Erratum: Screening statins for possible carcinogenic risk: up to 9 years of follow-up of 361 859 recipients.

摘要

We have learned that the sensitivity analyses in our paper, Screening statins for possible carcinogenic risk: up to nine years of follow-up of 361,859 recipients, (Pharmacoepiderniol Drug Saf 2008; 17:27-36) were done incorrectly. For the external adjustment, instead of applying the percent bias to the difference between the observed relative risk and 1.0, we should have applied it simply to the observed relative risk. Thus, using the maximum estimated percent bias of 29%, external adjustment for smoking by dividing by 1.29 reduced the relative risk of lung cancer from 1.06 to 0.82 (95% confidence interval [CI] 0.68-0.99) instead of to 1.05 for men taking statins for at least 5 years. For women, the corresponding reduction of 1.17 should have been to 0.88 (95% CI 0.70-1.10) instead of to 1.13, using the percent bias of 33%. Applying 18% bias, the corrected maximally externally adjusted relative risks of cancer for at least 5 years of statin use by men were: esophagus, 1.44 (95% CI 0.89-2.33); kidney 1.01(95% CI 0.67-1.52); bladder 1.36 (95%CI 1.09-1.67)