Clinical Risk Adjustment of Patient Outcomes when Sample Size Is Small

• Activity Number: 152
• Type: Luncheons
• Date/Time: Monday, August 8, 2005 : 12:30 PM to 1:50 PM
• Sponsor: Section on Health Policy Statistics
• Author(s): James Cubellis*+ and Martin Kleinbart
• Companies: Hoag Memorial Hospital Presbyterian and Hoag Memorial Hospital Presbyterian
• Keywords: risk-adjustment ; quality ; hospital outcomes ; spinal fusion

Abstract

Hoag Hospital sought to capture the extent to which surgical complication outcomes resulted from quality/processes versus patient severity and comorbidities. Clinical risk adjustment was the solution, but traditional regression methods were not available due to the absence of existing models and the small patient population. The risk stratification method was employed instead, and the steps involved included conducting a literature review on spine surgery outcomes, patient risk factors, and statistical methodology; identifying key risk factors and comorbidities; and ranking them by clinical and statistical importance as related to complications. Finally, patients were stratified into high- and low-risk groups according to the presence/absence of these factors and complication rates were trended over time using statistical process control methods. Results: High-risk patients experience complication rates 2-3 times higher than low-risk patients, Hoag's "true" adjusted complication rates are comparably lower than other study populations, and risk-adjusted outcomes have enabled Hoag to focus performance improvement activities on high-risk patients, where the effort is needed.