Mustafa Raoof1, Syed Nabeel Zafar2, Philip H G Ituarte3, Robert S Krouse4, Kurt Melstrom3. 1. Department of Surgery, City of Hope National Medical Center, Duarte, California. Electronic address: mraoof@coh.org. 2. Department of Surgery, MD Anderson Cancer Center, Howard University, Washington, DC, Houston Texas. 3. Department of Surgery, City of Hope National Medical Center, Duarte, California. 4. Department of Surgery, University of Pennsylvania and Surgical Service Line, Michael J. Crescenz Veterans Affairs Medical Center, Philadelphia, Pennsylvania.
Abstract
BACKGROUND: Investigating methods to assess the quality of cancer surgery and then benchmarking hospitals on these quality indicators can lead to improvements in cancer care in the United States. We sought to determine the utility of lymph node count as a quality metric. METHODS: We performed a retrospective analysis of the California Cancer Registry database (2004-2011) merged with Office of Statewide Health Planning and Development inpatient database. Patients with locally advanced rectal adenocarcinoma who underwent neoadjuvant therapy and resection were included. Hospital quality score was defined as the proportion of patients at a particular hospital that had adequate examination with at least nine lymph nodes. High-quality score hospitals were those that retrieved nine or more nodes among ≥25% of operations. A multivariate Cox proportional hazards (standard and shared frailty) model was used to determine differences in overall survival adjusting for age, hospital volume, race, sex, insurance, comorbidity, T-stage, response to neoadjuvant therapy, adjuvant chemotherapy, and teaching hospital status as covariates. RESULTS: A total of 2704 patients were treated at 228 hospitals (low-scoring hospital = 85 and high-scoring hospital = 143). Patient- and disease-specific characteristics were similar between the groups. Socioeconomic status and hospital characteristics were strongly associated with score status. High-scoring hospitals had higher sphincter preservation (P = 0.004), lower complications (P = 0.021), and a trend toward lower mortality (P = 0.079). Care at high-scoring hospitals independently predicted overall survival (hazard ratio: 0.74; 95% confidence interval: 0.61-0.90; P = 0.003). CONCLUSIONS: This study demonstrates that hospital quality score based on lymph node count can be used to identify underperforming hospitals.
BACKGROUND: Investigating methods to assess the quality of cancer surgery and then benchmarking hospitals on these quality indicators can lead to improvements in cancer care in the United States. We sought to determine the utility of lymph node count as a quality metric. METHODS: We performed a retrospective analysis of the California Cancer Registry database (2004-2011) merged with Office of Statewide Health Planning and Development inpatient database. Patients with locally advanced rectal adenocarcinoma who underwent neoadjuvant therapy and resection were included. Hospital quality score was defined as the proportion of patients at a particular hospital that had adequate examination with at least nine lymph nodes. High-quality score hospitals were those that retrieved nine or more nodes among ≥25% of operations. A multivariate Cox proportional hazards (standard and shared frailty) model was used to determine differences in overall survival adjusting for age, hospital volume, race, sex, insurance, comorbidity, T-stage, response to neoadjuvant therapy, adjuvant chemotherapy, and teaching hospital status as covariates. RESULTS: A total of 2704 patients were treated at 228 hospitals (low-scoring hospital = 85 and high-scoring hospital = 143). Patient- and disease-specific characteristics were similar between the groups. Socioeconomic status and hospital characteristics were strongly associated with score status. High-scoring hospitals had higher sphincter preservation (P = 0.004), lower complications (P = 0.021), and a trend toward lower mortality (P = 0.079). Care at high-scoring hospitals independently predicted overall survival (hazard ratio: 0.74; 95% confidence interval: 0.61-0.90; P = 0.003). CONCLUSIONS: This study demonstrates that hospital quality score based on lymph node count can be used to identify underperforming hospitals.