文档介绍：T h e ne w engl a nd jour na l o f medicine
spe c i a l r e p or t
Statistics in Medicine — Reporting of Subgroup
Analyses in Clinical Trials
Rui Wang, ., Stephen W. Lagakos, ., James H. Ware, ., David J. Hunter, ., .,
and Jeffrey M. Drazen, .
Medical research relies on clinical trials to as- coronary events with the use of pravastatin was
sess therapeutic benefits. Because of the effort examined in a diverse population of persons who
and cost involved in these studies, investigators had survived a myocardial infarction. In sub-
frequently use analyses of subgroups of study group analyses, the investigators further examined
participants to extract as much information as whether the efficacy of pravastatin relative to pla-
possible. Such analyses, which assess the heter- cebo in preventing coronary events varied accord-
ogeneity of treatment effects in subgroups of pa- ing to the patients’ baseline low-density lipopro-
tients, may provide useful information for the care tein (LDL) levels.
of patients and for future research. However, sub- Subgroup analyses are also undertaken to in-
group analyses also introduce analytic challeng- vestigate the consistency of the trial conclusions
es and can lead to overstated and misleading among different subpopulations defined by each
‑7 This report outlines the challenges as- of multi