On Evaluating EHRs

There's an interesting discussion going on over at the Journal of American Medicine about the different types of monitoring and evaluation required to mitigate potential errors in e-health records.

In an article[subscription required] dated Feb. 3, Dean Sittig, of the Memorial Hermann Center for Healthcare Quality and Safety at the University of Texas at Houston, and Dr. David Classen, doctor of internal medicine at the University of Utah School of Medicine, outlined a framework for keeping EHRs safe. In addition to certification standards for EHRs, two recommendations highlighted the need for measures to address serious adverse events caused by EHRs -- the creation of a national hazard reporting system and the formation of a safety board to investigate these events.

But Ross Koppel, a professor of sociology at the University of Pennsylvania, wrote in a letter on May 19 that those measures might not address the real problem: design flaws that cause day-to-day mistakes.

Sittig and Classen called for reporting structures for catastrophic system-caused errors, for example, in which all patients prescribed one drug receive another. However, it is likely that there are far more errors and near misses associated with poor but not obviously catastrophic design (for example, EHR screen displays that obscure critical information, drop-down menus reflecting inappropriate options, or oversensitive drug-drug interaction alerts). These types of near misses and errors may be unrecognized, under reported and attributed to inadequate physician training or commitment.

Sittig and Classen respond:

We agree that newer approaches to detect less serious and perhaps more common EHR-engendered errors are also needed. Ongoing work to refine and implement EHR-based triggers to detect and investigate these errors might have some potential. These triggers could be operationalized using a more population-based approach rather than alerting already overwhelmed clinicians at the point of care. We believe that all 5 components of our EHR monitoring and evaluation framework are necessary to identify these errors, reduce associated adverse events and create a safer and more effective health care delivery system.

Clearly there's a lot of work to be done.