CDS malfunctions widespread, often undetected

Clinical decision support (CDS) systems malfunction more often than expected and are frequently undetected before they reach users, which could lead to major patient safety problems, according to a new study in the Journal of the American Medical Informatics Association (JAMIA).

CDS is known to improve healthcare safety, quality and efficiencies, however, it does not always work as intended, creating potential risk of patient harm. The researchers conducted an investigation of CDS systems at Brigham and Women's' Hospital to identify the prevalence and type of CDS system malfunctions. They identified four such malfunctions. The first was found by happenstance during a demonstration; after that they reviewed all 201 alert firings and discovered three other anomalies:

  • An alert for monitoring thyroid function in patients receiving amiodarone stopped working when an internal identifier for amiodarone was changed in another system
  • An alert for lead screening for children stopped working when the rule was inadvertently edited
  • A software upgrade of the electronic health record software caused numerous spurious alerts to fire a malfunction in an external drug classification system caused an alert to inappropriately suggest antiplatelet drugs, such as aspirin, for patients already taking one.

They also found that some of the malfunctions had been occurring for some time. For instance, since the alert for amiodarone only fires after a patient is on the drug for at least one year, the malfunction was not identified in the first year that there was a discrepancy between the identifiers in the EHR and the other system.

To determine if other hospitals had encountered similar malfunctions, the researchers surveyed chief medical information officers at other facilities. Of 29 respondents, 27 reported at least one malfunction of their CDS systems; 11 respondents reported that they have had four or more malfunctions a year.

The researchers recommended, among other things, that CDS systems be tested after any CDS-related change is made and after major EHR upgrades; that there be more proactive monitoring of CDS systems; and that they be tested by an analyst different from the one that built the content.

Other studies have shown both the benefits and risks of CDS. Even those that are operating correctly are often overly intrusive or are so frequent that clinicians end up ignoring them.

To learn more:
- here's the abstract