Can routinely collected primary care data be used to predict future risk of morbidity and mortality in newly-diagnosed type 2 diabetes mellitus?

Ryan, Ronan Paul (2014). Can routinely collected primary care data be used to predict future risk of morbidity and mortality in newly-diagnosed type 2 diabetes mellitus? University of Birmingham. Ph.D.

Preview

Ryan14PhD.pdf
PDF
Download (5MB)

Abstract

Background/clinical context: Type 2 diabetes (T2DM) is associated with an increased risk of adverse outcomes. Data routinely recorded in general practice electronic patient records could be used to develop risk prediction models to identify those at higher risk and target preventative treatment.
Objective: To develop models to predict the 5-year risk of coronary heart disease (CHD), stroke, chronic kidney disease (CKD), and all-cause mortality following a diagnosis of T2DM.
Methods: Newly diagnosed T2DM patients registered at a practice contributing data to a large UK general practice database were included in the analyses. The models included clinical predictors routinely recorded following diabetes diagnosis plus cardiovascular preventative treatments.
Results: 20041 patients diagnosed with T2DM were included. The proportion of variation explained by each model (R2) was: CHD 0.09; stroke 0.35; CKD 0.34; and mortality 0.58. Hazard ratios for modifiable risks in the mortality model were: current smoking 1.65; blood pressure (high/treated) 1.07; and glycaemic control (HbA1C/%) 1.09 (p<0.01 apart from BP).
Conclusion: The models were predictive, particularly for mortality, and suggest that older, male, smokers, those with poor blood pressure and glycaemic control and those with cardiovascular co-morbidity are at highest risk and should be targeted at the point of diagnosis.

Type of Work:

Thesis (Doctorates > Ph.D.)

Award Type:

Doctorates > Ph.D.

Supervisor(s):