Health-related preferences of older patients with multimorbidity: an evidence map

Objectives To systematically identify knowledge clusters and research gaps in the health-related preferences of older patients with multimorbidity by mapping current evidence.

Design Evidence map (systematic review variant).

Data sources MEDLINE, EMBASE, PsycINFO, PSYNDEX, CINAHL and Science Citation Index/Social Science Citation Index/-Expanded from inception to April 2018.

Study selection Studies reporting primary research on health-related preferences of older patients (mean age >= 60 years) with multimorbidity (>= 2 chronic/acute conditions).

Data extraction Two independent reviewers assessed studies for eligibility, extracted data and clustered the studies using MAXQDA-18 content analysis software.

Results The 152 included studies (62% from North America, 28% from Europe) comprised 57 093 patients overall (range 9-9105). All used an observational design except for one interventional study: 63 (41%) were qualitative (59 cross-sectional, 4 longitudinal), 85 (57%) quantitative (63 cross-sectional, 22 longitudinal) and 3 (2%) used mixed methods. The setting was specialised care in 85 (56%) and primary care in 54 (36%) studies. We identified seven clusters of studies on preferences: end-of-life care (n=51, 34%), self-management (n=34, 22%), treatment (n=32, 21%), involvement in shared decision making (n=25, 17%), health outcome prioritisation/goal setting (n=19, 13%), healthcare service (n=12, 8%) and screening/diagnostic testing (n=1, 1%). Terminology (eg, preferences, views and perspectives) and concepts (eg, trade-offs, decision regret, goal setting) used to describe health-related preferences varied substantially between studies.

Conclusion Our study provides the first evidence map on the preferences of older patients with multimorbidity. Included studies were mostly conducted in developed countries and covered a broad range of issues. Evidence on patient preferences concerning decision-making on screening and diagnostic testing was scarce. Differences in employed terminology, decision-making components and concepts, as well as the sparsity of intervention studies, are challenges for future research into evidence-based decision support seeking to elicit the preferences of older patients with multimorbidity and help them construct preferences.