Abstract
Background
Few treatments are available to directly address employment or work functioning among individuals with bipolar disorder (BD) and currently available treatment models have not been evaluated to examine their impact employment outcomes. We examined impact of affective symptoms and health-related quality of life (HRQoL) on longitudinal employment outcomes in a community-based sample of individuals with bipolar disorder who completed the Life Goals-Collaborative Care (LG-CC) intervention.
Methods
Participants (N=178) were assessed based on HRQoL, employment status, affective symptoms (depressive/manic), and work hours at baseline, 6-, 12- and 24-months after initiation of LG-CC. Frequency of LG-CC sessions and number of care-manager contacts also were ascertained.
Results
At baseline, 21% were employed, 29.5% were unemployed, and 49.6% were on disability. Improvement in affective symptoms was seen over the 24-month period, but not in HRQoL. Lower depression symptoms, but not mania, at baseline predicted greater likelihood of employment status in 24-months. Degree of LG-CC participation was associated with a reduced likelihood of becoming disabled/unemployed and increased number of hours worked in 24-months.
Limitations
The study was originally designed to compare implementation strategies and not the effectiveness of LG-CC on employment outcomes. Further, it was unclear whether improvement in work functioning were personal goals of the participants of this study.
Conclusions
Fewer depressive symptoms were associated with positive employment outcomes over time. Collaborative Care Models that are already implemented by existing providers that focus on management of affective symptoms show promise in positively impacting employment outcomes.
Keywords: Bipolar Disorder, Employment, Longitudinal, Outcomes, Affective Symptoms, Care Models
Introduction
Bipolar disorder (BD) is a serious mental illness that profoundly affects functioning. It is the 6th leading cause of disability world-wide in ages 18-44 year olds (1). Medication and psychosocial treatments have shown benefits for outcomes including reducing hospitalizations and reoccurrence of mood episodes. Despite these improved outcomes, functional impairment remains high for many people experiencing BD (2-4). Hence, BD is associated with substantial work impairment (e.g., reduced work productivity, poor performance) (5) and instability in the job market.
Nonetheless, few treatments are available to address employment or work functioning among individuals with BD. Further, few longitudinal studies examining change in employment status have been conducted in clinical or community samples. Specific employment interventions for persons with serious mental illness, including supported employment, can help individuals compete for competitive employment and increase hours worked (6). However, many individuals with serious mental illness do not have access to employment services (7, 8), and supported employment may not be readily available for under-resourced community-based practices. Understanding how currently available treatment models can impact employment outcomes for persons with serious mental illness can help providers better tailor treatment options for this group.
Collaborative care models (CCMs) are evidence-based practices implemented by existing providers that have shown to improve medical and psychiatric outcomes for persons with chronic mental disorders (9). Central to CCMs is self-management and care management that focus on helping individuals manage their symptoms as a means of maximizing quality of life and functioning. Life Goals Collaborative Care (LG-CC), a particular type of CCM developed for individuals with BD, has shown in several randomized controlled trials to improve reduce affective symptoms and improve quality of life in community-based practices (10). While CCMs such as LG-CC do not specifically focus on employment, it might be assumed that improvements in symptoms will have an effect on employment outcomes, but this has not been formally examined.
This study examines the longitudinal course of employment outcomes during and after a 12-month LG-CC program, and after 24-months post initiation of LG-CC. We hypothesize that those who show improved symptoms and improved HRQoL will show improved work outcomes (increases in employment rates, increased number of hours worked) after a 12-month follow up period. We further hypothesize that employment outcomes will change as mood symptoms and HRQoL improves.
Methods
Participants and Procedure
This study involves longitudinal data analyses from the Recovery-Oriented Collaborative Care (ROCC) study. Details regarding the ROCC study design are described elsewhere (11). In brief, ROCC compared two different implementation strategies (standard vs. enhanced) designed to promote the use of LG-CC among 384 participants receiving care from seven community-based mental health or primary care clinics sites (12). Because no significant changes were found between the implementation strategies in changes in mood symptoms, HRQoL, or functioning, the data from both implementation arms were combined for this study. A subset of participants who were employed, unemployed, and disabled were included in this analysis (N=353). Those who self-identified as being employed or self-employed were included in the employed group. Number of hours worked per week was collected. Those who were unemployed or “unable to work,” but not on currently disability were included in the unemployed group. Those who identified as being “disabled” were included in the disabled group. Those who were retired (n=8), students (n=9), or considered themselves as homemakers (n=12) at baseline were excluded from the analyses.
Participants completed outcome measures at baseline, 6-, 12- and 24-month post initiation of the LG-CC. A demographics survey ascertained information (age, education, race, etc.) along with current employment status, disability status, and number of hours worked. Outcome measures included health-related quality of life (HRQoL; SF-12) and psychiatric symptoms (Internal State Scale-ISS, PHQ-9, see (13)). LG-CC participation, which is total number of LG-CC group sessions attended plus the number of care manager contacts completed during the 12 months by the LG-CC participant, was obtained.
Statistical Analyses
We compared baseline to 24-month outcome variables using two-tailed paired sample t tests. Linear mixed-effects regression models were conducted to examine the association of predictors with the number of hours worked (continuous variable); whereas generalized linear mixed effect models with logit link function were used to examine the association of predictors with categorical work status variables (employed vs. unemployed or disabled). These models were used to account for the clustering effects of multiple measurements of the outcome variables from the same individual on different follow ups, as well as to allow for time-dependent covariates. Covariates including age, education, HRQoL, psychiatric symptoms, and LG-CC participation (13) were included in the models. To assess if employment status changes were mediated by changes in mood symptoms, depressive symptom scores were included and excluded from the multivariable work status model. SAS 9.3 (Cary, NC, USA) was used for all the statistical analyses of this study.
Results
Those in the disabled group were older (M=44.2, SD=10.5) than the employed group (M=38.0, SD=11.3), but no different from the unemployed group (M=39.9, SD=10.7). The disabled group were less likely to be college educated (N=22; 12.6% college educated) compared to the employed group (N=16, 21.9%) (Table 1). There was a difference between employment status (employed vs unemployed) when comparing baseline employment status to employment status at 24-months follow up, χ2(1)=6.12, p=.0214. Specifically, a large proportion of those unemployed at baseline (N=478) were employed at 24-months (N=16). For the individuals employed at baseline (N=40), there were 5 who were unemployed at 24-months follow up. Further, most of the individuals who were employed at baseline (N=34) maintained employment at 24-month follow up (N=29). The majority of those who were disabled at baseline (N=99) continued to remain on disability at 24-month follow-up (N=85).
Table 1. Demographic and clinical characteristics of sample, Results of Random Effect Linear Mixed Regression.
| Employed | Baseline Unemployed | Disabled | Employed | 6 month Unemployed | Disabled | Employed | 12 month Unemployed | Disabled | Employed | 24 month Unemployed | Disabled | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Count (N/%) | 74 (21) | 104 (29.5) | 175 (49.6) | 62 (25.2) | 63 (25.6) | 121 (49.2) | 60 (27) | 44 (19.8) | 118 (53.2) | 56 (28.1) | 28 (14.1) | 115 (57.8) |
|
| ||||||||||||
| Clinical Characteristics | Mean (SD) | Mean (SD) | Mean (SD) | Mean (SD) | Mean (SD) | Mean (SD) | Mean (SD) | Mean (SD) | ||||
|
| ||||||||||||
| PHQ-9 a | 12.1 (6.2) | 13.3 (6.7) | 13.6 (6.3) | 9.3 (7.0) | 11.6 (6.7) | 10.7 (6.5) | 9.5 (6.7) | 11.1 (6.9) | 11.1 (6.5) | 9.4 (6.9) | 7.6 (6.4) | 10.9 (6.5) |
| ISS b | 20.9 (13.0) | 21.4 (12.9) | 20.6 (12.5) | 16.9 (10.8) | 17.7 (11.7) | 16.7 (12.1) | 18.1 (13.1) | 17.7 (10.9) | 17.0 (12.4) | 14.7 (12.4) | 17.8 (14.5) | 17.1 (12) |
| PCSc | 38.4 (6.8) | 38.3 (6.6) | 34.6 (7.7)e* | 39.4 (7.3) | 38.4 (6.3) | 34.8 (8.2) a | 38.9 (7.3) | 38.6 (8.5) | 34.5 (8.0) a | 37.9 (7.2) | 37.5 (8.2) | 35.2 (7.4) |
| MCSc | 32.6 (8.7) | 31.2 (8.4) | 31.8 (8.4) | 33.0 (8.0) | 32.3 (8.6) | 32.7 (8.2) | 33.9 (6.8) | 34.3 (9.0) | 34 (8.3) | 34.0 (8.3) | 36.4 (7.6) | 33.2 (7.4) |
| N. hours workedd | 24.6 (13.0) | 24.6 (12.7) | 27.5 (14.3) | 28.7 (13.1) | ||||||||
| Mixed Effect Regression Models | Employedf vs Unemployed Vs. Disabled Parameters estimates (standard errors) | Number Hours worked Parameters estimates (standard errors) | ||||||
|---|---|---|---|---|---|---|---|---|
|
| ||||||||
| Beta | 95% CI | T (df=268) | P | Beta | 95% CI | T | P | |
| Age | 0.00594 | -0.02, 0.03 | 0.48 | 0.6346 | -0.1960 | -0.47, 0.08 | -1.41 | 0.1644 |
| 1Unemployed | ||||||||
| 2Disabled | 0.07715 | 0.05, 0.11 | 5.29 | <.0001 | ||||
| Female | -0.5062 | -6.51, 5.50 | -0.17 | 0.8694 | ||||
| 1Unemployed | -0.2061 | -0.74, 0.32 | -0.76 | 0.4475 | ||||
| 2Disabled | 0.4760 | -0.15, 1.10 | 1.49 | 0.1387 | ||||
| White | -0.2751 | -0.86, 0.31 | -0.92 | 0.3572 | 7.2498 | 0.56, 13.94 | 2.12 | 0.0384 |
| 1Unemployed | ||||||||
| 2Disabled | -0.0897 | -0.76, 0.58 | -0.26 | 0.7928 | ||||
| College educated | 4.9480 | -1.67, 11.57 | 1.46 | 0.1490 | ||||
| 1Unemployed | -0.1703 | -0.79, 0.45 | -0.54 | 0.5902 | ||||
| 2Disabled | -1.3963 | -2.16, -0.64 | -3.60 | 0.0004 | ||||
| PHQ-9a | 0.05467 | 0.01, 0.10 | 2.24 | 0.0262 | -0.3564 | -0.79, 0.09 | -1.58 | 0.1204 |
| 1Unemployed | 0.03180 | -0.02, 0.08 | 1.27 | 0.2040 | ||||
| 2Disabled | ||||||||
| ISSb | -0.00034 | -0.02, 0.02 | -0.03 | 0.9731 | 0.07182 | -0.12, 0.26 | 0.75 | 0.4575 |
| 1Unemployed | ||||||||
| 2Disabled | -0.01393 | 0.04, 0.01 | -1.27 | 0.2047 | ||||
| MCSc | 0.01719 | -0.02, 0.05 | 0.95 | 0.3442 | -0.1250 | -0.45, 0.19 | -0.76 | 0.4493 |
| 1Unemployed | ||||||||
| 2Disabled | -0.02833 | -0.06, 0.01 | -1.55 | 0.1218 | ||||
| PCSc | 0.01148 | -0.02, 0.05 | 0.63 | 0.5279 | -0.0258 | -0.41, 0.36 | -0.13 | 0.8951 |
| 1Unemployed | ||||||||
| 2Disabled | -0.05420 | -0.09, -0.02 | -2.87 | 0.0045 | ||||
| LG-CC Sessionsg | -0.1051 | -0.20, -0.01 | -2.11 | 0.0354 | -1.4948 | -2.68, -0.31 | -2.47 | 0.0169* |
| 1Unemployed | ||||||||
| 2Disabled | -0.1394 | -0.25, -0.03 | -2.43 | 0.0157 | ||||
PHQ-9=Patient Health Questionnaire-9; a higher score reflects more symptoms related to depression
ISS=Internal State Scale; scores range from 0 to 50 and a higher score indicates worse symptoms related to mania
Health-related quality of life (SF-12) includes a mental component scale (MCS) and physical component scale (PCS). Possible scores range from 0 to 100, with higher scores indicating better health. For both summary scores, the population Mean ± SD is 50 ± 10
Number of hours worked per week
Disabled>Employed
Employed group is reference group
Total of number of LG-CC sessions attended and number of care manager contacts completed by LG-CC participant
A generalized linear mixed model showed that the likelihood of being unemployed (compared to employed) decreased over time (beta=-0.0373, OR=0.96, p=0.0044), but there was no effect of time on the likelihood of becoming disabled (compared to employed, p=0.9371). Time was not significantly associated with number of hours worked per week (p=0.0717).
After adjusting for demographic variables (age, education, race), a generalized linear mixed effect model showed that PHQ-9 was significantly associated with employment status during the follow-up period (when considering baseline, 6-, 12-, and 24-month employment status). A higher PHQ-9 score was significantly related to a higher likelihood of being unemployed (OR=1.06, p=0.0262), but PHQ-9 was not significantly associated with becoming disabled vs. employed (p=0.2040) or disabled vs. unemployed (p=0.3929).
Increased LG-CC participation was significantly associated with reduced likelihood of either becoming disabled or unemployed (ref: employed, respectively OR=0.87, and 0.90, p=0.0157, and 0.0354) after adjustment for demographic variables (age, education, gender). LG-CC sessions continued to be significantly associated with reduced likelihood becoming disabled or unemployed (ref: employed: respective ORs=0.88, and 0.91, p=0.0215, and 0.0614) when adding the effects of depressive symptoms. Adjusting for demographic variables, results also revealed that LG-CC session participation was associated with increased number of hours worked across follow-up periods among those who held employment during the follow-up periods (p=0.0384, p=0.0099, respectively).
Discussion
We investigated correlates of employment status and number of hours worked after initiation of a well-established LG-CC program. We found that employment status improved over time, and reduction of total PHQ-9 severity scores (indicating a reduction in depression-related symptoms) was associated with improvement in employment status. However, total PHQ-9 severity scores were not associated with changes in disability status during the follow-up period. Greater number of LG-CC sessions completed was also associated with greater likelihood of employment status and number of hours worked over the 24-month period.
CCMs, such as LG-CC, are designed to emphasize patient self-management through patient education and ongoing coordination of medical and mental health by current practitioners. Research on LG-CC's impact on outcomes traditionally focus on improvements in quality of life, symptoms, and health, but to date have not focused on other aspects of daily functioning that may provide a more global perspective on functional recovery, such as employment status. These findings, based on a large longitudinal sample of individuals with bipolar disorder, suggest that LG-CC can impact both symptom burden and employment status (employed vs unemployed) over time, thus highlighting the importance of focusing on affective symptoms throughout the course of interventions. However, improvements in total PHQ-9 severity scores did not relate to disability status, suggesting that changes in depression-related symptoms may only impact more transient employment status during a 24 month period rather than disability status, which is considered more a permanent or unmalleable status. Increased LG-CC attendance improved number of hours worked per week, suggesting that participants who were more engaged with the treatment provider and more focused on their wellness goals, were able to attend work more frequently.
This study has limitations that may affect overall generalizability. First, the study was designed to compare implementation strategies and not the effectiveness of LG-CC versus other treatments, so there was no comparison group to assess the differential effect of LG-CC on employment outcomes. Lastly, it was unclear whether improvement in work functioning were personal goals of the participants of this study and may be a driving force to increases in employment rates over the course of LG-CC.
Conclusions
Overall, improvement in affective symptoms can positively affect employment outcomes, and increased number of LG-CC sessions has a positive influence on number of hours worked per week. These results suggest that LG-CC is a promising treatment model to influence more global indicators of functioning in bipolar disorder, beyond self-reports of affective symptoms and quality of life. Future investigations into the LG-CC program could capture overall level of functioning within the work environment to determine if more hours worked per week translates into better employment outcomes or overall quality of work functioning. Implementation of LG-CC, or similar CCMs, may hold promise in helping individuals with bipolar or other serious mental disorders improve mood symptoms which can have positive effects on competitive employment, increasing number of hours worked, and ultimately may reduce the cost of disability and lost wages.
Highlights.
Improvements in affective symptoms over the 24-month period predicted likelihood of employment status.
Changes in health related quality of life did not predict employment status at 24-month follow up.
Participation in the collaborative care model (frequency of sessions and number of care manager contacts) reduced likelihood of becoming disabled or unemployed and increased number of hours worked.
The Life Goals Collaborative Care model is a promising treatment model to influence more global indicators of functioning in bipolar disorder, beyond self-reports of affective symptoms and quality of life
Acknowledgments
Research reported in this publication was supported by the National Center for Advancing Translational Sciences of the National Institutes of Health under Award Number 2KL2TR000434, the National Institute of Mental Health, the University of Michigan Comprehensive Depression Center (Director's Innovation Fund), the Heinz C. Prechter Bipolar Research Fund at the University of Michigan Depression Center and the Richard Tam Foundation. We would like to express appreciation for the contributions of partners from the following agencies: (from Colorado) Behavioral Healthcare, Inc.; Colorado Access; Community Reach Center; Mental Health Centers of Denver; and University of Colorado School of Medicine Depression Center, Departments of Psychiatry and Family Medicine; (from Michigan) Community Support and Treatment Services; Genesee Health System; Packard Health; University of Michigan Departments of Psychiatry, School of Public Health, and Depression Center; and Washtenaw Community Health Organization; (Others) Centers for Disease Control and Prevention; Harvard Medical School; National Institute of Mental Health; and Veterans Health Administration. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or the Department of Veterans Affairs.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Murray CJL, Lopez AD. Global mortality, disability, and the contribution of risk factors: Global burden of disease study. The Lancet. 1997;349(9063):1436–42. doi: 10.1016/S0140-6736(96)07495-8. [DOI] [PubMed] [Google Scholar]
- 2.Coryell W, Scheftner W, Keller M, Endicott J, Maser J, Klerman GL. The enduring psychosocial consequences of mania and depression. The American Journal of Psychiatry. 1993;150(5):720–7. doi: 10.1176/ajp.150.5.720. [DOI] [PubMed] [Google Scholar]
- 3.Goldberg JF, Harrow M, Grossman LS. Course and outcome in bipolar affective disorder: A longitudinal follow-up study. The American Journal of Psychiatry. 1995;152(3):379–84. doi: 10.1176/ajp.152.3.379. [DOI] [PubMed] [Google Scholar]
- 4.Gitlin MJ, Swendsen J, Heller TL, Hammen C. Relapse and impairment in bipolar disorder. Am J Psychiatry. 1995;152(11):1635–40. doi: 10.1176/ajp.152.11.1635. [DOI] [PubMed] [Google Scholar]
- 5.Sanchez-Moreno J, Martinez-Aran A, Tabares-Seisdedos R, Torrent C, Vieta E, Ayuso-Mateos JL. Functioning and disability in bipolar disorder: An extensive review. Psychotherapy and psychosomatics. 2009;78(5):285–97. doi: 10.1159/000228249. [DOI] [PubMed] [Google Scholar]
- 6.Cook JA, Leff HS, Blyler CR, Gold PB, Goldberg RW, Mueser KT, et al. Results of a multisite randomized trial of supported employment interventions for individuals with severe mental illness. Arch Gen Psychiatry. 2005;62(5):505–12. doi: 10.1001/archpsyc.62.5.505. [DOI] [PubMed] [Google Scholar]
- 7.Lehman AF, Steinwachs DM. Translating research into practice: the Schizophrenia Patient Outcomes Research Team (PORT) treatment recommendations. Schizophrenia bulletin. 1998;24(1):1–10. doi: 10.1093/oxfordjournals.schbul.a033302. [DOI] [PubMed] [Google Scholar]
- 8.Banks SM, Simon MM, Pandiani JA. Employment services and employment outcomes fo radults iwht serious mental illness. Research on Employment for Persons with Severe Mental Illness. :83–105. [Google Scholar]
- 9.Woltmann E, Grogan-Kaylor A, Perron B, Georges H, Kilbourne AM, Bauer MS. Comparative effectiveness of collaborative chronic care models for mental health conditions across primary, specialty, and behavioral health care settings: systematic review and meta-analysis. Am J Psychiatry. 2012;169(8):790–804. doi: 10.1176/appi.ajp.2012.11111616. [DOI] [PubMed] [Google Scholar]
- 10.Bauer MS, McBride L, Williford WO, Glick H, Kinosian B, Altshuler L, et al. Collaborative care for bipolar disorder: Part II. Impact on clinical outcome, function, and costs. Psychiatr Serv. 2006;57(7):937–45. doi: 10.1176/ps.2006.57.7.937. [DOI] [PubMed] [Google Scholar]
- 11.Waxmonsky J, Kilbourne AM, Goodrich DE, Nord KM, Lai Z, Laird C, et al. Enhanced fidelity to treatment for bipolar disorder: results from a randomized controlled implementation trial. Psychiatr Serv. 2014;65(1):81–90. doi: 10.1176/appi.ps.201300039. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Kilbourne AM, Post EP, Nossek A, Sonel E, Drill LJ, Cooley S, et al. Service delivery in older patients with bipolar disorder: a review and development of a medical care model. Bipolar Disord. 2008;10(6):672–83. doi: 10.1111/j.1399-5618.2008.00602.x. [DOI] [PubMed] [Google Scholar]
- 13.Kilbourne AM, Goodrich DE, Nord KM, Van Poppelen C, Kyle J, Bauer MS, et al. Long-Term Clinical Outcomes from a Randomized Controlled Trial of Two Implementation Strategies to Promote Collaborative Care Attendance in Community Practices. Administration and policy in mental health. 2014 doi: 10.1007/s10488-014-0598-5. [DOI] [PMC free article] [PubMed] [Google Scholar]