Skip to content
This repository has been archived by the owner on Jan 5, 2021. It is now read-only.

Latest commit

 

History

History
53 lines (42 loc) · 2.07 KB

ARCHITECTURE.md

File metadata and controls

53 lines (42 loc) · 2.07 KB
Raw

MIPS - Claims-to-Quality Analyzer - Open-Source - Architecture

This is a quick overview of how the C2Q Analyzer works.

About

claims-to-quality-analyzer is intended to calculate claims-based quality measures for providers. The analyzer works as follows: * Read provider data and compute the 74 claims-based quality measures for each provider. * Resulting measurements are submitted to the QPP Submissions API.

Project Structure

  • audit: contains scripts relating to verification of measure processing.
  • bin: contains scripts employed during setup of the C2Q system.
  • claims_to_quality: contains the bulk of the source code. Broken down into the following three components:
    • analyzer: calculating the quality measures
    • lib: shared resources
  • data: contains data sources used for analysis and exploration.
  • tests: contains automated unit tests.

Single Source

TODO - Explain the process around the Single Source document and how we use it.

Measure Definitions

Measure Classes

The 74 measure definitions can be split into 5 main measure classes and a few measure specific classes:

  • DateWindowEOCMeasure

  • IntersectingDiagnosisMeasure

  • MultipleEncounterMeasure

  • PatientProcessMeasure

  • PatientIntermediateMeasure

  • ProcedureMeasure

  • VisitMeasure

  • CTScanMeasure

  • Measure46

  • Measure407

Measure Mapping

We then use a measure mapping - claims-to-quality/analyzer/measure_mapping.py to use the correct class for each measure.

Measure Processing

The overall measure processing looks like this. For each measure:

  • Filtering claims to those containing quality codes.
  • If required, infer the performance period for the submission and filter claims by date.
  • Calculate measure:
    • Find relevant claims based on eligibility criteria
    • Filter claims by date if needed
    • Extract eligible instances, grouping claims together based on characteristics like beneficiary ID or beneficiary ID + date of service.
    • Score eligible instances using "most advantageous claims"
  • Return eligible population and performance information
  • (Send to the submission API)