PITCH: Does the negative correlation between satisfactory glycaemic control and deprivation and non-white ethnicity still persist in English general practices?

Problem

Following the introduction of the Quality and Outcomes Framework (QOF) in 2004, glycaemic control, as measured by glycated haemoglobin (HbA1c), improved in patients with diabetes, but there were ethnic and social disparities in the proportion of those with satisfactory control (<64mmol/mol or 7.5%) in 2009; overall, these were in the region of 50%.

We could find no recent studies examining this area. Having up to date knowledge is relevant to better understanding and addressing the current pressures on primary care, including the contribution of the increasing prevalence of diabetes on practices' workloads.

Our research question was to determine whether these disparities have persisted in English general practices, 13 years after the introduction of OOF.

Approach

Cross-sectional study in 7,281 English general practices. Published practice level data was obtained from NHS Digital (QOF), General Practice Survey and Public Health England.

Using variables for the nearest corresponding year, partial correlations were carried out in each of the four most recent financial years (2013-14 to 2016-17) between the study variable (% of the practice diabetes practice register with most recent HbA1c <64mmol/mol - DM08) and 3 predictors, the practice's 2015 Index of Multiple Deprivation (IMD) score, % practice list with Black ethnicity, and % practice list with South Asian ethnicity.

The following confounder variables were included: % on diabetes register, list size, % aged 75 years or more, numbers of GPs and practice nurses per 1,000 patients, adjusted payments per registered patient, % smokers, % with chronic condition, % claiming benefits, % having and able to see preferred GP, % with good experience of practice.

Findings

The partial correlations between satisfactory glycaemic control and the 3 predictors were slight and negative (all statistically significant): IMD -0.092 (2013-14), -0.13 (2016-17); Black ethnicity -0.048 (2013-14), -0.050 (2016-17); South Asian ethnicity -0.11 (2013-14), -0.051 (2016-17). Thus, social and ethnic disparities in the proportions of those with satisfactory glycaemic control are still present, but small.

The achievement of satisfactory control improved further since the introduction of QOF, but the mean % of the practice diabetes practice register with a most recent HbA1c of <64mmol/mol has remained fairly level over the last 4 years at between 68 and 69%.

Consequences

As diabetes complications are expensive, social and ethnic disparities between populations and the recent plateauing in achievement for this important intermediate outcome are of concern, especially as the prevalence of diabetes continues to rise.

Further research could investigate why this is happening and what other factors might predict variations between practices. A better understanding of these may help to inform policies that seek to further improve standards of care and to reduce disparities between local populations.