Can routinely collected primary care data determine the prevalence of chronic kidney disease and predict clinical events in patients with stages 3 of the disease?

Jain, Poorva (2015). Can routinely collected primary care data determine the prevalence of chronic kidney disease and predict clinical events in patients with stages 3 of the disease? University of Birmingham. Ph.D.

Preview

Jain15PhD.pdf
PDF - Accepted Version
Download (10MB)

Abstract

Introduction
Chronic Kidney Disease (CKD) is common and associated with cardiovascular morbidity and mortality. Few studies have assessed the real prevalence of CKD and the predictors of morbidity and mortality in this cohort.
Methods
Using THIN data, the prevalence of CKD 1-5 was ascertained .This was compared to the prevalence on the practice register using QOF Read codes in 2009. Cox proportional hazard models using routinely collected primary care were used identify predictors of i) all-cause mortality and ii) the composite of cardiovascular disease and all-cause mortality.
Results
The prevalence of CKD 1-5 was 5.01%. Patients with CKD not on the practice register were associated with worse management than those on the practice register. Increasing age and co-morbidity were associated with worse outcomes. Continuous variables such as hypertension, BMI, haemoglobin and cholesterol were associated with an inverse J shaped relationship with log relative hazard ratio. Antihypertensives and lipid lowering drug usage, and non white ethnicity was associated with improved outcomes. Blood thinning agents and diuretics were associated with worse outcomes.
Conclusions
CKD is common in UK. Many practices mis-label CKD which impacts upon management. Some predictors of mortality and morbidity is different to previous reports and this requires further investigation.

Type of Work:

Thesis (Doctorates > Ph.D.)

Award Type:

Doctorates > Ph.D.

Supervisor(s):