Does coding of chronic kidney disease (CKD) improve disease risk management and slow disease progression?

Problem

Aims: To examine risk factor management of Chronic Kidney Disease in patients with Quality and Outcomes Framework (QOF) coded and uncoded CKD 3-5 in primary care using an ethnically diverse population.Chronic Kidney Disease (CKD) affects approximately 6% of adults in the UK, increasing due to associated comorbidities such as hypertension, diabetes, cardiovascular disease and age. Most patients with CKD are identified and managed in primary care, supported by Quality and Outcomes Framework (QOF) incentives which were first introduced into the QOF in 2006, and recently revised in 2016. QOF indicators are based on CKD NICE guidance and used to assess risk management in these patients to help determine the effectiveness of CKD management in primary care.This study uses Lambeth DataNet a primary care electronic health record from the London borough of Lambeth in South London. An ethnically diverse population only 61% Caucasian and the 5th highest internal migration of all the boroughs in England. Patients of South Asian and African-Caribbean heritage have higher rates of renal replacement but have been found to have a lower prevalence of CKD than Caucasians. Understanding some of the factors that contribute to disease progression in ethnic communities is an important aim of this study.

Approach

Methods: We examined EHRs from 44 Lambeth practices with a total of 286,162 patients, in SE London (Lambeth DataNet, 2013). For CKD3-5 classification in patients with hypertension, diabetes and cardiovascular disease, we compared risk management using multivariate logistic regression analysis in QOF (READ) coded patients with patients where coding was absent “uncoded CKD 3-5”, calculated using MDRD eGFR equation with ethnicity correction (eGFR < 60mls/min/1.73m2) and .

Findings

We found 73% of hypertensive patients with CKD 3-5 were coded (n= 4655/6380). When stratified by age; we found significant differences in proportions between coded and uncoded CKD3-5 for statin use (p<0.0001 in under 75’s; p=0.0002 in over 75’s); and ACEI or ARB use (p<0.0001 in under 75’s; p=0.014 in over 75’s). For all CKD hypertensive patients, age, gender and ethnicity adjusted OR (95% CI) for coded versus uncoded patients were: use of statins 1.47 (1.29-1.67), ACEI or ARB medications 1.27 (1.12-1.44), systolic BP control 1.17 (1.02-1.34), and African 1.62 (1.32-1.99) or Afro-Caribbean ethnicity 2.74 (2.27-3.30).

Consequences

Our findings suggest suboptimal risk factor management in uncoded patients with CKD, which may reflect poorer quality of care. Patients with uncoded CKD have potentially higher risks for disease progression, and lower uptake of statins and ACEI/ARB medications which could be improved.