A. N. Kothari1, S. A. Brownlee1, E. C. He1, J. Rambo1, D. M. Hayden1, J. M. Eberhardt1, P. C. Kuo1, T. L. Saclarides1 1Loyola University Chicago Stritch School Of Medicine,Surgery,Maywood, IL, USA
Introduction:
Hospital quality is commonly measured using postoperative mortality rates. However, recent data demonstrate that patients may perceive several states of functional disability as worse than death. These include incontinence, ventilator dependence, being bed bound, confusion, and reliance on a feeding tube. The objective of this study was to compare how hospital performance would be ranked using traditional or patient-preferred outcomes for patients undergoing elective colon resections.
Methods:
The Healthcare Cost and Utilization Project State Inpatient Databases for California, Florida, Iowa, and New York were used to identify patients undergoing open colon resection. Traditional measures of postoperative outcomes included 1-year mortality. Patient-preferred measures of postoperative outcomes were based on prior study identifying states of functional disability that patients define as being as bad as or worse than death. Separate risk-adjustment models were constructed for each outcome using mixed-effects logistic regression models fit using patient-level covariates including diagnosis and procedure, and a random-effect term for the hospital. Expected event rates were derived from final models, with observed over expected ratios calculated for each hospital (with 95% confidence interval).
Results:
A total of 42 116 patients at 663 hospitals met our inclusion criteria. Overall mortality of the study population was 11.0% at 1 year, while 3.5% of patients had functional disability at 1 year. At the hospital level, unadjusted rates of mortality ranged from 0.0 – 24.1% across hospitals, while rates of functional disability ranged from 0.0 – 11.8%. Following risk adjustment, hospitals were ranked based on observed over expected ratios for both mortality and functional disability. There was poor agreement between ranks assigned using each outcome (r=0.22, ICC=0.22 [0.14 – 0.29], kappa=0.006, p<0.001). A total of 69 hospitals were high performing for both outcomes (top quartile) and a total of 72 hospitals were low performing (bottom quartile) for both outcomes.
Conclusion:
Hospital rankings based on traditional outcomes, including mortality, may not adequately assess hospital performance for patient-preferred outcomes related to long-term postoperative functional disability. Transparency regarding the metrics used for publicly reported ranking systems and consideration of functional outcomes are important for adequately measuring the quality of a hospital’s surgical care.