JAMIA: Filtering can reconcile medications
Collaborative filtering can be a valuable tool for reconciling medication lists, but context and consequences should be included in the process, rather than a one-size-fits-all approach, according to research in the July edition of the Journal of the American Informatics Association.
Sharique Hasan, PhD, of the Graduate School of Business at Stanford University in Stanford, Calif., and colleagues investigated the application of collaborative filtering methods to medication reconciliation. Given a current medication list, the authors employed collaborative filtering approaches to predict drugs that patients could be taking but were missing from their observed list.
“Online retailers use collaborative filtering to recommend relevant products using retrospective purchase data,” wrote Hasan and colleagues, who argued that patient information in EMRs, combined with artificial intelligence methods, can enhance medication reconciliation.
The authors formulated the detection of omissions in medication lists as a collaborative filtering problem. They used several machine-learning approaches to detect omissions. These approaches' eficacy was evaluated using medication data from three long-term care centers.
Collaborative filtering identified a missing drug in list of top 10 medications taken about 40 percent to 50 percent of the time and the therapeutic class of the missing drug 50 percent to 65 percent of the time at the three clinics in the study.
These findings suggest that a collaborative filtering approach to medication reconciliation holds promise, the authors concluded. “We anticipate improvements as additional information is used and as more and diverse clinical settings are evaluated using this methodology. We do not envision a one-size-fits-all solution to this problem, since clinical settings are heterogeneous, and different collaborative filtering approaches may work better in different scenarios.”
Sharique Hasan, PhD, of the Graduate School of Business at Stanford University in Stanford, Calif., and colleagues investigated the application of collaborative filtering methods to medication reconciliation. Given a current medication list, the authors employed collaborative filtering approaches to predict drugs that patients could be taking but were missing from their observed list.
“Online retailers use collaborative filtering to recommend relevant products using retrospective purchase data,” wrote Hasan and colleagues, who argued that patient information in EMRs, combined with artificial intelligence methods, can enhance medication reconciliation.
The authors formulated the detection of omissions in medication lists as a collaborative filtering problem. They used several machine-learning approaches to detect omissions. These approaches' eficacy was evaluated using medication data from three long-term care centers.
Collaborative filtering identified a missing drug in list of top 10 medications taken about 40 percent to 50 percent of the time and the therapeutic class of the missing drug 50 percent to 65 percent of the time at the three clinics in the study.
These findings suggest that a collaborative filtering approach to medication reconciliation holds promise, the authors concluded. “We anticipate improvements as additional information is used and as more and diverse clinical settings are evaluated using this methodology. We do not envision a one-size-fits-all solution to this problem, since clinical settings are heterogeneous, and different collaborative filtering approaches may work better in different scenarios.”