Alexandra Kogan

Inferring Monitoring Recommendation Frequencies for Multimorbidity Patients

Alexandra Kogan1; Mor Peleg1; Samson Tu2
1Department of Information Systems, University of Haifa, Haifa, Israel;
2Stanford Center for Biomedical Informatics Research, Stanford University, Stanford, CA, USA;

Multimorbidity patients are often managed according to multiple Clinical Practice Guidelines (CPGs). Conflicts or interactions between recommendations from different guidelines may result in adverse events. Hence, monitoring that medications are working as intended (toward the clinical goal) and that they do not result in known (or unknown) adverse events is important. However, due to the overwhelming amount of patient data and the need to pay attention to all the required treatments, monitoring tasks are often abandoned until a problem arises that could have been prevented otherwise.

In previous research [1], we developed a Clinical Decision Support System (CDSS) that detected conflicting recommendations and suggested mitigation strategies to clinicians. Our evaluation showed that medical students benefited from the system, which helped them provide more complete management plans for patient cases. Our current work addresses generation of monitoring plans for multimorbidity patients. The novelty of our methods is that they recommend the frequency in which monitoring actions for intended effects and adverse effects should be monitored. The knowledge about the monitoring frequency is often not found in a single guideline thus a combination of multiple knowledge source is needed.

Based on the patient’s data from the medical record, the monitoring plans in the PROforma computer-interpretable guidelines are traversed to identify the patient-specific monitoring plans and their frequencies, depending on the patient’s state of each disease. The algorithm also consults with external knowledge sources (RxNav, DDINTER, DrugCentral) regarding potential drug-drug interactions, occurring adverse events and their severity to determine how to mitigate interactions.


[1] Kogan A, Peleg M, Tu SW, et al. Towards a goal-oriented methodology for clinical-guideline-based management recommendations for patients with multimorbidity: GoCom and its preliminary evaluation. Journal of Biomedical Informatics. 2020 Dec 1;112:103587.