Skip to main content

Completeness of EHR Data

By April 18, 2016Commentary

Some days finding material for this blog is just too easy.  Research published in the Journal of the American Medical Informatics Association finds that electronic medical records are missing a lot of important data.   (JAMIA Article)   The federal government now penalizes hospitals, physicians and other providers who fail to have EMRs and to use them in a “meaningful” way.  Notwithstanding government and academic hectoring, a number of providers have resisted, because the costs imposed are too great and because of disruption of efficient work patterns.  But even when providers have sophisticated EMRs, their utility is limited because the patient often sees multiple providers and those providers rarely share data across EMRs.  In the current study, researchers looked at the adequacy of data contained in a large multi-specialty practice EMR.  The researchers looked at completeness of data and also looked at fragmentation, especially to the extent it is caused by patients visiting multiple providers, with an emphasis on mental health services.  They compared data in the EMR with data in the claims system for the patient’s health plan.

Looking at patients with a mental health diagnosis, the researchers found that many of the patients received mental health services outside of their primary care provider and that information from these visits was not contained in the primary care EHR, in fact data on 45% of visits was not in the primary care EHR.  The researchers found similar issues with missing visit and diagnosis data in other therapeutic areas.  Recent research has suggested that problems with EMRs have contributed to serious patient harm, including death.  An individual provider system’s EMR likely has errors and missing information; when the patient visits several providers, the problem is compounded.  Clinicians who assume they have all the relevant data because they have an EMR, or even worse, who rely on computerized decision support from an EHR with incomplete data, are at risk of making mistakes.

Leave a comment