Top Provider Questions – Overpayment Recovery
Click on an item to expand or Show All / Close All
- What is the Overpayment Recovery Department?
-
The Overpayment Recovery Department's goal is to assist you with overpayment refunds, offset, lost checks, and other overpayment issues.
Reviewed 12/20/2022
-
- When would I contact Overpayment Recovery?
-
You will contact Overpayment Recovery when you identify a Medicare overpayment.
Reviewed 12/20/2022
-
- How do I submit a request to Overpayment Recovery?
-
You will use the Overpayment Refund Form to submit a voluntary refund. This will ensure we properly record and apply your check.
Reviewed 12/20/2022
-
- What information is needed when submitting a voluntary refund?
-
Always complete the refund portion and use the reason codes listed on the bottom of the Overpayment Refund Form to identify the reason for your refund. The date of service (DOS), beneficiary Medicare ID number, and the reason you are submitting a refund is also required.
- When refunding for multiple beneficiaries, please be sure to include sufficient documentation to show how much money is being refunded for each claim.
- We recommend that you refund no more than 20 beneficiaries per check.
- We recommend that you not combine voluntary refunds with MSP involvement and voluntary refunds for NMSP in the same request.
- We recommend that you not combine voluntary refunds with demand check refunds.
Reviewed 12/20/2022
-
- Can you send me a stop payment form?
-
The Stop Payment Affidavit form is available at https://cgsmedicare.com/hhh/financial/forms/stop_payment_affidavit.pdf, and is located from the Forms icon under the Overpayments & Refunds topic on the left side navigation menu.
NOTE: A stop payment cannot be processed until the check has been outstanding for 30 days.
Reviewed 12/20/2022
-
- Where was my check applied?
-
Please contact our Customer Service Department.
Reviewed 12/20/2022
-
- I sent in a check and you offset. How will I get my money back?
-
In this scenario, you have sent in a payment to satisfy an Accounts Receivable listed in a demand letter sent to you. However, the demand has offset before the check is processed.
- If you have an open Accounts Receivable, we will send a letter to your office to inform you of where the money will be returned (the remittance notice it will be applied to.)
- If you have no open Accounts Receivables, we will send the funds back to you.
- If you want to avoid duplicate recoupment, your check must be received and posted within 30 days from the date of the Demand letter.
Reviewed 12/20/2022
-
- Did you receive my check?
-
Please contact our Customer Service Department.
Reviewed 12/20/2022
-
- Why am I not receiving payment?
-
The answer to this question varies. Please contact our Customer Service Department.
Reviewed 12/20/2022
-
- I need to update my address you are sending information to.
-
Please visit the Provider Enrollment section of this site or contact our Provider Enrollment Department by calling 1.877.299.4500 (option 3).
Reviewed 12/20/2022
-
- What information is needed when submitting an extrapolated refund?
-
A CGS statistician reviews for acceptability all voluntary refunds based on statistical sampling. In order for the statistician to replicate the sampling and validate the refund amount, with each refund check submission, providers must send CGS the details of the sampling and extrapolation performed to estimate the overpayment. The following is a listing of the information that must accompany any voluntary refund submitted that is based on statistical sampling for overpayment estimation. Note that the overpayment amount to be refunded is at the discretion of the MAC in coordination with CMS. Use of a methodology that yields the following elements is required and supersedes any methodologies that may have been approved by audit firms, statistical consultants, or the Office of Inspector General (OIG).
- Description of the Review Population (universe)—including an explanation of the methodology used to develop the population and the basis for this determination. [Provide any criteria that would warrant exclusion from the population. For example, claims for which the original payment was $0.]
- Electronic spreadsheet of the Sampling Frame—List the sampling frame, which is the totality of the sampling units from which the sample was selected. [The frame must be sorted in a way that is replicable (by Claim Number, for example) prior to random selection and the unique sort order identified even if the frame is serially enumerated to aid in matching to randomly generated numbers. CGS must receive the entire listing even if claims of other government or private payers were included in the universe. This is allowable under TPO of HIPAA, as it is required for validation.]
- Sample Size—Indicate how an estimate of the size for the sample was conducted. If RAT-STATS was used, provide the RAT-STATS output. [The Point Estimate (i.e., average overpayment of sample X # of units in the universe) is the best estimate of the overpayment and is the required refund for self-audits based on sampling. Providers should attain whatever level of precision will allow them confidence in submitting the point estimate amount as the refund. Stratification by paid amount is often used to lower the number of sampling units needed to achieve a desired level of precision. Stratification by HCPCS or CPT code will not aid you in this respect and should be avoided. Stratification by year should also be avoided unless separate analyses per year are needed for cost report purposes. Use of a certainty stratum for a few universe claims that are much higher in payment than the rest can also reduce sample size when estimating using paid amounts as surrogates for overpayment amounts.]
- Random Numbers—Either provide the RAT-STATS or other random number generator output in electronic format or indicate the seed number so that the generation of the random numbers can be replicated and applied against the frame to obtain the sample. Indicate the source of the seed number. [Note that the RAND() function in Excel is not sufficient as it does not provide a seed number that can be used to reproduce the random numbers for validation. The use of random tables within statistical texts on sampling is acceptable as long as the title, author, page number(s), and seed number are provided.]
- An electronic listing of the claims in the sample with the results of the review for each — Provide the overpayment for each item in the sample. [If the sampling unit is a claim with services per claim line, provide the overpayment per line prior to totaling per claim even if the overpayment is $0. Voluntary refunding providers are not allowed to net out underpayments. Please confirm this was not done. If possible, indicate the amount Medicare paid and amount that should have been paid for each item.]
- Estimation Methodology—The RAT-STATS results or if RAT-STATS was not used for the extrapolation, provide the formulas used to obtain the results along with the calculations. Be sure to include the resulting point estimate, standard error, and precision percent of your data as well as your level of confidence in the resulting point estimate. [When calculating precision, note that that the lower bound of the one-sided 90% confidence interval is equal to the lower bound of the two-sided 80% confidence interval. Still, as indicated previously, the Point Estimate of the confidence interval, not the lower bound, is the required refund for self-audits based on sampling.]
Reviewed 12/20/2022
-