Arch Systems demonstrated innovation by developing a Measure-Specific Reporting Error Prediction Engine (MSREPE) that analyzes clinical quality measure specifications to predict the extent of reporting errors that are expected for that measure which led to the Federal Health IT Innovation Award in 2017.
The project was focused on developing complex scientific models and applying intensive methodologies to validate the procurement, translation, transmission, and submission of data to Centers for Medicare and Medicaid’s (CMS) Physician Quality Reporting System and Electronic Prescribing Incentive Program Data Validation (PQRS and eRx) Incentive Programs. The methodologies developed were in the areas of interview protocol, sampling, data analytics, data validation, and incorrect payments analysis.
The Data Validation efforts were comprised of distinct phases – Survey, Data Analytics, Predictive Modeling, Primary Source Verification (PSV), and Reporting. Our processes included sampling, surveys, data validation of large datasets against enormous clinical data repositories, and development of data analytics reports. Arch performed in-depth analyses of disparate data sources and reports, including national benchmarks and HHS data, to define the business rules for measure submission and validation. Arch used the Cross Industry Standard Process for Data Mining (CRISP-DM) model during the Data Analytics phase to govern the data lifecycle.
