A close look at lay-led self-management programs for chronic diseases and health care utilisation: A systematic...

Introduction: Chronically ill people are confronted with significant challenges when dealing with health care services. Lay-led self-management programs aim to improve self-management skills and might enable patients to make appropriate decisions as to when to use health care services. Contrary to the general attitude that self-management programs reduce health care utilisation, we suspect that better self-management skills lead to increased or possibly more appropriate health care utilisation. Our review and meta-analysis sheds light on the relationship between lay-led self-management programs and health care utilisation. Methods: In March 2017, we searched 7 databases (CDSR, DARE, HTA, Medline, CINAHL, PsycInfo, and SSCI) to perform a systematic review and meta-analysis to identify studies that reported empirical data on lay-led self-management programs and health care utilisation. We extracted the characteristics of all primary studies and the data of four indicators of utilisation (physician visits, emergency department visits, hospital admissions, and length of stay in hospital), and analysed the role of health care utilisation in these studies. We present the results in frequency tables and as a conventional meta-analysis with the standardized mean difference (SMD), 95% confidence intervals (CI), and pooled overall effect sizes using RevMan 5.3.5. The manuscript follows the PRISMA checklist. Results: Overall, we include 49 primary studies; 10 studies provided sufficient data for the meta-analysis. Health care utilisation played a different role in the studies; 15 studies reported a clear explicit hypothesis on the influence of a lay-led self-management program on health care utilisation, and 17 studies assumed an implicit assumption. 8 studies discussed the possibility that a lay-led self-management program could lead to more appropriate health care utilisation. The meta-analysis showed mixed results, and many effect sizes were not statistically significant. The participants of a lay-led self-management program had fewer emergency department visits (SMD: -0.08; 95% CI: -0.15 to -0.01; p=0.02) than the control group. Conclusion: Although the statistically significant effects of the meta-analysis are low, our overall findings show that only a small number of the included studies tackled the task of comprehensively investigating self-management skills in the context of health care utilisation. This fails to do justice to the potential of self-management programs. It is essential to consider the appropriateness of health care utilisation. We propose the term self-management-sensitive utilisation for this purpose.