Did practice location and patient-perceived access predict achievement of good diabetes glycaemic control in English practices recovering from COVID lockdown?

Problem

Better glycemic control leads to better health outcomes for patients with diabetes. Practices offered fewer appointments during lockdown. The subsequent backlog of demand exacerbated the huge pre-existing pressures on practices and adversely affected their capacity to provide planned care for long-term conditions. Our aims were to:1. describe practice achievements of good diabetes glycaemic control post-lockdown.2. investigate whether practice location and patient-perceived access independently predicted variations in these achievements, after adjustments.

Approach

We studied English general practices in two post-lockdown years, 2021-22 and 2022-23, using published data from NHS Digital, Department of Local Government and the GPPS (2020-21 data were incomplete but we used 2019-20 data for comparison). We excluded practices with <750 patients or average NHS payments >£500/patient. Each year’s outcome was %patients aged 17 years or more on a QOF diabetes register whose last glycosylated haemoglobin (HbA1c) was 58mmol/mol or less. We fitted 12 relevant non-correlated independent variables into two ordinary least squares regression models for each year. These variables represented baseline (previous year’s outcome), practice location (deprivation, NHS region, rurality), demography (%white ethnicity), practice characteristics (list size, % on QOF diabetes register, average adjusted NHS payment/patient, FTE GPs and nurses/10,000 patients), and patient-perceived access (% with regular GP, %good experience making an appointment).

Findings

Our study populations were 6381 (2021-22) and 6326 (2021-22) practices [92.2% and 91.4%, respectively, of 6920 practices]. The means and (standard deviations) of %patients <58mmol/mol were 51.6% (7.5%) in 2021-22 and 55.4% (6.8%) in 2022-23, compared to 56.9% (6.6%) in 2019-20.Positive predictors were: • in both years higher previous % with HbA1c <58mmol/mol (2021-22 estimate 0.66 [CI 0.63-0.68, p<0.001]; 2022-23 - 0.65 [0.63-0.67, p<0.001])higher % on diabetes registers (0.15 [0.06-0.24, p=0.001]; 0.08 [0.02-0.15, p=0.016]) %good experience making appointments (0.06 {0.04-0.07, p<0.001], 0.03 [0.02-0.04, p<0.001] • in 2021-22 higher payments (0.003 [0.00-0.01, p=0.037])more GPs (0.06 [0.04-0.07, p=0.017])• in 2022-23 higher %white ethnicity (0.02 [0.01-0.03, p<0.001]) Negative predictors were: • in both years higher IMD (-0.07 [-0.09 to -0.06, p<0.001]; -0.04 [-0.05 to -0.03, p<0.001])living outside London (except South West in 2021-22)• in 2021-22 higher %white ethnicity (-0.02 [-0.03 to -0.01, p<0.001])The other variables were not predictors.Adjusted R-squared values were 0.438 (2021-22) and 0.589 (2022-23).

Consequences

Compared with 2019-20, the %achieving good glycaemic control dropped post-lockdown before partially rebounding. Higher previous achievement, populations living in London or less deprived areas, and better perceived access independently predicted higher achievement in both years. More funding and GP numbers were predictors only in 2020-21, when practices were under even greater pressure and overall achievement was lower. Building on previous achievement, improving access, and considering where patients live should inform strategies to increase the %achieving good glycaemic control. Optimising management of long-term conditions could help to reverse declining life expectancy.