A variation in the value of colectomy for cancer across hospitals: mortality, readmissions, and costs.

INTRODUCTION: Although hospital variation in costs and outcomes has been described for patients undergoing operation, the relationship between them is unknown. The purpose of this study was to evaluate this relationship among patients undergoing colon resection for cancer and identify characteristics of "high-quality, low-cost" hospitals.

METHODS: We identified adult patients who underwent colon resection for cancer in California, Florida, and New York from 2009 to 2010. We estimated hospital-level, risk-standardized 30-day hospital costs, in-hospital mortality rates, and 30-day readmission rates by using hierarchical generalized linear models. Costs were compared between hospitals identified as low, average, and high performers.

RESULTS: The final sample included 14,790 patients discharged from 389 hospitals. After adjusting for case mix, variation was noted in risk-standardized costs (median = $26,169, inter-quartile range [IQR] = $6,559), in-hospital mortality (median = 1.8%, IQR = 2.3%), and 30-day readmission (12.2%, IQR = 0.7%) rates. Minimal correlation was noted between a hospital's costs and outcomes, with similar costs noted across hospital performance groups (low = $25,994 vs average = $26,998 vs high = $25,794, P = .19). High-quality, low-cost hospitals treated a greater percentage of Medicare beneficiaries, approached fewer cases laparoscopically, and trended toward greater volume.

CONCLUSION: Hospital costs are not correlated with outcomes in this population. More work is needed to identify means of providing high-quality care at lesser costs.