The following is a summary of “Composite outcome measures in high-impact critical care randomised controlled trials: a systematic review,” published in the May 2024 issue of Critical Care by Walker et al.

Composite outcome measures (COMs) are gaining traction in clinical trials, offering advantages but also raising concerns. Critical care research, however, needs to understand specific applications and potential drawbacks better.

Researchers conducted a retrospective study investigating composite outcomes in high-impact critical care trials and compared trial characteristics like sample size, statistical significance, and effect estimates between studies with composite and non-composite outcomes.

They searched 16 high-impact journals for RCTs (2012 to 2022). The studies focused on patients with critical illness and reported outcomes that were considered essential.

The result showed an analysis of 8,271 trials, and 194 used COMs. Notably, the use of COMs increased over time, with 39.1% of trials employing them. A concerning finding was that only half (52.6%) of the outcomes were composite. The median COM comprised 2 components (IQR 2-3). Trials using COMs enrolled fewer participants (median 409 (198.8–851.5) vs. 584 (300–1566, P= 0.004) for non-COM trials) but did not achieve specific significant results. Predicted effect sizes were overestimated in nearly 6 COM trials. Furthermore, consistency of effect was only observed in 43.4% of all COM components. Most importantly, 93% of COMs included elements not directly relevant to patient well-being. 

Investigators concluded that critical care trials increasingly relied on composite outcomes, but the measures taken often lacked consistency across components, leading to unclear results.  Studies using composite outcomes had smaller sample sizes and no increased likelihood of significant findings, indicating a need to reevaluate the use in critical care research.

Source: ccforum.biomedcentral.com/articles/10.1186/s13054-024-04967-3