Test shows hospitals how CDS reduces malpractice losses

Clinical decision support (CDS) tools can reduce hospitals’ exposure to malpractice suits, but few hospitals seem to use the functions that would prevent the steepest losses, according to researchers who developed and tested a risk assessment system.

Adam Wright, PhD, an associate research scientist at Brigham and Women’s Hospital in Boston, and colleagues had previously determined that the CDS at Partners Healthcare System in Boston potentially could have prevented 123 claims that cost the system more than $40 million. They built on those findings in a study published online May 27 in the Journal of the American Informatics Association with a simulation-based self-assessment instrument that assessed CDS capabilities and their potential to prevent malpractice damages.

They previously had identified 41 types of CDS that might prevent these malpractice events. To build the instrument, they reviewed closed malpractice claims from seven hospitals and outpatient practices in the Partners system for events that occurred between 2000 and 2007 and were found in favor of the claimants by the end of 2008. They used that data to calculate losses paid.

The instrument included a questionnaire with vignettes to evaluate if a CDS type could be completed and if it led to the expected results, with the ability to account for time lags and EHR-related workflows. Test takers received a report card after completing the questionnaire with a score on the “potential indemnity loss prevented” with each CDS type.

Four Partner hospitals and three hospitals in other states participated in the study by taking the self-assessment. One hospital answered “yes” to 40 of the 57 questions asked (the highest) while one answered “yes” to only 19 questions (the lowest). The potential for damages prevented ranged from 16.5 percent to 73.2 percent.  

The most commonly used CDS types were not the most potentially cost saving, though. All sites used CDS for automated interpretation of electrocardiograms, for instance, and most sites used a CDS for checking drug allergies. But Wright et al found that the CDS was least adopted in scenarios where the potential costs were high: alerts for clinically significant test abnormalities, diagnostic decision support and informed consent.

“We found that a number of key CDS types are not widely used but are associated with significant malpractice loss,” they wrote. “We also identified considerable variability among sites in the breadth and depth of CDS coverage.”

The assessment tool is publicly available at https://isrisk.partners.org. The research team plans to expand the study in its next iteration.