Prediction of time to reflux using accelerated failure time model of Weibull distribution in children with antenatal hydronephrosis

Abstract

Background & Aim: Prediction of time to reflux can aid healthcare providers and preparation programs. We constructed a risk prediction instrument for occurrence reflux in children with antenatal hydronephrosis. Methods & Materials: Demographic and clinical information was collected retrospectively in children with the antenatal hydronephrosis and mostly with reflux, followed at least 5 years. Results: Accelerated failure time model of data from 333 children was developed to assess the risk of time to reflux. Likelihood ratio tests of statistical significant were used to identify best fitting predictive function. Variables “gender”, “Sr”, and “severity of ANH (in severe level)” were highly significant (p<0.05) in multivariate model, adjusting for some traditional risk factors. Conclusion: This proposed risk probability model allows prediction of time to reflux for children with antenatal hydronephrosis to better inform parents from possible time of occurrence reflux and treatment strategies.

Keywords

• Antenatal Hydronephrosis
• Reflux
• Risk factor
• Survival model
• Weibull distribution