Depression monitoring and patient behavior in the Clinical Outcomes in MEasurement-Based Treatment (COMET) trial.

OBJECTIVE: In this secondary analysis of results of the Clinical Outcomes in MEasurement-Based Treatment (COMET) trial, patient behaviors that might account for the differences observed in clinical outcomes were examined. METHODS: Patients (N=914) diagnosed as having major depressive disorder participated in telephone interviews either monthly for six months (intervention) or at three and six months (usual care) asking about antidepressant medication-taking, use of psychotherapy or counseling, and participation in depression support groups. Physicians (N=83) in the intervention arm received monthly feedback regarding their patients' depression severity. RESULTS: A total of 664 (73%) patients completed the month 6 interview. The adjusted odds of current antidepressant use at six months were 85% greater (p=.01) for patients in the intervention (N=380) versus usual care (N=284) arms, according to multivariate regression analyses. CONCLUSIONS: More frequent measurement of depression symptoms was associated with greater medication persistence, which in turn may have mediated clinical improvements.

Clinical Outcomes in Measurement-based Treatment (Comet): a trial of depression monitoring and feedback to primary care physicians.

BACKGROUND: Despite the availability of effective treatments for depression, many patients under the care of primary care physicians do not achieve remission. Clinical Outcomes in Measurement-based Treatment (COMET) was designed to assess whether communicating patient-reported depression symptom severity to primary care physicians affects patient outcomes at 6 months. METHODS: Nine hundred fifteen patients (intervention: n = 503; control: n = 412) diagnosed with major depressive disorder were enrolled in a prospective trial in which physician practice sites were assigned to either the intervention or control study arm. Only patients who were prescribed an antidepressant by their physician were eligible, but medication type was independent of the study protocol. Intervention-arm physicians received monthly updates on their patients' depression severity, which was determined with the nine-item Patient Health Questionnaire (PHQ-9) administered during telephone interviews. Remission was defined as a PHQ-9 score <5 at 6 months; response was defined as a score reduction ≥50%. RESULTS: Among patients with baseline PHQ-9 score ≥5, 45.0% achieved remission (46.7% intervention versus 42.8% control) and 63.9% responded (67.0% intervention versus 59.7% control) at 6 months. After adjusting for baseline demographic and clinical variables, odds of remission (odds ratio [OR], 1.59 [95% CI, 1.07-2.37]) or response (OR, 2.02 [95% CI, 1.36-3.02]) were significantly greater for the intervention group than for control patients. CONCLUSIONS: This study demonstrated that regular patient symptom monitoring with feedback to physicians improved outcomes of depression treatment in the primary care setting. Determining reasons for the high observed nonremission rates requires further investigation.

Effect of communicating depression severity on physician prescribing patterns: findings from the Clinical Outcomes in MEasurement-based Treatment (COMET) trial.

OBJECTIVE: In this secondary analysis from the Clinical Outcomes in MEasurement-based Treatment trial (COMET), we evaluated whether providing primary care physicians with patient-reported feedback regarding depression severity affected pharmacological treatment patterns. METHOD: Intervention-arm physicians received their patients' 9-item Patient Health Questionnaire scores monthly. Odds of having no change in antidepressant treatment during the 6-month study period were calculated. Relationships between depression symptom status (partial or nonresponse) at month 3 and treatment changes in months 3 through 6 were assessed. RESULTS: Among 503 intervention and 412 usual care (UC) patients with major depressive disorder, most received antidepressant monotherapy at baseline (79.4% UC vs. 88.4% intervention; P=.047). Few switched their baseline antidepressant (17.4%), increased their dose (12.4%) or augmented with a second medication (2%). Odds of having no change in antidepressant therapy did not differ significantly between study arms (odds ratio 1.21; 95% confidence interval 0.78-1.88; P=.392). Few month 3 partial or nonresponders had a regimen change over the following 3 months; the study arms did not differ significantly (partial responders: 4.1% UC vs. 7.7% intervention; P=.429; nonresponders: 14.6% UC vs. 15.9% intervention; P=.888). CONCLUSIONS: Among depressed patients treated in primary care, little active management was observed. The lack of treatment modification for the majority of partial and nonresponders was notable.