US · guidance
CMS Pub. 100-18, ch. 9, § 50.6.9
Use of Data Analysis for Fraud, Waste and Abuse Prevention and Detection
42 C.F.R. §§ 422.503(b)(4)(vi)(F), 423.504(b)(4)(vi)(F)
Sponsors must perform effective monitoring in order to prevent and detect FWA.
Sponsors may accomplish this through the use of data analysis. Data analysis
should include the comparison of claim information against other data (e.g.,
provider, drug or medical service provided, diagnoses or beneficiaries) to identify
unusual patterns suggesting potential errors and/or potential fraud and abuse. Data
analysis should factor in the particular prescribing and dispensing practices of
providers who serve a particular population (e.g., long term care providers, assisted
living facilities, etc.). Use of data analysis may include monitoring pharmacy and
medical billing to detect unusual patterns. Sponsors may invest in data analysis
software applications that give them the ability to analyze large amounts of data to
detect FWA both internally and externally. Data analysis should:
• Establish baseline data to enable the sponsor to recognize unusual trends,
changes in drug utilization over time, physician referral or prescription
patterns, and plan formulary composition over time;
• Analyze claims data to identify potential errors, inaccurate TrOOP
accounting, and provider billing practices and services that pose the greatest
risk for potential FWA to the Medicare program;
• Identify items or services that are being over utilized;
• Identify problem areas within the plan such as enrollment, finance, or data
submission;
• Identify problem areas at the FDR (e.g., PBM, pharmacies, pharmacists,
physicians, other health care providers and suppliers); and
• Use findings to determine where there is a need for a change in policy.
Sponsors should develop indicators that will be used to identify norms,
abnormalities, and individual variables that describe statistically significant time-series trends. Examples include:
• Standard deviations from the mean;
• Percent above the mean or median; and
• Percent increase in charges, number of visits/services from one period to
another.
Sponsors should routinely generate and review reports on pharmacy billing,
medical claims, etc., based upon the data analysis performed to identify pharmacies
and other FDRs that require further review.
History
(Chapter 9 - Rev. 15, Issued: 07-27-12, Effective: 07-20-12; Implementation: 07-20 12)
Provenance
- Source
- cms.gov
- Retrieved
- 2026-09-17
- Edition
- iom-2026-09-17
- Content hash
1aa094bda28a0cca82ea93f3460a8f3c7a5c440f9a6b84a7dff2ee46c4b49bd0
The link goes to the issuing authority’s own document — the one we read to produce this record. Where a source publishes whole titles rather than sections, your browser may need a moment to jump to the provision.
Unofficial copy of government-published law, reproduced from official sources with full provenance. Not an official publication; verify against official sources before relying on it in a filing. Records in the 'guidance' corpus, and only that corpus, are sub-regulatory (interpretive guidelines, survey procedures) and are not binding law. Validity bounds follow each jurisdiction's declared temporalBasis.