Mj. Glesby et Dr. Hoover, SURVIVOR TREATMENT SELECTION BIAS IN OBSERVATIONAL STUDIES - EXAMPLESFROM THE AIDS LITERATURE, Annals of internal medicine, 124(11), 1996, pp. 999-1005
Unlike patients in a randomized, clinical trial, patients in an observ
ational study choose if and when to begin treatment. Patients who live
longer have more opportunities to select treatment; those who die ear
lier may be untreated by default. These facts are the essence of an of
ten over-looked bias, termed ''survivor treatment selection bias,'' wh
ich can erroneously lead to the conclusion that an ineffective treatme
nt prolongs survival. Unfortunately, misanalysis of survivor treatment
selection bias has been prevalent in the recent literature on the acq
uired immunodeficiency syndrome. Approaches to mitigating this bias in
volve complex statistical models. At a minimum, initiation of therapy
should be treated as a time-dependent covariate in a proportional haza
rds model. Investigators and readers should be on the alert for surviv
or treatment selection bias and should be cautious when interpreting t
he results of observational treatment studies.