American journal of kidney diseases : the official journal of the National Kidney Foundation 2018 02 21() pii S0272-6386(17)31129-0
Clinical trials are most informative for evidence-based decision making when they consistently measure and report outcomes of relevance to stakeholders. We aimed to assess the scope and consistency of outcomes reported in trials for hemodialysis.
SETTING & POPULATION
Adults requiring maintenance hemodialysis enrolled in clinical trials.
All Cochrane systematic reviews of interventions published by August 29, 2016, and the trials published and registered in ClinicalTrials.gov since January 2011.
Any hemodialysis-related interventions.
Frequency and characteristics of the reported outcome domains and measures.
From the 362 trials, we extracted and classified 10,713 outcome measures (a median of 21 [IQR, 10-39] per trial) into 81 different outcome domains, of which 42 (52%) were surrogate; 25 (31%), clinical; and 14 (17%), patient reported. The number of outcome measures reported significantly changed over time. The 5 most commonly reported domains were all surrogates: phosphate (125 [35%] trials), dialysis adequacy (120 [33%]), anemia (115 [32%]), inflammatory markers (114 [31%]), and calcium (109 [30%]). Mortality, cardiovascular diseases, and quality of life were reported very infrequently (73 [20%], 44 [12%], and 32 [9%], respectively).
For feasibility, we included a sampling frame that included only trials identified in Cochrane systematic reviews or ClinicalTrials.gov.
Outcomes reported in clinical trials involving adults receiving hemodialysis are focused on surrogate outcomes, rather than clinical and patient-centered outcomes. There is also extreme multiplicity and heterogeneity at every level: domain, measure, metric, and time point. Estimates of the comparative effectiveness of available interventions are unreliable and improvements over time have been inconsistent.