Assessment of methods and outcomes: using modified inpatient ciprofloxacin criteria in community-based drug use evaluation

OBJECTIVE: To determine if published drug use evaluation (DUE) criteria for inpatients could be modified to describe and evaluate drug therapy in outpatients. DESIGN: Retrospective review of drug profiles and diagnostic codes in outpatients included in the Iowa Medicaid Management Information System database. METHODS: Criteria specifying clinical indication, process indicators, complications, and outcomes were modified from existing inpatient DUE criteria for ciprofloxacin. The Iowa Medicaid database provided demographics, drug profiles, and diagnostic codes for outpatients prescribed ciprofloxacin between March 1 and March 15, 1993. RESULTS: 539 patients were evaluated. Mean (+/- SD) age was 62.1 +/- 23.8 years; 70.9 percent were women, 44.9 percent were nursing home patients who received lower doses (p = 0.04). Of 146 patients (27.1 percent) with infection-related International Classification of Diseases-Clinical Modification, 9th Revision (ICD-9CM) codes, 43.8 percent did not meet indication for use criteria, 15.8 percent met criteria, and 40.4 percent were equivocal/other. Process indicators in 539 patients revealed that 2.4 percent were < 18 years old and potential interactions existed for theophylline (9.3 percent), iron (9.7 percent), warfarin (4.8 percent), sucralfate (2.2 percent). Two patients (0.4 percent) received prenatal vitamins. Dosage and duration of therapy could not be linked to disease severity. Complications of therapy were not evaluable. Outcomes were assessed indirectly by using concurrent antibiotic histories. These histories showed that patients received a mean 2.7 +/- 2.4 (median 2, range 1-16) ciprofloxacin prescriptions during the study period. There was frequent concurrent use of antibiotics, and 29 percent of prescriptions were preceded or followed by additional quinolone therapy within 30 days. CONCLUSIONS: Evaluating appropriateness of therapy with this methodology is feasible if validated ICD-9CM codes are reported consistently in sufficient numbers of patients. Low levels of reporting of relevant disease codes require that large numbers of patients be screened to use this method effectively. Descriptive use data are readily obtained. Data obtained by this method indicated several apparent deficiencies in ciprofloxacin therapy.