Establishing an Indication Prioritization Framework for the Saudi Private Health System Using Mahalanobis Distance on Claims Data
Ibrahim Abdulrahman Aljuffali, Nada Alagil, Shabab Alghamdi, Razan Aljaser, Muhammed Aljumah, Mirna Matni, Omar Aldibasi, Ghassan SleilatyAbstract
Background:
Drug formulary management requires timely and equitable indication reviews, ensuring optimal resource utilization. This study describes the development and deployment of an Indication Prioritization Toolkit using data science to support review of the Saudi Daman Drug Formulary.
Materials and Methods:
A methodological approach with four phases—criterion establishment, data analysis, scoring methodology, and stakeholder panel—was developed. Criteria were identified through literature review and included epidemiological and economic considerations. De-identified reports from the National Public Health Information Exchange System (September 2022 and February 2023) were analyzed using metadata and descriptive statistics. To evaluate multiple alternatives simultaneously while accounting for inter-criteria correlations, a scoring approach based on the Mahalanobis distance (MD) was applied using variance–covariance matrices. Results were summarized using a scorecard with heatmap visualization.
Results:
Out of the 1,795 indications, the MD prioritized 164 indications (9.1%) having the highest percentiles on all six criteria: macro-cost, volume of patients, macro-cost per member per month (PMPM), medication cost, medication cost PMPM, and ratio of medication cost to macro-cost. The top three prioritized indications were multiple sclerosis, malignant neoplasm of breast, bronchus, and lung. The priority set was consistent, clinically meaningful, and validated by policy stakeholders, providing a data-driven output to guide formulary review.
Conclusion:
The devised MD prioritized indications for expedited review based on validated criteria, with a dynamic and flexible process that can be iterated with various criteria, in different contexts, and for specific subsets. Future research should support Council of Health Insurance key stakeholders in identifying relevant criteria to address clinical concerns related to disease severity.