Population-based epidemiologic analysis of acute pyelonephritis

Background. Acute pyelonephritis is a potentially severe disease for which there are few population-based studies. We performed a population-based analysis of trends in the incidence, microbial etiology, antimicrobial resistance, and antimicrobial therapy of outpatient and inpatient pyelonephritis. Methods. A total of 4887 enrollees of Group Health Cooperative, based in Seattle, Washington, who received an International Classification of Diseases, Ninth Revision, Clinical Modification, diagnosis of acute pyelonephritis from 1997 through 2001 were identified using computerized records. Diagnoses were linked to urine culture and antibiotic prescription data. Case patients (n = 3236) included subjects who had received an inpatient or culture- confirmed outpatient diagnosis of acute pyelonephritis. Results. Among the female population, annual rates of outpatient and inpatient pyelonephritis were 12-13 cases per 10,000 population and 3-4 cases per 10,000 population, respectively; among the male population, the rates were 2-3 cases per 10,000 population and 1-2 cases per 10,000 population, respectively. Rates were relatively stable from year to year. Incidence was highest among young women, followed by infants and the elderly population. The ratio of outpatient to inpatient cases was highest among young women (ranging from 5:1 to 6:1). Escherichia coli caused 80% of cases of acute pyelonephritis in women and 70% of cases in men and was less dominant in older age groups. Among E. coli strains, the rate of ciprofloxacin resistance increased from 0.2% of isolates to 1.5% of isolates (P = .03), and the rate of trimethoprim-sulfamethoxazole resistance decreased from 25% of isolates to 13% of isolates (P = .01) from 1997 to 2001. Among outpatient cases, the rate of fluoroquinolone use increased from 35% to 61%, whereas the rate of trimethoprim-sulfamethoxazole use decreased from 53% to 32% over the 5-year period (P = .01). Conclusions. This comprehensive, population-based analysis adds to our limited knowledge of the epidemiology of acute pyelonephritis, especially among outpatients, in whom the majority of cases now occur.