March 2011
Reviewer: Jeff C. Huffman, MD
Collaborative care for patients with depression and chronic illnesses
Katon WJ, Lin EH, Von Korff M, et al
New Engl J Med 2010; 363(27):2611-2620
Background: Depression is common in patients with chronic medical illnesses, and has been linked to adverse medical events and mortality, independent of other risk factors. Prior depression ‘collaborative care’ programs have utilized a non-physician care manager to longitudinally manage depression by coordinating care between patient, primary physician, and consulting specialist. These programs have generally been effective in improving depression but have not substantially improved medical outcomes.
Methods: This was a randomized trial of a ‘blended’ collaborative care management model for primary care outpatients with diabetes or coronary artery disease who met criteria for both depression (diagnosed via PHQ-9 score ≥10) and poor disease control (defined as blood pressure >140/90 mm Hg, LDL-C >130 mg/dL, or hemoglobin A1c > 8.5%). Intervention subjects had coordination of care through a study team of a psychiatrist, primary care physician, psychologist, and nurse care manager for 12 months. The nurse care manager directly intervened with patients to assist with PCP implementation of psychiatrist-recommended pharmacotherapy for depression and/or provide problem-solving therapy (PST) for depression. In addition, the care managers helped to longitudinally monitor blood pressure and blood sugar, and they used motivational interviewing techniques to improve adherence to recommendations around health behaviors. When subjects continued to fall short of targets for their illnesses (e.g., having residual depression or elevated blood pressure), additional recommendations/steps were made by the care manager, in consultation with the study team, to optimize subjects’ outcomes in each illness domain. As with other collaborative care models, subjects’ care remained with their primary medical physicians, who wrote all prescriptions for subjects. Usual care subjects’ physicians were notified of the subjects’ depression and poor disease control at baseline and (as applicable) at 6 and 12 months, but received no other intervention. Study outcomes included depression scores, blood pressure, LDL cholesterol, and hemoglobin A1c levels, were measured at 6 and 12 months and compared between groups using mixed regression/generalized estimating equation (GEE) models.
Results: A total of 214 subjects (n=106 in intervention group) were enrolled; mean age was 57 years, approximately one-half were women and one-quarter were minorities, and 80% had diabetes. Over 12 months, subjects in the collaborative care arm had greater improvements of depression (60% vs. 30% with depression response; p<.001; effect size of depression improvement on SCL-20 scale =.67), systolic blood pressure (estimated between-group difference= 3.4 [95% CI= 6.9 to -0.1] mm Hg), LDL cholesterol (between-group difference= 9.1 [17.5 to 0.8] mg/dL), and hemoglobin A1c scores (0.56% [0.85%-0.27%]). Patients in the intervention group also had greater improvement in quality of life, more satisfaction with care received for depression and medical illness, and were more likely to have an A1c decrease of 1.0% or greater and a decrease of 10 or more mm Hg on systolic blood pressure.
Commentary: This is a landmark study in psychosomatic medicine: it is a study that focused on patients with a psychiatric illness and ongoing medical illnesses, utilized an intervention to manage depression and medical health, and resulted in improvements not only in mood but also in objective medical outcomes. The fact that the intervention was integrated with patients’ other medical care, occurred within primary care clinics, and that such inventions have been cost-effective/cost-reducing in other studies also suggests that this is an intervention that could be adapted to non-research clinical care settings (indeed collaborative care interventions for depression have been used throughout the country as part of usual clinical treatment). The innovative aspect of this intervention—having the care manager not only focus on depression but also manage chronic medical illness—also resulted in a very interesting finding: the effect size for depression improvement was more than double the effect size seen in most depression-only collaborative care programs, suggesting that a focus on both depression and physical illness results in better depression care.
The somewhat sobering finding of this positive trial was the relatively small absolute improvements in objective medical outcomes. Collaborative care subjects had improvements of 3 mm Hg of systolic BP, 9 points of cholesterol, and ~0.6% on A1c after all of this care management over the course of a whole year, and it is worth noting that subjects really got the ‘Cadillac’ of collaborative care: high quality care providers in a supervised research setting, performed by the group that essentially invented collaborative care depression interventions, and have performed multiple large trials of such interventions in the past. One would expect that a ‘real-world’ clinical adaptation of this intervention would not be as well-coordinated and expertly-provided and that improvements would be even less. This underscores the persistent and generally tough-to-treat nature of chronic illnesses—especially for patients who have comorbid psychiatric and medical illnesses. At the same time, these improvements are not inconsequential, and when projected for large populations of people over long periods of time, would represent a substantial improvement in health, quality of life, and survival for a very large population of patients seen in primary care clinics in the U.S. and abroad.