Keywords

data cleaning, electronic health record, evidence-based practice, fitness for use, informatics, nursing research, quality improvement

 

Authors

  1. Lyons, Ann M. PhD, RN
  2. Dimas, Jonathan PhD, RN
  3. Richardson, Stephanie J. PhD, RN
  4. Sward, Katherine PhD, RN, FAAN

Abstract

ABSTRACT: Data from electronic health records (EHRs) are becoming accessible for use in clinical improvement projects and nursing research. But the data quality may not meet clinicians' and researchers' needs. EHR data, which are primarily collected to document clinical care, invariably contain errors and omissions. This article introduces nurses to the secondary analysis of EHR data, first outlining the steps in data acquisition and then describing a theory-based process for evaluating data quality and cleaning the data. This process involves methodically examining the data using six data quality dimensions-completeness, correctness, concordance, plausibility, currency, and relevance-and helps the clinician or researcher to determine whether data for each variable are fit for use. Two case studies offer examples of problems that can arise and their solutions.