Objectives:

To assess the effectiveness of quality improvement, implementation, and dissemination strategies that seek to improve the mental health care of children and adolescents; to examine harms associated with these strategies; and to determine whether effectiveness or harms vary in subgroups based on system, organizational, practitioner, or patient characteristics.

Data sources:

Searches from inception through January 14, 2016, of MEDLINE^®, Cochrane Library, PsycINFO^®, CINAHL^® (Cumulative Index to Nursing and Allied Health Literature), and gray literature; additional studies from reference lists and study authors.

Review methods:

Dual selection, data extraction, and risk of bias assessment of relevant trials and observational studies, followed by analysis, synthesis, and grading the strength of evidence for each outcome. We also employed qualitative comparative analysis (QCA) to examine set relationships between combinations of strategy components and improvements in outcomes.

Results:

We found 17 studies testing overall effectiveness of 16 strategies, of which 1 reported on harms and 4 on moderators of effectiveness. The evidence base includes 13 randomized controlled trials (RCTs), 2 controlled clinical trials, 1 cohort, and 1 interrupted time series. The strategies included in this review were complex and heterogeneous. We found 7 studies (6 strategies) that comprised only professional components and 10 studies (10 strategies) that consisted of one or more financial or organizational components, although many of these included professional components as well. Twelve studies included multiple active components; 5 had a single active component.

We found evidence that a majority of strategies had at least some evidence of effectiveness. Twelve studies (11 strategies) had at least one outcome rated as low for benefit. We graded the strength of evidence of one outcome for one strategy as moderate: one RCT reported that provider financial incentives improve practitioner implementation competence. Our QCA revealed inconsistent evidence on strategies with educational meetings, materials, and outreach: these strategies appeared to be successful in combination with reminders or providing practitioners with newly collected clinical information. We also found low strength of evidence of no benefit for strategies that included educational materials only, educational meetings only, educational materials and meetings only, and educational materials and outreach components only.

We were unable to judge the overall potential for harms associated with these strategies that may mitigate benefits based on the single included study with information on harms. The available evidence from four studies on two moderators does not permit us to make general conclusions about the conditions under which these strategies might work optimally.

Conclusions:

Our findings suggest that several approaches can improve both intermediate and final health outcomes and resource use. Twelve of the 17 included studies (11 of the 16 strategies) significantly improved at least one such outcome or measure. The evidence does not permit us to have a high degree of confidence about the efficacy of any one strategy because we generally found a single study testing each strategy. We found inconsistent evidence involving strategies with educational meetings, materials, and outreach; programs appeared to be successful in combination with reminders or providing practitioners with newly collected clinical information. We also found low strength of evidence for no benefit for initiatives that included only educational materials or meetings (or both) or only educational materials and outreach components.