In Reply: Dr Loew and PhRMA contend that industry-sponsored studies are not at higher risk of bias and thus do not require “special scrutiny” and that “vigorous government oversight . . . ensures the integrity of data and results.” Despite these assurances, scientific and ethical lapses involving industry and industry-sponsored studies strongly indicate otherwise, as a few examples illustrate.
In the pivotal celecoxib trial, only 6-month data were reported when 12-month data showing no benefit in reducing gastrointestinal toxicity were actually available.1 In the key rofecoxib trial, data showing increased cardiovascular toxicity were downplayed and erroneously dismissed as due to a cardioprotective effect of naproxen.2 Evidence from the Vioxx litigation suggests that the manufacturer may have failed to disclose information about higher drug-related death rates in 2 other rofecoxib studies.3 Results of trials of paroxetine that demonstrated increased risk of teenage suicide or lack of efficacy were suppressed and only revealed after a lawsuit was brought against the manufacturer.4 The sponsor of a human immunodeficiency virus immunogen trial withheld key data and sued the investigators to try to block publication of the study results.5 Recent evidence indicates that the manufacturer of an implantable cardioverter-defibrillator failed to report critical, potentially fatal design defects for more than 3 years.6
Conflicts of Interest and Independent Data Analysis in Industry-Funded Studies—Reply. JAMA. 2005;294(20):2575-2577. doi:10.1001/jama.294.20.2576