Publication Abstract

Authors: Wang K, Sheets NC, Basak R, Chen RC

Title: Ascertainment of postprostatectomy radiotherapy for prostate cancer in the Surveillance, Epidemiology, and End Results database.

Journal: Cancer 122(19):3069-74

Date: 2016 Oct

Abstract: BACKGROUND: Surveillance, Epidemiology, and End Results (SEER) data are frequently used to examine receipt of adjuvant radiotherapy (RT), but to the authors' knowledge the accuracy of data regarding second-course treatments is unknown. METHODS: Using SEER-Medicare-linked data, the authors identified a cohort of men who underwent radical prostatectomy for localized prostate cancer with indications for RT due to adverse pathologic risk factors. Receipt of RT was compared between the SEER database and Medicare claims, with the latter considered to be the "gold standard." Multivariable logistic regression was used to assess factors associated with ascertainment of RT in SEER. RESULTS: A total of 3842 men were analyzed, 749 of whom were found to have Medicare claims for RT within 1 year of undergoing prostatectomy. SEER ascertainment of postprostatectomy RT was 56% overall: 76% among patients who received RT within 2 months of prostatectomy, 73% among patients who received RT between 2 to 4 months after prostatectomy, 63% among patients who received RT between 4 to 6 months after prostatectomy, 44% among patients who received RT between 6 to 8 months after prostatectomy, and 21% among patients who received RT between 8 to 12 months after prostatectomy. On multivariable analysis, increasing time from prostatectomy to RT was found to be significantly associated with decreased SEER ascertainment (odds ratio, 0.70 per month; P<.001). There also was variation noted by SEER region and urban/rural locale. CONCLUSIONS: SEER underascertains the receipt of postprostatectomy RT compared with Medicare claims, and the magnitude of the underascertainment increases with longer time between prostatectomy and RT. These findings have direct implications for the use of SEER data alone to assess patterns of care and guideline concordance for second-course treatment. Cancer 2016;122:3069-3074. 2016 American Cancer Society.