Assessment of hospital‐level adjusted breast cancer sentinel lymph node positivity rates

Background/Objectives Proficiency of performing sentinel lymph node biopsy (SLNB) for breast cancer varies among hospitals and may be reflected in the hospital's SLNB positivity rate. Our objectives were to examine whether hospital characteristics are associated with variation in SLNB positivity rates and whether hospitals with lower‐than‐expected SLNB positivity rates have worse patient survival. Methods Using the National Cancer Data Base, stage I to III breast cancer patients were identified (2004‐2012). Hospital‐level SLNB positivity rates were adjusted for tumor and patient factors. Hospitals were divided into terciles of SLNB positivity rates (lower‐, higher‐, as‐expected). Hospital characteristics and survival were examined across terciles. Results Of 438 610 SLNB patients (from 1357 hospitals), 78 104 had one or more positive SLN (21.3%). Hospitals in the low and high terciles were more likely to be low volume (low: RRR, 4.40; 95% CI, 2.89‐6.57; P