Program_Memos / 2003 / B-03-022

Why This Topic Matters

Accurate statistical sampling is central to projecting overpayments in large claims populations and documenting defensible review results. This memo helps readers understand the procedural framework CMS expected for sampling-based administrative reviews and related recovery actions.

Article Sections

1. Introduction

   Overview of the memorandum’s purpose, the context for statistical sampling in Medicare Part B administrative reviews, and the scope of the guidance.

2. Probability Sampling

   General requirements for the sampling process and the characteristics of a probability sample.

3. Selection of Period To Be Reviewed and Composition of Universe

   How the review period, universe, sampling unit, and sampling frame are defined and organized.

4. Sample Selection

   Sampling design options, random number use, sample size considerations, and documentation of methodology.

5. Calculating the Estimated Overpayment

   Approaches for estimating overpayment from sample results and handling of related payment determinations.

6. Actions To Be Performed Following Selection of Physician or Supplier and Sample

   Notification, review-site decisions, review conduct, documentation, and timing considerations after a sample is selected.

7. Overpayment Recovery

   Post-review recovery steps, required informational content in demand letters, and related coordination requirements.

8. Corrective Actions

   Examples of additional administrative actions that may follow a review outcome.

9. Changes Resulting From Appeals

   How appeal outcomes can affect the sampling-based overpayment determination and subsequent actions.

10. Appendix A

    Reference materials listed as background resources for statistical sampling and survey methods.

11. Appendix B

    Supplemental discussion of stratified sampling and cluster sampling concepts in the context of review methodology.