Infrastructure for Automated Charge Reconciliation

Charge reconciliation requires explicit, current documentation of which clinical activities map to which billable charges — OR implant usage to CDM supply codes, medication administration to pharmacy charge codes, nursing procedure documentation to revenue codes. These mapping rules must be formalized beyond billing department tribal knowledge. When 'we charge for this implant only when the surgeon documents placement' exists only in experienced charge capture staff's heads, the AI misses charges or generates erroneous automated corrections.

Automated charge reconciliation requires systematic capture of EHR clinical activity (procedure documentation timestamps, medication administration records, supply usage) and corresponding charges posted in the billing system — with sufficient metadata to link them. Template-required fields like provider ID, patient encounter ID, item code, and documentation timestamp enable the AI to perform item-level matching. Without systematic capture, the reconciliation engine operates on incomplete EHR event logs and misses charge discrepancies.

The reconciliation engine must join EHR clinical activity records to CDM charge items using consistent schema: ClinicalActivity.procedureCode maps to CDM.revenueCode under EncounterID. CDM items are highly structured (revenue codes, charge amounts, quantity fields). EHR procedure documentation follows standard CPT taxonomy. Consistent schema across these two systems enables algorithmic matching. Full formal ontology isn't required — the matching logic operates on field-level joins rather than relationship graph traversal.

Automated charge reconciliation must query EHR clinical documentation (procedures, medications, supply usage) and the billing system (charges posted) via API, then write discrepancy alerts and automated corrections back to both systems. The baseline confirms EHR-to-billing integration is standard — charge triggers flow from EHR to billing. Extending this to bidirectional API access for reconciliation matching is achievable and necessary for automated charge correction on high-confidence matches.

CDM charge descriptions are updated quarterly when services change and annually when CPT codes update. Charge mapping rules between clinical activities and CDM items follow the same schedule. The baseline confirms 'charge description masters updated quarterly or when services change.' For charge reconciliation, scheduled periodic review of mapping rules is sufficient — new service lines or implant categories added mid-quarter require manual CDM update before the reconciliation engine picks them up, which is an acceptable lag given the quarterly CDM update cycle.

Charge reconciliation requires API-based integration between the EHR (clinical documentation source), the billing system/CDM (charge records), and potentially the pharmacy system (medication charges) and supply chain system (implant usage). The baseline confirms EHR-to-billing integration is standard. Adding pharmacy and supply chain API connections constitutes L3 — most systems connected via APIs. This enables the AI to validate OR charges, medication charges, and supply charges in a single reconciliation run rather than siloed department checks.