Related Experiment Videos
Should data monitoring committees assess efficacy when considering safety in trials in acute stroke?
P M W Bath1, L J Gray, N-G Wahlgren
1Institute of Neuroscience, University of Nottingham, Nottingham, UK. philip.bath@nottingham.ac.uk
Background:
The primary role of a trial's data monitoring committee (DMC) is to ensure the safety of enrolled patients. In stroke trials, safety is monitored typically by comparing death and stroke-specific events between treatment groups. DMCs may also have the remit for monitoring efficacy depending on the aims of the trial. We hypothesised that functional outcome at the end of follow-up, a measure of efficacy, is also a powerful measure of safety and tested this in a systematic review.
Methods:
Acute stroke trials with a negative outcome or which were stopped prematurely on the grounds of safety were sought systematically from searches of electronic databases and published reviews. Information on early and late death, impairment and functional outcome, and the presence of a DMC, were recorded for each trial. The results for each outcome measure were ranked within each trial to determine which was most statistically efficient in detecting hazard.
Results:
Fourteen trials were included. The most efficient outcomes for detecting hazard were late death or disability, six trials; early death, four trials (two of which tested thrombolysis); late death, three trials and late death or impairment, one trial. Early death was insensitive to hazard in all six trials where late death or dependency was most sensitive. Two trials (both phase II) did not report the presence of a DMC.
Conclusions:
Functional outcome at end of follow-up can be sensitive to hazard and should be included in all assessments of safety in stroke trials, whether or not efficacy itself is being assessed.
