On March 6th ICER released their final Evidence Report on Multiple Sclerosis medications. The conclusion was that every MS medication, except for Lemtrada, do not provide enough value to justify their prices. As with every ICER report that is released, the pharmaceutical manufacturers refute the findings and tout their R&D expenses and the very real burden of bringing life saving drugs to markets.

I’m not looking to start a debate on the merits of either ICER’s or the manufacturers’ arguments in this post. Taking my position on the fence, I can say that I believe ICER provides a valuable, independent view into value, but at the same time pharmaceutical companies should expect to be able to be reasonably compensated for their efforts.

When ICER started producing their reports a couple years ago, I was very curious to see how they were going to account for adverse events in their models. Although they touched upon the implications of side effects, in the reports that I examined, it certainly wasn’t a focus and even deemed a non-factor in some. In the most recent report however, I was pleasantly surprised to see them state,

“…we included associated serious adverse events (SAEs), as severe events tend to differ between treatments and have effects on costs and/or health outcomes.”. 

Now we’re talking! However, as I continued to read, I was disappointed to see that they only included SAEs where rates were provided in the drugs’ label and calculated SAE rates for each DMT where that rate occurred in at least 1% of patients in clinical trials. As you can see from their the table that was included on page 206 of 253 in the Appendix, only 4 of the 14 drugs (Betaseron/Extavia, Aubagio, Tecfidera, and Tysabri) had any AE cost applied to them in the model, ranging from an additional $3 to $154 a patient per year.


ICER MS AE Costs.png 

While we understand the resource constraints that exist when putting together a complex model like this, we believe, based on data and analysis found in our Evidex™ platform that ICER drastically underestimated the impact that adverse events have on downstream medical costs for three main reasons:

  1. The ICER model limited the number of AEs included in their analysis by only looking at AEs that have rates provided on the collective labels of the drugs.
  2. The ICER model only used data from clinical trials and not what is being seen real-world.
  3. The ICER model did not take into account AEs that are occurring at disproportionate rates in the real-world that are not yet on drug labels.

To provide a comparison, and highlight the degree of underestimation, the below table provides downstream cost estimations for the DMT’s using Advera Health’s RxCost® analytic.

Brand Name

Generic Name

RxCost® per Patient








Interferon B-1a



Interferon B-1a



Interferon B-1b






Peginterferon B-1a



Interferon B-1b



Glatiramer Acetate



Dimethyl fumarate




 Not Enough Data



 Not Enough Data


Glatiramer Acetate

 Not Enough Data



 Not Enough Data


The methodology for RxCost® is peer-reviewed and published in the Journal of Managed Care and Specialty Pharmacy. It is also featured as a complimentary analytic on the AMCP eDossier® platform. Furthermore, a poster presentation will be held at the AMCP Annual Meeting on March 28-20 2017 in Denver, CO helping to validate the use of RxCost by payers. The poster is titled, Adverse Event Data as Proxy to Determine Total Medical Costs for TNF-alpha Inhibitors. This study was done in collaboration with our partner and client WEA Trust, a not-for-profit company based in Madison, WI offering high-quality health plans to all WI public employees.

The data behind RxCost is fully transparent on the Evidex platform.

We also provide the ability to quickly get the AE rates from clinical trials for the medications as well. A quick screenshot of our clinical trial evidence on Tysabri is below. You’ll note that in our pooled analysis, PML was reported at a rate of 0.43% vs. 0% in control groups, and being reported to FAERS at a rate of 1.0256%. Both higher than ICER’s estimation of 0.03%.

tysabri ace.png

Without access to ICER’s underlying model it is difficult to know whether these increased numbers would change their assessment. However, even if the underlying conclusions were to be the same, we believe that these differences in data are important to all healthcare decision makers.

If you are going to be at the AMCP meeting, I encourage you to stop by and talk with us at booth #237. Otherwise, feel free to contact us to discuss further.


Topics: Drug / Indication Information

Jim Davis

Written by Jim Davis

As Executive Vice President, Jim is responsible for the commercialization strategy for Advera Health Analytics.