Selection of Controls in Testing for Alcoholism-Reply

In Reply.—  It is unlikely that hospital status explained our results, because inpatients in the alcoholism treatment program were ambulatory, and were attending group therapy and physical therapy, as well as participating in a number of other physical activities. Moreover, they walked to the clinical chemistry laboratory to have their blood drawn.The remaining concern of Dr Dufour and Mr Irving is our comparing results from two clinical laboratories that used the same multiple analysis instruments and the same chemistry methodologies, but employed different reference values. We acknowledge that this is a difficult problem and welcome the opportunity to clarify further the rationale for our approach.Clinical laboratories adjust manufacture-suggested reference values to take into account biases and variances inherent in a specific instrument, as well as unique characteristics of their patient population and other situations that might influence laboratory values. These considerations are demonstrated in our study by the