The use of eGFR and ACR to predict decline in renal function in people with diabetes.

BACKGROUND: There have been few attempts to estimate progression of kidney disease in people with diabetes in a single large population with predictive modelling. The aim of this study was to investigate the rate of progression of chronic kidney disease in people with diabetes according to their estimated glomerular filtration rate (eGFR) and presence of albuminuria. METHODS: Data were collected on all people with diabetes in Salford, UK, where an eGFR could be calculated using the four-variable MDRD formula and urinary albumin-creatinine ratio (uACR) was available. All data between 2001 and 2007 were used in the model. Classification of albuminuria status was based on the average of their first two uACR measurements. A longitudinal mixed effect dynamic regression model was fitted to the data. Parameters were estimated by maximum likelihood. RESULTS: For the analysis of the population, average progression of eGFR, uACR and drug prescribing were available in 3431 people. The regression model showed that in people with diabetes and macroalbuminuria, eGFR declined at 5.7% per annum, while the eGFR of those with microalbuminuria or without albuminuria declined at 1.5% and 0.3% per annum, respectively, independently of age (P < 0.0001). CONCLUSIONS: The longitudinal effect of time on eGFR showed that people with diabetes and macroalbuminuria have an estimated 19 times more rapid decline in renal function compared with those without albuminuria. This study demonstrates that the progression of kidney disease in diabetic people without albuminuria is relatively benign compared with those with albuminuria.

The languages of health in general practice electronic patient records: a Zipf's law analysis.

BACKGROUND: Natural human languages show a power law behaviour in which word frequency (in any large enough corpus) is inversely proportional to word rank - Zipf's law. We have therefore asked whether similar power law behaviours could be seen in data from electronic patient records. RESULTS: In order to examine this question, anonymised data were obtained from all general practices in Salford covering a seven year period and captured in the form of Read codes. It was found that data for patient diagnoses and procedures followed Zipf's law. However, the medication data behaved very differently, looking much more like a referential index. We also observed differences in the statistical behaviour of the language used to describe patient diagnosis as a function of an anonymised GP practice identifier. CONCLUSIONS: This works demonstrate that data from electronic patient records does follow Zipf's law. We also found significant differences in Zipf's law behaviour in data from different GP practices. This suggests that computational linguistic techniques could become a useful additional tool to help understand and monitor the data quality of health records.