Electronic health records may threaten blinding in trials of statinsBMJ 2014; 349 doi: https://doi.org/10.1136/bmj.g5239 (Published 20 August 2014) Cite this as: BMJ 2014;349:g5239
- Paul v Nguyen, internal medicine specialist, Centre Hospitalier Universitaire de Montréal, 3840 St-Urbain, Montreal, Quebec, Canada QC H2W 1T8
Newly released US guidelines on cholesterol that promote more use of statins1 have been criticised as being over-reaching2 and biased.3 These guidelines are based on a meta-analysis of individual data from many randomised trials.4 But a previously overlooked problem may also have inflated the perceived benefits of statins: a lack of concealment of patient allocation.
In randomised controlled trials of the clinical efficacy of cardiovascular drugs in general—and statins in particular—a key control measure is the double blinding of patients, caregivers, and especially investigators and outcome assessors. A lack of blinding may significantly bias results in favour of the drug under study,5 particularly if clinical efficacy is based mostly on “soft outcomes” (those influenced by judgment or perception), such as the diagnosis of angina, or decisions on admission to hospital or on cardiac catheterisation or revascularisation.6
“Unblinding” can occur inadvertently in clinical trials when patient data are available in electronic health records. For example (and it has happened to me more than once), records could show a …