Impact of shared decision making on healthcare in recent literature: a scoping review using a novel taxonomy

Abstract

Abstract Background Numerous articles and reviews discussed the effects of shared decision making (SDM) on concept-specific and direct outcomes, showing great variety in methodology and results. Objectives This scoping review accentuates effects of shared decision making interventions on more distal and distant outcomes related to the healthcare experience of patients and physicians, the economy and treatment parameters. Eligibility criteria The search considered randomized controlled trials (RCTs), their secondary analyses and follow-up reports comparing shared decision making interventions against control conditions. Sources of evidence MEDLINE (through PubMed) and reference lists of included articles were systematically appraised. Charting methods First, relevant outcome effects were extracted following the authors’ conclusions. Second, all outcomes were sorted into one of five different effect levels: individual, interactional, organizational, systemic and clinical. Results The search process identified 120 eligible reports, representing 116 randomized controlled trials and four follow-up reports with a variety in research topics, intervention types, outcome measurements and effects. Most of the 296 extracted outcomes were reported as not affected by shared decision making (205). While some outcomes improved at least slightly (81), few tended to decline (7) or revealed mixed results (3). Considering the five outcome effect levels, individual and clinical outcomes were reported more frequently than interactional, organizational and systemic ones. However, many individual outcomes could be counted as systemic and vice versa. Conclusions Shared decision making can improve distal and distant outcomes depending on the healthcare context. Individual, systemic and clinical outcomes have been more frequently appraised than interactional and organizational ones. Single database search and limited assessment of articles’ risk of bias and effect size narrow reliability of our results.