Protopathic bias in observational studies on statin effectiveness

摘要

A meta-analysis of 14 randomized controlled trials (RCT) of statins in prevention of major vascular events suggests that almost all diabetic patients benefit from statin therapy, with a 10% relative reduction in event rates for every millimole per litre reduction in low-density lipoprotein cholesterol seen in 1 year [1]. The results of the RCTs including only patients with diabetes are, however, inconclusive. For ethical reasons, long-term placebo-controlled RCTs of statins in diabetic populations are no longer possible. Therefore observational studies are an essential tool for future research [2], but they are prone to various biases. Protopathic bias occurring when the drug is preferentially prescribed to persons with early symptoms of the outcome under study is one of them [3].