Introduction: It is customary to consider deaths that occur within 90 days of surgery as caused by that surgery. However, such practice may overestimate the true short-term mortality rates after radical prostatectomy (RP). Indeed, treatment-unrelated events might affect short-term mortality rates. We assess RP-specific excess short-term mortality.
Methods: We performed a retrospective analysis of a population-based cohort of 59 010 patients (RP = 28 281 and external beam radiation therapy [EBRT] as reference group, n = 30 729) who were treated between 1998 and 2005 for non-metastatic prostate cancer. Using univariate and multivariate logistic regression analyses, we assessed the rates of 30-, 60- and 90-day mortality after either RP or EBRT.
Results: Within the cohort, 30-, 60- and 90-day mortality rates were 0.2, 0.5 and 0.6%, and 0.1, 0.4 and 0.6% for RP and EBRT patients, respectively. This resulted in overall 30-, 60, and 90- day mortality differences of 0.1, 0.1 and 0%, respectively. After stratification according to age and Charlson comorbidity index (CCI), the magnitude of these differences increased up to 3.2% in favour of EBRT in patients aged >75 years with CCI ≥2. In multivariable analysis, rates of 30-, 60- and 90- day mortality were 5.2-, 1.8- and 1.3-fold higher after RP than EBRT, respectively. Our study is limited by its non-randomized design.
Conclusion: Overall, absolute short-term mortality rates after RP are comparable to those of EBRT. The difference decreases over time: 90 days <60 days <30 days. Nonetheless, their magnitude is far from trivial in the elderly and sickest patients.