Data Validation Overview
Assuring data accuracy is vital to public reporting programs. The Centers for Medicare & Medicaid Services (CMS) assesses the accuracy of chart-abstracted and Healthcare-Associated Infections (HAI) data submitted to the Hospital Inpatient Quality Reporting (IQR) Program through the validation process. CMS verifies on a quarterly basis that hospital-abstracted data submitted to the Clinical Warehouse and data submitted to the National Healthcare Safety Network (NHSN) can be reproduced by a trained abstractor using a standardized protocol.
CMS performs a random and targeted selection of Inpatient Prospective Payment Systems (IPPS) hospitals on an annual basis. The random selection of 400 hospitals for Fiscal Year (FY) 2018 occurred in November of 2015. In May of 2016, an additional targeted provider sample of 200 hospitals was selected, based on CMS targeting criteria as outlined in the Final Rule. The quarters included in FY 2018 validation are Third Quarter 2015 (3Q15), Fourth Quarter 2015 (4Q15), First Quarter 2016 (1Q16), and Second Quarter 2016 (2Q16).
Beginning with FY 2018, as described in the FY 2015 IPPS/LTCH PPS final rule (79 FR 50266), all chart-abstracted measure topic areas included in the Hospital IQR Program, with the exception of the Perinatal Care topic area, are automatically included in the validation process. This topic area is not included because the Elective Delivery PC-01 (NQF#0469) measure is reported in aggregate form, which is not consistent with the rest of the patient-level validation process.
CMS will validate up to eight cases for clinical process of care measures per quarter per hospital. Cases are randomly selected from data submitted to the warehouse by the hospital. The measures available for validation across the entire validation fiscal year are: Acute Myocardial Infarction (AMI), Stroke (STK), Emergency Department (ED), Immunization (IMM), Venous Thromboembolism (VTE), and Sepsis (SEP). Below is a list of the clinical process of care measures available to be validated in each quarter of FY 2018:
- 3Q15 - AMI, STK, ED, IMM, VTE
- 4Q15 - AMI, STK, ED, IMM, VTE, SEP
- 1Q16 - STK, ED, IMM, VTE, SEP
- 2Q16 - STK, ED, IMM, VTE, SEP
CMS will validate up to 10 candidate HAI cases total per quarter per hospital. CMS will validate candidate cases sampled for the following HAI measures: central line-associated blood stream infection (CLABSI), catheter-associated urinary tract Infection (CAUTI), Methicillin-Resistant Staphylococcus Aureus (MRSA) laboratory-identified (LabID) events, Clostridium difficile Infection (CDI) LabID events or Surgical Site Infection (SSI).
Hospitals selected for FY 2018 validation will provide two of four lists of positive cultures each quarter. Hospitals will submit either (but not all four):
- CLABSI and CAUTI Validation Templates or
- MRSA and CDI Validation Templates
CMS will randomly assign half of the hospitals selected for FY 2018 validation to submit CLABSI and CAUTI Validation Templates and the other half of hospitals will be assigned to submit MRSA and CDI Validation Templates. CMS will select up to four candidate HAI cases per hospital from each of the assigned Validation Templates. CMS will also select up to two candidate SSI cases from Medicare claims data for patients who had colon surgeries or abdominal hysterectomies that appear suspicious of infection. When there are not enough candidate cases for any one specific infection to meet the targeted number of cases, CMS will select candidate cases from other infection types to meet sample size targets.
Each quarter, CDAC will send hospitals a written request to submit a patient medical record for each case and candidate case that CMS selected for validation. For more information regarding medical records requested by the Clinical Data Abstraction Center (CDAC), refer to CDAC Information.
Beginning with FY 2018 payment determination, the finalized validation process for chart-abstracted measures includes two separate validation strata:
- Healthcare-associated infection (HAI), weighed at 66.7%; and
- Other/Clinical Process of Care, weighed at 33.3%.
CMS calculates a total score reflecting a weighted average of two individual scores for the reliability of the clinical process of care and HAI measure sets. After the scores are combined, CMS computes a confidence interval around the combined score. If the upper bound of this confidence interval is 75 percent or higher, the hospital will pass the Hospital IQR Program validation requirement.
Refer to Data Validation Resources for more information regarding the validation process.