Sample size and selection for a coding audit

March 30th, 2009 - Codapedia Editor
Categories:   Compliance  
 

In 2001, the OIG published a compliance plan recommendation for physician practices. This is one in its series of compliance plans: hospitals, labs, billing services, etc had already been published.  A copy of it  from the Federal Register is attached in the resource page.

In it, the OIG recommended an annual audit of each practitioner's claims of either ten encounters selected randomly or 2-5 encounters per government payer. Government payers include Medicare, Medicaid, TriCare and Federal BCBS.  Many groups find it simpler to select ten notes, randomly, without regard to patient insurance.

The OIG did not say how to select those services for audit.  One option is to print a detailed list of patient encounters for a one or three month period.  Note the total number of encounters.  Use a random number generator (available at numerous websites, for free) to generate 10 random numbers.  Apply those numbers to the list, and you have randomnly selected ten encounters.

A second option is to focus on high volume or high risk services.  This might include claims with certain modifiers.  For a primary care provider, it might be services billed as 99213 or 99214, since those are high volume services for primary care.  For a hospitalist, it might be focusing on 99233, since CMS finds a high error rate for level three subsequent hospital visits. 

Another method to select an audit focus is to review the OIG Work Plan for this year.  If the OIG is focusing on a service that is performed at a high frequency in your practice, include those services in your audit.

Feel free to change your audit focus from year to year.  If you find that your group is accurate in their established patient visit coding, select another focus for audit.  This gives you the opportunity to educate your physicians while checking for compliance.


###   Printed:  June 12th, 2026
Source:  http://www.codapedia.com//article_301_Sample_size_and_selection_for_a_coding_audit.cfm