Study type: Prevalence (prospective cohort with good follow up).
Level of evidence: 1a.
Objective: To examine contemporary (1989-2004) trends in partial nephrectomy (PN) within the Surveillance, Epidemiology and End Results (SEER) database, as among other considerations, a survival benefit due to avoidance of surgically induced renal insufficiency distinguishes PN from radical nephrectomy (RN).
Patients and methods: Diagnostic, stage and surgical codes of patients with T1-2N0M0 renal cell carcinoma treated with either PN or RN were assessed. Proportions, trends and multivariable logistic regression models tested the predictors of the use of PN.
Results: Of 19 733 assessable patients, 2614 (13.2%) and 17 119 (86.8%), respectively, had PN or RN. The use of PN decreased with increasing tumour size, was more frequent in younger patients and increased with more contemporary years of surgery (all P < 0.001). Intriguingly, there was important geographical variability (P < 0.001), e.g. in the San Francisco-Oakland Metropolitan Area the absolute PN rate was 16.4%, vs 7.6% in New Mexico (P < 0.001). In multivariable analyses, tumour size, age, year of surgery, gender and SEER registries were independent predictors of PN use.
Conclusion: Although as expected the rate of PN use increased over time, unexplained variability remained. For example, gender and SEER registries affected the likelihood of PN. These variables warrant further analyses to reduce unnecessary variability and to maximize PN use and its benefit.