Claims Anomaly Detection - Rules

Financial and count data cannot be negative. Detects when any numeric field (payments, claims, service lines) contains negative values indicating data entry errors or system bugs.

Medical service lines cannot be fewer than medical claims. A claim must have at least one service line. Indicates data inconsistency or orphaned records.

RX service lines cannot be fewer than RX claims. Each claim requires at least one service line. Suggests data integrity issues.

Incentive payments cannot exceed plan payments (overpayment). Indicates financial processing errors or unauthorized bonus payments that exceed contractual limits.

RX incentive payments cannot exceed RX plan payments (overpayment). Suggests financial control breach where bonuses exceed actual claim payments.

Medical claims are recorded but have no payment amount. Indicates orphaned claims not processed or payment system failures for medical services.

RX claims are recorded but have no payment amount. Suggests unprocessed pharmacy claims or payment processing gaps in the RX pipeline.

Medical service lines are recorded but no claims reference them. Indicates orphaned service line records detached from parent claims (data cleanup needed).

RX service lines are recorded but no claims reference them. Suggests orphaned pharmacy service line records not linked to actual prescription claims.

Claims, payments, or service lines exist when member count is zero or missing. Impossible scenario indicating missing member data or orphaned activity records.

Medical claims per member exceeds 50 (threshold). Indicates potential over-utilization, data duplication, or outlier members with extraordinary claim volume.

Medical service lines per claim exceeds 50 (threshold). Suggests complex claims with many line items or possible data entry errors creating excessive granularity.

RX to medical claim ratio is extreme (>10x or <0.1x). Indicates imbalanced claim distribution suggesting population mix changes or pharmacy data issues.

Required fields (client ID, name, month, members, payments) are null or missing. Critical for data quality; prevents accurate analysis and reporting.