Discriminant Patterns of Consistency in Health Data Dictionaries: Implications for Evidence-Based Medicine

Abstract

An increasing amount of the medical information in health provider organizations is derived from electronic patient records (EPRs). As such, the quality of information maintained by healthcare organizations becomes a key component of the healthcare delivery process, with a corresponding need for consistent information collection and management methods. The objective of this study was to examine and compare adoption rates of EPR data reliability within clinical support systems, identifying regional variation across the USA. In a nationwide study of all accredited US health information managers, reported levels of data dictionary adoption in electronic patient records were examined. Results show that the majority of health information managers have widely disparate policies and procedures related to data consistency. Despite increasing reliance on consistency of data for comparative purposes, most organizations have not exhibited regular adoption of data dictionaries to uniformly define system-based information within their organizations. This occurs in an era where government-mandated data standards require greater uniformity of medical information. Utilization of EPR data in evidence-based medical decision making must be undertaken with caution, especially when outcomes or other decision support data are compared across states or regions.