Insight into a Payer’s Perspective: BCBS Claim Upcoding by Hospitals

This week Blue Cross Blue Shield published an article titled “Study suggests AI is boosting hospital billing.” The study was based on data analyzed by Blue Health Intelligence® (BHI®) which looked at claims data from “tens of thousands of maternity admissions.”

Not all clinical documentation integrity (CDI) departments review obstetrics cases, but when they do, as reflected in this report, acute blood loss anemia (ABLA) is an often-recorded secondary diagnosis, especially in women who deliver by cesarean section.

The report states that ”there was a significant increase in cases coded for acute posthemorrhagic anemia, a serious condition that typically signals severe blood loss and a need for extra medical attention and treatment, like blood transfusion.” BHI® used this assumption to identify cases where ABLA was coded, but the patient did not receive a blood transfusion as the basis of this analysis. Their conclusion was that a misalignment existed between patients coded with ABLA and “the level of care we would have expected.” The report continued by suggesting the use of artificial intelligence (AI) has contributed to “the number of patients coded with complex conditions,” resulting in a practice commonly known as “upcoding.”

The authors of this report demonstrate a lack of understanding of the Medicare Severity Diagnosis Related Group (MS-DRG) reimbursement methodology and the Uniformed Hospital Discharge Data Set (UHDDS) requirements for reporting an additional diagnosis. The report claims that “each secondary diagnosis allows hospitals to bill insurance companies more.” Their stance is that “costs are artificially increased, however, when secondary diagnoses are added without evidence of treatment.” This statement alone shows a gross misunderstanding of healthcare reimbursement.

Most Network Contracts Favor the Payor

First, we must acknowledge that the payor has all the power in this dynamic. Most network contracts allow the payor to change their policies with a notice of material change – and with passive acceptance. The payor must notify the hospital of changes to an existing policy or issuance of a new policy that significantly changes the agreement. For example, it results in additional requirements for coverage, reduced coverage, or reduced payment. If the hospital does not respond to the planned change as outlined in the notice, it has passively agreed to the change. This is the strategy likely used by Aetna Medicare Advantage (MA) when it developed its Level of Severity Payment Policy. Basically, hospitals must accept the change, or they cannot remain part of the payor’s network. This is becoming a common scenario, whereby hospitals must decide if the cost of being in network outweighs the potential volume losses from being out-of-network. In relation to MA plans, when a hospital is out-of-network, the MA plan must follow traditional Medicare payment requirements, which would exclude use of the Aetna Level of Severity Payment policy.

MS-DRG Reimbursement Methodology

Not all secondary diagnoses increase the hospital bill. Under the MS-DRG reimbursement mechanism, about 25 percent of the more than 73,000 diagnosis codes within the ICD-10-CM code set are classified by the Centers for Medicare & Medicaid Services (CMS) as a complication/comorbidity (CC) or major complication/comorbidity (MCC). Complications/comorbidities are evaluated in addition to the patient’s age and discharge status, in relation to hospital resource consumption. Diagnoses that meet this requirement are then reviewed by a physician panel, which determines if the impact was substantial – as defined by a condition whose presence with a specific principal diagnosis “would cause an increase in the length of stay by at least one day in at least 75 percent of patients.”

The CC list, which accounts for about 21 percent of all diagnoses in the 2026 fiscal year (FY), describes “significant acute disease, acute exacerbations of significant chronic diseases, advanced or end-stage chronic diseases, and chronic diseases associated with extensive debility.”

In general, most chronic diseases were not included unless there was a “significant acute manifestation of the chronic disease.” Currently, ABLA is classified as a CC under MS-DRG methodology. When originally established, MCCs accounted for 12 percent of all diagnosis codes, but that has dropped over the years as new codes are added, and MCCs now only account for almost 5 percent of diagnosis codes.

Requirements for Reporting Secondary Diagnoses

All payors must adhere to the UHDDS to allow the collection of uniform health data. Other diagnoses, which include secondary diagnoses classified as CCs or MCCs, are defined as “all conditions that coexist at the time of admission, that develop subsequently, or that affect the treatment received and/or the length of stay. Diagnoses that relate to an earlier episode which have no bearing on the current hospital stay are to be excluded.”

All covered entities, including health plans, are required to use ICD-10-CM and ICD-10-PCS on electronic inpatient claims under the Health Insurance Portability and Accountability Act (HIPAA). Therefore, they are also required to adhere to the ICD-10-CM Official Guidelines for Coding and Reporting, which further define secondary diagnoses as “significant conditions that affect patient care” in terms of requiring:

• Clinical evaluation;
• Therapeutic treatment;
• Diagnostic procedures;
• Extended length of hospital stay; or
• Increased nursing care and/or monitoring.

Only allowing ABLA to be reported if the patient is transfused is inconsistent with official coding guidelines.

Objective Insight into Maternal Hemorrhage

I do agree that some consulting companies have pushed the boundaries of what qualifies for reporting ABLA as a secondary diagnosis, and it may be over-reported by some facilities and in some situations, but we also have healthcare statistics that support the validity of the severity of risks associated with the birth process. A 2019 newsletter reported that “the most frequent cause of severe maternal morbidity and preventable maternal mortality is obstetric hemorrhage — excessive blood loss from giving birth. Recent data indicate that rates of maternal hemorrhage are increasing in developed countries, including the U.S.”

Again, this was from 2019. This report led the Joint Commission to create safety measures regarding this topic. It may be that this has led to increased recognition of the danger of maternal hemorrhage, and that is also contributing to an increasing volume of cases being reported.

This same report includes the American College of Obstetricians and Gynecologists (ACOG) definition of maternal hemorrhage as “cumulative blood loss of greater than or equal to 1,000 mL, or blood loss accompanied by signs or symptoms of hypovolemia, within 24 hours after the birth process.” This is a definition that was reaffirmed in 2024. Another article references “excessive bleeding” as > 500 ml blood loss for a vaginal delivery or > 1000 mL for a cesarean delivery.

This is a starting point for CDIs and coders when determining if there is clinical evidence to support the diagnosis of ABLA in an obstetrics patient, but there may be variation among physicians and independent practitioners regarding when they feel comfortable making the diagnosis. Additionally, initial management includes administration of uterotonic agents and fluid resuscitation, not blood transfusion. Use of these medications during and following delivery can be an indicator of treating ABLA, making it reportable. There are also tools to help providers identify women at elevated risk for maternal hemorrhage, who require closer monitoring and evaluation compared to other women, which would also support the reporting of ABLA as an additional diagnosis.

A Conservative Approach

Personally, I am uncomfortable querying for ABLA in surgical patients when, in my opinion, the surgeon is preventing ABLA by addressing blood loss during surgery. No additional resources are being used when it is part of the standard of care. Additionally, I would not query if there were routine monitoring of a patient’s complete blood count (CBC), which often occurs daily. Therefore, I would only query when the health record demonstrates that monitoring (either through diagnostics or nursing care), evaluation, or treatment is beyond routine care. Often, this requires collaborating with surgeons to ask them what is routine. But I also encourage hospitals not to succumb to payors that deny ABLA unless the patient is treated, which they often equate to a blood transfusion.

Conclusion

BCBS concluded that AI is leading to upcoding, but we know that many payors are adopting AI faster and more comprehensively than hospitals, allowing them to challenge more claims. What is truly driving up healthcare costs is not “upcoding.” Yes, there may be some hospitals that push the envelope of clinical evidence, but payors are also imposing stricter criteria in an effort to reduce payments, as the cost of healthcare continues to rise. This article demonstrates that payors are inaccurately applying coding criteria, and likely disparaging compliant hospital coding and billing practices.

The healthcare industry is entering a dangerous negative feedback loop, wherein payors deny more claims and hospitals must appeal to recover earned revenue. This vicious cycle is one factor that is rapidly increasing the cost of healthcare. And it is not sustainable.