Skip to main content

What We Have Learned From the Prior Authorization Model for Non-Emergent HBOT

Taking a closer look at the Prior Authorization Project for Non-Emergent Hyperbaric Oxygen Therapy (HBOT), this author details what wound clinicians can learn from the utilization and expenditures of HBOT.

In 2015, the Centers for Medicare and Medicaid (CMS) implemented a “Prior Authorization Project” for Non-Emergent Hyperbaric Oxygen Therapy (HBOT Prior Auth). According to the Medicare website, CMS wished to test whether the prior authorization process could “reduce expenditures while maintaining or improving quality of care.”

Most healthcare providers are accustomed to obtaining authorization for specific treatments from private payers, but the process is relatively new to traditional Medicare and has been implemented on a trial basis only for select services. CMS selected 3 states for the HBOT Prior Auth and the program was overseen by the states’ respective Medicare Administrative Contractors (MACs). The selected states were: Illinois (National Government Services [NGS]), Michigan (Wisconsin Physician Services [WPS]), and New Jersey (Novitas Solutions). The claims review process started at slightly different times for each, but all began in 2015, and the model ended on February 29, 2018.  

The model included 5 non-emergent HBOT indications for which a Prior Authorization Request (PAR) was required:
1. Chronic refractory osteomyelitis (CROM)
2. Osteoradionecrosis of bone (ORN)
3. Complications of soft tissue radionecrosis (STRN)
4. Actinomycosis
5. Diabetic wounds of the lower extremity (DWLE)

“Demonstration projects” like this are subject to review, which allows CMS to determine whether the projects have accomplished their purpose; in this case, reducing expenditures for HBOT. After issuing an interim report in 2018, CMS released a final Evaluation of the Prior Authorization Model for Non-emergent Hyperbaric Oxygen on June 19, 2019.1,2 Mathematica Policy Research generated this report. What is clear is that CMS accomplished at least one of its goals with the HBOT Prior Auth experiment. HBOT utilization was reduced by 34% based on the number of HBOT treatments provided to Medicare beneficiaries. This translated to a reduction of approximately $307 million in HBOT charges, or $56.00 per Medicare beneficiary in quarterly HBOT expenditures. However, what is both confusing and fascinating is that despite the reduction in HBOT utilization and HBOT related expenditures, there was no change in overall Medicare expenditure.2

Key Points From the Report

There were several interesting observations in the final report that are well worth examining:

1. What was the most common condition treated with HBOT in the Prior Auth states? The most common indication for HBOT was DWLE. Over 80% of the cases submitted for PAR were DWLE. Approximately 6% of patients with DWLE received HBOT.
    a. Interestingly, the analysis showed that the lag time between the diagnosis of a DWLE and the start of HBOT was 220 days.

2. How did Prior Auth affect the initiation of HBOT when providers feared limb loss? If limb loss was of major concern, 28% of providers provided the hyperbaric therapy before the PAR was completed, and they were fully prepared to absorb the financial burden if the PAR was not affirmed.2

3. How did Prior Auth impact the number of HBOT providers? The number of HBOT providers increased in all of U.S. states, but less in the three “model” states when compared to similar regions (1% vs. 8%). However, during the Prior Auth timeframe, the number of facilities seemingly providing hyperbaric services dropped significantly. CMS reports claims data for facilities and providers that treat at least ten unique Medicare beneficiaries for a specific service. In New Jersey alone, there was a year-to-year drop of 19 reporting facilities from 2015–2016. When the facilities that no longer appeared in the CMS dataset were contacted, they had not closed their doors; they had chosen not to treat fee for service Medicare beneficiaries. This has important implications because it means that Prior Auth reduced access to HBOT for traditional Medicare beneficiaries.

4. What was a major barrier to implementation of the PAR? Staff from the 3 MACs reported that one of the consistent barriers to implementation of the PAR was HBOT National Coverage Decision 20.29. Input from the MACs suggested that the National Coverage Determination (NCD) for HBOT is outdated and that CMS should revisit the NCD. Novitas has a hyperbaric oxygen Local Coverage Determination (LCD). It is interesting to note that the largest reduction in HBOT services was seen in New Jersey (a Novitas state), as well as the most consistent percentage of “non-affirmations” (treatments that were not approved) (Table IV.2).

5. What type of providers supervise HBOT treatments? It was difficult to determine the type of provider that provided hyperbaric supervision (e.g., credentials such as MD, NP, DPM, etc.). Since the 2000 report by the U.S. Department of Health and Human Services (HHS) Office of the Inspector General (OIG) linked quality and appropriate utilization to physician presence, it is of concern that Mathematica could not identify the type of practitioner providing hyperbaric oxygen supervision.  

Some questions remain unanswered by the Mathematica analysis. If there was no reduction in overall Medicare expenditures despite a reduction of over $300 million in hyperbaric services, what Medicare charges increased to make up the difference? The lack of decrease in the total cost was not addressed in terms of a compensatory increase in  outpatient services; i.e., infusions, CTPS, more HOPD visits from debridements, etc. Mathematica also did not compare the cost/outcomes of those that were non-affirmed to those that were affirmed within the model states. That calculation would have answered the question: Did  those that were not affirmed ended up costing more Medicare dollars, and and did they have worse outcomes?

The length of time from initial presentation to the request for HBOT in the DWLE exceeded the 30 days of standard care. That leads us to ask whether there was a difference in outcome or cost among patients referred early versus late for adjunctive HBOT. Are we waiting too long to implement HBOT? Since some physicians are treating those at significant risk of limb loss with hyperbaric oxygen regardless of the financial repercussions did those patients who were non-affirmed but still received HBOT have better outcomes than their peers?

The impact of HBOT Prior Auth on the utilization of HBOT has not been limited to the three states in which it was performed. HBOT Prior Auth has resulted in a significant decrease in hyperbaric services across the U.S. In 2015, 681,121 claims for 99183 were submitted to FFS Medicare. By 2017, that number had dropped to 524,0653.3 The decrease cannot be explained solely due to the reduction of services in the three affected states.

CMS has indicated that it will not extend the HBOT Prior Auth model. It has been ended in the three states initially targeted and is not being expanded to other states currently. However, CMS has implemented a program across nearly all MACs called a “Targeted Probe and Educate” or TPE audit for HBOT. The impact of TPE is yet to be determined. The difference is that TPE occurs after the HBOT claims have been submitted and can result in paybacks to Medicare. The TPE program will be discussed in a future article. However, it would seem reasonable to predict a further reduction in hyperbaric services.

In Conclusion

This analysis provides a window into the process used by Medicare to assess the “appropriateness of use” of medical therapies. It is to our benefit that we continue to monitor and learn from the process, especially if we are facing the possibility of a single payer system with Medicare for All.

Helen B. Gelly is emeritus medical director of Hyperbaric Physicians of Georgia and chief executive officer of HyperbaRXs, Marietta, GA.



1. Mathematica Policy Research. Interim report for the evaluation of Medicare prior authorization model for non-emergent hyperbaric oxygen (HBO). Available at  Published June 2018.
2. Mathematica. Evaluation of the Medicare prior authorization model for non-emergent hyperbaric oxygen (HBO): final report. Available at Published May 2019.
3. CodeMap. Available at

Back to Top