Letters Section Editor: Robert M. Golub, MD, Senior Editor.
In Reply: Dr Shapiro postulates that detection bias in women who developed vaginal bleeding after randomization to CEE + MPA may have led to an overestimate of their HRs for CHD. Several considerations argue against this interpretation.
If women developed persistent bleeding, the clinic gynecologist was unblinded, but not the participant or clinic staff. Even if some participants correctly guessed randomization status, the central adjudication of hard outcomes was done by blinded adjudicators. A greater likelihood of CHD identification in the CEE + MPA group is plausible only if there was both differential unblinding and an expectation of harm. To the contrary, during the initial years of the study when most of the excess CHD events occurred, the expectation was that hormone therapy would reduce the risk of CHD. After participants were warned of an excess risk at about 3 years of follow-up, this expectation changed. This warning was in response to the excess risk; it did not cause it.
Rossouw JE, Prentice RL, LaCroix AZ, Wu L, Manson JE, Barad D, Barnabei VM, Ko M, Margolis K, Stefanick M. Hormone Therapy and Cardiovascular Risk—Reply. JAMA. 2007;298(6):623-625. doi:10.1001/jama.298.6.624