Practice patterns and individual variability of surgeons performing radical prostatectomy at a high volume academic
Hiten D Patel1, Elizabeth Humphreys2, Bruce J Trock2
1Center for Surgical Trials and Outcomes Research, Department of Surgery, Johns Hopkins Medical Institutions, Baltimore, Maryland.
Purpose:
Regional and local variation in radical prostatectomy rates contribute to overtreatment of low risk prostate cancer. We hypothesized that individual practice variability would be minimal among urologists practicing at a high volume academic center.
Materials And Methods:
We assessed the percent of patients at low risk treated with radical prostatectomy in a given year and comorbidity adjusted life expectancy in an institutional database accounting for temporal trends and disease characteristics. Multivariable linear, spline and logistic models were applied with a hierarchical random effects model to estimate the proportion of variance due to surgeon and temporal effects.
Results:
Of the 20,655 men included in study 11,873 (57.5%) had low risk disease. The Gleason score leading to radical prostatectomy increased with time. Overall the percent of patients at low risk treated with prostatectomy in a given year increased 3.49% yearly from 1991 to 2001 and then decreased by 1.73% yearly from 2001 to 2013. Greater surgeon experience was associated with a higher percent of patients at low risk treated with prostatectomy in a given year from 1991 to 2001 (0.46% per year of experience). High volume surgeons (total more than 1,000 radical prostatectomies) operated on a slightly greater percent of patients at low risk (3.54%). Substantial practice variation existed among surgeons for operating on men 65 years old or older at low risk (OR 3.15, 95% CI 1.62-6.11). There was similar variation when operating on older patients with a life expectancy of less than 15 years. Surgeon level and temporal effects explained 24% and 70%, respectively, of the variance in the percent of patients at low risk treated with radical prostatectomy in a given year.
Conclusions:
At a high volume academic center substantial practice variation exists among surgeons when selecting patients with prostate cancer to undergo radical prostatectomy based on risk and life expectancy even among older patients. In addition to patient decision support tools, publicly reporting individual practice patterns at the provider level could decrease the overtreatment of low risk prostate cancer.
