Estimating Median Hospital Charge in the CODES Crash Outcome Data Evaluation System

摘要

The Crash Outcome Data Evaluation System (CODES) is a data system that links crash data with medical and financial outcome data, such as hospital and emergency medical services data. This allows for improved studies of medical and financial outcomes of motor vehicle crashes. Because of the lack of personally identifiable information, crash and outcome data are linked probabilistically using a technique called multiple imputations. One of the variables in CODES is the hospital charge that was charged to each crash victim. This report derives the maximum likelihood estimator of the median hospital charge as well as its confidence interval. The estimator has desirable theoretical properties and performs well in simulations. The estimator is designed for non-negative positively skewed variables, such as cost, charge, price, income, wealth, and so on. It can be used with multiply imputed data, such as the data available in CODES, as well as with non-multiply imputed data.