Primary Care Coding Alert

With the COVID-19 public health emergency (PHE) flexibilities slated to end on May 11, you are probably already re-evaluating your audit program to ensure consistent compliance and optimal revenue going forward. And even though your practice may be doing okay, now is a good time to check if you’re reaching all your performance goals.

Identify the tools and/or resources you will use to perform the audit within the scope and engagement letter/agreement to be signed. If you are utilizing a tool from a particular vendor, you should identify that tool by name. Possible resources include:

• ICD-10-CM book for identified years, or in the unlikely event you want to audit records going back before Oct. 1, 2014, ICD-9-CM;
• CPT^® professional edition for a particular year, especially if performing a retrospective review;
• CPT^® Assistant;
• AHA Coding Clinic;
• Centers for Medicare & Medicaid Services’ (CMS) guidance and applicable transmittals; and
• Fee schedules, if revenue variance is to be calculated.

Ascertain Which Type of Audit Fits the Bill

Prospectively performing a review is a “pre-bill” audit, which is performed on claims after coding is completed but prior to being submitted to the insurance payer. Conducting audits prospectively results in claims being put on hold or suspended until after the review has been completed.

“We do very few prospective reviews just because we’re holding up the accounts receivable [AR] to do that,” says Kim Huey, MJ, CCS-P, PCS, CPC, CPCO, COC, CHC, a coding and reimbursement specialist in Alabaster, Alabama. But, if you do the review pre-bill, you don’t have to do corrected claims because you are going to correct them going out.

On the other hand, a retrospective audit is “post-bill” — after the claim has been submitted for payment and subsequently paid or denied. If issues are identified with retrospective claims, those claims should be rebilled as corrected claims.

Consider Proactive Compliance Audits

If you do a proactive compliance audit, you need to decide if it’s prospective or retrospective. In this case, you don’t know if there’s a particular problem, but you do know that you should be auditing.

First, decide how many encounters you are going to look at. “I personally don’t like to look at less than 10 because I think less than 10 doesn’t give you a good picture,” says Huey.

Second, you need to determine the focus or scope. What are the services or situations you want to audit? When you are making decisions regarding sample selection, you have different ways to think about this question. Industry experts suggest a blend of the following:

• Office of Inspector General (OIG) work plan
• Comprehensive Error Rate Testing (CERT) issues
• Recovery Audit Contractor (RAC) issues
• Top 10 denials for the practice
• Top 10 services billed for this practice
• Specific issues brought to your attention

For more detailed advice, check out these internal auditing sample checklists put out by CMS at www.cms.gov/files/ document/ehrinternalmonitoringja062816pdf.

Pinpoint Time Frame for the Audit

The timeframe you choose for the review will depend on the reason for the audit.

If this is a proactive or compliance audit, it may be more helpful to choose recent claims. When the purpose is education, it is better to work with recent visits the provider may remember.

However, when the audit is for a specific problem, you want to look at claims over a specific timeline for which that problem is suspected. For example, let’s say you brought a new physician into your practice two years ago, and you recently found out they have been billing every patient they saw as a new patient. But what if the patient had been seen by someone else in the practice? You may have a significant number of claims that need to be corrected. You will need to go back and review those submissions from the last two years.

Select Audit Sample in These Ways

Your audit sample will depend on the type of audit.

If you are not investigating a specific problem, our experts suggest looking at 10 encounters per provider. This is called a “judgmental sample,” where you are using a small sample to make a judgment.

Keep in mind: A judgmental sample cannot be extrapolated to a larger population since it is not truly random. The OIG recommends five judgmental samples per provider per federal payer per year.

Conversely, if you are investigating a specific problem, you may consider a statistically valid random sample, which will give you an idea of what you are going to have to refund to the federal government if you verify a problem exists. This type of sample is a probe sample followed by a larger sample with targeted confidence and precision. A probe usually involves 30-50 items.

Note: For self-disclosure, CMS requires that the sampling methodology be reviewed by a statistician or someone with equivalent experience, according to Huey.