This essay was submitted to Open Philanthropy's Cause Exploration Prizes contest.
If you're seeing this in summer 2022, we'll be posting many submissions in a short period. If you want to stop seeing them so often, apply a filter for the appropriate tag!
Introduction
As the disappearance of food and waterborne epidemics in many parts of the world following the introduction of certain hygienic measures regarding food, water and sewage treatment has shown, humanity is well equipped to prevent the pandemic spread of novel food- and water borne diseases. Given the lack of such epidemics and the relative successes of medical hygiene practices in clinical settings we are also probably moderately well equipped to prevent the pandemic spread of novel contact- and fomite transmitted diseases by using pathogen agnostic hygienic interventions. On the other hand, Covid has demonstrated that we are sorely lacking the means to do so for pathogens with airborne- and droplet based transmission modes. In the same vein the continued threat from malaria combined with the utter lack of preparation in non-malaria areas suggests that we would not fare much better (or possibly even worse) against a novel disease using a parasitic insect or other arthropod as a vector.
This situation combined with the rise of more and more capable bioengineering tools that make the creation of extremely infectious and extremely lethal pathogens - either through inadvertent lab leaks or through intentional weaponization - progressively more likely this state of affairs represents a significant threat to civilisation and even to the continued existence of humanity as a species.
This paper argues that there is significant opportunity for philanthropic intervention to facilitate the development, production and storage of both the physical means of protection from such threats (such as effective, easy to use and convenient respiratory and eye protection and effective, easy to use and convenient insect bite protection) as well as public education materials to ingrain proper practices in the public to use said equipment to its full potential.
Motivation
While pandemic prevention is a traditional cause area in the EA community I have never seen a comprehensive pathogen agnostic proposal for the prevention of the spread of a novel pathogen. Before the Covid pandemic many people, myself included, had an informal understanding that state sponsored health agencies deal with the topic and there is little room for improvement by private actors.
Our abject failure to contain Covid in spite of heroic public interventions was a rude wake up call in this regard. Political decisions were made on a week by week basis with the main goal being the prevention of an overload of the public health system while waiting for the vaccination rollout as a silver bullet. At least western societies mostly gave up on trying to contain Covid after vaccination turned out to be insufficient. While in this case of Covid this is not a threat to the continued function of society, history (e.g. the devastation of Native American populations and the Black Death in Europe) and a look at diseases in other species (e.g. the near eradication of rabbits in Australia by Myxomatosis and the ongoing devastation of many amphibian populations by Batrachochytrium dendrobatidis) tells us that there are many potential future diseases where a similar outcome would end our civilization.
With expanding biotech capabilities the probability of both lab leaks and weaponization by rogue actors rises steeply, ensuring that over the next decades this threat will be enhanced greatly.
Lessons from Covid
Our inability to contain the spread of Covid was not fate or a lack of the requisite technology. From a technical perspective, containing Covid would have been quite possible and our failure to do so is entirely explained by institutional incompetence.
Early on in the pandemic I have written extensively on how masks could be used to halt and reverse the spread of Covid. But while official resistance against mask use quickly faded and even mask mandates were put into place, the specific implementation of these measures ensured that only marginal effects were achieved. First, no one bothered to develop and distribute high quality masks for different facial shapes ensuring a good seal while maintaining ease of use as well as low breathing resistance and low impact on communication ability. Second, mask mandates were always location or situation specific. (e.g. In Germany in public transport, grocery stores and public facing businesses masking was almost universal while in many workshops and other businesses almost no one was masked.) Third, there was a lack of public education campaigns on when (all contacts with people not from your household, yes also family and friends!) and how (propper seal!) to use masks. These problems combined to ensure that there were always enough unmasked (or improperly masked) contacts for infection chains to persist.
With the developement of antigen self tests we had yet another tool that could in theory have been used to contain Covid, but again, incompetence prevailed.
Test sensitivity varies widely among commercial tests, and regulation (talking about Germany again) is lax enough that tests are approved (even for use in official test centers) that are about a thousandfold less sensitive than the most sensitive tests commercially available. This problem alone ensures that many infectious individuals are not caught by those tests. To contain and reverse Covid spread we would need to test most people who are going to have contacts on a given day every day for two or three weeks. But instead of doing so after saving up enough test kits, test regimes were again highly situational and again ensured that enough unbroken infection chains remained for continued spread.
Yes, probably none of these measures could have eradicated Covid due to lack of resources and institutions to coordinate such policies and individual noncompliance, but they would have been sufficient to locally nip infection chains in the bud and keep infection numbers low enough that society would continue to function even with a hypothetical 100% lethal Covid.
A remark on vaccination: Given the historical record of successes, vaccination is of course the gold standard to deal with a novel infectious disease. However, vaccination may not provide sufficient protection from a given pathogen and may well come too late especially in the current regulatory environment which is in my opinion unlikely to change anytime soon. Additionally vaccination can’t be fully pathogen agnostic, can’t really be prepared in advance and is probably not neglected, thus I will not further discuss it here.
Detection
To do anything about the spread of a novel pathogen obviously early detection is key.
Luckily our early warning systems seem to mostly work well already since novel infections (SARS, MERS, EHEC, problematic flu strains…) have mostly been caught quickly. In the last twenty years, identification has become mostly trivial due to genetic techniques.
The only exception I can see where current methods (detection of conspicuous clinical cases followed by genetic and conventional characterisation of the pathogen responsible) probably are insufficient are pathogens with very long asymptomatic periods that still spread quickly during their asymptomatic period. This may especially become a concern with weaponized pathogens in the future. To detect such threats early, a genomic waste water monitoring program possibly supplemented by indoor air monitoring at big gathering points and arthropod monitoring in population dense areas may be a reasonable albeit currently low priority improvement.
Surveillance
In theory, continuous real time surveillance of the infection status of the whole population trivially solves any pandemic since it allows to quarantine hosts (Note that this may mean different things for different pathogens, e.g. for a pathogen that uses an insect vector that would mean avoiding insect bite which can be nontrivial in itself.) before the pathogen can spread.
This is of course currently technologically impossible but we may be able to come close enough. To do so we would need to develop and hold in reserve a mechanism for quick production, approval and deployment of antigen tests for arbitrary pathogens. Judging from the Covid pandemic current market and regulatory mechanisms take about a year to deploy such tests in limited numbers which would be too slow and too little with a really dangerous pathogen. Outreach, awareness building and political lobbying are probably the prime avenues for EAs to improve the situation here. Paying a pharma company to build the requisite capabilities may also be a possible way to improve at least the production side of things but is probably cost prohibitive for current EA orgs. If monitoring systems as discussed above are in place they may help to provide guidance on where to deploy limited test capabilities.
The political and regulatory side is probably even more challenging. First, public health authorities must be able and willing to refuse sub standard tests and redirect production resources to more sensitive tests. Why this is not already the case is not clear to me as it seems pretty trivial since the data is already there. Second they need to be able to accumulate produced tests and distribute them in affected areas ideally at the beginning of an infection wave. Third, this needs to be supported by a public information campaign on when to use the tests according to the properties of the pathogen (Every morning before leaving the house for the next 20 days for example in the case of Covid.) and to ensure the cooperation of the public. Again, doing these things directly is probably beyond the means of any EA org, so EAs are probably limited to outreach, awareness building and political lobbying. Logistically these things seem to be well within the capabilities of at least most developed nations. My intuition is that the biggest problem here is ensuring sufficiently wide spread cooperation of the public, since personal incentives in most cases are aligned against testing properly / reporting on positive tests / quarantining.
Thus, while this is certainly a promising approach to limit the pandemic spread of a novel pathogen, I think that due to logistical challenges and especially the perverse incentives for the general population this approach is probably less tractable than the hygiene based transmission mode specific approaches I will mainly argue for in this paper.
Hygiene based transmission mode specific approaches to pandemic containment
Historically, apart from vaccination, hygienic interventions, especially those targeted at water- and food borne transmission (water- and sewage treatment, various food safety and -hygiene regulation) and contact transmission of pathogens (hand- and general hygiene in both medical and community settings), were the main drivers to limit infectious disease burden and many philanthropic organizations continue to pursue these goals today.
More recently in countries with high malaria burdens hygienic interventions targeted at insect vector transmission (mainly bed nets, but also habitat destruction and other vector reduction methods) have shown similar success.
Until the Covid pandemic such interventions were virtually absent for air and droplet borne infections. With the rise of Covid, several such techniques were tried (lockdowns, mask use, room air filters and a few others), which unfortunately failed to contain the infection for several reasons discussed above and below.
Therefore, I think the most promising strategy to contain a future society threatening pandemic is to develop and pre-produce hygiene based transmission mode specific transmission prevention “packages” and hold them in reserve as part of a national and international disaster prevention preparations.
What these packages should look like and how the EA community can help develop and deploy them will be discussed below.
Droplet- / Airborne transmission
Droplet or airborne transmission should probably be our main focus since most epidemics at least in the industrialized world during the 20th and 21st century (Spanish flu, SARS, Covid) as well as the most common endemic diseases (flu, common cold) have been airborne and droplet driven, demonstrating that the currently established hygienic practices are insufficient to prevent widespread pathogen transmission via this route. This also makes sense from first principles since breathing is a non-optional activity that creates a permanent exposure to respiratory aerosols created by other people in the immediate vicinity. To compare the situation to for example blood borne diseases, this is comparable to exchanging a couple of µl of blood with every person you meet during the day. Obviously this is not a very safe practice regarding the transmission of pathogens.
While lockdowns are obviously efficient they are similarly obviously unsustainable because they halt most economic and other activity in a given area. Other methods like air filters or UV disinfection are typically only usable in specific situations and can’t reliably prevent infection since exchange of respiratory droplets is still possible depending on the exact situation and airflow patterns.
This leaves wearing a well fitting (FF)P3 equivalent mask (depending on the pathogen with eye protection if unmasked people are around) which - at least from a spherical cow perspective - trivially solves the problem. While of course no mask can provide perfect protection and individual infections remain possible, (FF)P3 masks have viral (and probably also bacterial) filtration efficiencies of 98+% (page 6 Fig. 1 and 2) for aerosols and offer almost perfect protection from larger droplets. Even assuming a linear no threshold model for infection probability (which is unreasonably pessimistic given immune action of the potential host) this should reduce R0 at least 50 fold for one sided mask wearing vs. unmasked interactions and 50² = 2500 fold for both parties wearing masks. In practice, with a more reasonable sigmoid infection probability model, these values should be even higher thus driving the R0 of even the most highly infectious diseases comfortably below 1.
The failure to control Covid with mask use is mostly attributable to 3 problems: usage of low quality masks, low compliance and improper use.
Low quality masks: Instead of purposefully developed masks somewhat fitted to the face of the user or at least standard respiratory protection masks used in industry and trade, “medical grade” FFP2/N95 masks or even surgical masks were used. These masks are, to put it bluntly, an extremely low quality product that that can just barely pass the N95/FFP2 standard (or not even that in case of surgical masks), designed to be donned quickly before a brief patient visit and then discarded with minimal fuss as evidenced by their characteristic ear loops and the missing dedicated seal. While their stated filter efficiency (95%) would be just about sufficient against Covid (20-fold reduction of infection probability in linear model), they offer relatively high breathing resistance which makes them uncomfortable under exertion or with prolonged use and, even more damning, they are very difficult to properly seal to the point that a study found that 65% of health care workers on a dedicated tuberculosis isolation ward had not donned their mask properly. This neatly ties into…
Low compliance: While - at least in Germany - mask mandates were generally observed, during the whole pandemic mask use in non-consumer facing businesses, workshops, offices and especially at private gatherings, even indoors, was spotty at best and nonexistent at worst. I suspect that most other countries fared no better in that regard.
Improper use: Time and time again I have observed people symbolically complying with the mask mandate by wearing their mask over their mouth only while leaving their nose exposed. Others did not bother (or even know) to bend the nose wire of their mask, all the while rendering their mask mostly useless.
To use particle filtering masks to their full potential in a future pandemic we need to address these problems.
Better masks: Looking outside of the commonly used masks different commercial solutions are certainly available but they are far from perfect. Elastomeric masks with separate filters, while offering a superior, easily achievable seal and low breathing resistance are heavy, impair downward vision as well as communication ability and have unfiltered exhaust ports negating most protection for other people. Also they are a very uncommon sight and thus socially awkward. In my opinion, currently the best mask available for this purpose is the M3 Aura FFP3 variant as it combines good filter performance (FFP3) with low breathing resistance, a decent seal (if properly formed), little bulk and weight and little voice muffling. Downsides are the necessity to form the nose wire and a less perfect seal than elastomeric masks offer. There are also many startups who developed masks during the Covid pandemic but they failed to catch on and the two products I evaluated failed on poor seal and high breathing resistance.
I think there is a significant opportunity for philanthropic money to develop a better product that better balances the technical trade-offs (filter efficiency vs. breathing resistance vs weight and bulk vs. money) specifically for pandemic use and offers a better, easy to use seal possibly with multiple models for different facial shapes.
Another aspect that needs more research is what specifically makes masks so aversive for many people to reduce noncompliance as far as possible.
Once an optimized product is developed, ideally sufficient masks to protect each citizen for several months should be stored as part of the national pandemic preparedness plan. While this is of course nothing the EA community can achieve on its own, political lobbying combined with intermediary steps such as marketing these masks as individual pandemic preparedness, for use against Covid, other respiratory infections and allergies are an actionable course towards widespread deployment and may even turn out to be a business case.
Better information: As outlined above, mask use against respiratory infections is a cultural technique that needs to be accepted and learned just as much as hand hygiene or not defecating close to drinking water. Otherwise people have no good grip on when wearing a mask is necessary, that a good seal is necessary and how to achieve it. Thus, along with the masks described above, a public information campaign should be developed covering at least the following points. 1. Friends, family and colleagues present just as much an infection risk as strangers, thus a mask must be worn for all contacts outside of the household 2. Indoor gatherings with lots of talking and / or singing are especially dangerous. 3. A mask with a poor seal is useless. 4. A room that was recently visited by other people may still be dangerous.
Again, ideally this should be implemented into national pandemic preparedness plans but spreading the information via (social) media is a possible backup plan.
(Arthropod) vector transmission
While in recent history arthropod vector transmission caused no epidemic in the developed world, malaria, and to a lesser extent trypanosomiasis, serve as a stark reminder of the dangers of this transmission mode. There is no reason to think that a much more dangerous pathogen with this transmission mode could not arise. For example, one of the more lethal hemorrhagic fevers evolving (or being engineered) to be transmitted by one of the ubiquitous mosquito species could make for a true nightmare scenario where even lockdowns would do little since mosquitos rarely obey human laws.
To counter such a threat stocking insecticide treated bed nets and window screens for every household is an obvious first step though probably not sufficient, since getting mosquito or tick bites outside the house is quite common. To counter this insect protection clothing as well as insecticides and insect / tick repellents should be part of the package as well as an information campaign on how to use these products to maximum effect (Screens must cover all openings, protective clothing must be worn whenever screen protected areas are left, Insecticide / repellent treatment must be renewed every x hours…). Unfortunately I know way less about insect bite prevention than about respiratory protection, so I can’t provide any insight about the pros and cons of specific systems without extensive research.
That being said, to prevent predictable problems with low compliance and improper use a strong emphasis on ease of use, convenience and social acceptability is certainly necessary.
To foster adoption of these measures, marketing of these products for personal preparedness as well as everyday use against mosquito- and tick bites flanked by political lobbying is again probably the most actionable way forward.
Contact and fomite transmission
In many cases, contact and fomite transmission is pretty limited by the almost perfect barrier function of intact skin, the difficulty many microorganisms have to survive on dry surfaces for a long time and already ingrained hygienic practices such as hand washing and limited touching among strangers. I could also not find a single large epidemic that is primarily contact or fomite based though I may have overlooked something.
While those factors limit the importance of this transmission route for the rapid spread of a disease outside of a clinical setting where staff functions as a vector within a dense, vulnerable population, it should not be neglected completely. Contact and fomite transmission logically is always a possible secondary route for airborne pathogens if somehow contact is made with a proper target tissue such as the back of the throat. Note that this contact can happen very indirectly such as hand to eye to throat via tear duct or hand to mouth to throat via saliva if the pathogen manages to stay viable for long enough.
In many cases, this is already prevented by the body's own defense mechanisms such as various defensive enzymes secreted onto the skin surface and into bodily fluids such as tears and saliva or by the commensal microflora. In most other cases hand washing and not touching the face with contaminated hands is sufficient. Unfortunately there are a few pathogens such as dormant body forming bacteria and certain fungal spores which are resistant to most of this.
To counter such threats, it would be useful to stock a variety of different disinfectants (alcohol based, hydrogen peroxide based, and peracetic acid based (as A / B solution)) to cover almost all potential pathogens, of course combined with instructions on how and when to use them.
In case one of the harsher disinfectants is needed or if the pathogen is especially prone to enter via minor wounds, some additional protective equipment may be needed, such as easily disinfectable or single use gloves, coats and shoes / shoe covers.
Interestingly, the public information campaign in this regard worked pretty well in the beginning of the Covid pandemic, since advice to wash and disinfect hands, avoid touching of the face and touching of strangers was given quickly and thoroughly. Why this is, I can only speculate but I suspect the outsized role of this transmission route for nosocomial infections in a clinical setting and the corresponding firm establishment of the corresponding memeplex in the health care community is responsible.
I may be biased by Covid, but I think that all of this indicates that this transmission route is less important for the spread of novel diseases in a community setting and that many low hanging fruits have already been picked, so I would advise against concentrating too much of the initial effort here.
Water- and food borne transmission
While water- and food borne disease continues to be a problem in developing countries, larger outbreaks of diseases with this transmission mode are basically unknown in regions with modern food safety and water treatment practices. This indicates that probably no new practices are necessary to prevent the spread of a novel pathogen with this transmission mode and that the continued deployment of established practices is sufficient.
Should a pathogen somehow evolve a means to evade current practices we can always fall back on cooking or, if all else fails, autoclaving (i.e. pressure cooking) the food (or water) in question to inactivate all potential pathogens. Since most households already possess the means to do so or can trivially acquire them no action seems necessary regarding this transmission mode.
Acknowledgement
Many thanks to Simon Fischer for his encouragement and valuable input.