Chronic diseases: studying and understanding outcome using routine data: chronic kidney disease (CKD), an example

Abstract

Internationally, chronic disease represents a major healthcare challenge for the 21st century. Prognostic tools that streamline and target care have been developed for cardiovascular disease. This study illustrates how routine data can be used to develop tools for other chronic diseases; using chronic kidney disease (a precursor of renal replacement therapy (RRT: dialysis or transplantation) requirement) as a model. Routine clinical data—serum creatinine (a measure of kidney function), RRT initiation and death registration were used to identify a CKD cohort, and follow them over 6 years. Mortality was compared to the general population. 3426 persons were identified with CKD (median age 79 years, 56% female). RRT initiation rates decreased with age from 14.3 to 0.7 per 100 person-years among those aged 15–25 and 75–85 years at baseline respectively (absolute numbers 6 and 34). Mortality rates increased with age from 1.9 to 33.8 per 100 person-years for those aged 15–45 and over 85 years at baseline—a 19 and 2 fold increase in mortality risk compared to the general population respectively (2 and 17 excess deaths per 100 person-years). CKD has been labelled a public health concern, and provides a typical pattern for chronic disease. Personal risk is low for the majority, but they represent a high societal cost; whereas those with high personal risk are few, with lower societal cost. Exploitation of routinely collected data are an efficient way of following up health outcome, and informing the development of prognostic tools for a chronic disease cohort.