Skip to main content

Research Repository

Advanced Search

The pattern of testing for glycosylated haemoglobin (HbA 1c ) in people with diabetes is linked to the long term trajectory of blood glucose control

Fryer, A. A.; Holland, D.; Duff, C. J.; Heald, A.

Authors

D. Holland

C. J. Duff

A. Heald



Abstract

Background and aims: American Diabetes Association Guidelines (ADA) advocate testing “at least two times a year in patients who are meeting treatment goals” and “quarterly in patients who are not meeting glycaemic goals”. We previously described links between HbA1c testing frequency and outcome to show that many people with DM are not tested at the recommended frequency. Here we describe how variability in test interval over time relates to the deterioration of HbA1c control.

Materials and methods: Laboratory HbA1c level data, number of tests, interval between tests, variability in this interval and demographic data from a large UK hospital between June 2012-July 2019 was analysed. The outcome measure was the change in HbA1c over a 5-year period, focusing on those patients with a HbA1c within the first two years of the study period who also had a HbA1c 5 years±3 months after the starting HbA1c. To examine the link between variability in HbA1c testing interval, we focused on those with at least 6 tests during this period (n=23,582 patients). To minimise the impact of test numbers on interval SD, we separated the individuals into groups based on the number of tests between t0 and t0+5yrs and calculated the SD decile for each group. We then examined the effect of the combined SD deciles, together with other variables, on the change in HbA1c level between t0 and t0+5yrs, stratifying by starting HbA1c category.

Results: We show greater variability in testing frequency is associated with deterioration in HbA1c control over the 5-year period. This effect was most pronounced in those with lower starting HbA1c levels. In those with a starting HbA1c of <59mmol/mol, the lowest SD decile was associated with an increase in mean HbA1c of 3.9mmol/mol while those with the highest decile was more than double this 7.9mmol/mol. Similarly, in those with a starting HbA1c of 59-75mmol/mol, the lowest SD decile was associated with a mean reduction of 3mmol/mol, while those in the highest decile demonstrated a 4mmol/mol rise in HbA1c. In those with starting HbA1c values of 76-86mmol/mol and >86mmol/mol, the same trends were observed but were less marked. These effects were independent of the interval between tests. Mean HbA1c level also increased with increasing SD decile, an effect that was apparent irrespective of starting HbA1c.

Conclusion: We show that increased variability in the interval between
HbA1c monitoring tests is linked to increasing HbA1c over time, irrespective of the interval between tests. These findings indicated that regular monitoring, not just HbA1c level or the number of tests/year, is important in order to maintain diabetes control, especially in those people with relatively well-maintained diabetes control. This has implications for the ongoing management of those patients who attend sporadically for testing and suggests that general practitioners may need to develop

Citation

Fryer, A. . A., Holland, D., Duff, C. . J., & Heald, A. (2021, September). The pattern of testing for glycosylated haemoglobin (HbA 1c ) in people with diabetes is linked to the long term trajectory of blood glucose control

Presentation Conference Type Speech
Start Date Sep 27, 2021
End Date Oct 1, 2021
Deposit Date Jun 28, 2023
Publisher URL https://link.springer.com/article/10.1007/s00125-021-05519-y