Dealing with non-detects for dietary exposure assessment using Kaplan-Meier methods and Bootstrap

AIM:To combine Kaplan-meier method and Bootstrap estimation in dealing with the left-censored data for exposure assessment. METHODS:The cumulative probability distribution for the cadmium residue concentration in rice was fitted by Kaplan-Meier method,basing on which standard errors and interval estimates of mean,median et al. were gained by bootstrap method. RESULTS:The Kaplan-Meier method can be used simply and conveniently by software. The intervals estimated by bootstrap were consistent with those by approximate method. CONCLUSION:Kaplan-Meier method combined with bootstrap is efficient method worth to recommend in dealing with data including left-censored values.