Clinician self-reported data entry in imaging clinical decision support (CDS) was found to be 90.7 percent accurate, according to a study published in the January 2014 issue of the Journal of the American Medical Informatics Association. Quality improvement methods could be implemented to optimize adoption of evidence demonstrated through CDS.
Despite federal mandates to implement CDS, little evidence is available regarding the accuracy of clinician self-reported data entry. “Inaccurate computerized provider order entry (CPOE) data could lead to erroneous CDS recommendations and inappropriate testing, which might, in turn, result in suboptimal care or even potential patient harm,” wrote lead author Anurag Gupta, MD, MBA, of the Brigham and Women’s Hospital in Boston, and colleagues.
Gupta and colleagues examined the accuracy and downstream effects of clinician data entry-dependent CDS that guides the use of CT angiography (CTA) for the evaluation of patients with suspected pulmonary embolus (PE) in the emergency department (ED). Of 1,296 patients with CTA for suspected PE during 2011, 90.7 percent had accurate D-dimer values entered into the CDS.
Fifty-five orders, or 4.2 percent of the cohort, had incorrectly entered data in order to prevent intrusive computer alerts—a practice known as “gaming.” Forty of the 121 errors potentially resulted in inappropriate CTA performance. Of the 121 inaccurate imaging requests, ten percent involved patients who should have undergone CTA. No PEs were found in the cohort.
While the results demonstrated that the great majority of self-reported data entry is accurate, Gupta and colleagues suggested “quality improvement strategies including retrospective sampling of clinical data entered into CDS and academic detailing when appropriate are probably needed to minimize the small portion of erroneous data entries that may be perceived to be motivated by the desire to avoid intrusive computer interactions and alerts.”
Future studies should evaluate the nature of gaming errors to determine if they are intentional or random data entry errors, added the study’s authors.