Correlates of Emergency Department Use among Individuals with Bipolar Disorder enrolled in a Collaborative Care Implementation Study

Abstract

Objectives

The study assessed correlates of emergency department use among individuals with bipolar disorder receiving a collaborative care program.

Methods

Community-based clinics from two states implemented the evidence-based Life Goals-Collaborative Care Model (LG-CC), which included self-management sessions and care management contacts. Logistic regression analysis was used to determine participant factors associated with 12-month emergency department use after LG-CC implementation.

Results

Among 219 participants with non-missing baseline and 12-month data, 24.2% (n=53) reported at least one emergency department visit. Participants with homelessness history (OR=3.76; 95% CI=1.39–10.2; p=.01) or 5 or more care management contacts (OR=2.62; 95% CI=.1.00–6.95; p=.05) had higher probability of emergency department visit, adjusting for demographic and clinical factors including physical health and hospitalization history.

Conclusions

Individuals with bipolar disorder and history of homelessness were more likely to use the emergency department despite enrollment in a Collaborative Care program, suggesting a greater need for more intensive care coordination.

INTRODUCTION

It is well-documented that persons with serious mental illness (e.g., schizophrenia, bipolar disorder), have substantial health care costs and poor outcomes (1–2). Collaborative care models (CCMs) have been shown to improve medical and psychiatric outcomes for persons with serious mental illness (3). CCMs are typically implemented by a clinical nurse or social worker (care manager) who provides individual self-management support, coordinates care between providers (care management), completes outcomes assessments, and links patients to specialists in collaboration with the individual’s physician.

CCMs have not been widely implemented for persons with serious mental illness due to limited research on their value, especially in regards to their impact on preventable critical service encounters such as emergency room use. Cost-effectiveness of CCMs may not be realized in the short-term (<1 year) (4). Moreover, CCMs might need to be further enhanced for persons with serious mental illness, especially given the propensity for these individuals to experience greater clinical severity and lower socio-economic status compared to the general population (1). Bipolar disorder, associated with wide variations in functioning and substantial individual and societal health care costs (5), is an appropriate index condition in which to better understand the key drivers of critical service encounters. This study assessed factors associated with increased emergency department use among individuals with bipolar disorder from community-based practices who received a collaborative care model program in a multisite implementation study.

METHODS

This study involved a secondary outcome analysis of the Recovery-oriented Collaborative Care Study (ROCC), a cluster-randomized controlled implementation trial from 2009–2014 involving five community-based clinics from Michigan or Colorado that was designed to improve the uptake of a CCM for bipolar disorder, Life Goals-Collaborative Care model (LG-CC). The study received approvals from local institutional review boards. Details regarding ROCC are available elsewhere (6–7). In brief, the study compared implementation strategies provided at sites which offered LG-CC to patients diagnosed with bipolar disorder and who were able and willing to provide informed consent to complete assessments. LG-CC is an evidence-based CCM (9) that included four weekly self-management sessions focused on managing depressive and bipolar symptoms that occurred during the first month, and ongoing telephone-based care management contacts occurring up to 6 months after patient enrollment that were implemented by clinical social workers at each site. Providers from participating clinics received an LG-CC toolkit and training, but were randomized at the clinic level to receive either additional coaching via proactive, regular contacts to address barriers to LG-CC uptake (10) or as-needed technical support.

Prior to attending LG-CC sessions, participants gave informed consent and completed baseline, 6, and 12-month assessments. The primary outcome was self-reported emergency department visits ascertained between 6 and 12 months. As many of the clinics lacked comprehensive medical record data, self-reported data were used in order to capture consistent information across all sites.

Factors thought to influence emergency department visits were also ascertained from the survey and included in the analyses as covariates, including age, gender, race/ethnicity (white, non-white), education (high school or more versus less than high school), employment (unemployed versus employed/retired), recent homelessness history (past-month-yes, no), depression symptoms based on the Patient Health Questionnaire-9 (PHQ-9) (11), anxiety symptoms (Generalized Anxiety Disorder symptom scale, >=10 cutpoint indicates moderate anxiety symptoms) (11), alcohol use (Alcohol Use Disorders Identification Test 3-item survey score) (12), illicit substance use, current smoking status, prior hospitalization history, number of self-reported medical conditions, and overall physical health, assessed based on the Short Form-12 (SF-12) physical health component score (13). For the PHQ-9, a cutpoint score of 10 or more indicated clinically significant depressive symptoms (11). Finally, number of LG-CC care management contacts was ascertained from provider record reviews. Participants were categorized as having >=5 care management contacts vs. less based on previous analyses describing variations in LG-CC completion during the first 6 months (6).

Descriptive statistics were calculated for baseline characteristics and the outcome variable. Multiple logistic regression analysis (SAS 9.3) was used to determine participant factors associated with probability of an emergency department visit between 6–12 months after the end of the LG-CC intervention period, adjusting for the aforementioned demographic and clinical variables from the survey Sensitivity analyses were conducted to determine if additional organizational factors such as clinic type (primary or mental health care) influenced probability of emergency department use, resulting in virtually identical results.

RESULTS

Of the 385 study participants with complete baseline surveys, 139 did not complete the 12-month survey. Of those who completed both surveys, 27 had either missing outcome or covariate data. The final analytical sample included 219 participants (57% retention), excluding 166 participants. The 166 participants excluded from the analysis were not significantly different from participants in the analytical sample (N=219) in characteristics surveyed at baseline with the exception that those with a homelessness history had a higher odds to drop from the study (OR=2.16; 95% CI=1.30–3.59; p=.003). Baseline homelessness history has been included as a covariate in the analyses.

Overall 24.2% (N=53) reported at least one emergency department visit during 6–12 months. At baseline, the mean age in years of the analytical sample was 42.6±11.0, 32.4% (n=71) were male, 72.1% (n=158) were white, 88.1% (n=193) had a high school education or less, 73.1% (n=160) were unemployed, 46.6% (n=102) were current smokers, 15.1% (n=33) had a recent history of homelessness, 38.8% (n=85) had PHQ-9 score >=10, 84.5% (n=185) had moderate anxiety symptoms, 29.2% (n=64) reported substance abuse at baseline, and 24.2% (n=53) had a history of recent hospitalization at baseline. The mean physical health component score (PCS-12) from the Short Form-12 health-related quality of life assessment was 36.3±7.3, the mean AUDIT-C score was 1.2±2.1 and the mean number of comorbidities was 1.6±1.2.

After adjusting for participant demographic/clinical factors and site, participants with a recent history of homelessness (OR=3.76; 95% CI=1.39–10.2; p=.01) or >=5 care management contacts (OR=2.62; 95% CI= 1.00–6.95; p=.05) had a higher probability of an emergency department visit in 6–12 months; see Table 1. Additionally, patients with a higher PCS-12 score had lower probability of an emergency department visit (OR=.94; 95% CI= .89–1.00; p=.05).

Table 1.

Probability of Emergency Room Use among Patients with Bipolar Disorder from Community Practices Receiving the Life Goals Collaborative Care Model: Multiple Logistic Regression Analysis

Independent Variables	OR	95% CI	P-value
Age (years)	.99	.95–1.03	.70
Male (vs. Female)	.84	.36–1.94	.69
White (vs. Non-White)	2.39	.96–5.92	.06
>=High school educ. (vs. less than HS)	.95	.31–2.91	.93
Unemployed (vs. employed/retired)	1.50	.58–3.89	.40
Current smoker (baseline)	1.46	.66–3.26	.35
Homelessness-recent history (baseline)	3.76	1.39–10.20	.01
PHQ-9 score >=10	1.85	.87–3.94	.11
Moderate anxiety symptoms (baseline)	2.09	.59–7.4	.25
# Medical comorbidities (baseline)	.95	.67–1.33	.75
Alcohol Use AUDIT-C Score	1.10	.91–1.33	.31
Substance abuse (baseline)	.67	.27–1.65	.38
LG-CC Care Management Contacts (>=5)	2.62	1.00–6.95	.05
SF-12 physical health component score	.94	.89–1.00	.05
Hospitalization history (baseline)	1.59	.68–3.68	.28
Site (fixed effects)
Site 1	.24	.07–.81	.02
Site 2	.80	.25–2.54	.71
Site 3	.19	.04–.97	.05
Site 4	.27	.07–1.14	.08

DISCUSSION

This study determined participant factors associated with probability of emergency department use after the implementation of a collaborative care model (Life Goals Collaborative Care) for bipolar disorder in community practices. We found that participants with a recent history of homelessness were more likely to have an emergency department visit compared to those without these factors, after adjusting for other clinical and socio-demographics factors. Because homelessness history was associated with missingness, it is possible that those with a history of homelessness who needed to be seen in the emergency department were more likely to have participated in the 12-month survey than those with a history of homelessness who did not need ED visits. In prior research, bipolar disorder was associated with increased emergency department use and subsequent hospitalization, representing 60–70% of their total direct costs (14), and much of the increased critical service encounters is associated with instability from homelessness and poverty (1).

This study also found that participants completing five or more care management contacts also had a greater probability of an emergency department visit. Perhaps these individuals may have had more unmet health services needs that were not observed in the study assessments. Most of the individuals enrolled in the study, while primarily recruited from community mental health programs, were not eligible for more intensive case management services such as assertive community treatment, and hence, a subset (e.g., those with homelessness history) might have required more case management than provided in the relatively brief LG-CC intervention. Perhaps collaborative care models that emphasize psycho-education and health care coordination may not fully address social determinants of health, notably homelessness or lack of employment, and for this population, different types of services such as social services, housing, or other support systems might be needed for this group.

There are several limitations of this study that warrant discussion. The primary utilization measure was based on participant self-report and not verifiable with actual utilization data from electronic medical records or claims. However, evidence suggests that patients are more accurate in reporting low frequency events such as emergency room visits than high frequency events such as office visits (15). Because LG-CC was implemented for individuals between their baseline and 6-month assessments, his study assessed medical care utilization between 6 and 12 months. While the optimal recall period for self-report is typically 6 months (15), we were unable to reliably assess utilization beyond the 12-month assessment period. Finally, there was limited information on the types of care management services provided and lack of information on the mechanisms for how enhanced REP might have influenced the probability of emergency department use at the participant level beyond LG-CC contacts.

CONCLUSIONS

The increased desire by health care payers to promote more efficient care has led to a growing desire to implement more proactive models of care to better manage patients with chronic physical and mental health conditions. At the same time, these models of care may not be adequate especially given socioeconomic determinants of health beyond the practice setting such as poverty and homelessness. In this study we found a greater likelihood of emergency room use among a patient population with bipolar disorder, especially those a history of homelessness despite enrollment in a Collaborative Care Model-based program. These individuals may have unmet service needs that require additional management beyond standard care management. Healthcare systems deploying Collaborative Care Model-based programs should invest in more resources that address these social determinants of health that are important predictors of emergency department utilization. Future studies may also require the further intensification of evidence-based Collaborative Care Models for individuals with serious mental illness in order to potentially mitigate critical service encounters and improve outcomes.