Send to

Choose Destination
Diabetes Obes Metab. 2018 Jan;20(1):185-194. doi: 10.1111/dom.13067. Epub 2017 Sep 8.

Improving type 2 diabetes mellitus glycaemic outcomes is possible without spending more on medication: Lessons from the UK National Diabetes Audit.

Author information

School of Medicine and Manchester Academic Health Sciences Centre, University of Manchester, Manchester, UK.
Department of Endocrinology and Diabetes, Salford Royal Hospital, Salford, UK.
Department of Blood Sciences, Walsall Manor Hospital, Walsall, UK.
Robotics and Control Systems, Technical University of Wroclaw, Wroclaw, Poland.
York Health Economics Consortium, York, UK.
RES Consortium, Andover, UK.
Keele University School of Medicine, Newcastle-under-Lyme, UK.
Department of Clinical Biochemistry, University Hospital of North Staffordshire, Stoke-on-Trent, UK.



To determine the factors at general practice level that relate to glycaemic control outcomes in people with type 2 diabetes (T2DM).


Data were accessed from 4050 general practices (50% of total) covering 1.6 million patients with T2DM in the UK National Diabetes Audit 2013 to 2014 and 2014 to 2015. This audit reported characteristics, services and outcomes in the T2DM population, including percentage of patients who had total glycaemic control (TGC), defined as glycated haemoglobin (HbA1c) ≤7.5% (58 mmol/mol), and the percentage who were at higher glycaemic risk (HGR), defined as HbA1c >10% (86 mmol/mol); the respective figures were 67.2% and 6.2%. The medication data were examined in terms of annual defined daily doses (DDDs). Multivariate linear regression analysis was used to identify associations between DDD and patient and practice characteristics.


Over the period 2012/2013 to 2015/2016, patient numbers grew 4% annually and annual medication expenditure by 8%, but glycaemic control outcomes did not improve. The main findings were that practices with better outcomes: had a higher percentage of patients aged >65 years; provided more effective diabetes services (including case identification, care checks, patient education, percentage of patients with blood pressure and cholesterol under control and more patients with type 1 diabetes achieving target HbA1c levels); spent less overall on prescribing per patient with T2DM; and on average, prescribed fewer sulphonylureas, less insulin (for patients with T2DM), fewer glucagon-like peptide-1 agonists, more metformin, more dipeptidyl peptidase-4 inhibitors, and more blood glucose monitoring strips. Ethnicity and social disadvantage and levels of thiazolidinedione (glitazone) prescribing had no significant impact on outcomes. Sodium-glucose co-transporter-2 inhibitor use was too low for an effect to be observed in the period examined.


If all practices brought their service and medication to the level of the top decile practices, they could achieve 74.7% compared with the median of 67.3% of patients achieving TGC, showing an increase of 213 000 in patients achieving TGC, while reducing the number at HGR to 3.8% compared with 6.1%, benefiting 62 000 patients. This could have a major impact on the overall consequent healthcare costs of managing diabetes complications with their attendant mortality risks.


HbA1c; big data; diabetes; outcome; prescribing; primary care

[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Wiley
Loading ...
Support Center