Economic consequences of two drug-use control systems in a teaching hospital.

Length of stay (LOS), total cost per admission (TCA), and pharmacy cost per admission (DCA) were determined for two drug-use control systems in a 1058-bed university hospital; a centralized unit dose drug distribution system served as a control. The two study systems were (1) pharmacist monitoring of drug therapy in the patient-care area and (2) centralized pharmacist monitoring of computerized patient profiles. LOS data were collected retrospectively for 659 patients admitted during a seven-month control interval. LOS, TCA, and DCA data were collected prospectively for 496 patients admitted during a five-month experimental interval. Each study system was assigned to one of three teams making rounds among intact patient groups. LOS differences were compared between intervals and by month. After corrections were made for differences in patient mix, the drug-use control system in which pharmacists were assigned to the patient-care area yielded a 1.5-day-shorter average LOS, $1293 lower average TCA (p less than 0.05), and $155 lower average DCA than under the unit dose system. The drug-use control system in which pharmacists were assigned to monitor patients' drug therapy from a central location was associated with a 0.13-day-shorter average LOS, $235 lower average TCA, and $55.13 lower average DCA than under the unit dose system. No systematic differences between teams, other than drug-use control system, appeared to explain the differences in LOS, TCA, and DCA. A drug-use control system based in a patient-care area, overseen by clinically experienced pharmacists, may result in shorter LOSs and lower total costs than centralized systems for general-medical inpatients of teaching hospitals.