We thank Dr. Slater for his comments and congratulate him on the successes he has had in improving the quality of prescribing in nursing homes. We too are looking at interventions to improve prescribing.

As Dr. Terplan notes, our study [1] did not examine outcomes of inappropriate drug use, although we soon plan to report data on selected outcomes. Using explicit criteria without clinical outcomes, however, does not distort the results of evaluating prescribing quality. It may be, as he suggests, that many of these medications were not needed, but we did not study the indications for therapy. Such studies are needed, but the clinical information generally available in nursing-home records makes this task difficult, if not impossible. Dr. Terplan is also concerned about the time spent in nursing homes recording data. Some might disagree that doctors spend too much time on this activity, given the inadequacies of clinical information currently available in those records. Our study required no additional data collection. One advantage of computerized drug utilization review is that it uses existing data without expensive additional technology.

Dr. Ashburn is concerned that explicit criteria limit the freedom that doctors have in prescribing in nursing homes, and he states that we have done a disservice to front-line physicians. Several of the investigators are front-line physicians, and this study grew out of our observations in nursing homes. In contrast to Dr. Ashburn's perspective, we believe our study shows the disservice that some doctors are doing to their nursing-home patients. Further, drug utilization review does not limit prescribing options. Although space did not allow us to list preferred substitutions in our Annals article, examples were given in our previous publication [2]. Dr. Ashburn's own examples indicate that doctors have many choices available when prescribing. Our explicit criteria do not say that nonsteroidal anti-inflammatory agents should be avoided, but rather that specific ones should be. Our criteria do not say that antacids or histamine-blocking agents should be avoided, but rather that their duration and dose should be modified. Clearly, chronic depression requires treatment in older persons: Our criteria simply state that use of some antidepressants should be avoided in older persons. Dr. Ashburn also comments on the frequent use of medications prescribed on an as-needed basis. No current data show that such medications are used frequently because they are needed. In fact, many such prescriptions are written routinely without evaluation of need. Finally, we agree with Dr. Ashburn that the reimbursement for attentive care in nursing homes is inadequate. Liability, however, is not the issue. The issue is quality of care; periodic assessment is essential to understand how well we are doing and what we need to improve.

We appreciate the basis of Dr. Segal's pessimism, but we are more hopeful than he is about the future.