Stephen J.LurieMD, PhD, Senior EditorIndividualAuthorJody W.ZylkeMD, Contributing EditorIndividualAuthor
In Reply: Dr Altman is correct about the differences
between allocation concealment and double-blinding, which we, unfortunately,
did not clearly differentiate. However, 18 of the 20 studies were placebo-controlled.
In placebo-controlled trials it is unlikely that allocation is unconcealed.
Only 1 of the remaining studies described the randomization process, which
used sealed envelopes.1 Envelopes are often
judged an "adequate" method of concealed allocation.2
However, envelopes can be tampered with, as demonstrated in a surgical trial
in which 44% of patients allocated using envelopes had suspected spoiled randomization
envelopes.3 Importantly, the RRs between
placebo and nonplacebo trials were not materially different.
Torgerson DJ, Bell-Syer SEM. A Meta-analysis of Hormone Replacement Therapy for Fracture Prevention—Reply. JAMA. 2001;286(17):2096-2097. doi:10.1001/jama.286.17.2091