By: Matt Fischer
Since the implementation of the ZPIC audit and RAC audit programs, healthcare providers and suppliers have experienced increased scrutiny in the pursuit of overpayments and fraud. Medicare’s most vital tool in its progressive search is the use of statistical sampling. In theory, statistical sampling offers a reliable and low cost approach to addressing large volumes of claims. However, this process gives the government a huge advantage as it places a heavy assumption on a large number of claims without actual review of the claims. Thus, it is important for providers and suppliers to understand the process and know how to challenge such studies in order to minimize potential repayment obligations and retain their revenue.
What is statistical sampling?
Statistical sampling draws a random sample from a universe of claims and extrapolates or projects the results of the sample to the entire universe of claims. In other words, the Medicare contractor will select a sample of claims to review from a look back period or examination period of typically two or three years. For this example, let’s say that the review finds a 40 percent error rate in the sample, meaning 40 percent were not found to meet Medicare requirements for payment. In this case, a contractor will apply the 40 percent finding to the entire two years’ worth of claims and deny these claims based on the sampling results.
ZPIC Audit Weak Spots
This process can be challenged. No provider or supplier should assume that the government is always right. If a challenge is successful, the projected amount is thrown out and the provider or supplier will only be liable for the actual denied claims in the sample. However, blanket arguments such as the contractor failed to follow one or more requirements of the Medicare Program Integrity Manual or that there is a more precise method available rarely prevail. In contrast, challenges should be concentrated in these areas: (1) attack the technique (i.e., lack of randomness, selection bias, small sample size, or unrepresentative sample); (2) attack the conclusions (i.e., degree of confidence or precision level); or (3) attack the contractor’s expert (i.e., challenge the sampling creator’s credentials).
The ability to develop a successful appeal is imperative for healthcare providers seeking to hold reimbursements to which they are entitled. If statistical sampling is utilized, a heavy burden is placed on the provider or supplier to rebut. When faced with this scrutiny, providers and suppliers need to be aware of all options and be ready to counter the sampling process which will contribute to their bottom line.