Identifying poorly performing general practices in England: a longitudinal study using data from the quality and outcomes framework

OBJECTIVE: to determine the characteristics of general practices which perform poorly in terms of Quality and Outcome (QOF) performance indicators in England's NHS.

METHOD: retrospective, four year longitudinal study, 2005 to 2008. Data were obtained from 8515 practices (99% of practices in England) in year 1, 8264 (98%) in year 2, 8192 (98%) in year 3 and 8256 (99%) in year 4.

OUTCOME MEASURES: QOF performance scores; social deprivation (IMD-2007) and ethnicity from the 2001 national census; general practice characteristics.

RESULTS: we identified a cohort of 212 (2.7%) practices which remained in the lowest decile for total QOF scores in the four years following the introduction of the QOF. A total of 705,386 patients were registered at these practices in year 4. These practices were more likely to be singlehanded (odds ratio [OR], 13.8), non-training practices (OR, 3.9) and located in deprived areas (OR, 2.6; most vs least deprived quintiles). General practitioners (GPs) in these practices were more often aged ≥ 65 years or more (OR, 7.3; mean GP age ≥ 65 years vs <45 years), male (OR 2.0), UK qualified (OR 2.0) with small list sizes (OR 3.2; list size <1000 vs 1500-2000 patients). We identified individual QOF indicators which were poorly achieved. The reported prevalence of most chronic diseases was lower in the poorly performing cohort.

CONCLUSIONS: a small minority of practices have remained poor performers in terms of measurable performance indicators over a four-year period. The strongest predictors of poor QOF performance were singlehanded and small practices, and practices staffed by elderly GPs.