Time Series Analysis for Selected Clinical Indicators from the Quality and Outcomes Framework - 2001-2006
This statistical publication was pre-announced as ˜Quality and Outcomes Time Series Analysis in QRESEARCH 2001“2006'.
This report describes a five year time series of nineteen selected quality indicators from general practice based on the national GMS contract for GPs (the Quality and Outcomes Framework).
The data was collected from the QRESEARCH database, which collates anonymous electronic medical records. The present analysis was restricted to 498 practices in England and to those with complete data for the five year study period 2001-2006. The population covered was 3.4 million patients.
QRESEARCH is a joint partnership between the University of Nottingham and EMIS (supplier of IT systems to UK general practices). The quality indicators which have been selected for this release reflect partial outcome measures in several major disease groups (coronary heart disease, stroke, hypertension, chronic kidney disease, epilepsy and diabetes).
It spans the period during which the QOF was introduced into clinical General Practice and therefore offers an opportunity to examine changes before and after the introduction of the new GMS contract.
there was an increase in the average percentage achievement of all the quality indicators included in the study across the five year study period in all ages for both men and women e.g. the proportion of patients with CHD who have a total cholesterol level lower than or equal to the nationally recommended target (National Service Framework for CHD), increased from 27 per cent in 2001 to 69 per cent in 2006.
Patients with CHD who had a blood pressure reading below or equal to nationally agreed targets (NICE clinical guideline), has increased from 57 per cent in 2001 to 85 per cent in 2006
there was also a very good correspondence of prevalence rates between the data derived from the sample practices and national QOF data for England 2005/06. The prevalence of CHD in both data sources is 35.7 per 1000, hypertension had slightly lower prevalence in QRESEARCH, April 2006, (116 vs. 119 per 1000) and stroke prevalence was slightly higher (16.8 vs.15.7 per 1000)
as expected, the recorded prevalence of all six diseases (CHD, stroke, hypertension, chronic kidney disease, epilepsy and diabetes) increased over the five year study period. The increase in the prevalence of hypertension was the most marked whilst the increase in the prevalence of epilepsy was only marginal
some of the increases in the recorded disease prevalence are likely to be due to the changes in the â€˜Read' codes used to record diseases. For example, the Read codes to denote chronic kidney disease have only recently been introduced and so the increase which is apparent in 2006 will partly reflect the change in usage of Read codes within clinical practice
however, some of the increase of the recorded disease prevalence is likely to be due to the ageing population, improved screening and recording, changes in diagnostic thresholds and increased incidence or improved survival
the increases of disease prevalence across the five year study period were consistent across different age bands for both men and women.