“Upcoding” is a term applied to a category of fraud that can be perpetrated by healthcare providers on payers, or by carriers of Medicare Advantage (MA) plans when billing Centers for Medicare and Medicaid Services (CMS). It usually involves submitting billing codes (or hierarchical condition codes in the case of MA plans) that reflect a more complex service than was actually provided or a more severe diagnosis than was medically warranted.
Some life insurance carriers have asked us whether upcoding could skew underwriters’ risk assessments when using Irix Medical Data. We’ve seen no evidence of that, and a closer look at upcoding helps to explain why.
In practice, most upcoding is perpetrated by submitting current procedural terminology (CPT) codes for higher-level evaluation and management services than were provided, e.g., billing for more time spent with patients in routine office visits. Providers may also bill for more expensive procedures, such as claiming for a CT scan with contrast when no contrast was used. The practice of billing separately for services that should be billed as a single comprehensive code is another form of upcoding.
Fraudulent billing for services provided is a serious concern for healthcare payers and the United States government, but it rarely results in changed diagnoses and, as such, has insignificant impact on the use of billing data for underwriting.
A less common type of upcoding is when providers bill for more serious diagnoses, such as submitting diagnosis-related group (DRG) codes for more severe or complex conditions than were actually encountered. This type of fraud shouldn’t usually skew risk assessments because our tools are calibrated to downplay anomalous codes.
The Irix Rules Engine has special logic to verify the credibility of diagnosis codes entered on claims for diagnostic tests (e.g., a head CT) to ensure accuracy. The Rules Engine allows for different interpretation if a code is found without corroborating information or is only found once. And because Prescription Data and Medical Data are synergistic in nature, they corroborate treatment of conditions found in Medical Data with fills found on Prescription Data, increasing underwriter confidence in the final underwriting decision
Our internal data confirms that upcoding is not significantly influencing our clients’ mortality risk assessments. Our very low FCRA revision rate—about 0.04% on Medical Data—suggests that upcoding does not result in a significant number of unwarranted conditions appearing in applicant data.
So, while upcoding is a legitimate concern for payers, it primarily affects procedure codes, and there’s no evidence that it has a measurable impact on the results returned by Irix Medical Data, or on the decisions made by underwriters using the tool.
Nonetheless, it is always wise to rigorously vet data quality and relevance. Underwriters and medical directors can be confident that this form of fraud is not compromising their data-driven decisions.
