Corporate Compliance

Tip: Choosing an audit sample size

Compliance Monitor, July 30, 2008

If you want reliable audit results, choose an effective population sample. This is one of the most critical steps in every audit—it influences the findings by outlining which records you will review. Choosing a sample isn’t as easy as closing your eyes and grabbing random medical records. Mistakes during the sample selection process can materially distort or even invalidate the sample results, particularly if the mistake is not detected. You can, however, correct mistakes when estimating the sample size.            

Using statistical sampling allows you to make valid inferences about the characteristics of the population you want to measure. Consider the following factors when determining a sample size:
  • Professional judgment—how good is your knowledge of the characteristics of the population?
  • Characteristics of the population, such as outliers and the number and size of dissimilar items.
  • Confidence level (tolerable error) desired. For example, how sure do you want to be that the desired statistic is within a certain range, plus or minus, of the real population mean? (This is the number you would get if you inspected every item in the population.)
  • Strength of controls, like preventing billing errors that lead to overpayments.
  • Type of sampling method to be used, such as simple random sampling, stratified random sampling, multistage sampling, discovery sampling, and dollar unit sampling.
This tip was adapted from The Healthcare Auditor’s Handbook. For more information about the book or to order your copy, click here.


0 comments on “Tip: Choosing an audit sample size


Most Popular