To assess rates of treatment-related complications after radical prostatectomy or radiotherapy monotherapy, using propensity score matching to account for baseline differences between these patient populations.

On the basis of a population-based study of men undergoing surgery or radiotherapy for prostate cancer in Ontario between 2002 and 2009, we undertook a propensity score–matched analysis including age, comorbidity, and year of treatment to assess treatment-related complication end points. These included hospital admission; urologic, rectal, or anal procedures; open surgeries; and secondary malignancies.

From the original cohort of 32,465 patients, 15,870 (48.9%) had surgery and 16,595 (51.1%) had radiation. Propensity score matching produced 8797 pairs (17,594 patients). Among these, when compared with patients treated with surgery, those treated with radiation experienced fewer admissions to hospital (hazard ratio [HR], 0.85; 95% confidence interval [CI], 0.78-0.92) and urologic procedures (HR, 0.50; 95% CI, 0.46-0.53) at year 1 but higher rates at year 3 (HR, 5.65; 95% CI, 4.61-6.91 and HR, 1.86; 95% CI, 1.62-2.13, respectively) and year 5. Although there was no significant difference in open surgeries at year 1, patients undergoing radiotherapy were at higher risk by year 3 (HR, 2.06; 95% CI, 1.23-3.47) and this rose by year 5. Over the study period, patients undergoing radiotherapy experienced more rectal-anal procedures (HR, 2.64; 95% CI, 2.37-2.95) and were diagnosed with more secondary malignancies (HR, 2.44; 95% CI, 1.16-5.14). Direct matching produced similar results.

From a propensity score–matched analysis, we found that patients undergoing radiation therapy for prostate cancer had higher rates of long-term complications in all 5 categories studied than patients undergoing surgery.