Volpicelli et al1 presented an interesting study on the use of naltrexone in the treatment of alcohol- dependence. We make two observations that although seemingly minor nevertheless have an impact on design and reporting in research.
First, we fail to understand why the Brief Psychiatric Rating Scale (BPRS),2 which is customarily used to rate psychosis, was applied to assess psychiatric status; did the authors expect their alcoholic subjects to become psychotic during naltrexone therapy? The futility of having used this scale is apparent from the mean scores in the two groups: 24.3 and 23.3 at baseline and 21.6 and 19.9 at end point. Considering that the BPRS version that was used scores the subject at 17 in the total absence of psychopathology, this scale probably measured emotionality with little sensitivity. More appropriate would have been specific anxiety, depression, and other scales. Also curious is that interrater reliability for the