UC Davis

The Anesthesia Quality Institute (AQI) promotes improvements in clinical care outcomes by managing data entered in the National Anesthesia Clinical Outcomes Registry (NACOR). Each case included in NACOR is classified as "performance met" or "performance not met" and expressed as a percentage for a length of time. The clarity associated with this binary classification is associated with limitations on data analysis and presentations that may not be optimal guides to evaluate the quality of care. High compliance benchmarks present another obstacle for evaluating quality. Traditional approaches for interpreting statistical process control (SPC) charts depend on data points above and below a center line, which may not provide adequate characterizations of a QI process with a low failure rate, or few possible data points below the center line. This article demonstrates the limitations associated with the use of binary datasets to evaluate the quality of care at an individual organization with QI measures, describes a method for characterizing binary data with continuous variables and presents a solution to analyze rare QI events using g charts.