A Randomized Clinical Trial Assessing the Cost-Effectiveness of Generalist Care Managers for the Treatment of Depression in Medicaid Recipients in Primary Care Settings

Suzanne Landis, MD, MPH, Principal Investigator
Joseph P. Morrissey, PhD, Co-Investigator
Bradley Gaynes, MD, MPH, Co-Investigator
Marisa Domino, PhD, Co-Investigator
Alan R. Ellis, MSW, Project Analyst

Related Publications

This project, funded by the Robert Wood Johnson Foundation, is a collaborative effort among four agencies located in Western and Central North Carolina—the Mountain Area Health Education Center (MAHEC) Family Health Center; the Hot Springs Health Program; Access II Care of Western North Carolina, Inc.; and the Cecil G. Sheps Center for Health Services Research at the University of North Carolina at Chapel Hill—as well as the North Carolina Division of Medical Assistance (Medicaid). The goal is to assess the cost-effectiveness of generalist care managers vs. usual care in the treatment of depression in Medicaid patients seen in community-based primary health care practices. The use of generalist care managers is believed to have wide applicability and value for state Medicaid programs.

A total of 45 patients with depression were recruited at two predominantly rural community-based primary care practices in Western North Carolina and randomly assigned to either generalist care management (GCM, n=22) or usual care (n=23). Patients in each condition were assessed at baseline and six-month follow-up. Measures include the nine-item depressive symptom measure from the Patient Health Questionnaire (PHQ-9), the Hamilton Depression Rating Scale (HAM-D), the Medical Outcomes Study Short Form 12 (MOS-SF12), the Hopkins Symptom Checklist (SCL-20), and prior six-month service utilization and cost from Medicaid claims data. Primary outcomes at 6 months will include percentage of patients with a 50% decrease in depressive severity, percentage of patients with remission of depressive episode, adequacy of current dose and duration of treatment according to depression guidelines, number of depression-free days, number of disability days in prior 6 months, and service utilization and cost.

The Sheps Center team participated in the study design, monitored progress, is conducting statistical analyses of six-month outcomes and cost-effectiveness, and will collaborate in the dissemination of study findings. Some preliminary findings follow:

The baseline sample included 43 men and 2 women. Of the participants, 62% were white only and 24% black only. Fourteen percent classified themselves as having another race, multiple races, Hispanic ethnicity, or Russian/Ukrainian ethnicity. The mean age was 40 (SD=11); the mean scores on the PHQ-9 and HAM-D were 16.4 (SD=4.9) and 20.1 (SD=4.6) respectively. Patients in both conditions improved, with 22.2% of GCM patients and 37.5% of usual care patients achieving at least a 50% decrease in severity. The PHQ-9 score decreased below 10 for 38.9% of GCM patients and 43.8% of usual care patients.

Based on medical records, Medicaid pharmacy data, and patient self-report, there is some evidence that GCM (compared to usual care) is associated with higher levels of prescription, prescription filling, and/or use of adequate medication dosages. GCM appears to have influenced prescribing more than it was able to influence prescription filling or medication use.

This study provides a valuable opportunity to learn about the implementation of GCM in a primary care setting. A detailed description is being developed based on administrative data and physician and staff surveys and focus groups. Care managers had extensive contact with patients, not only to address core treatment issues but also to address issues that complicate depression treatment, such as comorbid conditions (mentioned in the record for 30% of contacts) and social issues (22%). Of the 22 GCM patients, 18 had a total of 69 comorbid conditions recorded in care management records; asthma accounted for 14% of these and diabetes accounted for 9%. Seventeen patients had a total of 73 social issues recorded that could be expected to affect depression outcomes. Family issues accounted for 16% of these. Other social issues included transportation, money, Medicaid, legal, housing, linkage with community services, and inability to locate the patient. GCM patients with certain chronic medical conditions (congestive heart failure, coronary artery disease, chronic obstructive pulmonary disease, asthma, diabetes) received more contact than did other GCM patients. The contacts for these patients were more likely to be in person and focused more on clinical issues and core treatment issues.

This project is currently in the analysis stage. All results reported here are preliminary.