Using quality indicators to predict inspection ratings: cross-sectional study of general practices in England

BACKGROUND: The Care Quality Commission regulates, inspects, and rates general practice providers in England. Inspections are costly and infrequent, and are supplemented by a system of routine quality indicators, measuring patient satisfaction and the management of chronic conditions. These indicators can be used to prioritise or target inspections. AIM: To determine whether this set of indicators can be used to predict the ratings awarded in subsequent inspections. DESIGN AND SETTING: This cross-sectional study was conducted using a dataset of 6860 general practice providers in England. METHOD: The indicators and first-inspection ratings were used to build ordered logistic regression models to predict inspection outcomes on the four-level rating system (‘outstanding’, ‘good’, ‘requires improvement’, and ‘inadequate’) for domain ratings and the ‘overall’ rating. Predictive accuracy was assessed using the percentage of correct predictions and a measure of agreement (weighted κ). RESULTS: The model correctly predicted 79.7% of the ‘overall’ practice ratings. However, 78.8% of all practices were rated ‘good’ on ‘overall’, and the weighted κ measure of agreement was very low (0.097); as such, predictions were little more than chance. This lack of predictive power was also found for each of the individual domain ratings. CONCLUSION: The poor power of performance of these indicators to predict subsequent inspection ratings may call into question the validity and reliability of the indicators, inspection ratings, or both. A number of changes to the way data relating to performance indicators are collected and used are suggested to improve the predictive value of indicators. It is also recommended that assessments of predictive power be undertaken prospectively when sets of indicators are being designed and selected by regulators.