CMS Pub. 100-16, ch. 21, § 50.6.9

and Detection

(Chapter 21 - Rev. 109, Issued: 07-27-12, Effective: 07-20-12; Implementation: 07- 20-12)

(Chapter 9 - Rev. 15, Issued: 07-27-12, Effective: 07-20-12; Implementation: 07-20- 12)

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.