Meta-analysis does not allow appraisal of complex interventions in diabetes and hypertension self-management: a methodological review

Abstract

Common methodologies used in systematic reviews do not allow adequate appraisal of complex interventions. The aim of the present study was to describe and critically appraise current methods of systematic reviews on complex interventions, using diabetes and hypertension patient education as examples. PubMed, the Cumulative Index to Nursing and Allied Health (CINAHL), the Cochrane Library and Health Technology Assessment databases were searched. Systematic reviews focusing on diabetes or hypertension patient education were included. Authors were contacted. Two investigators independently evaluated the reviews. The available evidence of three patient education programmes of diabetes and hypertension self-management implemented in Germany was used as a reference. We included 14 reviews; 12 reviews mentioned that the included education programmes were multidimensional. Reviews on comparable topics identified different publications of the same programme. Identical programmes were classified differently within and between reviews. Education programmes were dissected to analyse effects of single components. Different components of identical programmes were used. Interdependencies between components were not explored. Six reviews performed meta-analysis across programmes with heterogeneous educational or organisational approaches. The complexity of efficacy measures was disregarded: e.g. HbA(1c) was used as an isolated outcome variable without considering treatment goals, effects on hypoglycaemia, body weight or quality of life. Our results indicate that methods of current systematic reviews are not fully equipped to appraise patient education and self-management programmes. Since these are complex and heterogeneous interventions, consideration of aggregated evidence is necessary.