In her editorial, Dr Fitzgerald1 discusses the limitations to the generalizability of the findings reported in our article.2 Dr Fitzgerald correctly stresses the important issue that an individualized treatment regimen does not apply to the usual population of alcoholic patients admitted to most hospitals in the United States. However, it is not correct to say that only 6% of eligible patients (117 of 2000) were ultimately included in the study since the final inclusion rate was 48.1% (117 of 243, see the article's "Results" section, p 1119).
Daeppen J. Percentage Set Straight. Arch Intern Med. 2002;162(20):2379. doi: