Infrastructure for Medication Therapy Management (MTM) Identification

MTM identification requires explicitly documented eligibility criteria—CMS and payer-specific thresholds for polypharmacy count, chronic condition combinations, and annual drug cost minimums—to be current and findable. These criteria are not tribal knowledge; they are regulatory requirements that change with annual CMS rule updates. The ML cannot generate compliant MTM-eligible patient lists without documented, current eligibility logic that maps to specific payer contract terms and program requirements accessible to program administrators.

MTM identification requires systematic capture of medication fill history, adherence patterns, diagnosis codes, and hospitalization events through pharmacy systems, EHR, and claims data. CPOE, eMAR, and pharmacy dispensing systems log medication events through defined workflows requiring patient ID, medication, fill date, and dispensed quantity. This systematic capture creates the longitudinal medication and clinical history the ML needs to identify polypharmacy patients and detect non-adherence patterns from refill gaps.

MTM candidate identification requires consistent schema: patient ID, active medication list with RxNorm codes, diagnosis codes, refill dates, days supply, proportion of days covered (PDC), and eligibility flags by payer. Standardized drug codes (RxNorm, NDC) and diagnosis taxonomies (ICD-10) provide structured identifiers. The ML needs all patient medication records to share these defined fields to compute polypharmacy counts, PDC scores, and payer-specific eligibility criteria reliably across the enrolled population.

MTM identification requires the ML to access patient medication histories, diagnosis data, refill records, hospitalization events, and payer eligibility files. Existing pharmacy-EHR integration and API connections to drug databases provide access to internal medication and clinical data. The system can query active medication lists, recent fill history, and chronic condition diagnoses through connected systems to generate MTM candidate lists without manual data extraction for routine program operations.

MTM eligibility criteria are updated on scheduled annual cycles aligned to CMS rule updates and payer contract renewals. Medication fill history and adherence data are maintained through ongoing systematic capture. For the identification model, scheduled updates to eligibility parameters are sufficient since CMS program criteria change predictably on annual cycles. Between updates, the eligibility logic is stable, and the ML operates on continuously updated patient data to identify current program candidates.

MTM identification requires integration between pharmacy management, EHR clinical data, claims data for adherence metrics, and outreach management systems. Existing closed-loop medication use integration—EHR to pharmacy to ADC to eMAR—with emerging payer data connections provides API-based connectivity for the core identification workflow. The ML can access medication, diagnosis, and hospitalization data through connected systems to generate prioritized MTM candidate lists and route outreach recommendations to care coordinators.