Hospital Data Validation

CCME works with hospitals in the Carolinas in abstracting data related to the quality measures established by the Centers for Medicare & Medicaid Services (CMS), and then submitting the information to the QIO Clinical Data Warehouse.

For each hospital, cases from across the inpatient topics are randomly selected for validation review by the CMS Clinical Data Abstraction Center (CDAC). The CDAC requests copies of the medical records for re-abstraction and then submits data from that validation review to the Quality Improvement Organization (QIO) Clinical Data Warehouse. A hospital’s data is considered to be “validated” if its overall validation (agreement) score is greater than or equal to 80 percent. If the overall validation rate is less than 80 percent, the hospital has 10 business days (from the date the result was posted) to submit an appeal to its QIO.

All data that has been successfully submitted and is in the QIO Clinical Data Warehouse is subject to the hospital data validation process. An overview of the process is described below:

• A simple random sample of five charts per quarter will be identified from all hospitals with a minimum of six discharges in the QIO Clinical Data Warehouse. The sample is selected from all the cases submitted and is not topic-specific.
• The CDAC will request the paper medical records for each of the sampled charts from the hospitals.
• The CDAC will re-abstract the charts using the CMS Abstraction & Reporting Tool (CART). The relevant differences will be identified, and the CDAC will assign a reason code to each difference noted.
• The results of the re-abstraction will be stored in the QIO Clinical Data Warehouse and made available to the QIO to provide feedback to each hospital.
• Hospitals will receive educational feedback, including an overall reliability rating and case details on each abstraction.
• Based upon the CDAC re-abstraction, the percent agreement at the element level will be calculated. Hospitals that reach or exceed the 80 percent pre-determined threshold will be considered to be supplying valid data for that quarter.

Visit the QualityNet website for more information on validation. For questions regarding hospital data validation, contact Lynn Page, at lpage@scqio.sdps.org.