Incidence and impact of withdrawal of life-sustaining therapies in clinical trials of severe traumatic brain injury: A systematic review.
Level 1 - systematic review of randomized trials
Systematic review of randomized controlled trials
PubMed 29865897 · doi:10.1177/1740774518771233
What was done
A systematic review of randomized controlled trials (RCTs) published between January 2002 and August 2015 in 18 high-impact general medicine, critical care, and neurocritical care journals. Eligible studies included adult severe traumatic brain injury patients (Glasgow Coma Scale ≤ 8) reporting mortality. The primary objective was determining the proportion of trials reporting withdrawal of life-sustaining therapies (WLST), alongside assessing overall mortality, WLST-associated death proportions, and the modeled impact of WLST on trial results.
What was found
From 5,987 retrieved citations, 41 RCTs were included (total n = 16,364 patients; range 11 to 10,008). Overall mortality across trials was 23% (range 3% to 57%). WLST was reported in 20% of trials (8/41; 932 patients), and the crude number of deaths due to WLST was reported in 17% of trials (7/41; 884 patients). In the reporting trials, 63% of deaths (105/168) followed WLST. Imputing a 4% differential rate of WLST between trial groups changed the results and conclusions in one-third of the trials.
Why it matters
Withdrawal of life-sustaining therapy accounts for the majority of deaths in severe TBI trials yet is rarely reported, introducing unmeasured confounding that can flip trial conclusions. Standardized reporting of WLST is necessary to interpret trial outcomes accurately.
Limits
The review was restricted to 18 specific high-impact journals published through August 2015, potentially missing trials published in other venues or more recently. The assessment of impact on trial outcomes relied on hypothetical imputation rather than observed individual patient data, and crude WLST mortality was available from only 7 trials.