Reporting and handling of missing participant data in systematic reviews of kidney transplant studies.

BACKGROUND Participant withdrawal from clinical trials occurs for various reasons, predominantly adverse effects or intervention inefficacy. Because these missing participant data can have implications for the validity, reproducibility and generalizability of study results, when conducting a systematic review, it is important to collect and appropriately analyze missing data information to assess its effects on the robustness of the study results. METHODS/DESIGN In this methodological survey of missing participant data reporting and handling in systematic reviews, we included meta-analyses that provided pooled estimates of at least 1 dichotomous intervention outcome of a randomized controlled trial performed in adult kidney transplant subjects. RESULTS Eighty three systematic reviews (17 Cochrane and 66 non-Cochrane reviews) met the inclusion criteria. The most common intervention was drugs (80%), with the majority involving immunosuppressant drugs 55% (n=46), followed by surgery in 14% (n=12). The median follow-up duration was 12 months (maximum, 240 months). Intention-to-treat (ITT) or modified ITT analysis was reported in 24% (n=20) of the reviews (76% of Cochrane and 10% of non-Cochrane). Overall, the majority of systematic reviews did not quantify [90% (n=60) non-Cochrane and 29% (n=5) Cochrane)] or include the reasons for missing participant data [88% (n=58) non-Cochrane and 24% (n=4) Cochrane)]. 11% (n=9) handled missing participant data, 5% (n=4) justified the analytical method(s) used to handle it, and 2%(n=2) performed a sensitivity analysis for it. DISCUSSION Systematic reviews of kidney transplantation provide inadequate information on missing participant data and usually do not handle or discuss the associated risk of bias with it. SYSTEMATIC REVIEW REGISTRATION PROSPERO CRD42019114235.