female doctor in surgery

Using Data Analytics to Identify Physician Coding Opportunities

Bryan Beaudoin, Associate Director Health Information Management Solution Lead
Kelsey Cavanaugh, Associate Director Healthcare Innovation — Analytics

The big picture: Provider organizations face increasing regulatory scrutiny on professional fee billing practices but have finite auditing resources to address these challenges.

What’s new: Employing data analytics when executing professional fee billing (PB) medical coding audits allows physician groups to rank coding-compliance risk areas and reduce audit sample sizes while providing assurance to leadership that coding practices are compliant.

Between the lines: Some of the top physician billing coding risks include inaccurate evaluation and management (E/M) selection, incorrect use of modifiers 25 and 59, impossible days, and incorrect billing of split/shared services.

Why it matters: Given the vast array of compliance concerns and limited resources, it is increasingly important that compliance functions leverage data to pinpoint specific populations for auditing and develop robust and agile auditing plans.

___  ____  ___

With increasing regulatory scrutiny on professional fee billing (PB) practices and finite auditing resources available for physician groups to leverage, agile PB coding functions are turning to data analytics to identify and mitigate coding compliance risks. Employing data analytics when executing PB medical coding audits allows physician groups to rank coding-compliance risk areas and reduce audit sample sizes while providing assurance to leadership that coding practices are compliant. Below are some examples of top PB coding risks, as well as methods by which an organization can leverage data to address these risks more effectively.

  • Inaccurate evaluation and management (E/M) selection: The Comprehensive Error Rate Testing (CERT) program noted in 2023 that level 5 E/M codes (the highest level) are frequently coded incorrectly. For example, CPT 99205 (office-based, new patient E/M for Medicare) had roughly a 19 percent billing-error rate. These level 5 E/M codes are often audited by governmental and commercial payors to ensure medical decision making and/or time spent with the patient supports the high level selected.

An effective way to identify claims with a higher risk of incorrect coding is to benchmark providers’ use of E/M codes against their peers within the same specialty. This can be done using a box-and-whisker plot, or box plot, to easily isolate outlier providers who are coding level 5 E/M more frequently than their peers. While these claims will require further review to determine whether the codes are supported by provider documentation, this analytic allows organizations to better monitor (and better target audits of) the use of level 5 E/M codes.

  • Incorrect use of modifiers 25 and 59: Modifiers 25 and 59 are two payment modifiers that result in additional reimbursement when appended to certain CPT codes by unbundling two CPT codes that are typically bundled into one payment, allowing payment of both.
    • Modifier 25, “Significant, Separately Identifiable Evaluation and Management Service by the Same Physician or Other Qualified Health Care Professional on the Same Day of the Procedure or Other Service,” is used when distinct services are performed on the same day as an E/M service, such as an annual wellness visit. This service must be above and beyond the other service provided or beyond the usual preoperative and/or postoperative care associated with the procedure or service performed on that same date. Additionally, it must be substantiated by documentation in the patient’s record that satisfies the relevant criteria for the respective E/M service reported.
    • Modifier 59 is used to identify procedures and services, other than E/M services, that are not normally reported together but are appropriate under the circumstances. Documentation must support that a separate procedure or surgery was performed in a different session, on a different site or on a different organ system than the other procedure or service for which the provider is also billing.

Analytics can be employed to identify providers who bill these modifiers more often than their peers (based on overall volume and percentage of total visits) and focus on CPT codes on the Office of Inspector General (OIG) workplan, as they have a higher likelihood of being audited. Once the target population is identified, organizations should review these claims further to ensure the documentation supports the use of these modifiers.

  • Impossible days: Enforcement agencies have focused on identifying providers who bill an unlikely volume of time-based CPT codes, such as E/M codes, prolonged-care codes and therapy codes. Enforcement agencies analyze claims data to identify providers who have billed more time-based CPT codes than is reasonable within a day, week or month time frame. Once this population is identified, agencies perform a record review to validate that documentation in the medical record supports the billing of such services. Additionally, through this record review, auditors may identify instances in which additional billed services were not supported, such as those billed with modifiers 25 or 59, as outlined above.

Provider organizations can build similar reporting that sums time billed by provider and date of service by leveraging claims data and the list of time-based CPT codes from the Centers for Medicare & Medicaid Services (CMS). Specific thresholds based on number of hours billed can be set to identify claims that may be at a higher risk of being incorrectly or even potentially fraudulently billed.

  • Incorrect billing of split/shared services: Split (or shared) E/M visits are evaluation and management visits performed in part by a physician and in part by other non-physician practitioners in hospitals and other institutional settings. The work provided may be both face-to-face and non-face-to-face. Payment is made to the practitioner who performs the substantive portion of the visit. When a physician who did not perform the substantive portion of the visit bills for the service under his or her national provider identifier (NPI) and receives the full Medicare Physician Fee Schedule (MPFS) payment, the claim is noncompliant.

Regulatory bodies audit these records, both through their routine physician billing audits and by targeting claims with the fee schedule (FS) modifier appended to the CPT, to verify that the physician truly performed the substantive portion of the visit and appropriately received the full MPFS payment. Recently, the CMS created modifier FS, which practitioners must append to the E/M CPT to notate that the service is split/shared. Split/shared services can only be provided in a facility setting, such as a hospital, and not in a clinic setting. As such, monitoring can be established to identify instances in which the FS modifier is being used in a clinic setting (designated by place of service). Additionally, trending by provider can identify providers who leverage these modifiers more often than their peers in the same specialty. Exceptions and outliers from these analytics should be reviewed to ensure documentation within the medical record sufficiently documents that the physician performed the substantive portion of the visits billed under his or her provider number.

  • Social Determinants of Health (SDoH): Although not yet a bona fide compliance concern, accurate capture of these codes is increasingly important from a strategic and data-quality perspective. Additionally, missing SDoH documentation and subsequent coding could impact patient care as nonclinical interventions (e.g., connecting a patient with social workers or assisting them with finding transportation to check-ins) may be missed.

As the Biden administration and the CMS focus on addressing health disparities, the agency has placed increasing emphasis on the accurate coding of Social Determinants of Health Z-55 through Z-65 code ranges. These are codes that indicate nonmedical factors that may impact a patient’s access to effective healthcare treatment and/or ability to manage their conditions effectively. Examples include codes for factors such as food insecurity, homelessness, transportation insecurity, etc. Unlike most diagnosis and procedure codes, which can only be coded based on provider documentation, coders can assign these codes if the associated nonmedical conditions are documented in the medical record by any individual on the care team or self-reported with provider acknowledgment in the medical record.

PB coding leadership should ensure processes are in place to capture these codes, including auditing and monitoring activities, given the importance for health equity efforts. Coding leadership can utilize claims data to identify patients for whom there is no ZIP code assigned on the CMS 1500, focus on ICD-10 CM codes associated with behavioral health or substance use disorder diagnoses – typically codes in the F chapter (e.g., F31 – code range for bipolar disorders), and conduct a spot check to determine if any Z codes associated with social determinants of health were captured.

The above is not an exhaustive list of PB compliance topics that coding and compliance leadership should consider in their audit plans. There are many other areas of focus, such as incident-to billing, adhering to local and national coverage determination (LCD/NCD) requirements for medical necessity, telehealth services and use of locum tenens. Given the vast array of compliance concerns and limited resources, it is increasingly important that compliance functions leverage data to pinpoint specific populations for auditing and develop robust and agile auditing plans.

Add comment