Hospital Variations in Clinical Complications and Patient-reported Outcomes at 2 Years After Immediate Breast Reconstruction.

OBJECTIVE: Our objectives were to investigate case-mix adjusted hospital variations in 2-year clinical and patient-reported outcomes following immediate breast reconstruction.

BACKGROUND: Over the past few decades, variations in medical practice have been viewed as opportunities to promote best practices and high-value care.

METHODS: The Mastectomy Reconstruction Outcomes Consortium Study is an National Cancer Institute-funded longitudinal, prospective cohort study assessing clinical and patient-reported outcomes of immediate breast reconstruction after mastectomy at 11 leading medical centers. Case-mix adjusted comparisons were performed using generalized linear mixed-effects models to assess variation across the centers in any complication, major complications, satisfaction with outcome, and satisfaction with breast.

RESULTS: Among 2252 women in the analytic cohort, 1605 (71.3%) underwent implant-based and 647 (28.7%) underwent autologous breast reconstruction. There were significant differences in the sociodemographic and clinical characteristics, and distribution of procedure types at the different Mastectomy Reconstruction Outcomes Consortium Study centers. After case-mix adjustments, hospital variations in the rates of any and major postoperative complications were observed. Medical center odds ratios for major complication ranged from 0.58 to 2.13, compared with the average major complication rate across centers. There were also meaningful differences in satisfaction with outcome (from the lowest to highest of -2.79-2.62) and in satisfaction with breast (-2.82-2.07) compared with the average values.

CONCLUSIONS: Two-year postoperative complications varied widely between hospitals following post-mastectomy breast reconstruction. These variations represent an important opportunity to improve care through dissemination of best practices and highlight the limitations of extrapolating single-institution level data and the ongoing challenges of studying hospital-based outcomes for this patient population.