Original Clinical ScienceçGeneral

Posttransplantation Outcomes in Veterans With Serious Mental Illness Lianna D. Evans, PsyD,1,2 Eileen M. Stock, PhD,1,2,3 John E. Zeber, PhD,1,2,3 Sandra B. Morissette, PhD,1,2,4 Andrea A. MacCarthy, BS,5 Edward Y. Sako, MD,6 Jacqueline Lappin, MD,2,7 Valerie A. Lawrence, MD,6 Daniel J. MacCarthy, BS,6 and Laurel A. Copeland, PhD1,2,3 Background. Anticipating poor recovery due to impaired self-management and appointment-keeping, clinicians may consider

serious mental illness (SMI) a significant concern in organ transplantation. However, little empirical evidence exists regarding posttransplantation outcomes for patients with SMI. Methods. This study analyzed health services data to evaluate posttransplantation 3-year survival by SMI status in a nationwide cohort of patients in the Veterans Health Administration (VHA). Results. A total of 960 recipients of solid organ or bone marrow transplants were identified from Veterans Health Administration administrative data extracts for fiscal years 2006 to 2009. Of these, 164 (17%) had an SMI diagnosis before transplantation (schizophrenia, posttraumatic stress, major depressive, and bipolar disorders); 301 (31%) had some other mental illness diagnosis (such as anxiety, adjustment reactions, or substance abuse); and 495 (52%) had no mental health diagnosis. Twenty-two patients (2%) required retransplantation and 208 patients (22%) died during follow-up. Data on whether these were primary or repeat transplantations were unavailable. Rates of attendance at postoperative outpatient visits and number of months for which immunosuppressive drugs fills were recorded were similar among mental illness groups, as were rates of diagnosed immunological rejection. Threeyear mortality was equivalent among mental health groups: no mental health (19%) versus other mental illness (23%) versus SMI (27%; χ2 = 5.11; df = 2; P = .08). In adjusted survival models, no effect of mental health status was observed. Conclusions. Serious mental illness diagnosis does not appear to be associated with adverse transplantation outcomes over the first 3 years; however, a potentially diverging survival curve may portend higher mortality at 5 years. (Transplantation 2015;99: e57–e65)


he National Organ Transplant Act of 1984 (revised and implemented in 1988) established the Organ Procurement and Transplantation Network (OPTN) and Scientific Registry of Transplant Recipients in the United States1 to streamline availability and coordination of donated organs. The OPTN is the only national patient waitlist, linking all professionals involved in the transplantation system with the aims of increasing the efficiency and equity of organ sharing.

From its creation in 1988 through 2012, the OPTN reported a total of 532,591 adult solid organ transplants; with approximately 10% reported as repeat transplants and 1.5% as multiple transplants. Within the Veterans Health Administration (VHA), a member of OPTN, approximately 1800 veterans undergo transplant evaluation annually and approximately 450 undergo transplantation at an approved VHA Transplant Center.2 Patients undergoing organ transplant face high risks for complications related to surgical procedures and

Received 24 March 2014. Revision requested 4 April 2014. Accepted 21 November 2014. 1

Central Texas Veterans Health Care System, Department of Veterans Affairs, Temple, TX.


Texas A&M Health Science Center, College of Medicine, Temple, TX.


Center for Applied Health Research, Central Texas Veterans Health Care System jointly with Baylor Scott & White Health, Temple, TX.


Department of Veterans Affairs VISN 17 Center of Excellence for Research on Returning War Veterans, Waco, TX.


South Texas Veterans Health Care System, Department of Veterans Affairs, San Antonio, TX.


UT Health Science Center San Antonio, San Antonio TX.


Baylor Scott & White Health, Temple TX.

L.D.E. was supported by the Office of Academic Affiliations, Advanced Fellowship Program in Mental Illness Research and Treatment. This research was supported by the Veterans Health Administration, Health Services Research & Development grant IIR-09-335 (Copeland-PI), with additional support from the Center for Applied Health Research jointly sponsored by Central Texas Veterans Health Care System, Scott & White Healthcare, and the VA VISN 17 Center of Excellence for Research on Returning War Veterans.


August 2015

Volume 99

Number 8

V.A.L. is a perioperative clinician, works at UT Health Science Center San Antonio and was previously employed in the Veterans Health Administration. J.L. is a transplantation surgeon at Baylor Scott & White Health. L.D.E. participated as principal in research design and writing of the manuscript. L.A.C. participated in research design, writing and revising the manuscript, and data analysis interpretation. E.M.S. participated in writing of the manuscript and as principal in data analysis. J.E.Z. participated in research design and writing of the manuscript. A.A.M.C. developed variables, conducted analyses, and provided preliminary interpretations. E.Y.S. and V.A.L. provided input on how transplantation surgeons conduct and describe research on transplantation. D.J.M. C. participated in data analysis and variable development. J.L. provided input on how transplantation surgeons conduct and describe research on transplantation. S.B.M. participated in research design and writing of the manuscript. Correspondence: Lianna D. Evans, PsyD, Central Texas Veterans Health Care System 4800 Memorial Drive Waco, TX 76711. ([email protected]). Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved. ISSN: 0041-1337/15/9908-e57 DOI: 10.1097/TP.0000000000000616


Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.




August 2015

Volume 99

Number 8

posttransplantation recovery. Because each transplant requires some level of immunosuppressive aftercare to prevent rejection and promote successful grafting, it is essential for transplant candidates to maintain medication adherence, participate in follow-up care, and have a supportive environment in which to recuperate. Given the limited availability of organs and the associated costs and wait times for transplantation, it is critical that the transplant evaluation team identify those who would benefit most and suffer fewest posttransplant complications, while balancing considerations of equity for disadvantaged patients. To illustrate the great need and extent of the supply-and-demand problem, we note that among 117,391 current adult waitlisted transplant candidates, 14% (N = 16,435) have been waiting for more than 5 years with a pool of only 5355 donors available from January to August 2013.1 Mental health factors are prominent when conceptualizing pretransplantation surgical candidacy3 but seem to be understudied for some types of transplants and some types of mental disorders. One study on pretransplant mental health in 179 kidney recipients identified 28% with pre-existing depression, 8% with comorbid depression/anxiety, 3% with anxiety alone, 2% with depression/posttraumatic stress disorder (PTSD) or “other” mental health problems, and 2% with mental health symptoms without a corresponding diagnosis (such as “anger disorder”).4 Of 2 studies evaluating liver transplant candidates/recipients, the first (N = 73) identified 17% with symptoms of depression and 33% with symptoms of anxiety during pretransplantation assessment5; whereas the second (N = 71) identified 66% with a current or past psychiatric illness, though none with psychotic spectrum disorders, such as schizophrenia and schizoaffective disorders.6 A study evaluating heart transplant recipients indicated that “depression” (all forms) correlated positively with heart transplantation and with likelihood of posttransplant intervention needs.7 In our review of the literature, no information on pre-existing mental illness rates was found for intestinal, lung, bone marrow, or pancreas transplants, although much was available about posttransplant mental health symptom development (regardless of transplant type). Available literature suggests that the transplant patient community, at large, is at high risk for the development of psychiatric symptoms due to psychosocial stressors (eg, finances, work limitations), sequelae of long-term chronic illness, fear of transplant evaluation outcomes, and general health declines.3 Transplant evaluations have established guidelines for candidacy selection based on patients' current and historical medical conditions, treatment adherence, psychological status, substance use, social support, and medical team relationships.8-10 In practice, serious mental illness (SMI) is commonly considered a significant surgical concern, including active psychosis and certain symptoms, such as untreated severe depression, anxiety, or somatic complaints.8-13 Indeed, chronic conditions, such as schizophrenia, have been associated with worsened posttransplantation outcomes including death or complicated graft failures.3,14 Notably, there are only 4 published case studies in the area of SMI and organ transplantation in the past 20 years,15-19 with no information on base rates of survival by any mental health conditions (SMI or other). To date, no studies have examined the impact of SMI on transplant outcomes among military veterans receiving care


in the nation's largest integrated health care system, the VHA. Moreover, previous studies have been unable to speak broadly about the impact of SMI diagnoses on organ transplant outcomes due to lack of diagnostic capture and insufficient focus on pretransplant conditions in relation to a wide range of organ types. The current study evaluated 1- and 3-year posttransplant survival rates across major transplant categories among veterans enrolled in the VHA, among those with and without pretransplantation mental health diagnoses. The data relied on administrative data extracted from the electronic medical record for health services and outcomes analysis and do not include many of the clinical measures commonly collected on transplant patients. However, these data provide a unique overview of mental illness visà-vis transplantation in the VHA. Understanding transplant survival factors can inform the evaluation and follow-up process for patients with mental health diagnoses and potentially improve outcomes among mental health patients undergoing transplantation.

METHODS Sample and Data Sources

Veterans enrolled in the VHA during fiscal years 2006 to 2009 (October 2005-September 2009; N = 8.1 million) were assessed for inclusion in the current study. Data sources included medical record data sets stored at the VHA central data repository in Austin, TX, which are updated biweekly with transmissions from all VHA medical centers. Uniformly defined variables provide the same types of information on patients at every facility system wide, including demographic characteristics, types of care received, and diagnoses.20,21 Within the VHA, SMI is defined as “diagnoses that result in significant functional impairment and/or disruption in major activities in daily living,” including schizophrenia and other psychotic disorders, bipolar disorder, major depressive disorder (MDD), and severe PTSD.”22 The current study aggregated administrative data as part of the Surgical Treatment Outcomes for Patients with Psychiatric (STOPP) project in veterans with and without SMI and was further narrowed to the subset of VHA patients (defined by a valid VHA priority score [1–8]) who received a solid organ transplant per VHA hospital and Veteran Affairs (VA)paid non-VA hospital surgery/procedure files (STOPP Transplant cohort). The STOPP transplant recipients' first qualifying transplant procedure, or index surgery, during FY2006–2009 was grouped into the following: heart (ICD-9-A and CPT-4 heart transplant codes 37.5; 33945);2 lung (single/double lung codes 33.50–33.52; 32851–32852, 32854)3; liver (liver codes 50.51, 50.59; 47135–47136)4; liver with small intestine (intestine codes 46.97; 44135–44136)5; kidney (kidney codes 55.61, 55.69; 50360, 50365)6; liver with kidney (from the above); and7 bone marrow (autologous/allogeneic codes 41.02–41.03, 41.05, 41.08; 38240–38241). To provide greater rigor in survival outcome analyses, transplant procedures with the same codes occurring within ± 3 days of the index surgery were considered a single transplant and not transplant failure, retransplantation, or “multiple” transplant. Patients receiving “other” transplant (ICD-9-A codes 00.91– 00.93) without an associated specific transplant procedure

Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

Evans et al

© 2015 Wolters Kluwer

were excluded (N = 14). The final STOPP transplant cohort included 960 veterans identified with one of the described index surgeries. Demographic Measures

Sociodemographic variables were aggregated from administrative data to create indicators of ethnicity (Hispanic vs. other status), race (black, white, other, unknown), marital status (married vs other status). Disease and treatment-related variables included obesity diagnosis (yes/no) within 1 year before transplantation, age at transplantation, transplant year, and smoking history (assessed by ICD-9 diagnosis codes [305.1, V15.82], clinic visits relating to tobacco treatment, and use of nicotine replacement or abatement drugs). Age, sex, and VHA priority status were based on enrollment data. The VHA priority status has 8 categories to describe veteran eligibility for VHA services and co-pay status based on servicerelated disability, other catastrophic disability, annual income relative to the poverty threshold, and military service experiences, such as being a prisoner of war. The VHA priority status correlates with both socioeconomic status and severity of illness.23 Priority 1 designation denotes at least 50% disability from a military service-connected condition with no copays for associated treatment or medications; priorities 2 to 6 includes veterans with up to 40% service-connected disability incurring medication co-pays; priorities 7 to 8 veterans incur co-pays for both treatment and medications. As veterans with priority 1 status have no co-pays for treatments or medications, their access-to-care does not have the barrier of cost associated with transplant procedures or perioperative care. Diagnostic Measures

Diagnosis on 2 or more outpatient dates in the year before the qualifying organ transplant procedure identified preoperative schizophrenia (ICD-9 codes 295.x, excluding 295.5), bipolar disorder (296.0–296.1, 296.4–296.8), MDD (296.2x–296.3x, 311), and PTSD (309.81),24,25 the 4 types of SMI per VHA Directive 2012-002, “Re-engaging Veterans with Serious Mental Illness in Treatment” [http:// www1.va.gov/vhapublications/ViewPublication.asp?pub_ID=2476]. One inpatient diagnosis of schizophrenia or bipolar disorder was also considered sufficient to categorize a patient.26 Patients with any SMI diagnosis documented in outpatient care only once in the preceding year or with any other mental health diagnosis in the range of 290 to 311 (excluding substance use disorders) were categorized as having other (nonSMI) mental illnesses for use in multivariate models. Selected non-SMI mental health conditions were operationalized (substance use disorders [SUD] [291.xx, 292.xx, 303.xx–305.xx, excluding 305.1], smoking history previously described) defined by at least 1 outpatient diagnosis. For descriptive purposes anxiety (300.00, 300.02, 300.09), adjustment disorders (309.xx excluding 309.81) were also identified. Patients not meeting the previous diagnostic criterion for SMI or other mental disorder were considered to have no mental health disorders in the previous year. Patients were then categorized by previous-year mental health status as1: any SMI,2 other mental illness including SUD, and3 no mental illness. To understand the unique contribution of SUD, we also created indicators of other mental illness (OMI)-no SUD and of SUD alone for use in the multivariable models.


Comorbidity Risk Adjustment Measures

Summary measures of comorbidity were created using diagnosis codes from inpatient and outpatient records in the year before the qualifying organ transplant procedure for casemix adjustment. These included the Charlson comorbidity score comprised of 19 weighted indicators for conditions associated with 1-year posthospitalization mortality rates27-29 and the Selim comorbidity score of chronic conditions30 as operationalized in VHA administrative data. Selim's physical comorbidity score (range, 0–30) sums indicators of 30 chronic medical conditions to assess burden of chronic disease,30 whereas the Charlson comorbidity index (range, 0–37) sums weighted indicators of 19 conditions related to 1-year postdischarge mortality, including metastatic cancer (weighted 6), diabetes without complications (weighted 1), diabetes with complications (weighted 2), etc.27-29 These case-mix adjusters can be used together as they capture different aspects of disease burden. Correlation analysis of these summary measures produced a correlation of 0.46 (21% shared variance). Graft Outcomes

Variables to identify postoperative complications were defined to assess graft-versus-host disease and complications of transplanted organ: “graft problems” per ICD-9 codes 279.50–279.53 and 996.8x; “graft rejection” defined by inpatient or outpatient diagnosis of E87.80, 996.8, or V45.87; and infection diagnosed on an inpatient basis (ICD-9 codes 996–999). Posttransplantation Compliance

Compliance with posttransplantation care was approximated by fills for antirejection medication over the year after transplantation (number of months on antirejection drugs during the first year after transplantation) and by outpatient clinic visits (average time to first outpatient visit, average time to first outpatient visit for transplantation follow-up, number of post-transplantation months with outpatient visit, and days until first diagnosis of graft follow-up visit). Maintenance and induction medications included: corticosteroids (methylprednisolone, prednisolone, prednisone, hydrocortisone), calcineurin inhibitors (cyclosporine, tacrolimus [Prograf, FK506]); antiproliferatives (azathioprine, mycophenolic acid, mycophenolate mofetil [Cellcept]), mammalian target of rapamycin (mTOR) inhibitors (sirolimus [Rapamune], everolimus), antibody drugs, including monoclonal anti-IL-2Rα receptor antibodies (basiliximab, daclizumab), monoclonal and polyclonal anti–T-cell antibodies (muromonab-CD3 antithymocyte globulin [ATG]; thymoglobulin [rabbit ATG; Genzyme], Atgam [equine ATG], antilymphocyte globulin), and monoclonal antiCD20 antibodies (rituximab). “Successful graft” postoperative outpatient visits were indicated by diagnosis with ICD-9 code V58.44, V42.x excluding V42.9 or by a visit to a clinic identified as a Transplant Clinic. Regular follow-up care was defined as the number of months in the first 2 years after transplantation with at least 1 outpatient visit. Transplantation failure was determined and defined as the same transplantation procedure repeated at least 4 days after index transplantation (i.e., retransplantation of same organ or graft failures of organ tissue). Posttransplantation Mortality

Three-year mortality was assessed from the index transplantation date. Death was ascertained by the Mini-Vital

Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.



August 2015

Volume 99

Number 8

Status files per a validated algorithm from four sources: inpatient stays, veterans' death benefits claims through the Beneficiary Identification Records Locator Subsystem, Social Security Administration, and Medicare death files. The latter 2 sources have a known lag of 2 years between when a death occurs and when it is added to the database; therefore, we did not have 5-year death data for patients entering the study later than 2007. The Mini-Vital algorithm has been validated against the National Death Index with a sensitivity of about 98%.31 Data Analysis

Means and frequencies were calculated for patient characteristics and treatment measures by previous-year mental health status (SMI, OMI including SUD, no mental health [NONE]). Bivariate analyses assessed underlying differences by mental illness group. The χ2 analysis compared categorical characteristics and mental illness group. Analysis of variance was used for continuous measures with approximately normal distributions; otherwise, the nonparametric Kruskal-Wallis test assessed differences between patient groups. Multivariable survival analyses applied a Cox proportional hazards regression model to examine the association between mental health status and mortality, adjusting for transplant type (heart, lung, liver, kidney, kidney with liver, small intestine with liver, and bone marrow) in addition to clinical and demographic characteristics. In the model, age was divided by 10 to assess a decade effect. A type I error of α = 0.05 was used for all tests. All analyses were performed using SAS, Version 9.2. RESULTS Veterans who received an organ transplant in the VHA between FY2006 and FY2009 (October 2005–September 2009; N = 960) were identified from data developed in the STOPP disorders study. The STOPP study was funded to examine surgery experiences and outcomes of veterans with SMI using administrative data derived from the electronic medical records system. Patients diagnosed with mental illness in the year before transplantation were compared to those without mental illness; all data came from the VHA.32 Descriptive statistics of patient characteristics and transplantation outcomes, by mental health status, are reported in Table 1. Transplant recipients had a mean age of 56 years (SD = 9; range, 21–79), were primarily white (67%) or black (23%) with 5% other races and 5% unknown race, 11% were of Hispanic ethnicity (assessed independently of race). The most common type of transplant was kidney (396; 41%), followed by liver (344; 36%), bone marrow (138; 14%), and heart (49; 5%); 7 patients (1%) had combined kidney and liver transplantation; and 1 patient had combined liver and small intestine transplantation. The SMI diagnoses were present in 164 veterans (17%). The OMI disorders (e.g., minor depression or anxiety, SUD) were present in 301 veterans (31%). Those diagnosed with SUD (37%) most often had another non-SUD mental health condition as well (93%). No patients were diagnosed with drug overdose, either intentional or undetermined as to intent, within the year before transplantation; data were not available before September 2004, so earlier events may have occurred. Patients without any mental health diagnosis (henceforth “NONE”) before transplantation comprised 52% of the sample (N = 495).


History of tobacco use characterized 15% of the sample. A total of 19 (2%) patients required retransplantation, and 208 (22%) patients died during the first 3 years after transplantation. Characteristics by Mental Health Status

Mental health groups did not differ significantly by age. Race varied across mental health groups, with blacks comprising a smaller proportion of transplant recipients with any mental illness relative to non–mental health patients (SMI, 17%; OMI, 17%; NONE, 28%; χ2 = 22.49; df = 6; P < 0.01). Nearly half (47%) of patients with SMI were VHA priority 1 status (no co-pays/high disability; defined in Methods) compared to 27% OMI and 38% NONE (chi-square = 20.71; df = 2; P < 0.01). Higher rates of liver transplants were observed in the SMI and OMI groups than in the no mental illness group (52% SMI, 45% OMI, 25% NONE; χ2 = 86.51; df = 12; P < 0.01) consistent with the known association between mental illness and alcohol or other drug abuse. Veterans with no mental health diagnosis more frequently received a kidney transplant (53% vs 24% SMI, 32% OMI; P < 0.01). Characteristics by Transplant Type

The age distribution of veterans was similar across transplant types, although patients undergoing bone marrow transplantation were slightly younger (mean age: 53 vs 56–57 years for other transplant groups; P < 0.01). Kidney transplants were more frequent among black patients (64% of black patients vs 35% white, 30% other race, 35% unknown race patients; χ2 = 95.61; df = 18; P < 0.01). There were fewer black patients receiving a liver transplant than expected by chance: 17% black versus 41% white, 38% other race, 48% unknown race (χ2 = 95.61; df = 18; P < 0.01). Bone marrow transplantation was also more common among white or other race than black patients (15% white patients received bone marrow vs 9% black, 30% other races, 13% unknown race; χ2 = 95.61; df = 18; P < 0.01). Heart transplant recipients (5% of the sample) included a greater proportion of veterans with MDD (22% vs 6%–17% for other transplant groups; χ2 = 26.86; df = 6; P < 0.01). Kidney plus liver transplant recipients had greater rates of PTSD (29%) than expected; other organ groups had PTSD rates of 12% (lung), 9% (liver), 8% (heart), and 4% (bone marrow and kidney; χ2 = 18.43; df = 6; P < 0.01). Patients with SUD were common among liver recipients (37%) and kidney plus liver (14%) but were absent from the heart and lung groups (χ2 = 171.23; df = 6; P < 0.01). Graft Outcomes and Follow-up Care

Graft problems were diagnosed in a follow-up outpatient visit in 25% of transplant patients in the first year, and 1 person had immediate posttransplantation kidney graft loss but survived. Immunological graft rejection (per ICD-9 code E87.80, 996.8, or V45.87) was diagnosed in a posttransplantation outpatient visit or hospitalization for 21% of patients within 30 days, 45% within 1 year, 49% within 2 years (47% of 1-year survivors), and 52% within 3 years (51% of 2-year survivors). Graft rejection diagnosis did not differ by mental health status. Rates of attendance at postoperative outpatient visits and number of months for which immunosuppressive drug fills were recorded were similar among mental illness groups, as were rates of diagnosed organ rejection. Specific

Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

Evans et al

© 2015 Wolters Kluwer



Descriptive statistics of transplant recipients by smi group (n = 960) Patient characteristics

Demographics Female Age mean (SD) Median (min-max) Racea White African-American Asian/other Missing Hispanica Missing Marrieda Missing Census Region Midwest Northeast South West Puerto Rico/Virgin Islands VA priority 1 status Obese Smoking history Transplant clinic visit in the previous year Transplant clinic visit in first postoperative year Comorbidity measures Charlson mean (SD) Median (min-max) Selim physical mean (SD) Median (min-max) Days to first postoperative outpatient visit, mean (SD) Median (min-max) Count of postoperative months (of 24) with outpatient visit, mean (SD) Median (min-max) Count of postoperative months (of 12) on immunosuppressive drugs, mean (SD) Median (min-max) Mental health disorders Schizophrenia Bipolar PTSD MDD SUD Anxiety Adjustment Transplant measures Bone marrow Heart Kidney Liver Liver with kidney Liver with small intestine Lung

Serious mental illness (164, 17%) N (%)

Other mental illness (301, 31%) N (%)

No mental illness (495, 52%) N (%)

3 (1.8) 56.0 (7.3) 58.0 (27‐78)

3 (1.3) 55.7 (8.6) 57.0 (22‐79)

17 (3.0) 56.0 (10.9) 58.0 (21‐78)

120 (73.2) 30 (18.3) 7 (4.3) 7 (4.3) 23 (14.0) 7 (4.3) 96 (58.5) 0 (0.0)

221 (73.4) 51 (16.9) 18 (6.0) 11 (3.7) 34 (11.3) 11 (3.7) 149 (49.5) 0 (0.0)

298 (60.2) 138 (27.9) 25 (5.1) 34 (6.9) 44 (8.9) 34 (6.9) 301 (60.8) 8 (1.6)

31 (18.9) 32 (19.5) 51 (31.1) 47 (28.7) 3 (1.8) 77 (47.0) 25 (15.2) 37 (22.6) 102 (62.2) 131 (79.9)

71 (23.6) 49 (16.3) 91 (30.2) 81 (26.9) 9 (3.0) 80 (26.6) 42 (14.0) 90 (29.9) 177 (58.8) 237 (78.7)

88 (17.8) 52 (10.5) 217 (43.8) 120 (24.2) 18 (3.6) 186 (37.6) 65 (13.1) 20 (4.0) 305 (61.6) 407 (82.2)

5.1 (2.4) 5.0 (0‐14) 5.0 (1.9) 5.0 (0‐10) 22.8 (49.3)

4.8 (2.4) 5.0 (0‐14) 4.5 (2.0) 4.0 (0‐11) 15.5 (30.9)

4.1 (2.1) 4.0 (0‐13) 4.1 (1.9) 4.0 (1‐11) 15.8 (36.4)

8.5 (0‐331) 14.4 (8.6)

9.0 (0‐274) 14.8 (8.7)

7.0 (0‐357) 14.8 (8.3)

Posttransplantation Outcomes in Veterans With Serious Mental Illness.

Anticipating poor recovery due to impaired self-management and appointment-keeping, clinicians may consider serious mental illness (SMI) a significant...
308KB Sizes 0 Downloads 8 Views