The relationship between general practice characteristics and quality of care: a national survey of quality indicators used in the UK Quality and Outcomes Framework, 2004–5

0 Address: Department of General Practice & Primary Care, King's College London School of Medicine at Guy's, King's College and St Thomas' Hospitals , London , UK Background: The descriptive information now available for primary care in the UK is unique in international terms. Under the 'Quality and Outcomes Framework' (QOF), data for 147 performance indicators are available for each general practice. We aimed to determine the relationship between the quality of primary care, as judged by the total QOF score, social deprivation and practice characteristics. Methods: We obtained QOF data for each practice in England and linked these with census derived data (deprivation indices and proportion of patients born in a developing country). Characteristics of practices were also obtained. QOF and census data were available for 8480 practices. Results: The median QOF score was 999.7 out of a possible maximum of 1050 points. Three characteristics were independently associated with higher QOF scores: training practices, group practices and practices in less socially deprived areas. In a regression model, these three factors explained 14.6% of the variation in QOF score. Higher list sizes per GP, turnover of registered patients, chronic disease prevalence, proportions of elderly patients or patients born in a developing country did not contribute to lower QOF scores in the final model. Conclusion: Socially deprived areas experience a lower quality of primary care, as judged by QOF scores. Social deprivation itself is an independent predictor of lower quality. Training and group practices are independent predictors of higher quality but these types of practices are less well represented in socially deprived areas. - Background Prior to 2005, few measures of quality of care were routinely available in UK primary care. Routine data were only available for services attracting a separate fee such as cervical smear rates, vaccination rates, child health surveillance, minor surgery and contraceptive services while studies collecting more detailed information data had been restricted to a limited number of volunteer practices. From 2004, however, a new system of reimbursement linked to performance indicators (the 'Quality and Outcome Framework' (QOF)) made available a rich new vein of measures of quality of care [1]. The detail and breadth of descriptive information now available for general practice in Britain is unique in international terms and makes the UK a leader in international quality improvement initiatives in primary care [2]. The Quality and Outcomes Framework is comprised of 146 indicators drawn from the following domains: chronic disease management (76 indicators covering eleven chronic diseases), 'practice organisation' (56 indicators), 'patient experience' (4 indicators) and 'additional services' (10 indicators); there is an additional 'bonus' indicator, 'access' (1 indicator) [3]. Each indicator is weighted, contributing to an overall maximum quality score for each practice of 1050 points. The performance of individual practices in England has been made publicly available. Variability in the quality of care offered by different practices has been a cause of concern for many decades. In part it might be explained by the difficulty of providing good quality care to needy populations and in part by the intrinsic differences in the care offered. Thus when Tudor Hart described the 'inverse care law' some 30 years ago (which argued that the provision of health care was inversely proportional to the health needs of the population), his evidence from general practice was the relative newness of practice buildings in more affluent areas [4]. More detailed information about the quality of primary care provides an opportunity to revisit Tudor Hart's thesis in more depth. For example, a link between social deprivation and poorer quality of care as judged by QOF scores might arise because GPs in more deprived areas have larger, more unmanageable lists, or because they have a higher turnover of patients making it difficult to accumulate sufficient clinical successes. In other words, is there an accumulation of factors, clustering together, that hamper the delivery of high quality care in deprived areas? Overall QOF scores have already been found to be lower in areas of higher social deprivation [5]. However, the earlier survey did not explore possible confounding by other practice variables which may have borne a stronger predictive relationship with the total QOF score. We therefore aimed to gather a broad series of nationally available practice variables in order to explore whether the relationship between overall QOF scores and deprivation remained, allowing for the effect of confounding. Method Quality and Outcomes Framework data We obtained QOF data for all general practices in England from the Health and Social Care Information Centre, Leeds. Data for each of the domains within QOF were analysed collectively and individually: clinical, practice The QOF dataset also provided raw prevalence data for each of the eleven chronic diseases in QOF and the size of each chronic disease register enabled the proportion of 'exception reporting' to be calculated. 'Exception reporting' is the means by which certain patients are exempted from the requirements of clinical quality indicators on the basis of 'unsuitability'. For example, a patient might be exempted from the requirement to achieve a blood pressure indicator if they were unable to tolerate any additional hypotensive medication. There has been concern that some practices could use unduly high levels of 'exception reporting' as a means of achieving higher QOF scores [2]. Practice characteristics A detailed summary of practice characteristics was obtained from the Manchester Primary Care Research and Development Centre, University of Manchester. We obtained data on the following variables: practice list size, age/sex breakdown of registered population, number of full time equivalent GPs, training practice status, Personal Medical Services [6] or General Medical Services [6] status. The practice list turnover was calculated as the number of new patients joining the practice list over the year April 2004 to March 2005 (minus births) divided by the number of registered patients in March 2005. Data were obtained from the National Health Applications & Infrastructure Services Programme (formerly known as the 'Exeter system') which houses registration data for all general practices in England. Census based variables Based on the practice postcode, a list of Lower Layer Super Output Areas [7] was obtained for all practices. These are geographical, 'socially homogenous' areas which are arguably a better link to social measures than political units such as local government Wards. They contain an average population of around 1500. Super Output Areas form the basis of the Index of Multiple Deprivation (IMD), the most wide-ranging and (...truncated)