OBJECTIVES: The burden of type 2 diabetes mellitus (T2DM) is increasing worldwide, and real-world evidence (RWE) is needed to fill gaps in knowledge for payers.Unlike randomized controlled trials, lack of methodological rigor is a concern about RWE studies. The objective was to describe characteristics of RWE studies in T2DM.

METHODS: A literature review was performed using the PICOframework: Population consisted of T2DM patients, Interventions and Comparators were any intervention for T2DM care or absence of intervention, and Outcomes were resource utilization, productivity loss or utility. Only RWEstudies were included, which were defined as studies that were not clinical trials and that collected de novo data (no retrospective analysis).

RESULTS: Among the 493 studies published in 2017, 26 articles were included. Resource utilization was collected in 58% of studies, utility in 42% and productivity loss in 23%. Most outcomes were self-reported (70%), 15% were retrieved from medical files and 15% used both data sources. A high proportion (65%) were cross-sectional studies. Interventions were compared in only 12% of studies: 4% were comparative effectiveness research (CER) and 8% were cost-effectiveness studies (single group pre/post-intervention). A high proportion of studies did not present sources for selecting patients (27%) or sampling methods (27%) or sample size calculations (50%) or participation rates (66%). There was evidence of a formal protocol for 89% of studies, and only 8% reported public registration. Internal and external validity were discussed in 85% and 81% of studies respectively.

CONCLUSIONS: RWE from T2DM studies lacks uniformity. Despite the recent recommendations in CER, there is still a need for good practices for RWE studies. Standardized methodologies specifically adapted for RWE studies collecting pharmacoeconomic data for the management of T2DM could help future reimbursement decision-making in this major public health problem.