Identifying outlying proportions in the field of health care quality indicators

We will first walk through the short history of applying statistical quality control in health care. Then we will focus on the problem of identifying outliers among overdispersed proportions and random-denominator ratios. I will present a method which I have developed for that purpose (double-square-root chart with linear regression through the origin and a corresponding funnel plot) and the research (in collaboration with Rok Blagus) that compares the method with the established approaches (general-purpose, outlier detection tests and rules, Laney's p' control chart for cross-sectional data, and Spiegelhalter's multiplicative and additive regression approach).