Representation and reporting of kidney disease in cerebrovascular disease: A systematic review of randomized controlled trials

Patients with kidney disease (KD) are at increased risk for cerebrovascular disease (CVD) and CVD patients with KD have worse outcomes. We aimed to determine the representation of KD patients in major randomized controlled trials (RCTs) of CVD interventions. We searched MEDLINE for reports of major CVD trials published through February 9, 2017. We excluded trials that did not report mortality outcomes, enrolled fewer than 100 participants, or were subgroup, follow-up, or post-hoc analyses. Two independent reviewers performed study selection and data extraction. We included 135 RCTs randomizing 194,977 participants. KD patients were excluded in 48 (35.6%) trials, but were less likely to be excluded from trials of class I/II recommended interventions (n = 7; 15.9%; p = 0.001) and more likely to be excluded in trials with registered protocols (45.5% vs. 22.4%; p = 0.007). Exclusion was lower in trials supported by academic or governmental grants compared to industry or combined funding (21.2% vs. 42.0% and 47.8%; p = 0.033 and 0.028, respectively). Among trials excluding KD patients, 24 (50.0%) used serum creatinine, 7 (14.6%) used estimated glomerular filtration rate or creatinine clearance, 7 (14.6%) used renal replacement therapy, and 19 (39.6%) used non-specific kidney-related criteria. Only 4 (3.0%) trials reported baseline renal function. No trials prespecified or reported subgroup analyses by baseline renal function. Although 19 (14.1%) trials reported the incidence of acute kidney injury, no trial examined adverse event rates according to renal function. In summary, more than one third of major CVD trials excluded patients with KD, primarily based on serum creatinine or nonspecific criteria, and outcomes were not stratified by renal parameters. Therefore, purposeful efforts to increase inclusion of KD patients in CVD trials and evaluate the impact of renal function on efficacy and safety are needed to improve the quality of evidence for interventions in this vulnerable population.