Baseline Symptom Score and Flow Rate Can Predict Failure of Medical Treatment of Lower Urinary Tract Symptoms: Prospective 12-year Follow-up Study

Abstract

To assess predictors of failure of medical treatment of lower urinary tract symptoms (LUTS) and evaluate long-term outcome. Between January 1993 and September 1994, 178 men referred with LUTS were prospectively recruited. Assessments included maximum urine flow (Qmax), postvoiding residuals (PVR), transrectal ultrasound (TRUS) prostate volumes, American Urological Association symptom score, and validated quality of life (QOL) and bother scores. Treatment failure was defined as need for transurethral resection of the prostate (TURP). Data were collected at baseline, with final follow-up at 12 years. Univariate and multivariate analyses used Kaplan-Meier and the Cox proportional hazards regression model, respectively, to assess covariates on risk of failure and independent variable prognostic values. Median follow-up was 7.9 years (range, 0-12 years). The mean QOL baseline score of 7.1 improved to 3.6 at 6 years and to 3.3 at 12 years (P <.05 for all). Fifty patients (28%) underwent TURP, with 36 undergoing TURP within 3 years. By univariate analysis, Qmax, AUA and bother scores, and PVR were significantly associated with treatment failure. Independent predictors of failure by multivariate Cox regression were Qmax (>15 vs <15 mL/s; hazard ratio, 3.37; 95% confidence interval, 1.74-6.52; P <.0001) and bother score (<13 vs >13; hazard ratio, 2.37; 95% confidence interval, 1.29-4.35; P = .005). At 12 years, AUA, QOL, and bother scores statistically improved compared with baseline (13 vs 8, 10 vs 6, and 5 vs 2, respectively). Limitations included attrition bias from nonresponders. The beneficial effect of medical treatment persisted for up to 12 years. Treatment is more likely to fail within the first 3 years in patients with low baseline Qmax and high bother scores.