Abstract

OBJECTIVE:

To define and characterize risk factors for the failure of treatment for acute uncomplicated pyelonephritis, as there are few reports available to predict which patients will respond poorly to treatment.

PATIENTS AND METHODS:

Retrospective univariate and multivariate logistic regression analyses were used to assess data from two prospective clinical trials designed to evaluate antibiotic regimens for urinary tract infections. Data from 522 adult patients with acute uncomplicated pyelonephritis were analysed.

RESULTS:

The cure rate was 442/522 (85%) for patients with acute uncomplicated pyelonephritis. Significant independent predictors for treatment failure included hospitalization at baseline (P < 0.001), the presence of a resistant infecting organism (P < 0.001), diabetes mellitus (P = 0.001), and a history of kidney stones (P = 0.004). The cure rate was 35/74 (47%) for patients with at least one of these four risk factors. Of the 80 patients assessed as treatment failures, only 39 (49%) had at least one of the four risk factors.

CONCLUSION:

Four risk factors for a poor outcome after therapy for acute uncomplicated pyelonephritis were identified. The strongest predictors of failure were the need for hospitalization at baseline and the presence of an organism resistant to the antimicrobial agent used for therapy. Two other factors, diabetes mellitus and a history of kidney stones, might assist clinicians at the initial evaluation to decide which patients are at risk of subsequent treatment failure.