Missing and Incomplete Data Reduces the Value of General Practice Electronic Medical Records as Data Sources in Research

Abstract

The objective of this study was to identify limitations to the use of electronic medical records (EMRs) in general practice and to identify consequent limitations to the value of EMRs as data sources for research or case finding. Patient demographic, diagnostic, therapeutic and administrative data pertaining to osteoporosis management during 2003 were extracted from the EMRs of participating general practitioners (GPs) in Bendigo, Victoria. Data quality was measured primarily by the absence of data from fields. Twenty-three of 60 eligible GPs participated; 29, 356 patient records were searched. Patient age or sex was missing from 1.2 per cent of records. For those with a diagnosis of osteoporosis on the record, reason for prescription was missing for 54 per cent of patients receiving osteoporosis medications, recording of Authority numbers was inconsistent and only 31 per cent of records noted fracture site in the prescription free-text detail field. The study concluded that missing data compromises the value of general practice EMRs as data sources. Researchers wishing to use EMR data will benefit from assisting GPs to appreciate and use EMR capabilities and to appreciate the value of maintaining accurate and complete datasets.