Abstract
BACKGROUND: Congenital anomalies of the kidney and urinary tract (CAKUT) are the most frequent causes of childhood chronic kidney disease (CKD). Using a large CAKUT cohort, we sought to identify the predictors of CKD and to develop a prediction model that informs a risk-stratified clinical pathway. METHODS: This was a retrospective cohort study including cases with multicystic dysplastic kidneys (MCDK), unilateral kidney agenesis (UKA), kidney hypoplasia (KH), and posterior urethral valves (PUV). We identified risk factors for CKD (estimated glomerular filtration rate [eGFR] <60 ml/min/1.73 m(2)) and tested their performance in an adjusted multivariate binary regression model. Prediction probability scores for CKD were used to separate cases likely to develop complications from those not needing specialist follow-up. RESULTS: We identified 452 eligible cases of CAKUT with 22% developing CKD. Strongest associations with CKD included primary diagnosis (OR 3.5, 95% CI 2.6-4.6), preterm delivery (OR 2.3, 95% CI 1.2-4.4), non-kidney anomalies (OR 1.8, 95% CI 1.1-3), first eGFR<90 (OR 8.9, 95% CI 4.4-18.1), small kidney size (OR 9, 95% CI 4.9-16.6), and additional kidney anomalies (OR 1.6, 95% CI 1.2-2.8). PUV (OR 4.7, 95% CI 1.5-15.3), first eGFR <90 (OR 4.4, 95% CI 2-9.7), and kidney length to body length ratio <7.9 (OR 4.2, 95% CI 1.9-9.2) were independent predictors of CKD. The regression model had a prediction accuracy of 80% and a prediction probability c-statistic of 0.81. CONCLUSION: Using a large combined CAKUT cohort we identified risk factors for CKD. Our prediction model provides the first steps towards a risk-stratified clinical pathway. A higher resolution version of the Graphical abstract is available as Supplementary information.