The following is a summary of “Positive single-center randomized trials and subsequent multicenter randomized trials in critically ill patients: a systematic review,” published in the November 2023 issue of Critical Care by Kotani et al.
The reliability of survival benefits reported in single-center randomized controlled trials (sRCTs) for critically ill patients when evaluated in subsequent multicenter randomized controlled trials (mRCTs) remains uncertain.
They searched PubMed for sRCTs published in the New England Journal of Medicine, JAMA, or Lancet from inception (December 31, 2016). The inclusion criteria involved a statistically significant decrease in mortality with any intervention (drug, technique, or strategy) among adult critically ill patients. Afterward, searches were done for mRCTs addressing the same research question as the sRCTs.Results concordance between sRCTs and available mRCTs was compared. The systematic review was registered in PROSPERO. (CRD42023455362).
The results showed 19 sRCTs outlining a notable decrease in mortality among adult critically ill patients. For 16 of these sRCTs, at least one subsequent mRCT was identified (24 trials in total), leaving three sRCTs with interventions yet to be addressed in a subsequent mRCT. Merely one out of the 16 sRCTs (6%) was succeeded by an mRCT replicating a significant mortality reduction; conversely, 14 (88%) were grown by mRCTs displaying no mortality difference. A subsequent mRCT contradicted the positive outcome of one sRCT (6%) on intensive glycemic control, revealing a significant mortality increase. Six of the 14 sRCTs in international guidelines (43%) are now either removed or discouraged in the latest versions of relevant guidelines.
They concluded that sRCT-reported mortality reductions rarely persisted in mRCTs, suggesting sRCT findings should guide hypothesis generation rather than guideline formation.
Source: ccforum.biomedcentral.com/articles/10.1186/s13054-023-04755-5
