Cost of Delayed Action

Disability Adjusted Life Years (DALY) is a measure of the loss of life due to disease. It combines an estimate of the mortality in a population, life expectancy, and the reduction in quality of life caused by the long term health impacts. There are two free parameters in our estimates of the per capita loss in DALY due to Covid-19. The disability weight measures the product of the number of people impacted and the individual amount of suffering or disability caused by that consequence. For example, if you think that Covid-19 will produce chronic but mild respiratory issues with a disability weight of 0.17 for 20% of patients who experience severe symptoms, your per capita disability weight would be 0.03. The discount rate is estimate of how much more valuable years at present are compared to years in the future. A typical discount rate is 3%.

The Methods tab contains more detailed information on the calculations and assumptions that went into this figure.

Advanced Market Commitment contracts as a way to overcome COVID-19

Introduction

To overcome the pandemic we need a vaccine. Around 150 vaccines are currently being developed. However, successful development of the vaccine alone does not solve the global problem posed by SARS-CoV-2. For this, equitable access to the vaccine should be ensured for every member of the population.

Why could equitable access to the vaccine be a problem?

• Scaling production of the vaccine: Production capacity for vaccines is costly and may take a long time to install: it means that there are capacity constraints in these markets.
• Cold chain requirements to keep/transport the produced vaccines: it is easier to deliver the vaccine to consumers that live nearby, than to the consumers in remote countries.

Historically many vaccine markets have faced supply shortages. This is due to the demand uncertainty (e.g. uncertain compliance rate) and the production capacity constraints.

It is important to solve/mitigate the “scaling problem” of the production of vaccines, because these capacity constraints translate into delayed or denied access to the vaccine by some groups of the population. There are huge costs of this delayed access in terms of the human health and economic wellbeing of the planet.

Ways to solve/mitigate the “scaling problem”/“delayed access” problem

Advanced Market Commitments (AMCs) are agreements among a developer/producer of the vaccine (i.e. Pharma Company(ies)) and a Regulator/Donor in advance about the terms of trade. That is a Regulator/Donor serves as an aggregated buyer/customer at the vaccine market. The regulator/donor commits to purchase at least a certain amount of vaccines from the producer at a certain price. These agreements are called advanced because they are signed before the actual demand for the product is realized (is known).

AMC creates a more certain environment for the producers. They know that at least a specified amount of their product is going to be purchased. As a result, the producers are more encouraged to invest into their production capacities. This facilitates the production scaling of the vaccines and helps to tackle the problem of availability of the vaccine to the population.

Usually AMC contracts are used between the donors and the pharma companies for the sake of ensuring access to the vaccines for LMIC (low and middle income countries). This is a GAVI model. Serving as an aggregated buyer GAVI Alliance helps to procure the vaccines in the needed amounts for the LMIC.

AMC-type contracts during the pandemic

There are some peculiarities of using AMC contracts during the COVID-19 pandemic:

• The COVID-19 PANDEMIC demonstrates AMCs are used not only to procure the vaccines for the LMIC: during the pandemic HIC (high income countries) also use this type of contracts [1,2].
• Another peculiarity of the AMC-type contracts during pandemic times: the contracts are signed even before the product is developed. So, the regulators commit to buy certain amount of product that does not yet exist and that may never come to the market (since the development process of the vaccine involves some risks: not all the vaccine candidates will turn out being safe and effective and will manage to make it to the market).
• Why is it so? What is it done for?
  • The reason why the contracts are signed in advance is related to the capacity constraints in the vaccine production/ difficulties with scaling of the production/ problem of delayed access to the vaccine.
  • Since building the production capacity takes time, constructing the capacity in parallel (rather than in sequence) with the development of the vaccine will shorten the time when the vaccine is readily available to the public.

Our Question

Are AMC-type contracts signed even before the vaccine is actually developed a good thing to have during pandemic times? Do these types of contract help to solve/mitigate the problem with the delayed access (which is caused by the production capacity constraints)?

Advantages

• AMC contracts give incentives to the vaccine producers to invest into the construction of the production capacities: this is the case since AMCs make the return from the investments into the production capacities more certain (i.e. they remove the demand uncertainties) and the producers are more eager to spend money on costly production capacities.
• Since building the production capacity takes time, constructing the capacity in parallel (rather than in sequence) with the development of the vaccine will shorten the time when the vaccine is readily available to the public.

Disadvantages

The disadvantage of the AMC-type contracts mentioned earlier in the times of pandemic:

• While giving some certainty to the producers with whom AMC contract is signed, AMC contract can harm the investment incentives of the competing developers/producers.
• Assume there are two vaccine developers : A and B. AMC contract between the regulator and developer A removes demand uncertainties for developer A and it guarantees some market share to A. As a result, A is more eager to invest in to the development of its product and into the production capacities suitable for its product.
• On the other hand, since the market for any product (including that of the vaccine) is finite, that is the demand for vaccine is finite), stimulating the demand for one of the product (in our example, stimulating the demand for A) implies decrease in the potential demand for a competing product (i.e. the one produced/developed by B).
• Lower potential demand for the product that is in the development process by B discourages B’s investments. Lower expected return from the participation in the market discourages B’s investments not only into the capacity construction, but also into the development of the B’s product.
• So, the AMC contract between Regulator R and company A:

• Stimulates A’s investments into production capacity and development of A’s product
• Discourages B’s investments into production capacity and development of B’s product
• Overall effect is ambiguous
• It may well be the case that B’s product is of a better quality (i.e. safer/more effective) than A’s product. But because the contract between R and A discourages B’s incentives to invest. It may turn out that the AMC contract between A and R will disturb the development and scaling the production of B’s product. The society may lose from the implementation of the contract between A and R (Deadweight Losses).
• In a positive scenario: A is the best possible product, the AMC contract between A and R encourages A’s investments. AMC contract facilitates access and distribution of A’s product. The society gains from this AMC contract.

HOW to overcome the negative scenario? What is a possible solution?

• We believe that one of the possible solutions is to sign BOTH contracts: not only between R and A, but also between R and B.
• This will provide demand certainty for both developers and none of them will be discouraged to invest into the development of its product and into the production capacity.
• We think that under some circumstances it may turn out optimal even to commit to by the amount of doses of vaccine/product that exceeds the potential market size: on the one hand it is costly for R to do it, but on the other hand this approach gives higher potential return to the developers and stimulate them to expand their production capacity. In the uncertain environment of the development of the vaccine product, it may turn out optimal (ex post) to have the excess production capacity (in order to minimize the time to massive access to the best possible vaccine).

Recommendations

• Use AMC contracts to speed up the scaling of the production of the vaccine.
• BUT: these contract should be signed between the regulator and all the major developers in order to guarantee demand certainty for all of them and encourage their investments into development and capacity construction.
• It may turn out being optimal to have some excess production capacity. So, the regulator(s) may commit to buy the number of doses that exceeds the potential market size.

Further recommendations

• Encourage cooperation among different countries in order to ensure that the decisions that are undertaken are made for the benefit of the globe and not only in the interests of the certain population groups.
• Invest more into the innovation processes in the domain of the global health: this will help accelerating the process of finding solutions in unexpected and urgent circumstances as we are facing now (i.e. COVID-19 pandemic).

References

[1] “BARDA[ Biomedical Advanced Research and Development Authority] awarded a $1.2 billion grant to UK-based AstraZeneca, not just to speed up the research and development of a vaccine, but to reserve 400 million doses of it for the US.” (https://qz.com/1858682/the-us-just-bought-400-million-doses-of-a-coronavirus-vaccine-that-may-never-exist/)
[2] “THE US has purchased around 300 million COVID-19 vaccines from AsteaZeneca.” “After President Donald Trump demanded a vaccine, the US Department of Health and Human Services (HHS) agreed to provide up to $1.2 billion to accelerate British drugmaker AstraZeneca’s vaccine development and secure 300 million doses for the United States.” “This contract with AstraZeneca is a major milestone in Operation Warp Speed’s work toward a safe, effective, widely available vaccine by 2021.” (https://www.express.co.uk/news/world/1285541/Coronavirus-us-vaccine-millions-Donald-trump-AstraZeneca)

Assumptions

Due to the lack of data to guide effective policy during an ongoing pandemic, we have made certain assumptions. The most important of these are that:

1. It is technically possible to develop an effective vaccine,
2. Governments are rational and ethical actors.
3. Vaccine development is controllable through incentives.

Timelines

Google Trends of Vaccine Developers

Using Google Trends, region specific statistics on public search behaviour regarding the interest in companies with vaccine trials underway is extracted as a function of time. A time period from 01/01/2020 to 06/20/2020 and the following geological regions with in media prominent representative countries are chosen: World, Europe (Germany, France, Spain, Italy), Asia (India, Singapore), Africa (Nigeria, South Africa), America (USA, Brazil). As China is critically viewed in releasing information on COVID-19, it is not considered here. The vaccine manufactures names are separately entered as a general search term (no comparative study in Google Trends) and the search is covering the overall word wide web. Data is normalized by Google Trends such that every data point is divided by the total number of Google searches in the respective geological region over the selected time period to obtain relative values between 0 and 100 %.

Search Terms

In order to evaluate sentiments in a statistical manner using Google Trends, 11 keywords connotated with different emotions are chosen. The search terms are based on a recently published study ([1]) having collected 22 million Twitter messages with 25 different hashtags from March 1 to April 21 in 2020. A machine learning approach, Latent Dirichlet Allocation (LDA), was used to identify popular unigram, bigrams, salient topics and themes, and sentiments in the collected Tweets. The NRC Emotion Lexicon, which consists of eight primary emotions: anger, anticipation, fear, surprise, sadness, joy, disgust, and trust was used to classify the sentiments. According to the study, disgust, joy, sadness and surprise are much less significantly detected (<5%) than anger, fear, and anticipation (ca. 10-25%). Consequently, the following keywords from the later emotions are chosen for our Google Trend analysis:

• Anger and fear: deaths, people died, new cases, stop spread
• Anticipation: save lives, public health, rest world, social distancing
• Based on personal interest, we added the terms Category general interest shopping and toilet paper.

On Google Trends, statistics for selected keywords are generated for the same time period as for the vaccine manufacture search above, e.g. from 2020-01-01 to 2020-06-20. No country specific study is perfomed, but the overall worldwide trend is obtained.

[1] https://arxiv.org/ftp/arxiv/papers/2005/2005.12830.pdf

Cost of Delayed Action

To determine the net positive impact of early vaccine development over time we calculated the Cumulative Disability Adjusted Life Years (DALY) lost during the predicted development of the pandemic. We use a piecewise gaussian approximation to the Covid-19 epidemiological model from [2]. Beyond mortality, recovered Covid-19 patients may suffer long term declines in quality of life due to, e.g. lung scarring [3]. To estimate the total cost in human life and well being, we calculate the Cumulative Disability Adjusted Life Years (DALY) by assuming a range of a) long term health consequences and b) future discounting rates [cf 4]. Even under the most conservative assumptions studied, total impact on human well being amounts to millions of lost DALY and grows substantially over time [see also 5]. Moreover, even after excluding mortality, the long life expectancy of those suffering from post-covid lung damage results in significant suffering. We show Quality Adjusted Life Years Gained (QALY), a measure of the benefit gained by preventing a case. Both QALY and DALY can be converted to an economic estimate of the gain or loss to the economy, respectively, by empirically measuring healthcare expenditures. Since these expenditures are both highly uncertain and globally variable, we note only that even the most conservative estimates suggest total economic impacts well in excess of development and deployment costs. Our key conclusion is that governments should seek the rapid development of a vaccine, even at significant cost, and with special attention to the LMICs.

[2] Kissler, Stephen M.; Tedijanto, Christine; Goldstein, Edward; Grad, Yonatan H.; Lipsitch, Marc (2020): Projecting the transmission dynamics of SARS-CoV-2 through the postpandemic period. Science 368 (6493), pp. 860–868. www.doi.org/10.1126/science.abb5793
[3] Chang, Min Cheol; Hur, Jian; Park, Donghwi (2020): Chest Computed Tomography Findings in Asymptomatic Patients with COVID-19.
[4] Jonathan Corum, Denise Grady and Carl Zimmer: Coronavirus Vaccine Tracker. New York Times. Available online at https://www.nytimes.com/interactive/2020/science/coronavirus-vaccine-tracker.html, checked on 2020-06-21.
[5] AstraZeneca (2020): AstraZeneca to supply Europe with up to 400 million doses of Oxford University’s potential COVID-19 vaccine. Available online at https://www.astrazeneca.com/media-centre/articles/2020/astrazeneca-to-supply-europe-with-up-to-400-million-doses-of-oxford-universitys-potential-covid-19-vaccine.html, checked on 2020-06-21.

About

This is a Shiny web application to visualise the quantitative results of group Richardson during the Online Sciathon 2020. A report of our work is available here:

Team

• Nora Bach; Institute of Physics, Carl-von-Ossietzky University Oldenburg, Oldenburg, Germany
• Jyaysi Desai; Department of Rheumatology, Leiden University Medical Center, the Netherlands
• Gaurika Garg; Department of Molecular Biology, Max Planck Institute for Biophysical Chemistry, Germany
• Jana Huisman; Department of Environmental Systems Science, ETH Zurich, Switzerland
• Ruchira Mishra; Department of Physics, Indian Institute of Science Education and Research, Mohali, India
• Joseph (JJ) Richardson; Department of Chemical Engineering, University of Melbourne, Australia
• Olga Rozanova; RANEPA, Moscow, Russia
• Lena Schorr; Division of Genetics, Friedrich-Alexander University, Germany
• Bryan Scott; Department of Physics & Astronomy, University of California, Riverside, United States
• Melania Zauri; CNIO, Madrid, Spain

Contact

For questions or suggestions concerning the app please contact:
jana.huisman [at] env.ethz.ch
This app is still under development. If you notice any bugs, please report them to the email adress above.