Control charts to identify adverse outcomes in elective colon resection

Abstract

Background: Control charts have been proposed for the measurement of quality in surgical care.

Methods: For each of 181 study hospitals in the 2005 National Inpatient Sample of the Healthcare Cost and Utilization Project database, an average moving range control chart for risk-adjusted postoperative length of stay (RApoLOS) was created for patients discharged alive after elective colectomy. RApoLOS outliers using upper control limits of 2.0σ, 2.5σ, and 3.0σ were correlated to coded complications (CCs). Hospital costs were correlated to RApoLOS outliers and CCs.

Results: Of 13,118 live discharges, 902 (6.9%) were outliers using a 3.0σ upper control limit, 1,350 (10.3%) were 2.5σ outliers, and 2,053 (15.7%) were 2.0σ outliers. CCs were identified in 92.7% of 3.0σ outliers, in 81.3% of 2.5σ outliers, and 70.6% of 2.0σ outliers. Increased costs were associated with RApoLOS outliers and poorly with CCs.

Conclusions: Average moving range control charts for RApoLOS outliers are valid tools for measurement of surgical quality and costs.