Differential receipt of sentinel lymph node biopsy within practice-based research networks.

Abstract

6104 Background: Sentinel lymph node biopsy (SLNB) for breast cancer was introduced into clinical practice in the late 1990s as an alternative to axillary lymph node dissection (ALND). Provider-based research networks (PBRNs) are believed to promote diffusion of innovations like SLNB into clinical practice; however, evidence of this association is limited. This study examines the diffusion of SLNB for early-stage breast cancer through the Community Clinical Oncology Program (CCOP), a community-based PBRN. Methods: We identified women undergoing breast conserving surgery with axillary staging for stage I or II breast cancer between January 2000 and December 2003 using Surveillance Epidemiology and End Results-Medicare data (n=6,226). The primary outcome was receipt of SLNB vs. ALND, and exposure was care received from CCOP physicians or institutions between diagnosis and surgery. Exposure was quantified as both a binary measure of ever seeing a CCOP, and as a proportion of all their claims associated with a CCOP. Covariates included race, age, marital status, education, Medicaid eligibility, comorbidity, tumor grade, stage, estrogen receptor status, year of diagnosis, SEER region, and other institutional characteristics such as NCI center designation, cooperative group, and medical school affiliation. Multivariable generalized linear modeling with generalized estimating equations was used to measure association between CCOP exposure and receipt of SLNB. Results: Women who received a higher proportion of their care from a CCOP-affiliated physician or hospital were more likely to receive SLNB. A 10% increase in the proportion of CCOP-affiliated claims was associated with a greater odds of receiving SLNB (OR 1.14; 95% CI 1.08, 1.20), after controlling for covariates. Similarly, sensitivity analysis of the binary indicator of CCOP exposure also showed greater odds of receiving SLNB (OR 1.32; 95%CI 1.01, 1.74). Conclusions: The quality of cancer care delivered in community settings can be influenced by provider-based research networks. Our findings contribute to the growing body of evidence that community-based PBRNs can facilitate adoption of cancer innovations outside of academic medical centers.