Article type
Year
Abstract
Background: The vast majority of depression is (often sub-optimally) managed in primary care, which is where interventions should be targeted. The most promising approach to improving the quality of primary care for depression seems to be 'case management' [1]. However, substantial heterogeneity in the design and implementation of this intervention exists in terms of who should delvier the care (eg specialists vs non specialists); how care should be delivered (eg face to face or over the phone) and for how long.
Case management is therefore a good example of complex intervention, where heterogeneity is potentially informative in helping us understand what works for whom, and what are the 'active ingredients' that are essential for effective outcomes.
Objective: To examine the active ingredients of case management models for depression in primary care, using meta-regression as an exploratory technique.
Methods: Data on the content and outcomes of case management models of care were extracted from 36 RCTs , and meta-regression used to examine relationships between intervention content and outcomes.
Results: Case management has significant effects on both antidepressant use (OR = 1.944 95% CI 1.570 to 2.407) and depression outcomes (SMD 0.25 95% CI 0.184 to 0.324). Key predictors of the effect on depression outcomes include: patient recruitment through screening (p=0.114); case managers having a specific mental health background (p=0.024); provision of regular supervision for case managers (p=0.018); greater numbers of case management sessions (p=0.214), and the addition of a specific psychotherapeutic intervention to standard case management (p=0.092).
Conclusions: Although there are significant limitations with the methods presented [2], they provide a useful way of examining active and inactive ingredients in complex interventions. These 'active ingredients' have been used in designing the first UK trial of case management, where their predictive power will be examined prospectively.
References: 1. Gilbody S, Whitty P, Grimshaw J, Thomas R. Educational and organizational interventions to improve the management of depression in primary care: a systematic review. JAMA 2003;289:3145-51.
2. Thompson SG, Higgins JP. How should meta-regression analyses be undertaken and interpreted? Stat Med 2002;21:1559-73.
Case management is therefore a good example of complex intervention, where heterogeneity is potentially informative in helping us understand what works for whom, and what are the 'active ingredients' that are essential for effective outcomes.
Objective: To examine the active ingredients of case management models for depression in primary care, using meta-regression as an exploratory technique.
Methods: Data on the content and outcomes of case management models of care were extracted from 36 RCTs , and meta-regression used to examine relationships between intervention content and outcomes.
Results: Case management has significant effects on both antidepressant use (OR = 1.944 95% CI 1.570 to 2.407) and depression outcomes (SMD 0.25 95% CI 0.184 to 0.324). Key predictors of the effect on depression outcomes include: patient recruitment through screening (p=0.114); case managers having a specific mental health background (p=0.024); provision of regular supervision for case managers (p=0.018); greater numbers of case management sessions (p=0.214), and the addition of a specific psychotherapeutic intervention to standard case management (p=0.092).
Conclusions: Although there are significant limitations with the methods presented [2], they provide a useful way of examining active and inactive ingredients in complex interventions. These 'active ingredients' have been used in designing the first UK trial of case management, where their predictive power will be examined prospectively.
References: 1. Gilbody S, Whitty P, Grimshaw J, Thomas R. Educational and organizational interventions to improve the management of depression in primary care: a systematic review. JAMA 2003;289:3145-51.
2. Thompson SG, Higgins JP. How should meta-regression analyses be undertaken and interpreted? Stat Med 2002;21:1559-73.