Systematic evaluation of the effectiveness of innovative care concepts for diabetes, taking into account heterogeneity of study results.
Systematic review, meta-analysis and meta-regression analysis.
Medline, CINAHL and PsycINFO were searched for reviews and empirical studies into adult patients with diabetes in whom interventions were used that consisted of a minimum of 2 components from the Chronic Care Model. In the meta-analysis, a random-effects model was used to pool results on HbA1c and systolic blood pressure. Meta-regression analysis was used to examine the extent to which 3 differences in study design – (a) methodological quality, (b) length of follow-up, and (c) number of intervention components – could explain the heterogeneity in effect measures.
The reviews (n = 15) generally reported modest improvements in glycaemic control. The empirical studies (n = 56) differed greatly in the direction and magnitude of measured effects on HbA1c (mean: -5 mmol/mol; 95% CI: -7- -3.5) and systolic blood pressure (mean: -2.8 mmHg; 95% CI: -4.7- -0.95). Variation in methodological quality between the studies could not explain this. Short-term studies (< 1 year) reported more positive results than longer-term studies (≥ 1 year), but this difference was not significant. The variation in the number of intervention components explained 8-12% of the heterogeneity in effects of diabetes care innovations.
Diabetes care interventions that include more than 2 components have the most positive effects on clinical outcomes. Considering the connection between length of follow-up and results as well as the fact that the targets of these innovations focus primarily on the future, the impact of such interventions should be measured in the long term.
Conflict of interest: none declared. Financial support: none declared.