Master Patient Index

Match patients and members across data sources for a fully longitudinal patient view. Across multiple disparate data sources achieve 99% Specificity, 93% Sensitivity

A Master Patient Index (MPI) is a database that is used to identify, match, de-duplicate, and cleanse patient and member records to obtain a single view of an individual across multiple data systems.

An MPI can be used to identify individuals across member and patient records based on demographic information derived from multiple claims and clinical systems respectively. This single person view can then be used to collate the clinical history and health status of an individual by linking records back to the clinical systems and cross-verified against claims systems from a payer database. Matching and de-duplicating patient records allows you keep track of individual patients across the delivery network for the purposes of analytics and performance management, enabling better healthcare and customer service quality.

MPIs generally fall into one of two categories: (1) high-throughput transactional models, or (2) scalable batch models. Arcadia’s approach to creating a Master Patient Index is based on a batch-based five step process that runs on a nightly basis.

  • Primary Factors all must match.
  • Remaining factors are scored on similarity with thresholds set and tested to meet statistical standards.
  • Custom identifier allows for flexible configuration:
    • Forced Merging or Splitting based on source system.
    • Integration with existing eMPI solution that may already be deployed.
    • There is also a backend merge cross-walk.
  • Testing to over 95% Sensitivity, 98% Specificity.
  • 98% Recall based on information retrieval.

Member Attribution

Before engaging in a population health management or value based care programs, health systems and health plans must have an accurate foundation of provider-patient attribution. When primary care providers are auto-assigned to a new member by the claims system depending on geographic and demographic information, without patients and providers validation, PCP assignment is inaccurate resulting in poor quality reporting and care coordination.

Arcadia Analytics implements multiple views of provider-patient attribution to allow for focused alignment and ongoing maintenance.

  • Rendering Provider
    • 1 patient to many providers
    • The patient has been seen this physician in the specified lookback period
  • Clinically Assigned Provider
    • 1 patient to 1 provider per data source
    • Clinical data sources show assignment of this patient to this provider
  • Plan/Contract Assigned Provider
    • 1 patient to 1 provider per contract
    • For managed care, the assigned provider and location as assigned by the managed care plan
  • Historically Derived Provider
    • 1 patient to 1 provider
    • Based on AAFP four-cut analysis, Providers assigned based on analysis of patients 18 month encounter history
  • Functional Provider
    • 1 patient to 1 provider
    • Best-fit provider for most contexts.

Want to read more about our MPI solution and patient attribution?