Skip to main content

Publication, Part of

National Diabetes Audit, 2020-21, Type 1 Diabetes

Publication Date:
16 Jun 2022
Geographic Coverage:
England, Wales
Geographical Granularity:
Regions, Clinical Commissioning Groups, Integrated Care Boards
Date Range:
01 Jan 2020 to 31 Mar 2021

Previous Chapter

Additional information

Current Chapter

Appendix

View all

Future publication format

Please note that future publications will be web-based only. This is to improve the accessibility of our publications

16 June 2022 09:30 AM

Page contents

Appendix

HbA1c – specialist service variation in recording HbA1c values

Data was available from 110* specialist services in England and Wales.

As a result of low care process completion and incomplete HbA1c values in cases where the HbA1c check had taken place, the number of values recorded in 2020-21 was considerably lower than in 2019-20.

A logistic regression model** was used to produce expected numbers of HbA1c values recorded for each provider - the model c-statistic was 0.79 (strong).

Figure A1: Funnel plot of specialist service level standardised ratios of recording an HbA1c value, England and Wales, 2020-21
Download the data for this chart Figure A1: Funnel plot of specialist service level standardised ratios of recording an HbA1c value, England and Wales, 2020-21
  • Of 110* specialist services, 33 were more than 2 SD from the expected number of adults with a recorded HbA1c care process.
  • 29 specialist services were above expectation.
  • 4 specialist services were below expectation.
Logistic regression model used for standardisation. Variables used in the models:
  • Sex
  • Age group
  • Duration of diabetes
  • Ethnicity
  • Deprivation quintile
  • BMI group
  • Smoking status
  • Insulin regimen
  • CGM status
  • Frailty status

HbA1c – specialist service variation in proportion of adults meeting an HbA1c treatment target (HbA1c ≤ 58 mmol/mol, 7.5%)

A logistic regression model** was used to produce expected numbers of adults meeting the HbA1c ≤ 58 mmol/mol treatment target in each specialist service provider.

The model c-statistic was 0.64 (weak).

Figure A2: Funnel plot of provider standardised ratios of achieving HbA1c ≤ 58mmol/mol, England and Wales, 2020-21
Download the data for this chart Figure A2: Funnel plot of provider standardised ratios of achieving HbA1c ≤ 58mmol/mol, England and Wales, 2020-21
  • Of 110* specialist services, 38 were more than 2 SD from the number of adults expected to achieve the treatment target.
  • 16 specialist services were above expectation.
  • 22 specialist services were below expectation.

HbA1c – specialist service variation in proportion of adults meeting an HbA1c treatment target (HbA1c ≤ 86 mmol/mol, 10%)

A logistic regression model** was used to produce expected numbers of adults meeting the HbA1c ≤ 86 mmol/mol treatment target in each specialist service provider.

The model c-statistic was 0.73 (moderate).

Figure A3: Funnel plot of specialist service standardised ratios of achieving HbA1c ≤ 86mmol/mol, England and Wales, 2020-21
Download the data for this chart Figure A3: Funnel plot of specialist service standardised ratios of achieving HbA1c ≤ 86mmol/mol, England and Wales, 2020-21
  • Of 110* specialist services, only 4 were more than 2 SD away from the number of adults expected to achieve the treatment target.
  • 2 specialist services were above expectation.
  • 2 specialist services were below expectation.

* 1 service was excluded as the number of adults with type 1 diabetes was too small

** For explanations of odds ratios, logistic regression and forest plot interpretation please see Additional Information: ‘Definitions – Statistical terms’. 

HbA1c achievement – 2019-20 and 2020-21 direct comparison

Figure A4: Distribution of HbA1c achievement among adults who had a result in both years, England and Wales, 2019-20 and 2020-21
Download the data for this chart Figure A4: Distribution of HbA1c achievement among adults who had a result in both years, England and Wales, 2019-20 and 2020-21

Adults with type 1 diabetes who had an HbA1c result in both 2019-20 and 2020-21 were compared:

There were almost 160,000 adults with a result in both audit years.
This cohort has been analysed separately to test whether the overall improvement detailed in the main report might have been influenced by missing data.
The profile of HbA1c achievement was compared across the years and showed that there has been a significant improvement in achievement of targets.
•The percentage of adults in each of the 3 lower HbA1c groups (HbA1c ≤ 48, 48 < HbA1c ≤ 53 and  53 < HbA1c ≤ 58 mmol/mol) was significantly higher in 2020-21 than in 2019-20.
•The percentage of adults in each of the 2 highest HbA1c groups (69 < HbA1c ≤ 86 and  HbA1c > 86 mmol/mol) was significantly lower in 2020-21 than in 2019-20.
This provides reassurance that there has been a genuine improvement between 2019-20 to 2020-21.

HbA1c achievement – 2019-20 and 2020-21 sensitivity analysis

In order to further examine the issue of potential bias in 2020-21 achievement, the 2019-20 and 2020-21 cohorts were further compared using a sensitivity analysis.

First the number of people with HbA1c checks in 2019-20 and 2020-21 were both increased to the size of the overall 2020-21 cohort with every additional person assigned an HbA1c levels above 86 mmol/mol. The percentages in each HbA1c group were recalculated using the same denominator.

This analysis was repeated, this time building both cohorts up only to the size of the cohort who had the HbA1c care process in 2019-20.

In both cases the percentage of people in each of the HbA1c groups up to 58mmol/mol was significantly higher in 2020-21

Table A1: HbA1c achievement groups using overall 2020-21 cohort size, with the assumption that anyone added in falls into the highest HbA1c group England and Wales, 2019-20 and 2020-21.

HbA1c achievement 2019-20 2020-21
Count Percentage of total Count Percentage of total
HbA1c ≤ 48 15,630 7.1 17,170 7.8
48 < HbA1c ≤ 53 16,175 7.4 16,990 7.7
53 < HbA1c ≤ 58 22,940 10.4 22,490 10.2
58 < HbA1c ≤ 70* 58,395 26.6 49,130 22.4
70* < HbA1c ≤ 86 48,140 21.9 43,480 19.8
HbA1c > 86 58,395 26.6 70,415 32.1
Total 219,675   219,675  
         
Total with HbA1c ≤ 58   24.9   25.8

* 58<HbA1c<=69 and 69<HbA1c<=86 in 2020-21

Table A2: HbA1c achievement groups using 2019-20 HbA1c care process cohort size, with the assumption that anyone added in falls into the highest HbA1c group England and Wales, 2019-20 and 2020-21.

HbA1c achievement 2019-20 2020-21
Count Percentage of total Count Percentage of total
HbA1c <= 48 15,630 8.1 17,170 8.9
48 < HbA1c <= 53 16,175 8.4 16,990 8.8
53 < HbA1c <= 58 22,940 11.9 22,490 11.7
58 < HbA1c ≤ 70* 58,395 30.4 49,130 25.5
70* < HbA1c ≤ 86 48,140 25.0 43,480 22.6
HbA1c > 86 31,040 16.1 43,060 22.4
         
total 192,320   192,320  
         
Total with HbA1c <= 58   28.5   29.5

* 58<HbA1c<=69 and 69<HbA1c<=86 in 2020-21

Last edited: 3 October 2024 8:13 am