Impact of geography on the control of type 2 diabetes mellitus: a review of geocoded clinical data from general practice.
Objective: To review the clinical data for people with diabetes mellitus with reference to their location and clinical care in a general practice in Australia.
Materials and methods: Patient data were extracted from a general practice in Western Australia. Iterative data-cleansing steps were taken. Data were grouped into Statistical Area level 1 (SA1), designated as the smallest geographical area associated with the Census of Population and Housing. The data were analysed to identify if SA1s with people aged 70 years and older, and with relatively high glycosylated haemoglobin (HbA1c) were significantly clustered, and whether this was associated with their medical consultation rate and treatment. The analysis included Cluster and Outlier Analysis using Moran’s I test.
Results: The overall median age of the population was 70 years with more males than females, 53% and 47%, respectively. Older people (>70 years) with relatively high HbA1c comprised 9.3% of all people with diabetes in the sample, and were clustered around two ‘hotspot’ locations. These 111 patients do not attend the practice more or less often than people with diabetes living elsewhere in the practice ( p=0.098). There was some evidence that they were more likely to be recorded as having consulted with regard to other chronic diseases. The average number of prescribed medicines over a 13-month time period, per person in the hotspots, was 4.6 compared with 5.1 in other locations ( p=0.26). Their prescribed therapy was deemed to be consistent with the management of people with diabetes in other locations with reference to the relevant diabetes guidelines.
Conclusions: Older patients with relatively high HbA1c are clustered in two locations within the practice area. Their hyperglycaemia and ongoing cardiovascular risk indicates causes other than therapeutic inertia. The causes may be related to the social determinants of health, which are influenced by geography.
type 2 diabetes, diabetes mellitus, geocoded clinical data, Australia