Stuart R. Gallant, MD, PhD

This is the second post in a two-part series on the value of medications.  In this post, we bite the bullet and look at cost-effectiveness data in units of dollars per QALY, bearing in mind that there are some significant theoretical and practical problems of this type of analysis.

An Example:  Edoxaban Versus Apixaban

In one example of QALY analysis, 1000 Spanish patients were studied to compare edoxaban and apixaban, two direct factor Xa inhibitors used to prevent clotting in certain medical situations [1].  The development of both drugs somewhat overlapped in time, with apixaban licensed in the EU 2011 and edoxaban licensed in the EU 2015, both medicines are on the WHO List of Essential Medicines.  It would be hard to come up with a more “apples to apples” comparison (similar mechanisms, medical indications, and patient populations).

In this study, the authors looked at the use of these two medicines to prevent stroke and systemic embolus in patients with atrial fibrillation (AF).  Patients with atrial fibrillation have inefficient heart contraction which can allow blood to remain in the heart long enough to form clots.  Medications which act against factor Xa make clotting more difficult (a benefit in the case of AF), but they can also increase the risk of uncontrolled bleeding.  So, the treatment with these drugs entails risk.  The study of produced the following data:

	Apixaban	Edoxaban 60 mg	Edoxaban 30 mg
QALY (years)	6.924	6.882	6.844
NHS Cost (Euros)	€ 19,053	€ 18,651	€ 19,025

This table is exactly the kind of information that a patient, medical provider, regulator, or healthcare administrator might want.  The years of extended life for both medications are excellent (about 7 years in each case), and the medical costs are almost identical.  Presumably, the two pharmaceutical companies negotiated essentially the same supply agreement in each case.  The only difference is a difference in “major bleeds” (see original paper) which was higher for apixaban, compared to edoxaban 30 mg; however, based on the identical QALY values, the risk of bleeding did not significantly lower apixaban’s QALY value.

Looking at this data, a medical provider might conclude that the two medications are interchangeable (except for the lower major bleed count for edoxaban 30 mg, and the marginally cheaper cost of edoxiban), but that’s not what the authors did.  They went on to create some numerical salad.  The difference between QALYs for apixaban and edoxaban 60 mg is 0.042; the difference in NHS cost is €402.  The authors calculated the “incremental cost/QALY” ratio as €9639.33/yr.  Most analysists would realize that this ratio has no meaning—the difference in QALY is essentially zero and the difference in cost is essentially zero.  The ratio has no meaning because the numerator and particularly the denominator are for all intents and purposes zero—the resulting ratio is essentially numerical round-off error.

Big picture:  except for fumbling the ball on the 5-yard line, this is a good study and use of QALY analysis; however, it shows the dangers of QALY analysis because it is possible to wander off into numerical Neverland.

Institute for Clinical and Economic Review (ICER)

The Institute for Clinical and Economic Review (ICER) is an independent non-profit organization that frequently opines on the cost effectiveness of medications [2].  In its work, ICER calculates cost per QALY ($/QALY) and votes on the relative cost effectiveness of various treatments.  ICER doesn’t have a single value threshold for “high value,” “intermediate value,” or “low value,” but a pattern can be discerned in ICER votes (the following figure is taken from Cohen et al. [3]):

By examining the buckets, it is possible to get a sense of the range that ICER feels represents “high value” medications, with almost none occurring above $175,000 per QALY.  Similarly, almost all medications above $500,000 per QALY are reviewed as “low value.”  Once again, it is important to point out, as discussed in the first part of this series, that QALY analysis has significant flaws.  To ICER’s credit, the fact that the data does not show a simple threshold value shows that other factors were considered in the review process (other than $/QALY).

Four Quadrant Analysis

It is possible to tie together economic (investment) value and clinical cost effectiveness value.  For each pharmaceutical development project, there is an expected value based on:  1) cost of pre-clinical and clinical programs, 2) manufacturing costs, 3) overhead (management, regulatory, legal, staffing, marketing, fund raising, financing…), 4) return on investment, factoring in failed as well as successful development programs.  Comparing this expected market value to the actual market value of the drug (number of treatable patients x revenue per patient), the following equation applies:

Why is the ratio greater than unity?  If the expected value of the project, including return, is not greater than the actual value of the market, then the project does not go forward because investment capital will not be available.  Combining Cohen’s concept of a bucket plot and the market ratio above, a four-quadrant plot can be prepared.  (Since this is a purely conceptual device, drugs are partitioned into high and low cost-effectiveness buckets (rather than five buckets used by Cohen), but concepts discussed below apply even in the grey five-bucket world of reality.)

This simple plot contains a large number of lessons:

1. The red zone (Quadrants 3 and 4):  As mentioned above, drugs with low economic viability do not advance to approval because investment capital is not available (unless for example, a non-profit or government entity funds development).
2. Quadrant 1 (high cost effectiveness and high economic viability) is the dream of every pharmaceutical company—they want to heal patients and they need to please their investors.  There is an element of chance to ending up in the first quadrant—every idea for a drug sounds good before actual data begins to accumulate, but only the drugs favored by high pharmaceutical efficacy actually demonstrate high cost effectiveness (low $/QALY).
3. Quadrant 2 is an interesting quadrant.  Within this quadrant are what might be called borderline drugs (show in the blue region).  Some of these drugs in the blue region have reasonable efficacy (QALY) but too high cost.  A drop in price could make them more cost effective (in $/QALY) but might not destroy the economic value of the project.  One thing that can catalyze this type of move from Quadrant 2 to Quadrant 1 is a change of the economic basis of the drug.  The classic example is when a medication moves to generic—its price drops and prescriptions go up.  What allows this to happen is that the generic company isn’t burdened by the high costs of innovative drug development, so the generic company’s investors have a lower expected project value.  Another example is the sale of a product from one pharmaceutical company to another.  For example, Sollpura (liprotamase) was purchased by Anthera Pharmaceuticals from Eli Lilly [4]; the sale allowed Sollpura development to continue because essentially all the development costs prior to the sale were written off by Eli Lilly.  Ultimately, Sollpura failed in the clinic, but the Sollpura example nevertheless illustrates the point.
4. Another borderline area is the set of drugs in the crosshatched area of Quadrant 3.  These have high-cost effectiveness, but poor economics.  A modest price increase could improve economics without totally compromising cost effectiveness.  A good example of this type of drug are the numerous generics that form a bulwark of the pharmaceutical armamentarium.  Frequently, they are produced at aging plants which require significant capital investment to bring the drugs up to current manufacturing standards; however, it can be difficult to negotiate those price increases and a careful dialog with payors and formulary committees is often required.

The Future

Attempting to bring together this wandering two-part post, I would like to return to the three spheres figure from the beginning of the post.  Pharmaceuticals exist at the intersection of these three spheres:  science (what is technically possible), economics (what the individual and the society can afford), and patient care (what patients want for their health).

In that golden age of pharmaceuticals from the 1950s through the 1990s, medications were developed to treat the major epidemiological phenomena of industrial societies:  infectious diseases (antibacterials, antifungals, and antivirals), metabolic disease (statins, insulins, and others), cardiac (antihypertensives), psychiatric (antidepressants, mood stabilizers), and others.  For the most part those medications were celebrated as a triumph within all three spheres (science, economics, and patient care).

Now medicine is making bold leaps into the future:  not just bifunctional antibodies but trifunctional antibodies, viruses, mRNA, cell therapies, etc.  And, pharmaceutical companies are being asked to discuss the value of these treatments.  This post has been devoted to the idea that there are different ways of valuing pharmaceuticals.  A medication can be valued for its technical excellence of curing disease or for its ability to marginally improve the level of care as one step toward a cure.  A medication may be valued highly by patients for providing them freedom from one particular burden of disease or for recognizing their needs and attempting to provide care in a patient-centric way.

Hopefully, we will continue to look at medications in a complex way and recognize their many values to society and to individual patients.  And, hopefully pharmaceutical companies will continue to work with patients to design better treatments to disease and provide greater control to each individual patient.

[1] Oyagüez, I. et al.  “Cost‑Effectiveness Analysis of Apixaban Versus Edoxaban in Patients with Atrial Fibrillation for Stroke Prevention,” PharmacoEconomics – Open (2020) 4:485–497.

[2] ICER.  “2020-2023 Value Assessment Framework” 2020.

[3] Cohen, J.T., et al.  “Does the Institute for Clinical and Economic Review Revise Its Findings in Response to Industry Comments?,” Value in Health 22 (12), P1396-1401, Dec 01, 2019, www.valueinhealthjournal.com/article/S1098-3015(19)32345-9/fulltext.

[4] Accessed on Internet Archive:  investor.anthera.com/releasedetail.cfm?ReleaseID=859598

Disclaimer:  PharmaTopo^TM provides commentary on topics related to drugs.  The content on this website does not constitute technical, medical, legal, or financial advice.  Consult an appropriately skilled professional, such as an engineer, doctor, lawyer, or investment counselor, prior to undertaking any action related to the topics discussed on PharmaTopo.com.