[Skip to Content]
[Skip to Content Landing]
December 12, 2001

Preventable Deaths From Medical Errors—Reply

Author Affiliations

Stephen J.LurieMD, PhD, Senior EditorIndividualAuthorJody W.ZylkeMD, Contributing EditorIndividualAuthor

JAMA. 2001;286(22):2813-2814. doi:10.1001/jama.286.22.2808

In Reply: Previous studies1 estimating preventable deaths have assumed that the error assessment had unbiased interrater reliability and that if care had been optimal, the patient would have had a 100% chance of living. Our study found that both assumptions were incorrect and led to dramatic overestimation. We see no merit in Drs Barach and Mohr's concern that our study is too small. Our study's statistical power is comparable to that of the Harvard Medical Practice Study (HMPS),2 and the 95% confidence interval for our key finding is 1.0-1.5. Therefore, our 383 reviews of 179 deaths (including 111 active-care deaths) is more than adequate.