HERC Spotlight: November 2023


 Health Economics SpotlightHERC logo

Updates on VA data, health economics research, and analytic methods

November 2023


In This Issue

  1. Opinion: Psychedelics and mental health. Is there a role for economics?
  2. Updates to the Nosos Risk Scores
  3. 2024 Cost-Effectiveness Analysis Seminar Series: CEA and Beyond


Psychedelics and mental health. Is there a role for economics?

By Todd Wagner

Over the past 18 months, I have fielded a few questions on psychedelics as a treatment for mental health conditions, such as post-traumatic stress disorder (PTSD). I’m admittedly pretty ignorant about psychedelics, with most of my “knowledge” coming through podcasts like Peter Attia’s The Drive. I’m also not a mental health practitioner. What I am is a health economist, and so people usually come to ask for my advice on how to conduct a cost-effectiveness analysis (CEA) on a new treatment. In this case, the new treatment is the psychedelic, MDMA. My current opinion is that we don’t need a CEA on MDMA. My opinion has surprised some, so in this Spotlight, I thought I would explain my rationale.

If you are not familiar with CEA, here is a description on our webpage. And if you’d like to see a good example, let me point you to a recent CEA model for treating opiate use disorder. I think this is an excellent study, but full disclosure, the authors are my colleagues. A CEA essentially compares two or more treatments on their costs and benefits. So let’s look at the costs and benefits of MDMA.

MDMA is not an expensive drug. What makes it more costly is paring the drug with intensive psychotherapy. A recent randomized trial paired MDMA with 2-3 intensive 8-hour psychotherapy sessions. Psychotherapy may seem expensive and many people pay out of pocket for psychotherapy, where hourly rates can be upwards of $250 an hour. If you did 24 hours of psychotherapy at $250/hr then the total would be $6000. That is real money, especially if you are paying out of pocket. But it isn’t that expensive for insurers who routinely see much higher prices for many treatments. So I don’t think a CEA would be merited solely due to its cost.

What excites many about psychedelics is the ability to treat, and in some cases, cure conditions, for which we don’t have a cure. The effects from some of the trials, including the one mentioned above are large. Moreover, MDMA isn’t addictive and it is well tolerated. It almost sounds too good to be true. To be fair, the Food and Drug Administration (FDA) has not ruled on whether MDMA or other psychedelics are safe and effective, but the Agency is currently reviewing the evidence for MDMA.

The last time VA was faced with a game changing treatment was Sofosbuvir (brand name Sovaldi) in 2015 for Hepatitis C. That drug cost $1000 a pill, but it was considered by many to be a revolutionary treatment. In the end, VA, with some help from Congress, paid for the new treatment. In comparison, MDMA could be as big a game changer as Sofosbuvir and less expensive. That is why if MDMA is approved by the FDA, I fully expect VA to provide coverage.

A few weeks ago, I was invited to attend a VA State of the Art conference on using psychedelics for treating mental health conditions. The meeting in Denver was well attended by VA as well as other agencies (FDA, NIMH, NIDA). On the first day of the meeting, we broke into smaller working groups and my group was focused on implementation. Rumor had it that the FDA might rule on MDMA as early as Q2 of 2024. Other psychedelics are not as far along. That rumor was a wake up call for many of us in the room because most providers have not created a plan for dispensing this treatment. This is one of those moments where you can revel at the sheer power and will of VA as a health care provider. It has the ability to see these issues coming and develop implementation plan accordingly. The second day included summary presentations from the breakout sessions from the day prior and a keynote from the Undersecretary, who stressed the importance of using science to help Veterans.

I’m optimistic about MDMA as a treatment for PTSD and other conditions. I am eager to see the science play out, but if the treatments prove safe and effective, I remain skeptical that we need a CEA to decide whether to adopt this treatment.

Disclaimer: The views expressed in this article are those of the author and do not necessarily reflect the position or policy of the Department of Veterans Affairs or the United States government.


Updates to the Nosos Risk Scores

Nosos is the VA-specific risk score developed in 2016 to support performance measurement. Nosos is based on the Centers for Medicare and Medicaid Services (CMS) Hierarchical Conditions Categories (HCCs) risk adjustment system that was developed for Medicare Advantage (see HERC’s Guide to the Nosos Risk Adjustment Score for details). Nosos is available in the CDW and listed on the VHA Data Portal.

Over the past 8 months, we have been investigating ways to improve the production of the Nosos risk score. Given our limited budget, we were focused on areas where we could gain efficiencies and areas of scientific merit. Three areas came up: quarterly updates, the registry variable, and social risk factors.


The first noteworthy change is that we are going to stop production of the quarterly Nosos estimates. The annual updates are the best risk scores given that they reflect all the costs and care in the past year. We realized that few people were using the quarterly updates.

Next, we examined the impact of the VA registry variable included in Nosos. Since its inception, the Nosos risk score has included registry information from the VA Allocation Resource Center. The registry variable describes whether a Veteran is enrolled in one of 16 VA registries: spinal cord injury, chronically mentally ill, blind, post-traumatic stress disorder (PTSD), alcohol, stroke, hepatitis C virus, home health, domiciliary, long-term care, brain, AIDS, transplant, end stage renal disease, homeless, and care coordination home telehealth. The registry variable is predictive of costs, even though most of the illness captured in the registry are already included in Nosos through HCC and PsyCMS codes.

What is risk adjustment?

Risk adjustment is one way of adjusting for inherent differences in patients' underlying clinical needs

What risk scores are available in VA?

The most commonly used risk scores in VA are the Care Assessment Need (CAN) and Nosos risk scores.

CAN was designed to predict hospitalization and mortality.

Nosos was designed to predict costs.


We computed Nosos prospective and concurrent scores, both with and without the registry variable, for all Veteran enrollees in 2022 (n=6 million) and extracted two 20% random samples. To understand distribution effects, we examined changes in Nosos score among quartiles, where quartile 1 is the least expensive. We computed standardized mean differences (SMD) for the difference between the concurrent Nosos score with and without the registry variable, as well as the difference between the prospective Nosos score with and without the registry variable. We then compared veterans with different diagnoses to understand if excluding the registry had any effect on the mean score. Full details on the analyses are in Technical Report 43.

In most comparisons, and we made a lot of comparisons, removing the registry variable had little or no effect on the Nosos scores. Where the registry variable helped predict cost was among 12 high cost (quartile 4) HCCs: HIV/AIDS, Quadriplegia, Paraplegia, Amyotrophic Lateral Sclerosis and Other Motor Neuron Disease, Cerebral Palsy, Intracranial Hemorrhage, Ischemic or Unspecified Stroke, Hemiplegia/Hemiparesis, Monoplegia, Other Paralytic Syndromes, Dialysis Status and Chronic Kidney Disease – Stage 5. There was little impact on mental health conditions (PsyCMS) or VA drug class categories. Because we made so many comparisons, we use a second random sample data to confirm our results; the the results were highly robust.

Removing the registry variable would help us expedite the processing of Nosos. Removing the registry limits our ability to predict high costs among a handful of HCC groups, but we think the efficiency gain is worth it.

Next up, we know that people are interested in controlling for social risk factors. Based on the work of a recent team that examined social risk variables in VA, we do not think the data are currently available at scale to include them in Nosos, but we are tracking the issue and will let users know if and when we make changes to include social risk.

Nosos Resources


2024 Cost-Effectiveness Analysis Seminar Series: CEA and Beyond

The Cost-Effectiveness Analysis seminar series will kick off in January 2024. The purpose of this series is to introduce researchers to conducting economic evaluations of healthcare interventions in VA. Seminars will provide an overview of the decision analysis landscape and methods to operationalize economic analyses. We’ll look at cost-effectiveness analysis, budget impact analysis, and additional methods such as event study design. Topics will include estimating the cost of an intervention, operationalizing Markov models, and CEA alongside a clinical trial.

The full schedule will be released December 2023 and will be available on the HERC website.