Treatment choices for managing glucose control in impaired renal function: Interim Results IJPDS (2017) Issue 1, Vol 1:371 Proceedings of the IPDLN Conference (August 2016)

Main Article Content

Gareth Davies
Jeff Stephens
Sam Rice
James Chess
Published online: Apr 19, 2017


Patients with chronic kidney disease (CKD ≥3) and diabetes mellitus comprise approximately 25% patients with diabetes. These patients are at a higher risk of cardiovascular morbidity and mortality and furthermore therapies targeting glucose control are limited. The management of glycaemic control in type 2 diabetes and chronic renal disease is difficult with limited therapeutic choices. This issue has been a matter of longstanding debate. Following a number of joint Diabetes-Renal meetings between the Diabetes and Renal teams based in Hywel Dda and ABM University Health Boards, a proposal was put forward to the SAIL team to examine the relationship between diabetes therapies in relation to eGFR, as this may influence further practice and guidance for patients with type 2 diabetes and renal impairment.

Linkage and re-use of routinely collected anonymised clinical data held in the SAIL databank was employed, to identify a cohort of adult patients in Abertawe Bro Morgannwg Health Board (ABMU) having type 2 diabetes (excluding type 1 diabetes). Diagnosis of diabetes was achieved by use of National Health Service ‘Read’ codes. Creatinine, eGFR, age, gender, weight, height, cholesterol, LDL, HDL, TG, systolic blood pressure, diastolic blood pressure, diabetes medication prescriptions, the use of statins, ACEis, aspirin, CHD status, CVD status, duration of diabetes were identified in primary care GP and pathology datasets.

42170 (6.0%) of adults in ABMU were identified as having type 2 (excluding type 1) diabetes , 13369 of which had good GP registration coverage. The gender split was male 56%, female 44%. Duration of diabetes (years) was (mean/median/SD/IQR) 9.96/8.97/6.78/8.10; weight (Kg) was 86.94/85.00/21.23/28.19; age (years) 65.49/66.74/13.75/19.03; BMI 31.57/30.70/6.62/7.99. Incidence of CKD as defined by GP coded data was 24%, renal replacement therapy 0.4%, Ischaemic Heart Disease 22%. Prevalence of prescriptions during 2014 was: Anti-diabetic medication 72%, statins 75%, aspirin 34%, ACEi/ARB 61%. The import of pathology laboratory data into SAIL is currently pending, and is anticipated before April 2016. This will allow the accurate stratification of CKD status and detailed description of use of anti-diabetic agents.

The project methods and coding structure are well place to provide anticipated results as soon as pathology data arrives. The percentage of ABMU patients having type 2 diabetes is in line with other literature for adults in the UK.


Identify geographic variations in health and healthcare among US Veterans living in rural areas and understand the relationships between social determinants of health and these variations.


Data from 11 data sources will be leveraged to create the US Veterans Rural Health Atlas and chart book (VeRHA) patterned after the Dartmouth Atlas, The VeHRA will provide an interactive map and chart book can be used to efficiently examine a wide range of factors related to health and healthcare of rural Veterans. The analyses will assess the relationships between socioeconomic, cultural and environmental factors and geographical variation in access, utilization, quality, satisfaction and outcomes. Semi-structured qualitative interviews will be used to elicit the perspective of Veterans not using VA care and to identify non-governmental organizations who provide care and support to US Veterans. The project will also identify community, state, and federal entities with which ORH could form strategic partnerships to improve health and healthcare for Rural Veterans.

Initially, three maps will be created for Veterans who are not enrolled in care, those enrolled but not using care and those enrolled and using care. Areas where many Veterans live and use VA healthcare will be identified as ``hot spots'' while areas where Veterans live but do not use care will be identified as ``cold spots''. Metrics for determining ``hot and cold spots'' will include measures of temporal and geographic access as well as measures of quality of care. We will first calculate raw rates for outcomes across geographic areas (census tract, county, and market/regions) Exploratory Spatial Data Analysis (ESDA) will be conducted by mapping the geographic distribution of key measures and then calculate the values of the local and global Moran’s I measures of spatial autocorrelation.

The relationship between social determinants of health and geographical variation in access, needs, utilization, quality, satisfaction, and outcomes for rural Veterans will be assessed, focused primarily on the ``cold spots'' - areas of greatest need.


The project is a work in progress with initial maps created showing the density of Veterans across the United States. More extensive results will be available for presentation.


This work demonstrates the value of using large data sets to guide development of policies and programs at a national level.

Article Details