Assessing Data Accuracy in Electronic Medical Records

Author: Gary J. Lu

Primary Advisor:

Committee Members:

Masters thesis, The University of Texas Health Science Center School of Allied Health Sciences at Houston

Clinical data in computer systems should be accurate if they are to support patient care, research, and health service management. Inaccurate data may undermine the reliability, usefulness, and public trust in a database. This paper reviews published studies of data accuracy in electronic medical record systems (EMR). These studies report highly variable levels of accuracy. The causes of inaccurate data and interventions to improve data accuracy are discussed. The essential ways to deal with the problems of inaccurate data in EMRs are: 1) Decrease the amount of inaccurate data, and 2) increase the accuracy of information that is retrieved from EMRs. As EMR data are increasingly used for many purposes, standard methods for future studies of data accuracy in EMR systems are needed.