, the TCare-NP prescribed a few days of a medication that allowed him to sleep. He was less restless and anxious in subsequent days and maintained his residence. Another participant with bipolar and borderline personality disorder was started on lithium while hospitalized. After discharge, the participant frequently missed appointments and ran out of lithium. The TCare-NP managed her lithium medication, monitoring it until she was engaged with her provider.
Please cite this article as: Solomon, P., et al., Lessons Learned from Implementing a Pilot RCT of Transitional Care Model for Individuals with Serious Mental Illness, Archives of Psychiatric Nursing (2014), http://dx.doi.org/10.1016/j.apnu.2014.03.005
4
P. Solomon et al. / Archives of Psychiatric Nursing xxx (2014) xxx–xxx
Intensity of Physical Health Need Participants with immediate and pressing medical problems most actively utilized TCare. For example, a participant with bipolar disorder attended her first visit with an oncologist shortly after discharge, accompanied by the TCare-NP. The participant was highly anxious and had little social support as her fiancé was actively abusing alcohol and crack cocaine. When the oncologist informed her that she had rectal cancer and needed chemotherapy and radiation treatment, she had a severe anxiety attack. The TCare-NP used anxiety-reducing techniques to enable the patient to calm herself and continue to engage with the oncologist. When the participant was calmer, the TCare-NP translated the cancer treatment plan, and subsequently, coordinated multiple medical and psychiatric treatment appointments, medication changes, and radiology appointments. By the end of her TCare intervention, the participant successfully completed chemotherapy and radiation treatment and was connected with her primary care and outpatient mental health providers. The participant was not rehospitalized and did not use emergency services during this time. Another participant with chronic obstructive pulmonary disease had a co-morbid substance use disorder and required close monitoring as he was in terminal stages of his respiratory illness. He was receptive to assistance from TCare. In contrast, a woman with asthma was randomized to receive TCare, however, she was not having problems with her asthma and avoided contact with the TCare-NP. Conflict-Intense Relationships and Unstable Housing The TCare-NP frequently managed relational conflicts and housing problems. Approximately two-thirds of the hospital readmissions were associated with unstable living situations and relational conflicts. A third of participants changed their housing during the 90 days of the TCare intervention. When a participant with schizophrenia who had a toe amputated due to diabetes refused assistance from residential staff, his housing became jeopardized. He was irritable, aggressive, and threatened other residents and staff. The sutures from his surgery remained in place 3 months after the procedure because he was paranoid and refused to see the doctor. The TCare-NP established a relationship, ensured he was taking the prescribed dose of antipsychotic medication, and went with him to the surgeon's office to get the sutures removed. She worked with boarding home staff to understand the participant's health conditions and taught them strategies to de-escalate uncontrolled behavior and to care for his wound. His housing situation stabilized, and he was not rehospitalized during 90-days of TCare. Communication and Privacy Issues In the community in which our study was conducted, there are thirteen hospitals with psychiatric and drug and alcohol inpatient services. Each hospital had unique forms with their own interpretation of HIPPA regulations and policies concerning patient privacy. Study participants could be admitted to any of these hospitals after their index hospitalization. The TCare-NP attempted to communicate with participants when she learned of their admission to a hospital. However, she was often stopped by front line psychiatric hospital staff who would essentially say, “We can neither confirm nor deny that the person is a patient” and would invoke HIPPA regulations. Although the NP had a signed standard release form that indicated she could speak with the patient's providers, many facilities would not accept it as valid, stating that the patient had to sign their form to talk with her. This was a catch twenty-two as the NP could not contact the patient to ask him/her to sign the release and most were unwilling to give a message to the patient or to ask the patient to sign the needed release form. Repeatedly, she was denied contact with unit staff to explain her role. In sharp contrast for medical hospitalizations, the TCare-NP was
accepted and valued for the contribution she made to treatment. For example, the physician treating the participant with the recent cancer diagnosis was most appreciative of the NP assistance in calming the patient and translating medical information for her and assisting her with the necessary medical appointments. DISCUSSION The primary modifications made to the original TCM was the use of a psychiatric nurse practitioner with a collaborating psychiatrist and broadening the concept of caregiver to better fit the patient population. The TCare-NP was able to assess, diagnose, and treat the integrated physical and mental health needs of patients. She played a key role in translating and sharing information among both physical and mental health sectors when system barriers did not make this impossible. Physical health providers welcomed and valued the TCare-NP and were frequently relieved to have her expertise, time and attention to assist in managing emotional and behavioral responses from patients, as they did not have knowledge or time to address complex unmet social needs of these patients. With her prescribing ability, the TCare-NP was able to write prescriptions for lost discharge prescriptions, ensure that patients were taking newly prescribed medications, substitute medications that insurance would not cover, and write prescriptions that addressed behavioral problems. Furthermore, she was an ambassador for the patient, playing a key role in negotiating with and serving as a bridge among physical and behavioral health providers and systems. She translated information from patient to provider, provider to patient, provider to family/residential support staff, and provider to provider. However, we also recognize that issues of implementation cannot be separated from the intervention per se or the APN delivering the intervention. Since there was only one APN, there is no way to disentangle these two factors. Modifications to TCM: Eligibility Requirements Participants with an active need for medical services were most receptive to TCare, such as the participant who had a recent cancer diagnosis or the one with end-stage respiratory failure. In contrast, a young woman diagnosed with asthma who was having no problems did not perceive the need for medical attention and avoided the TCare-NP. However, there were ineligible patients who may have benefited from the TCare e.g., a woman who was 7 months pregnant with bipolar disorder who remained in the hospital during her pregnancy due to concerns about her inability to care for herself. Indeed, eligibility for TCare might be better assessed by asking the following questions rather than employing a major medical problem as the criterion: Does the patient have an active health need that is not being addressed? If so, can this health need be significantly addressed, managed, and impacted within the 90 days of TCare? Is TCare likely to improve the outcome for the patient? Issues Arising From TCare Implementation Patient Challenges Lack of Receptivity to the Intervention. Greater integration of the TCare-NP into the hospital discharge planning process may have resulted in greater receptivity to TCare. Forchuk et al. (2005) found that a transitional program from hospital to community with a similar population averaged 3 months but could take up to a year for community providers to establish a working relationship with the patient, and this relationship could be enhanced with pairing staff with peers. Utilizing a peer, as did Kelly and colleagues, may have facilitated engagement with TCare-NP.
Please cite this article as: Solomon, P., et al., Lessons Learned from Implementing a Pilot RCT of Transitional Care Model for Individuals with Serious Mental Illness, Archives of Psychiatric Nursing (2014), http://dx.doi.org/10.1016/j.apnu.2014.03.005
P. Solomon et al. / Archives of Psychiatric Nursing xxx (2014) xxx–xxx
Relational Conflicts. Relational conflicts were common and precipitated a higher level of care or lost housing which caused greater instability. The NP could have benefited from working with a social worker, as they are trained to assess patients within their environmental context and assist the client to adapt and manage their environments. Social workers are accustomed to handling conflict-ridden situations and are experts in managing social needs and making referrals to community resources. In some cases, the TCare-NP was able to engage the assistance of the patients' case managers. For example, with one patient, the case manager was initially receptive to working with the TCare-NP, but then became unresponsive. Other case managers were unresponsive to any overtures of TCare-NP. Accordingly, community mental health case managers could not be counted on for assistance. It is likely case managers having relatively heavy caseloads with high need clients are relieved to have someone else servicing their clients. Since much of their time is focused on dealing with crises (Hromco, Moore, & Nikkel, 2003), they likely would triage a client who is receiving services as a low priority compared to those with more pressing needs. A study by Hromco et al. (2003) found that the size of community mental health case managers' caseloads increased between 1992 and 2000, but the responsibilities and tasks were unchanged. The number of clients served by case managers was approaching 40 to 50. Basic Social Needs. Many of the patients had a need for assistance in maintaining stable housing situations. Without stable housing, it is difficult to focus on managing medical conditions. The lack of a stable address also meant that much of the TCare-NPs time was spent in tracking down patients. The TCare-NP did not have expertise in this realm, and her time would have been better spent on medical rather than social issues. Again, this is an area in which a social worker who has the relevant experience and knowledge would have been helpful to the TCare-NP. Furthermore, a peer may have better connections and ability to track down patients. Provider Challenges Poor Communication and Coordination With Other Services. A real challenge for operationalizing TCare for the severe psychiatric population is poor communication between providers and coordination of services. A standardized release of information form that was accepted system-wide would have been helpful. Although electronic health records are becoming more common, they are not necessarily interoperable between facilities making data difficult to access. Had such data been available to support clinical decisions by the TCare-NP, she would have been able to make more informed decisions and be in a better position to locate patients who tended to “game” the system by seeking admission at another hospital that was unaware of their recent hospitalization. Also, if the TCare intervention had been known system-wide with a plan for accessing psychiatric services in multiple hospitals, this may have helped TCare-NP to have access to other facilities. Finally, a consumer-driven communication process such as a psychiatric advance directive (PAD) may have cleared some obstacles to service coordination by documenting the patient's desire to have the NP and other providers included in treatment (Henderson, Jackson, Slade, Young, & Strauss, 2010). This would of course require a system infrastructure that recognizes and supports these PADs. CONCLUSION AND RECOMMENDATIONS Severe psychiatric patients with an acute medical condition and those who acknowledged a medical health need were more receptive to engaging with the TCare-NP. Many patients discharged from psychiatric services do not have urgent health concerns and thus, do not need or want the services of TCare. Making TCare also available on acute hospital medical service may lead to more timely discharge (as these co-morbid physical–behavioral health patients tend to
5
remain longer in inpatient than those without psychiatric conditions, Benzer, Sullivan, Williams, & Burgess, 2012) and to more effective treatment for individuals with behavioral health conditions who have active health issues. TCare may prove to be cost effective by reducing the hospital length of stay. Due to the complex social needs of this population, we recommend that TCare utilize a team approach with a psychiatric advanced practice nurse, a social worker, a peer specialist, and a consulting psychiatrist. While this is a change from the original TCM, it is necessitated by the greater social needs, more housing instability, fewer individuals with committed long-term caregivers and fewer having established relationships with physical or behavioral health care providers as compared to physically ill elderly population. These characteristics may well have been major contributing factors to the lack of effect between conditions in the pilot RCT. Naylor and her colleagues had noted that a team was created between NP, patient, caregiver and primary care physicians, but such a team was often not possible with this population. This adaptation to the model is also likely to be more cost effective, as each team member is functioning in their specific domain of expertise. The APN would not be spending time tracking down patients or working with housing issues. The peer provider could focus efforts on engaging patients in care and finding individuals, and the social worker could assist in accessing needed social services and benefits and helping patients to adapt to their environments. In implementing TCare, efforts need to focus not only at the program level, but also at multiple systems level. We recognize that confidentiality policies were created in part to protect privacy and mitigate some of the stigma surrounding behavioral health conditions. However, these policies can present challenges to the delivery of high quality care. A recent study of 2,000 psychiatric patients found that, in those cases where electronic records were shared with nonpsychiatrists, there were lower rates of readmission within 30 days (Kozubal et al., 2013). In Canada, a system known as the Circle of Care (Cavoukian, 2009) seeks to balance privacy concerns and the need for effective care coordination through the use of implied consent for health care providers. Circle of Care enables all who provide services to a given individual to have access to the same health care information. Thus, there are no barriers to a provider from a program to meeting with a patient when the patient is in another facility. Investigation of this implied consent approach in the United States is warranted. We intend to conduct a follow-up RCT of TCare incorporating the lessons learned from this translational project. Individuals with complex health and behavioral health needs will benefit from more effective integration of the behavioral and physical healthcare systems. Any solution must balance individual privacy concerns and providers' need for timely access to clinical data in order to achieve optimal health outcomes. Also clinical interventions need to be designed with the expressed needs of clients and with their environmental context in mind.
References Benzer, J., Sullivan, J., Williams, S., & Burgess, J. (2012). One-year cost implications of using mental health care after discharge from a general medical hospitalization. Psychiatric Services, 63, 672–678. Cavoukian, A. (2009, Septt). Circle of care: Sharing personal health information for health-care purposes. Ontario, Canada: Information and Privacy Commissioner (www.ipc.on.ca). Colton, C., & Mandersheid, R. (2006). Congruencies in increased mortality rates, years of potential life lost, and causes of death among public mental health clients in eight states. Preventing Chronic Disease, 3, 1–14. Davis, A., Dinitz, S., & Pasamanick, B. (1972). The prevention of hospitalization in schizophrenia: Five years after an experimental program. American Journal of Orthopsychiatry, 42, 375–388. Druss, B., Rohrbaugh, R., Levinson, C., & Rosenheck, R. (2001). Integrated medical care for patients with serious psychiatric illness. Archives of General Psychiatry, 58, 861–868.
Please cite this article as: Solomon, P., et al., Lessons Learned from Implementing a Pilot RCT of Transitional Care Model for Individuals with Serious Mental Illness, Archives of Psychiatric Nursing (2014), http://dx.doi.org/10.1016/j.apnu.2014.03.005
6
P. Solomon et al. / Archives of Psychiatric Nursing xxx (2014) xxx–xxx
Druss, B., Zhao, L., von Esenwein, S., Bona, J., Fricks, L., Jenkins-Tucker, S., et al. (2010). The health and recovery peer (HARP) program: A peer-led intervention to improve medical self-management for persons with serious mental illness. Schizophrenia Research, 118, 264–270. Forchuk, C., Martin, M., Chan, Y., & Jensen, E. (2005). Therapeutic relationships from hospital to community. Journal of Psychiatric and Mental Health Nursing, 12, 556–564. Forchuk, C., Reynolds, W., Sharkey, S., Martin, M. L., & Jensen, E. (2007). The transitional discharge model: Comparing implementation in Canada and Scotland. Journal of Psychosocial Nursing, 45, 31–38. Griswold, K., Homish, G., Pastore, P., & Leonard, K. (2010). A randomized trial: Are care navigators effective in connecting patients to primary care after psychiatric crisis? Community Mental Health Journal, 46, 398–402. Griswold, K., Servoss, T., Leonard, K., Pastore, P., Smith, S., Wagner, et al. (2005). Connections to primary medical care after psychiatric crisis. Journal of American Board of Family Practice, 18, 166–172. Henderson, C., Jackson, C., Slade, M., Young, A., & Strauss, J. (2010). How should we implement psychiatric advance directives? Views of consumers, caregivers, mental health providers and researchers. Administration and Policy in Mental Health, 37, 447–458. Hromco, J., Moore, M., & Nikkel, R. (2003). How managed care has affected mental health case management activities, caseloads, and tenure. Community Mental Health Journal, 39, 501–509. Kelly, E., Fulginiti, A., Pahwa, R., Tallen, L., Duan, L., & Brekke, J. (2014). A pilot test of a peer navigator intervention for improving the health of individuals with serious mental illness. Community Mental Health Journal, http://dx.doi.org/10.1007/ s10597-013-9616-4. Kozubal, D., Samus, Q., Bakare, A., Trecker, C., Wong, H., Guo, H., et al. (2013). Separate may not be equal: A preliminary investigation of clinical correlates of electronic psychiatric record accessibility in academic medical centers. International Journal of Medical Informatics, 82, 260–267.
Lawrence, D., & Kisely, S. (2010). Inequalities in healthcare provision for people with severe mental illness. Psychopharmacology, 23(Supplement 4), 61–68. Lester, H., Tritter, J., & Sorohan, H. (2005). Patients’ and health professionals’ views on primary care for people with serious mental illness: Focus group study. British Medical Journal, 330, 1122–1126. Naylor, M. (2000). A decade of transitional care research with vulnerable elders. Journal of Cardiovascular Nursing, 14, 1–14. Naylor, M., Aiken, L., Kurtzman, E., Olds, D., & Hirschman, K. (2011). The importance of transitional care in achieving health reform. Health Affairs, 30, 746–754. Naylor, M., Bowles, K., McCauley, K., Maccoy, M., Maislin, G., Pauly, M., et al. (2013). High-value transitional care: Translation of research into practice. Journal of Evaluation in Clinical Practice, 19, 727–733. Naylor, M., Brooten, D., Campbell, R., Jacobsen, B., Mezey, M., Pauly, M., et al. (1999). Comprehensive discharge planning and home follow-up of hospitalized elders: A randomized clinical trial. JAMA, 281, 613–620. Piatt, E., Munetz, M., & Ritter, C. (2010). An examination of premature mortality among decedents with serious mental illness and those in the general population. Psychiatric Services, 61, 663–666. Reynolds, W., Lauder, W., Sharkey, S., Maciver, S., Veitch, T., & Cameron, D. (2004). The effects of a transitional discharge mode for psychiatric patients. Journal of Psychiatric and Mental Health Nursing, 11, 82–88. Rose, L., Gerson, L., & Carbo, C. (2007). Transitional care for seriously mentally ill persons: A pilot study. Archives of Psychiatric Nursing, 21, 297–308. Scarpitti, F., Albini, J., Baker, E., Dinitz, S., & Pasamanick, B. (1965). Public health nurses in a community care program for the mentally ill. American Journal of Nursing, 65, 89–95. Woltmann, E., Grogan-Kaylor, A., Perron, B., Georges, H., Kilbourne, A., & Bauer, M. (2012). Comparative effectiveness of collaborative chronic care models for mental health conditions across primary, specialty, and behavioral health care settings: Systematic review and meta-analysis. American Journal of Psychiatry, 169, 790–804.
Please cite this article as: Solomon, P., et al., Lessons Learned from Implementing a Pilot RCT of Transitional Care Model for Individuals with Serious Mental Illness, Archives of Psychiatric Nursing (2014), http://dx.doi.org/10.1016/j.apnu.2014.03.005
Contents lists available at ScienceDirect
Archives of Psychiatric Nursing journal homepage: www.elsevier.com/locate/apnu
Lessons Learned from Implementing a Pilot RCT of Transitional Care Model for Individuals with Serious Mental Illness Phyllis Solomon a,⁎, Nancy P. Hanrahan b, Matthew Hurford c, d, Marissa DeCesaris e, LaKeetra Josey e a
University of Pennsylvania School of Social Policy & Practice, Center for Mental Health Policy and Services Research University of Pennsylvania School of Nursing, Center for Health Outcomes and Policy Research Philadelphia Department of Behavioral Health and Intellectual Disability Service d University of Pennsylvania School of Medicine, Center for Mental Health Policy and Services Research e University of Pennsylvania School of Nursing, Center for Health Outcomes and Policy Research b c
a b s t r a c t We adapted an evidence-based transitional care model for older adults being released from acute care hospitals for patients with serious mental illness and medical co-morbidities being discharged from two psychiatric units of an acute care hospital (TCare) and evaluated implementation issues. An advisory group (AG) of community stakeholders assessed barriers and facilitators of a 90-day T-Care intervention delivered by a psychiatric nurse practitioner (NP) in the context of conducting a pilot randomized controlled trial. Minutes of AG and case narratives by NP of 20 intervention participants were content analyzed. Patients with immediate and pressing physical health problems were most receptive and actively utilized the service. Provider barriers consisted of communication and privacy issues making it difficult to contact patients in mental health facilities. In contrast, the NP was accepted and valued in the physical health arena. Psychosocial needs and relationship issues were demanding, and we recommend a team approach for TCare with the addition of a social worker, peer provider, and consulting psychiatrist for severely mentally ill patients being released from an acute physical health hospitalization. © 2014 Published by Elsevier Inc.
Adults with severe psychiatric disorders have excess medical morbidity and mortality as compared to the general population (Colton & Mandersheid, 2006; Piatt, Munetz, & Ritter, 2010). This psychiatric population dies on average 25 years sooner than others, often due to preventable medical illnesses (Colton & Mandersheid, 2006). Integrated medical and behavioral health care has been found to improve outcomes for this population (Druss, Rohrbaugh, Levinson, & Rosenheck, 2001; Lawrence & Kisely, 2010; Woltmann et al., 2012); yet their ability to access and interact with the medical care system is often compromised due to personal, provider, and system factors which result in inferior quality healthcare and consequently, poor health outcomes. For example, adults with severe psychiatric disorders often face motivational and cognitive challenges to access and communicate their concerns to health providers and to act as their own advocates in the complex medical arena. Furthermore, the mental health providers serving them do not have the medical All authors note no conflict of interest. Funding: Robert Wood Johnson Foundation Interdisciplinary Nursing Quality Research Initiative. ⁎ Corresponding Author: Phyllis Solomon, PhD, Professor, University of Pennsylvania School of Social Policy & Practice, Center for Mental Health Policy and Services Research. E-mail addresses: [email protected] (P. Solomon), [email protected] (N.P. Hanrahan), [email protected], [email protected] (M. Hurford), [email protected] (M. DeCesaris), [email protected] (L. Josey).
knowledge, time, or resources to deal with their physical health problems or to assist them in accessing these needed medical services. Analogously, medical providers are uncomfortable dealing with this population as they have inadequate knowledge of their psychiatric disorders and also have limited time and resources to deal with patients with complex co-morbidities compounded by social needs (Druss et al., 2010; Lawrence & Kisely, 2010; Lester, Tritter, & Sorohan, 2005). To further complicate the situation, system-level factors such as financing policies and privacy requirements create challenges to integrating care across these systems. Care coordination to assist with appropriate connections between systems such as hospitals and community health services is widely accepted as a cost-effective strategy to improve access and care quality (Griswold, Homish, Pastore, & Leonard, 2010; Griswold et al., 2005). The point of discharge from a psychiatric inpatient stay for those with severe psychiatric illnesses is an opportune time to intervene with those with medical co-morbidities to assist them in connecting with medical and behavioral health systems of care and educating them about managing both their mental and physical health conditions. A model that has been shown to be effective, particularly with elderly patients being released from an acute medical hospitalization, is the transitional care model (TCM) (Naylor, 2000; Naylor, Aiken, Kurtzman, Olds, & Hirschman, 2011; Naylor et al., 1999, 2013). This evidenced- based model seemed appropriate to adapt for adults with severe mental illness and a comorbid major medical problem being discharged from an acute
http://dx.doi.org/10.1016/j.apnu.2014.03.005 0883-9417/© 2014 Published by Elsevier Inc.
Please cite this article as: Solomon, P., et al., Lessons Learned from Implementing a Pilot RCT of Transitional Care Model for Individuals with Serious Mental Illness, Archives of Psychiatric Nursing (2014), http://dx.doi.org/10.1016/j.apnu.2014.03.005
2
P. Solomon et al. / Archives of Psychiatric Nursing xxx (2014) xxx–xxx
psychiatric unit by employing a psychiatric advanced practice nurse to deliver the intervention. We conducted a feasibility assessment of the implementation of the adaptation of TCM within the context of a pilot RCT naming it transitional care (TCare) (so as not to conflict with targeted case management also referred to as TCM in the mental health arena). Here we report on the results of the qualitative assessment of the implementation of TCare. The outcomes of the pilot RCT (to be reported in another manuscript) of rehospitalizations, emergency room use of both psychiatric and medical, and appointment keeping did not differ by conditions, and therefore, the need to assess the implementation of the TCare intervention was all the more pressing to understand. BACKGROUND LITERATURE The use of nurses to assist in the coordination of care of psychiatric patients from hospital discharge to community placement has long been considered an effective strategy. Public health nurses working with patients and their families in the beginning of deinstitutionalization in the 1960s were found to effectively help in maintaining patients to remain in their homes (Davis, Dinitz, & Pasamanick, 1972; Scarpitti, Albini, Baker, Dinitz, & Pasamanick, 1965). Forchuk, Reynolds, Sharkey, Martin, and Jensen (2007) developed a transitional discharge model (TDM) which included peer support and inpatient providers (including nurses) to help with the linkages to community services post-discharge for patients being released from a psychiatric hospitalization. They found that it took an average of 3 months for discharged patients to form a relationship with community providers and that the model did enhance social relationships and did reduce length of hospital stay, but did not improve functioning or quality of life or reduce hospitalization or emergency room use (Forchuk, Martin, Chan, & Jensen, 2005). However, when the model was tested employing an RCT in Scotland, the researchers did find TDM patients reported being less symptomatic, having higher levels of functioning, better quality of life, and had fewer hospital readmissions within 5 months of discharge (Reynolds et al., 2004). Recently, a variety of care navigator models to assist in connecting this psychiatric population with medical services have been developed. A randomized pilot study employed peer navigators to teach patients in situ self-advocacy skills within the healthcare system to engage more effectively with the medical system. Preliminary results indicated that those in the experimental intervention which utilized peer navigators changed their views about seeking care from primary health care providers away from emergency room use (Kelly et al., 2014). Another study connected with patients at the point of a psychiatric crisis using care navigators with support of peers to assist patients in linking with primary care and found that after 1 year, access to primary care was greater for the intervention condition than the control (Griswold et al., 2010). TCM AND PILOT STUDY INTERVENTION TCM was designed as a 90 day intervention employing an advanced practice nurse who does a comprehensive discharge plan for chronically ill older adult patients prior to their discharge from the hospital. The APN then sees the patient within 48 hours of discharge in their home with the family caregiver and provides education and support to them regarding managing the patient's illness. The APN engages in outreach, goes to primary care appointments with the patient and is available 7 days per week by phone during the intervention period (Naylor et al., 1999). The essential elements of TCM as delineated by Naylor et al. (2013) are: 1. coordination of care by an APN; 2. a plan developed prior to hospital discharge; 3. home visits by APN for approximately 90 days post-hospital discharge and available 7 days a week; 4. coordination with physicians in community, including accompanying patient on these visits; 5. inclusive focus on health needs of patient; 6. involvement of both patient and family in patient care through educating and supporting them; 7. early
detection and quick “response to health care risks and symptoms”; 8. patient, family caregiver, and providers function as a team; 9. collaboration of nurse and physician; and 10. information sharing with and among all members of the team. TCM has been found to effectively reduce readmissions and lower healthcare costs in three randomized controlled trials and a recent quasi-experimental design within routine care (Naylor, 2000; Naylor et al., 1999, 2011, 2013). Rose, Gerson, and Carbo (2007) conducted a pilot study of TCM for psychiatric patients discharged from hospital to family settings, i.e., study eligibility required that the patient had to reside with a family member. TCM was selected due to its effectiveness with highrisk populations in preventing rehospitalizations and reducing cost. The researchers felt that TCM had the potential to assist patients who are often discharged with symptomatic behavior, have “complex and unpredictable illness course”, are frequently medication non-adherent due to difficulty handling poly-pharmacy and therefore require support and education to prevent hospital readmission (Rose et al., 2007). This study included 10 patients and found that the intervention needed to address the concerns of caregivers in relating to non-adherence issues and potentially violent threats of the patient; assistance in connecting and engaging patients with structured daily activities; help in obtaining and developing strategies to improve adherence with prescribed medication, and to manage symptomatic behaviors. In conclusion, Rose and colleagues found that the advanced practice nurse needed to be “fluid and responsive” and consequently, taking on functions that could have been provided in partnership with lay persons or peers as had been done in the TDM model discussed above. This study showed potential for the model, but required further study. In our pilot study TCare included all of the elements of TCM, but the advanced practice nurse was specialized and had experience in psychiatric nursing as in the pilot by Rose and colleagues. The APN hired had both inpatient and outpatient psychiatric experiences as well as research coordination background, which made her an ideal candidate to deliver this intervention. Coordination in our study was not only with health care providers but also with psychiatric care. Furthermore, a psychiatrist was available for consultation. Also, caregivers were more broadly defined than in Rose's study as often psychiatric patients do not live with family members but in residential care facilities, boarding homes, and homeless shelters. Therefore, the APN worked with others besides families with whom the patient resided, such as residential staff. Education and support was offered to any caregivers in the patient's environment who could assist the patient with managing the illness and offer support for adherence to treatment. TCare was similar in some respects to targeted case management in that it offered assessment, planning, assistance in accessing medical and social services based on the patient's needs, including attending appointments with the patient and monitoring receipt of services. Targeted case management is offered by community mental health agencies to Medicaid eligible clients. However, case managers in the public mental health system often have limited training regarding medical issues and generally focus their efforts on mental health and social service needs of clients. Case managers may have a bachelor's degree or less in any discipline. Thus, their clinical training is minimal, if any. A master's degree, such as in social work, is often required for supervisors of case managers. The psychiatric APN was clinically trained in both medical and psychiatric assessment and treatment as well as had medication prescribing privileges. Thus, she was in a better position to ensuring that other physical and behavioral health providers had accurate clinical information and that patients understood their physical health management needs and she could offer education and support in regard to these issues. In addition, the APN met with patients in the hospital to engage with them prior to discharge, while the usual system is a referral to a mental health agency upon discharge from the hospital, where they are often assigned a case manager. The patients are thus left to their own devices to connect with community mental health services.
Please cite this article as: Solomon, P., et al., Lessons Learned from Implementing a Pilot RCT of Transitional Care Model for Individuals with Serious Mental Illness, Archives of Psychiatric Nursing (2014), http://dx.doi.org/10.1016/j.apnu.2014.03.005
P. Solomon et al. / Archives of Psychiatric Nursing xxx (2014) xxx–xxx
METHODS An advisory group (AG) of key stakeholders met throughout the study to translate the original TCM and to assess barriers and facilitators to TCare implementation. The AG met once a month for about 1 year and included key stakeholders: two mental health consumers; the director of behavioral health programs for the city of Philadelphia; an administrator from the public behavioral health managed care organization who was also a psychiatrist and the recruitment hospital staff including a primary care physician, a psychiatrist, and a social worker; a visiting nurse from the community; and one private insurance representative. University IRB approval was obtained, and all AG and RCT participants were consented. At monthly meetings, the research team provided examples to the AG of barriers and facilitators from the implementation of the TCare RCT. Between meetings issues that arose during the implementation were discussed by the research team for presentation and discussion at the next AG meeting. Minutes of these meetings were taken. Additionally, the psychiatric nurse practitioner who provided TCare wrote detailed case narratives of study participants in the experimental arm of the study. RCT participants were forty adult patients with serious mental illness, i.e., schizophrenia, bipolar, and major affective disorders and a major medical condition who were assigned to TCare or usual care. The 20 case narratives from the experimental arm along with the AG meeting minutes were content analyzed and used to generate the findings described below. These materials were read by three authors (PS, NH, & MH) and discussed to come up with the themes with input of the APN (LJ). Consensus regarding themes was reached by the three research team members, a social worker, a nurse and a psychiatrist, respectively. Thus a qualitative approach was employed. RESULTS Sociodemographics Mean age was 44 years, 55% male, and 45% African American. Forty percent had less than a high school education, were poor (mean income = $717 per month), and single (60%). At baseline, a quarter lived in emergency shelters, halfway homes or had no place to live. Most were unemployed (see Table 1). Complex Health and Behavioral Health Needs On average, participants had 1.6 psychiatric diagnoses and 3.3 comorbid medical diagnoses. Half of the participants had a diagnosis of major depression (50%), a third had a diagnosis of schizoaffective disorder and/or substance use (30% and 30%, respectively), and a quarter of participants had a diagnosis of schizophrenia (25%). The following medical problems were common: endocrine disorders (60% with hypothyroid and/or type II diabetes mellitus); hypertension (40%); respiratory conditions (30%), and hyperlipidemia (20%). Medication regimens were complex, and medication problems were common. Participants took an average of 6 different types of medication each day. Three participants misplaced their hospital prescriptions, and the TCare-NP replaced the prescription until an appointment with a psychiatrist could be made. One participant with schizophrenia did not fill his new prescription and instead used his pre-hospital antipsychotic medication. Within a week he began to relapse with auditory hallucinations and paranoia. The TCare-NP assessed the problem, corrected his dose of medication and his symptoms remitted. Another participant's insurance would not pay for his prescribed medication so the TCare-NP worked with the primary care provider to find a compatible substitute. A participant with schizophrenia became physically aggressive with a boarding
3
Table 1 Characteristics of the Intervention Group (n = 20). Characteristic Mean (SD) 44.1 ± 11.2 $717 ± $458 6.1 ± 3.03
Age Income (total money received previous month) Mean # of daily medications
Gender as observed Male Female Race Black/African American White Asian More than one race Other Hispanic/Latino Education Less than high school High school Post-high school technical training Some college College degree Some graduate study Marital status Single Married or with consistent partner Divorced Separated Widowed Current living situation Home, hotel, or apartment With parents or other family members With friends Emergency shelter Halfway house or board and care No home or regular place to live Employed No Yes ⁎Mental disorder diagnoses Major depression Bipolar disorder Schizoaffective disorder Schizophrenia Psychosis NOS Personality disorder Substance use ⁎Comorbid medical problems Endocrine (diabetes, hypothyroid) Hyperlipidemia Hypertension Cardiovascular Respiratory Gastric-GERD Seizure disorder M/S-arthritis Infectious disease (hepatitis B, C; HIV) Cancer Traumatic brain injury
n
%
11 9
55 45
9 6 0 2 2 1
45 30 5 10 10 5
8 6 0 5 0 1
40 30 0 25 0 5
12 3 2 2 1
60 15 10 10 5
12 3 0 2 1 2
60 15 0 10 5 10
15 3
75 15
10 2 6 5 2 2 6
50 10 30 25 10 10 30
12 4 8 4 6 2 3 3 1 2 1
60 20 40 20 30 10 15 15 5 10% 5%
Data source: Hospital medical record. ⁎ Note: Individuals may meet criteria for more than one diagnostic category thus percentages exceed 100%.
home staff member, and his housing was in jeopardy. Given that he was having difficulty sleeping, the TCare-NP prescribed a few days of a medication that allowed him to sleep. He was less restless and anxious in subsequent days and maintained his residence. Another participant with bipolar and borderline personality disorder was started on lithium while hospitalized. After discharge, the participant frequently missed appointments and ran out of lithium. The TCare-NP managed her lithium medication, monitoring it until she was engaged with her provider.
Please cite this article as: Solomon, P., et al., Lessons Learned from Implementing a Pilot RCT of Transitional Care Model for Individuals with Serious Mental Illness, Archives of Psychiatric Nursing (2014), http://dx.doi.org/10.1016/j.apnu.2014.03.005
4
P. Solomon et al. / Archives of Psychiatric Nursing xxx (2014) xxx–xxx
Intensity of Physical Health Need Participants with immediate and pressing medical problems most actively utilized TCare. For example, a participant with bipolar disorder attended her first visit with an oncologist shortly after discharge, accompanied by the TCare-NP. The participant was highly anxious and had little social support as her fiancé was actively abusing alcohol and crack cocaine. When the oncologist informed her that she had rectal cancer and needed chemotherapy and radiation treatment, she had a severe anxiety attack. The TCare-NP used anxiety-reducing techniques to enable the patient to calm herself and continue to engage with the oncologist. When the participant was calmer, the TCare-NP translated the cancer treatment plan, and subsequently, coordinated multiple medical and psychiatric treatment appointments, medication changes, and radiology appointments. By the end of her TCare intervention, the participant successfully completed chemotherapy and radiation treatment and was connected with her primary care and outpatient mental health providers. The participant was not rehospitalized and did not use emergency services during this time. Another participant with chronic obstructive pulmonary disease had a co-morbid substance use disorder and required close monitoring as he was in terminal stages of his respiratory illness. He was receptive to assistance from TCare. In contrast, a woman with asthma was randomized to receive TCare, however, she was not having problems with her asthma and avoided contact with the TCare-NP. Conflict-Intense Relationships and Unstable Housing The TCare-NP frequently managed relational conflicts and housing problems. Approximately two-thirds of the hospital readmissions were associated with unstable living situations and relational conflicts. A third of participants changed their housing during the 90 days of the TCare intervention. When a participant with schizophrenia who had a toe amputated due to diabetes refused assistance from residential staff, his housing became jeopardized. He was irritable, aggressive, and threatened other residents and staff. The sutures from his surgery remained in place 3 months after the procedure because he was paranoid and refused to see the doctor. The TCare-NP established a relationship, ensured he was taking the prescribed dose of antipsychotic medication, and went with him to the surgeon's office to get the sutures removed. She worked with boarding home staff to understand the participant's health conditions and taught them strategies to de-escalate uncontrolled behavior and to care for his wound. His housing situation stabilized, and he was not rehospitalized during 90-days of TCare. Communication and Privacy Issues In the community in which our study was conducted, there are thirteen hospitals with psychiatric and drug and alcohol inpatient services. Each hospital had unique forms with their own interpretation of HIPPA regulations and policies concerning patient privacy. Study participants could be admitted to any of these hospitals after their index hospitalization. The TCare-NP attempted to communicate with participants when she learned of their admission to a hospital. However, she was often stopped by front line psychiatric hospital staff who would essentially say, “We can neither confirm nor deny that the person is a patient” and would invoke HIPPA regulations. Although the NP had a signed standard release form that indicated she could speak with the patient's providers, many facilities would not accept it as valid, stating that the patient had to sign their form to talk with her. This was a catch twenty-two as the NP could not contact the patient to ask him/her to sign the release and most were unwilling to give a message to the patient or to ask the patient to sign the needed release form. Repeatedly, she was denied contact with unit staff to explain her role. In sharp contrast for medical hospitalizations, the TCare-NP was
accepted and valued for the contribution she made to treatment. For example, the physician treating the participant with the recent cancer diagnosis was most appreciative of the NP assistance in calming the patient and translating medical information for her and assisting her with the necessary medical appointments. DISCUSSION The primary modifications made to the original TCM was the use of a psychiatric nurse practitioner with a collaborating psychiatrist and broadening the concept of caregiver to better fit the patient population. The TCare-NP was able to assess, diagnose, and treat the integrated physical and mental health needs of patients. She played a key role in translating and sharing information among both physical and mental health sectors when system barriers did not make this impossible. Physical health providers welcomed and valued the TCare-NP and were frequently relieved to have her expertise, time and attention to assist in managing emotional and behavioral responses from patients, as they did not have knowledge or time to address complex unmet social needs of these patients. With her prescribing ability, the TCare-NP was able to write prescriptions for lost discharge prescriptions, ensure that patients were taking newly prescribed medications, substitute medications that insurance would not cover, and write prescriptions that addressed behavioral problems. Furthermore, she was an ambassador for the patient, playing a key role in negotiating with and serving as a bridge among physical and behavioral health providers and systems. She translated information from patient to provider, provider to patient, provider to family/residential support staff, and provider to provider. However, we also recognize that issues of implementation cannot be separated from the intervention per se or the APN delivering the intervention. Since there was only one APN, there is no way to disentangle these two factors. Modifications to TCM: Eligibility Requirements Participants with an active need for medical services were most receptive to TCare, such as the participant who had a recent cancer diagnosis or the one with end-stage respiratory failure. In contrast, a young woman diagnosed with asthma who was having no problems did not perceive the need for medical attention and avoided the TCare-NP. However, there were ineligible patients who may have benefited from the TCare e.g., a woman who was 7 months pregnant with bipolar disorder who remained in the hospital during her pregnancy due to concerns about her inability to care for herself. Indeed, eligibility for TCare might be better assessed by asking the following questions rather than employing a major medical problem as the criterion: Does the patient have an active health need that is not being addressed? If so, can this health need be significantly addressed, managed, and impacted within the 90 days of TCare? Is TCare likely to improve the outcome for the patient? Issues Arising From TCare Implementation Patient Challenges Lack of Receptivity to the Intervention. Greater integration of the TCare-NP into the hospital discharge planning process may have resulted in greater receptivity to TCare. Forchuk et al. (2005) found that a transitional program from hospital to community with a similar population averaged 3 months but could take up to a year for community providers to establish a working relationship with the patient, and this relationship could be enhanced with pairing staff with peers. Utilizing a peer, as did Kelly and colleagues, may have facilitated engagement with TCare-NP.
Please cite this article as: Solomon, P., et al., Lessons Learned from Implementing a Pilot RCT of Transitional Care Model for Individuals with Serious Mental Illness, Archives of Psychiatric Nursing (2014), http://dx.doi.org/10.1016/j.apnu.2014.03.005
P. Solomon et al. / Archives of Psychiatric Nursing xxx (2014) xxx–xxx
Relational Conflicts. Relational conflicts were common and precipitated a higher level of care or lost housing which caused greater instability. The NP could have benefited from working with a social worker, as they are trained to assess patients within their environmental context and assist the client to adapt and manage their environments. Social workers are accustomed to handling conflict-ridden situations and are experts in managing social needs and making referrals to community resources. In some cases, the TCare-NP was able to engage the assistance of the patients' case managers. For example, with one patient, the case manager was initially receptive to working with the TCare-NP, but then became unresponsive. Other case managers were unresponsive to any overtures of TCare-NP. Accordingly, community mental health case managers could not be counted on for assistance. It is likely case managers having relatively heavy caseloads with high need clients are relieved to have someone else servicing their clients. Since much of their time is focused on dealing with crises (Hromco, Moore, & Nikkel, 2003), they likely would triage a client who is receiving services as a low priority compared to those with more pressing needs. A study by Hromco et al. (2003) found that the size of community mental health case managers' caseloads increased between 1992 and 2000, but the responsibilities and tasks were unchanged. The number of clients served by case managers was approaching 40 to 50. Basic Social Needs. Many of the patients had a need for assistance in maintaining stable housing situations. Without stable housing, it is difficult to focus on managing medical conditions. The lack of a stable address also meant that much of the TCare-NPs time was spent in tracking down patients. The TCare-NP did not have expertise in this realm, and her time would have been better spent on medical rather than social issues. Again, this is an area in which a social worker who has the relevant experience and knowledge would have been helpful to the TCare-NP. Furthermore, a peer may have better connections and ability to track down patients. Provider Challenges Poor Communication and Coordination With Other Services. A real challenge for operationalizing TCare for the severe psychiatric population is poor communication between providers and coordination of services. A standardized release of information form that was accepted system-wide would have been helpful. Although electronic health records are becoming more common, they are not necessarily interoperable between facilities making data difficult to access. Had such data been available to support clinical decisions by the TCare-NP, she would have been able to make more informed decisions and be in a better position to locate patients who tended to “game” the system by seeking admission at another hospital that was unaware of their recent hospitalization. Also, if the TCare intervention had been known system-wide with a plan for accessing psychiatric services in multiple hospitals, this may have helped TCare-NP to have access to other facilities. Finally, a consumer-driven communication process such as a psychiatric advance directive (PAD) may have cleared some obstacles to service coordination by documenting the patient's desire to have the NP and other providers included in treatment (Henderson, Jackson, Slade, Young, & Strauss, 2010). This would of course require a system infrastructure that recognizes and supports these PADs. CONCLUSION AND RECOMMENDATIONS Severe psychiatric patients with an acute medical condition and those who acknowledged a medical health need were more receptive to engaging with the TCare-NP. Many patients discharged from psychiatric services do not have urgent health concerns and thus, do not need or want the services of TCare. Making TCare also available on acute hospital medical service may lead to more timely discharge (as these co-morbid physical–behavioral health patients tend to
5
remain longer in inpatient than those without psychiatric conditions, Benzer, Sullivan, Williams, & Burgess, 2012) and to more effective treatment for individuals with behavioral health conditions who have active health issues. TCare may prove to be cost effective by reducing the hospital length of stay. Due to the complex social needs of this population, we recommend that TCare utilize a team approach with a psychiatric advanced practice nurse, a social worker, a peer specialist, and a consulting psychiatrist. While this is a change from the original TCM, it is necessitated by the greater social needs, more housing instability, fewer individuals with committed long-term caregivers and fewer having established relationships with physical or behavioral health care providers as compared to physically ill elderly population. These characteristics may well have been major contributing factors to the lack of effect between conditions in the pilot RCT. Naylor and her colleagues had noted that a team was created between NP, patient, caregiver and primary care physicians, but such a team was often not possible with this population. This adaptation to the model is also likely to be more cost effective, as each team member is functioning in their specific domain of expertise. The APN would not be spending time tracking down patients or working with housing issues. The peer provider could focus efforts on engaging patients in care and finding individuals, and the social worker could assist in accessing needed social services and benefits and helping patients to adapt to their environments. In implementing TCare, efforts need to focus not only at the program level, but also at multiple systems level. We recognize that confidentiality policies were created in part to protect privacy and mitigate some of the stigma surrounding behavioral health conditions. However, these policies can present challenges to the delivery of high quality care. A recent study of 2,000 psychiatric patients found that, in those cases where electronic records were shared with nonpsychiatrists, there were lower rates of readmission within 30 days (Kozubal et al., 2013). In Canada, a system known as the Circle of Care (Cavoukian, 2009) seeks to balance privacy concerns and the need for effective care coordination through the use of implied consent for health care providers. Circle of Care enables all who provide services to a given individual to have access to the same health care information. Thus, there are no barriers to a provider from a program to meeting with a patient when the patient is in another facility. Investigation of this implied consent approach in the United States is warranted. We intend to conduct a follow-up RCT of TCare incorporating the lessons learned from this translational project. Individuals with complex health and behavioral health needs will benefit from more effective integration of the behavioral and physical healthcare systems. Any solution must balance individual privacy concerns and providers' need for timely access to clinical data in order to achieve optimal health outcomes. Also clinical interventions need to be designed with the expressed needs of clients and with their environmental context in mind.
References Benzer, J., Sullivan, J., Williams, S., & Burgess, J. (2012). One-year cost implications of using mental health care after discharge from a general medical hospitalization. Psychiatric Services, 63, 672–678. Cavoukian, A. (2009, Septt). Circle of care: Sharing personal health information for health-care purposes. Ontario, Canada: Information and Privacy Commissioner (www.ipc.on.ca). Colton, C., & Mandersheid, R. (2006). Congruencies in increased mortality rates, years of potential life lost, and causes of death among public mental health clients in eight states. Preventing Chronic Disease, 3, 1–14. Davis, A., Dinitz, S., & Pasamanick, B. (1972). The prevention of hospitalization in schizophrenia: Five years after an experimental program. American Journal of Orthopsychiatry, 42, 375–388. Druss, B., Rohrbaugh, R., Levinson, C., & Rosenheck, R. (2001). Integrated medical care for patients with serious psychiatric illness. Archives of General Psychiatry, 58, 861–868.
Please cite this article as: Solomon, P., et al., Lessons Learned from Implementing a Pilot RCT of Transitional Care Model for Individuals with Serious Mental Illness, Archives of Psychiatric Nursing (2014), http://dx.doi.org/10.1016/j.apnu.2014.03.005
6
P. Solomon et al. / Archives of Psychiatric Nursing xxx (2014) xxx–xxx
Druss, B., Zhao, L., von Esenwein, S., Bona, J., Fricks, L., Jenkins-Tucker, S., et al. (2010). The health and recovery peer (HARP) program: A peer-led intervention to improve medical self-management for persons with serious mental illness. Schizophrenia Research, 118, 264–270. Forchuk, C., Martin, M., Chan, Y., & Jensen, E. (2005). Therapeutic relationships from hospital to community. Journal of Psychiatric and Mental Health Nursing, 12, 556–564. Forchuk, C., Reynolds, W., Sharkey, S., Martin, M. L., & Jensen, E. (2007). The transitional discharge model: Comparing implementation in Canada and Scotland. Journal of Psychosocial Nursing, 45, 31–38. Griswold, K., Homish, G., Pastore, P., & Leonard, K. (2010). A randomized trial: Are care navigators effective in connecting patients to primary care after psychiatric crisis? Community Mental Health Journal, 46, 398–402. Griswold, K., Servoss, T., Leonard, K., Pastore, P., Smith, S., Wagner, et al. (2005). Connections to primary medical care after psychiatric crisis. Journal of American Board of Family Practice, 18, 166–172. Henderson, C., Jackson, C., Slade, M., Young, A., & Strauss, J. (2010). How should we implement psychiatric advance directives? Views of consumers, caregivers, mental health providers and researchers. Administration and Policy in Mental Health, 37, 447–458. Hromco, J., Moore, M., & Nikkel, R. (2003). How managed care has affected mental health case management activities, caseloads, and tenure. Community Mental Health Journal, 39, 501–509. Kelly, E., Fulginiti, A., Pahwa, R., Tallen, L., Duan, L., & Brekke, J. (2014). A pilot test of a peer navigator intervention for improving the health of individuals with serious mental illness. Community Mental Health Journal, http://dx.doi.org/10.1007/ s10597-013-9616-4. Kozubal, D., Samus, Q., Bakare, A., Trecker, C., Wong, H., Guo, H., et al. (2013). Separate may not be equal: A preliminary investigation of clinical correlates of electronic psychiatric record accessibility in academic medical centers. International Journal of Medical Informatics, 82, 260–267.
Lawrence, D., & Kisely, S. (2010). Inequalities in healthcare provision for people with severe mental illness. Psychopharmacology, 23(Supplement 4), 61–68. Lester, H., Tritter, J., & Sorohan, H. (2005). Patients’ and health professionals’ views on primary care for people with serious mental illness: Focus group study. British Medical Journal, 330, 1122–1126. Naylor, M. (2000). A decade of transitional care research with vulnerable elders. Journal of Cardiovascular Nursing, 14, 1–14. Naylor, M., Aiken, L., Kurtzman, E., Olds, D., & Hirschman, K. (2011). The importance of transitional care in achieving health reform. Health Affairs, 30, 746–754. Naylor, M., Bowles, K., McCauley, K., Maccoy, M., Maislin, G., Pauly, M., et al. (2013). High-value transitional care: Translation of research into practice. Journal of Evaluation in Clinical Practice, 19, 727–733. Naylor, M., Brooten, D., Campbell, R., Jacobsen, B., Mezey, M., Pauly, M., et al. (1999). Comprehensive discharge planning and home follow-up of hospitalized elders: A randomized clinical trial. JAMA, 281, 613–620. Piatt, E., Munetz, M., & Ritter, C. (2010). An examination of premature mortality among decedents with serious mental illness and those in the general population. Psychiatric Services, 61, 663–666. Reynolds, W., Lauder, W., Sharkey, S., Maciver, S., Veitch, T., & Cameron, D. (2004). The effects of a transitional discharge mode for psychiatric patients. Journal of Psychiatric and Mental Health Nursing, 11, 82–88. Rose, L., Gerson, L., & Carbo, C. (2007). Transitional care for seriously mentally ill persons: A pilot study. Archives of Psychiatric Nursing, 21, 297–308. Scarpitti, F., Albini, J., Baker, E., Dinitz, S., & Pasamanick, B. (1965). Public health nurses in a community care program for the mentally ill. American Journal of Nursing, 65, 89–95. Woltmann, E., Grogan-Kaylor, A., Perron, B., Georges, H., Kilbourne, A., & Bauer, M. (2012). Comparative effectiveness of collaborative chronic care models for mental health conditions across primary, specialty, and behavioral health care settings: Systematic review and meta-analysis. American Journal of Psychiatry, 169, 790–804.
Please cite this article as: Solomon, P., et al., Lessons Learned from Implementing a Pilot RCT of Transitional Care Model for Individuals with Serious Mental Illness, Archives of Psychiatric Nursing (2014), http://dx.doi.org/10.1016/j.apnu.2014.03.005
