Home » Biosecurity » Benefit-Cost Analysis of Ebola Response Strategies

Benefit-Cost Analysis of Ebola Response Strategies

Sign up to the Barefoot Economist’s Newsletter

© Barefoot Economic Services

All content on this website (unless otherwise specified) is copyright of Barefoot Economic Services. Please acknowledge any content used in your work appropriately. Thank you for your consideration.

The spectre of an Ebola outbreak has predictably prompted knee jerk reactions from governments around the world in an attempt to demonstrate that they are in control. Certainly, governments have a responsibility to protect their citizens and to prevent a panic amongst their constituents. In a sense, Ebola is only one example of how perception has come to dominate policy effectiveness in governments around the world. This post is not a whinge-fest on the hopelessness of governments. Instead, I want to demonstrate how a benefit-cost framework can help governments understand the pros and cons of different Ebola response strategies taken from my experience working on biosecurity issues. I will do so in a qualitative way to show that benefit-cost analysis does not necessarily involve lengthy reports and expensive consultants – this is something that can be done quickly to give policy-makers a sense of what are the main drivers of the problem (Ebola) and what strategies can be used to deal with it.

For any good benefit-cost analysis (BCA), the technical facts drives the results. This may require inter-disciplinary cooperation between economists and infectious disease experts. The economists have to understand the facts in order to estimate the benefits and costs, even in a qualitative sense. For a disease, the key facts for a BCA on Ebola (or any infectious disease) are:

  • Infection rate or R from the epidemiological (i.e. disease science) jargon. If R>1, that means every infection will lead to more than one person being infected. In the absence of effective intervention, this means the disease will eventually infect every member of the population. The goal is to reduce R to less than 1 so that the disease ‘burns out’ by itself.
  • Transmission or how does the disease infect people (i.e. voters)? How does it get into a country or my electorate? This will inform the analyst on what strategies are likely to be effective in reducing R.
  • Fatality rate or how deadly is the disease? Again, this can inform the choice of strategy. If a disease is very deadly, prevention is better than cure.

Together, R and transmission can be used in quantitative models to estimate ‘spread’ of the disease. Having an understanding of both of these parameters can give policy-makers of how fast a disease will spread and groups most at-risk. How quickly a disease spreads will drive the economic costs of the disease such as by reducing labour productivity, increasing health care costs and reducing trade and commerce. The fatality rate is a key factor in determining the human cost of the disease which not only have short-term but also long-term economic costs by reducing overall demand and labour supply.

So where does Ebola stand with these three facts? For R, Ebola has estimates of 1.2 to 4, clearly a highly infectious disease but not as infectious as SARS. The reason for the variation in estimates is due to different contexts. For example, the more conservative estimates are from Sierra Leone and the higher estimates are for hospitals in the worst-affected African countries. Clearly, Ebola is not going to burn out by itself and will need an effective public health intervention to reduce R<1.

How is Ebola transmitted? Unlike other diseases, Ebola is only contagious when an infected person is displaying symptoms, such as fever, vomiting, diarrhoea and malaise. But to be contracted, there needs to be an exchange of bodily fluids (e.g. saliva or blood) between an infected to a healthy person. Even then, the level of Ebola virus needs to be high to be contagious. This happens 2-3 days after the patient is displaying symptoms. Quick and accurate tests are available to diagnose Ebola once an infected person is symptomatic. Furthermore, the nature of the transmission provides for a relatively targeted response. More on that later.

But once contracted, Ebola is highly fatal. This is probably where the potential for panic is greatest, that and the lack of commercially-available vaccine. Historically, Ebola has been estimated to have a fatality rate of 25-90%. The current outbreak has been estimated to have a fatality rate of 70%.

So now we know some epidemiological facts about Ebola, we can now bring a benefit-cost lens to evaluate some strategies for responding to it. First, we need to distinguish between national and international responses. The difference of the geographic scope of the analysis determines what benefits and costs to include (i.e. who to care about). Obviously, for a national analysis, we only include benefits and costs for our country. For an international analysis, you still have to make choices on which countries to include. For example, the Asian Development Bank may be concerned on how Ebola would affect it’s Asian clients so would only include those countries in it’s analysis.

Secondly, we would also need to know the cost of a specific strategy, preferably in financial terms. It is not uncommon for agencies to have only a vague idea on the cost of response strategies because of interdependencies between spread and cost.

Finally, we would also need some way of estimating benefits. A key benefit would be avoided deaths. This can be estimated by using estimates of the value of human life. There are many estimates available and the choice depends on the context. I won’t get into a discussion about which one to use, ultimately an analyst and policy-makers are accountable to their stakeholders on which one to use. Other benefits would include the avoided disruption to the economy and avoided health care costs. Economic disruption costs may require additional modelling depending on cost and time constraints. Health care costs could be based on similar incidents (e.g. intensive care costs). The act of going through this process helps analysts and policy-makers understand the scale of the economic costs that are at stake.

Let us now look at three national strategies that can be implemented to respond to Ebola: quarantine, isolation and travel bans. Quarantine strategies are a preventative measure that isolates at-risk individuals from the rest of the population. At-risk groups could include medical professionals returning from assisting with the Ebola effort in West Africa. Quarantined individuals need not be symptomatic, instead it is to see if they will develop symptoms. The main benefit of a quarantine strategy is that it prevents an outbreak occurring if well-targeted. It can also prevent a mass panic by removing at-risk individuals from the population. The costs of quarantine could be minimal if home quarantine is used. There are wider economic costs from removing people from their normal economic activities. Quarantine may inadvertently detain healthy individuals and disrupt economic activity. In some cases it can cause mental health issues by stigmatising quarantined individuals. Furthermore, it could reduce the pool of available medical professionals who would volunteer to assist with the effort to contain Ebola in West Africa. The medical community have criticised US States’ use of quarantines as a political response that is “…like driving a carpet tack with a sledgehammer: it gets the job done but overall is more destructive than beneficial“. Use of a quarantine strategies for Ebola may inadvertently send the wrong signals about the nature of the disease and may lead to stigmatisation of health workers and people from West Africa. While quarantine can reduce R, it may be a high-cost alternative.

Isolation strategies are responsive rather than preventative in nature. Isolation works by removing symptomatic individuals from the population. Given that Ebola requires 2-3 days when an individual displays symptoms to become contagious, if a symptomatic individual is isolated before their illness becomes contagious, isolation could effectively target individuals at risk of spreading Ebola for treatment. The benefits from isolation are in reducing spread (i.e. reduces R) by removing contagious individuals from the healthy population. The costs may be higher than quarantine because it requires the capability to adequately isolate and care for infected individuals to allow for their recovery. Furthermore, how infected individuals are found is a design question. In some US States, at-risk individuals were provided with financial assistance if they handed themselves in to authorities. Or do you rely on people’s self-interest who want to receive health care to recover as some infected health professionals have done. A combination of these two measures, could be used to design an ‘information-revelation’ mechanism that is ‘incentive-compatible’ for infected individuals to voluntarily submit to isolation for the good of the community. This strategy may not be appropriate for all countries given that it requires a well-functioning health system to effectively isolate and care for infected individuals. Countries such as Australia and Canada should be in a good position to implement an isolation strategies.

Travel bans to and from countries suffering from Ebola have been implemented in almost 30 countries. Like quarantines, travel bans are a preventative strategy to sever the physical transmission channels of Ebola between countries. The benefits are similar to quarantine. The costs are minimal to the country implementing it unless there are significant economic links between the infected and healthy country. This may be a concern for major importers of palm oil and cocoa which are key regional exports of West Africa. Travel bans may be a high-cost way of preventing the transmission of Ebola into a healthy country. Furthermore, most of the costs are imposed on the country suffering from Ebola through ‘aversion‘ of contact so may be ignored unless the BCA was for an infected country. Nevertheless, a travel ban strategy may be the only option for countries with poor health systems and share a border with an infected country.

What about international strategies? I’ll consider two strategies here: eradication and containment. The eradication strategy is essentially to wipe out Ebola at it’s source. The goal is to ensure it cannot establish and become an ongoing public health issue such as influeza or AIDS. The benefits from eradication is that it contains the human and economic cost while it is active. The avoided costs could be substantial given R and the fatality rate of Ebola especially if we expect the disease to persist in the absence of effective intervention. However, eradication is not cheap. It requires the use of highly-trained medical professionals in a large enough scale sufficient to reduce R<1. In the current West African outbreak, this is beyond the capabilities of the countries infected and requires the use of foreign professionals who are often more expensive than local counterparts. If there is international cooperation, the costs may be shared by local countries through the economic and human costs caused by Ebola and international partners providing the medical professionals. Even if there is sufficient resources to combat Ebola, there are no guarantees that an eradication strategy would succeed. In that case, the international community would have to consider a containment strategy.

A containment strategy attempts to isolate regions with a country that is suffering from Ebola so that it burns itself out. Essentially, it is an imposition of travel bans to and from the most worst-affected regions. At it’s most extreme it may impose travel bans on Guinea, Liberia and Sierra Leone. This would be the equivalent to trade sanctions because it would stop not only movement of people into and out of the affected countries but also halt international trade. The benefits of a containment strategy would be to prevent the spread of Ebola beyond West Africa. However, the human and economic costs will almost be borne exclusively by Guinea, Liberia and Sierra Leone. Given that R has been estimated at over 1, it is unlikely that Ebola will burn itself out until it has infected everyone in Guinea, Liberia and Sierra Leone. The human costs of a containment strategy is almost too much to bear thinking about. Personally, the only reason you would want to impose a containment strategy is if you have run out of options and if Ebola is likely to spread to a much larger country (e.g. Nigeria), the human and economic costs could be even greater in this case. As a stand-alone strategy, containment is unlikely to succeed in a cost-effective manner. A combination of containment and eradication may be a more preferable alternative.

So, hopefully this post has shown you how the dark arts of BCA can be used to weigh the pros and cons of Ebola response strategies. What do I think we should do? I’ll give you a typical economist’s response: it depends. The choice of strategy would depend on the local context. For example, if I was advising the Prime Minister of Australia or Canada, I would say use an isolation strategy at a national level because of the strength of our health systems. However, if I was advising the President of Nigeria, I would advise the use of travel bans because our health system isn’t set up to deal with an Ebola epidemic. As Deng Xiaoping said, “seek truth from facts”. The same could be said for a BCA because like any model, it relies on the parameters and assumptions you use and the best source for this are the scientists who are experts on Ebola. And don’t get too hung up on precision, as John Maynard Keynes said, “better to be vaguely right than to be precisely wrong”. Otherwise, by the time you have gathered all the necessary data for your brilliant economic model, thousands more people in West Africa might’ve died.


6 Comments

  1. Archie says:

    Hi Barefoot Economist, I’m curious as to why you would advise an Isolation over a Quarantine strategy? You mentioned that Quarantining is preventative and that Isolation could be more expensive. I’m also wondering if there is a psychological benefit: would the community feel safer (less panicked?) if anyone, particularly health professionals, returning/travelling from West Africa were quarantined for 21 days on arrival before being allowed back into the community?

    • arthurchha says:

      Hi Archie, thanks for the comment. Maybe I wasn’t clear enough on my reasons. Quarantining for Ebola may involve detaining more people than you need to. As I mentioned in the post, Ebola isn’t contagious until 2-3 days an infected individual starts displaying symptoms. This should be enough time for the infected individual to be diagnosed and isolated before infecting some one else. So a quarantining strategy would needlessly detained infected and healthy individuals. Yes, it could prevent panic as I mentioned. But it doesn’t seem to be more effective than isolation. As for costs, it could be higher than quarantine if home detention was used. Unfortunately, home detention hasn’t been implemented to prevent quarantine to prevent contagion. For example, family members are allowed to visit quarantined individuals in some US States. So if quarantine is implemented ineffectively, it could actually incur all the costs of quarantine but have no meaningful effect on reducing R. Isolation on the other hand is more targeted because you can effectively treat the individuals who have Ebola rather than individuals who might have it. Of course, it depends if you have the systems capable of detecting and isolating infected individuals. Even better, if there are systems to help people self-report their Ebola status. To me, it seems that a country with an excellent health care system such as Australia or Canada would be better off using isolation because they are capable of already dealing with such emergencies (i.e. the capability already exists and doesn’t require additional investment) without needlessly stigmatising at-risk individuals regardless of their Ebola status. Also, isolation doesn’t create a climate of panic that a quarantine strategy does which pre-judges all at-risk individuals regardless of their status. Remember, Ebola is only contagious only if an individual displays symptoms and even then, the infected individual should be diagnosed properly to determine if isolation is required. So I guess isolation is a more effective alternative for a country with a strong health system. It may not be appropriate for Nigeria or Mali. Hope that makes sense.

  2. Archie says:

    > “Of course, it depends if you have the systems capable of detecting and isolating infected individuals.”

    Yes, it appears it is not possible to distinguish whether someone has influenza or Ebola based on symptoms alone. A test (prescribed by a doctor) is usually required, by which time the person may well have been in close contact with those who may not know they have been put at risk . Every approach has some level of risk I guess. The risk with the Isolation approach is that it relies on the individual to know/guess if they have been with at-risk people and to get themselves properly checked out when flu-like symptoms occur.

    • arthurchha says:

      Yes, you could be right. But if we quarantined everyone with influenza symptoms the costs could be prohibitive to both government and the wider economy. For example, if Australia decided to quarantined anyone with Ebola-like symptoms (i.e. flu symptoms) in 2014, that means the government would’ve had to detain 62,918 people as of 10 October. How many Ebola cases would that have prevented? I think it would’ve prevented zero cases. So I think the quarantine approach in a country like Australia would be a massive over-reaction. I did point out, that isolation would not be feasible in countries without the robust systems that Australia or Canada has. I also did point out the Ebola response strategies depends on the strength of a country’s health system. So, the choice of strategy depends on what tools you have available.

      • Archie says:

        Yes that would be expensive. I actually thought quarantining would be of people coming to Australia from significantly affected West African nations (not those with flu-like symptoms), though I suspect the number would be just as high.

        I was discussing this topic with a friend and medical practitioner/trainer the other night and he was telling me that Australia doesn’t yet have adequate facilities to implement an Isolation approach, so that would be something we’d have to invest in quickly if we chose to go down this path. (I encouraged him to comment on your site with clearer details than what I could provide, but he may be a bit too shy/busy 🙂

        • arthurchha says:

          A quarantine strategy can be as broad as you want it. It doesn’t necessarily have to involve newly arrived people. It could get out of hand. You can imagine a situation where people wouldn’t report any flu-like symptoms because they know they would be quarantined for 21 days.

          It would be interesting to hear from your friend about isolation and other strategies discussed in this post. The more facts the better I say!

Comments are closed.