Think like a virus: pandemic preparedness

Now that we are in the midst of a pandemic, I thought it would be of interest to post a section of a report I wrote in 2009 which included pandemics. Given the date, you should assume the science is dated.

I want to highlight in this posting is the core message of this work: that pandemic preparedness is complex and may require new thinking, indeed, to ensure that in responding to a pandemic, we solve the right problem, hence the title “think like a virus”.

This pandemic will in time pass, but we must be alert to another disease whether within a known viral family such as coronaviruses, or a Disease X with unknown origins and potential to do harm. It is hard to justify preparing for rare events, but there are a number of factors that suggest we should do just that, and not just because the world is now so highly interconnected. The response through policies to guide actions by people, healthcare providers, companies, governments, etc. is absolutely critical to avoid knee-jerk decisions that may actually have unknown and unintended consequences. That’s what can happen if we solve the wrong problem, or fail to develop the right tools and techniques suitable for the pandemic ‘problem’. Being international in nature, exacerbations are possible simply because so many countries, acting in their own self-interest and protection, may taken together create circumstances that may be harmful. At this stage, we don’t know enough and lack the capability of harmonised actions on the public health front.

Coming out this pandemic, the least we need is a thoroughgoing root and branch assessment of what was done by countries, and not done and so on. A detailed and careful review is certainly called for. We also need to spend a lot more time understanding zoonoses, disease transmission people and animals and the determinants that are pandemogenic.

I believe this is important given the politicisation of the pandemic in some countries, (the US is shamefully included here), where data were not well organised (Canada is shamefully included here from work by Macleans https://www.macleans.ca/society/health/this-coronavirus-chart-shows-canada-is-not-flattening-the-curve-at-least-not-yet/) and inconsistent public messaging and confusion. As far as I can tell, there is no pandemic preparedness roadmap of what to do in advance, and perhaps conflicting sources of advice from experts.

The approach taken by China appears to be have worked, but many have said such measures are incompatible with liberal democracies. What we are in fact seeing is that strong measures, that do indeed restrict liberties, may be necessary for our species to survive (a point worth revisiting next time you watch the film Interstellar). Any analysis would also need to address the civil liberties aspects as well as how to harness human behaviour to enable effective pandemic response.

Certainly, we need to catalogue policy responses around the world, and the tools and techniques used, how they were implemented and whether or how they were effective, no not.

We could use more science, perhaps starting with a systematic review of what we already know from behavioural economics and any studies in the context of pandemics, epidemics or just wide spread infectious outbreaks. I found this paper of interest in that respect and we need more of that type of research: Liam Delaney, et al, Reflections on a Virtual Experiment Addressing Human Behavior During Epidemics [http://www.academia.edu/download/45949720/Reflections_on_a_Virtual_Experiment_Addr20160525-3206-b356e2.pdf]. And we need to have better understanding of the broader economic impact, such as N. Geard et al, Modelling the economic impacts of epidemics in developing countries under alternative intervention strategies, [https://www.gtap.agecon.purdue.edu/resources/res_display.asp?RecordID=4980] This pandemic is not a financial crisis, it is a health crisis, but by its nature, it will have real world economic impact. We certainly do not want the financial levers to be pulled in ways that makes the public health levers ineffective. Indeed, what are these levers, what circumstances necessitates their use so they are not used prematurely, in a panic, so to speak.

Responding to a pandemic is not as a result of anticipatory and prior planning, but done in real time, with planning dynamically responding to the evolving situation; the risk of this is the methods chosen vary by the experts consulted and they can be in disagreement. It the immediacy of the need to make decisions that raises risks in the absence of evidence or guidance.

In the US, the Obama administration’s pandemic team, a potential source of consistency at least for them, was shamefully and thoughtlessly abolished by the Trump administration which then discovered it needed this capability. [https://www.huffpost.com/entry/trump-coronavirus-threat_n_5e56e9a7c5b66137fb5d4445?guccounter=1&guce_referrer=aHR0cHM6Ly93d3cuZ29vZ2xlLmNvbS8&guce_referrer_sig=AQAAALHz6Bbsx-kkacxoiEl3fI62Pxg6sa97AlzyOl19IQtwZYyPpV3icN_TlprZ0DIkoCc5sNtCBmsSWmDYmkPdxuPHIA23qJprz8sa4W9WbZ-1Pod2nvEFxbNi8qXW5t7PTT9IW8j_xT3zbCQQQoqJz8xXhRAAzjAho8s9vWbkVYcp] The UK appears to have a better grasp of the science, but it remains to be seen whether they will be able to justify the plan on the premise of building herd immunity, in the absence of a vaccine, as a matter of policy. It is the plurality of responses that in this pandemic offers natural experiments to see what worked and what didn’t — is anyone keeping track?

I wrote elsewhere that the problem of quarantines is that to be effective they have to contain 100% of the infected population, but risk also confining uninfected people who are thereby put at risk. There are important ethical considerations that should also be guiding our actions.

So, please have a read of the following and a think and don’t hesitate to get in touch. Thanks and be safe.

NOTE: This is an edited extract from a report on a number of areas and what research was needed. The report was prepared for the Research Directorate, UK Department of Health, in 2009. The preparation of the material on pandemics coincided with the evolution of a zoonotic viral strain in Mexico. The focus of the work was research to guide policy development.

In order to understand pandemics, it is necessary to think like a virus and create pandemic capabilities that function like an immune system, namely a health ecological focus on all-risks with strong anticipatory capacity and real-time surveillance.

The current paradigm that we are seeing and which I noted in 2009 is to focus on the threat to humans, involving crisis management, which is episodic and weakly anticipatory, with a public health focus on high risk.

The infectious disease production system

Pandemics are a health problem embedded within an infectious disease transmission process of environmental, social, and geographic complexity. Knowledge involved comes from a wide variety of clinical and scientific disciplines, involving a host of different individuals in varying roles. Their effective engagement and integration is generally only required when there is a concern of a pandemic (which is a relatively rare event), and otherwise experts and resources operate largely on their own or with minimal interaction, despite the underlying causal factors existing at all times.

The general model within which we understand pandemics is based on a public health model. Pandemics are linked to the role animals (birds, farm animals, etc.) play in hosting and transmitting disease to humans, and humans to each other.

It may be appropriate to consider whether the prevailing public health paradigm is most appropriate in structuring research, research translation and response capabilities.

Infectious diseases as a whole kill millions of people annually. It is when the deaths are attributed to a specific infectious agent and there is widespread transmission of the disease from animals to humans, and across geographies that we think of these infectious diseases as pandemics. There is no necessity for the infectious agent to be particularly virulent or fatal, only that large numbers of people are affected.

The diagram shows the “infectious disease production process” seen as a system.

Two types of transmission are possible:

• Zoonosis: animal transmission to humans
• Anthropozoonosis: human transmission to animals.

Generally, pandemic models do not incorporate the view that humans and animals occupy a shared ecology. What this means is that public policy prioritises the lives of humans above animals, leading to frequent culling of animals to stem infections.

Some have written of a possible ‘turf war’ between animal health and public health scientists over the framework that should underpin research priorities and public policy objectives.¹ Part of this may arise from a scientific and policy disconnect between the expanding list of agents that can cause pandemics and the policy options, tools and instruments to manage particular pandemics. ²

There is also the view that the economic benefits when seen from a public health perspective may never be cost effective, but would be when seen from an ecological (societal, whole system) perspective.³ Where you stand depends on where you sit. This relativism feeds confusion.

To resolve some of these incompatibilities, the emerging discipline of ‘conservation medicine’endeavours to bring together human and animal health, but this is work in progress. ⁴

As a general point, pandemogenic viruses were only understood after the pandemic occurred.⁵ This suggests that while conceptual models can be useful in explaining the behaviour of pandemics and viruses, there is a need to understand pandemics and viruses in such as way to avoid the increased risk if the causal agent does not fit the model. The advice is that zoonoses must be dealt with at the interface of human and animal health.⁶

The conceptual model underlying influenza A virus is that avian viral strains have the capacity to infect humans only after undergoing genetic reassortment within pigs.⁷ Some suggest there may be merit in exploring other conceptual causal models, such as direct transfection of humans by avian-harboured viral genotypes based on the emergence in 1997 of H5N1 virus. This suggests more generally, that the evolutionary processes underlying influenza A and the models of human infection are linked in ways that are not fully understood.⁸

In addition to the prevailing conceptual model prioritising research on viruses, study of bacterial superinfections which may have actually caused the vast majority of the deaths in pandemics, is not a prominent area of research. This also suggests that the virulence of the virus may not be the key predictor of mortality, but the interaction with bacteria.^{9 10}

The 1918 influenza pandemic remains the world’s worst pandemic, with over 500 million people infected (30% of the global population at the time) and 50 million deaths. It remains important for study as all influenza A pandemics since then have been caused by descendants of the 1918 virus, which have mutated to form H1N1, H2N2 and H3N2 viruses. The origin of H1N1 is not known. Pre-1918 RNA-positive human samples of the virus are needed, along with samples from each of the three waves of the virus.

Influenza pandemics can occur at any time of the year, but eventually settle into annual rhythms. Without genetic drift by the viruses themselves, herd immunity of human populations would lead to the disappearance of the virus at levels where transmission was limited. The timing and spacing of pandemics is not well-understood, but is thought to be related to partial herd immunity which limits the virus’s effects in all but the most favourable circumstances, such as: lower environmental temperatures, human nasal temperatures, optimal humidity, increased indoor crowding, imperfect ventilation and suboptimal airflow.

What makes viruses pandemic? Are there good human models of pathogenicity? What specific host factors account for unique influenza mortality patterns? Are there pandemic precursor viruses? What is the evolutionary path that leads to the pandemic emergence of a virus? What is the nature of viral adaptation to efficient human-to-human spread and how do viruses acquire this capability? All this we do not appear to know.

It does appear that the transfer of viruses between hosts to create new self-sustaining epidemics is rare. ¹¹ But how this works is not known. The key host-specific amino acid mutations required for an avian influenza virus to function in humans are also unknown.¹²

While mortality profiles of pandemics typically are U-shaped, affecting the very young and very old, the 1918 pandemic exhibited a W-shaped pattern with a middle peak amongst young adults (20-40 years of age). This suggests that more than the features of the virus itself are relevant, but include environmental and host-specific factors (e.g. immunopathology or prior virus exposure). Prior exposure is an unproven hypothesis to account for apparent tempering of fatality during the 1918 pandemic of people born prior to 1889.

Disease agents and their hosts

Within the infection production chain, the role of zoonotic transmission is of particular importance as component of disease spread amongst humans. There are 5 vectors for zoonotic transmission:

• bacteria
• fungi
• parasites (protozoa, helminths/worms)
• prions¹³
• viruses.

There is a long list of agents that have caused human suffering by transmission from animals, many of which have pandemic potential if not checked, but not all are likely to cause pandemics. The main focus for pandemics is viruses, bacteria and parasites. The primary viruses of research interest include a wide range such as Influenza A (variants of the 1918 H1N1 virus), malaria, tuberculosis, while public health interest in pandemics tends to focus on influenza and in particular avian and swine variants. This is not to say that concern about TB or malaria etc. are not also important, though.

There is little understanding how viruses become causes of pandemics. While there is surveillance of new and emerging diseases, again there is little understanding how they might become pandemigenic.

Military research on the use of biological agents in combat offers a clue to the pandemic potential, as in the interests of warfare, an efficient and effective infectious agent would be desired, that could be transmitted by air and which lacked an effective antidote/vaccine. Military interest has included: Ebola, anthrax, cholera, brucellosis, Q fever, Psittacosis, Rift Valley fever, and smallpox.^{14 15} It is worth adding that the focus of public health interest and viral research largely focuses on pandemics as natural events, rather than the deliberate or accidental release of viral agents. We know that accidents happen, and we will worry about deliberate actions of a few.

Historically, pandemics have been caused by cholera, influenza, typhus, HIV/AIDS, bubonic plague and smallpox, while regional pandemics have been caused by measles, malaria, Ebola, and others.

Future pandemics are theorised as arising from influenza A, bird flu, H5N1, SARS, but we are not at this time able to know which ones have future pandemic potential – i.e. are easily transmissible.

Table 1: Some diseases transmitted to humans from animals, many of which have pandemic potential and some hosts
Viruses	Parasites	Fungi	Prions	Bacteria
Influenza A, Swine flu, Hanta virus, Avian flu, West Nile virus, Henipavirus (Nipah, Hendra), Monkey Pox, Yellow Fever	Oriental Liverfluke, Echinococcosis, Babesiosis, Malaria	Aspergilloses, Candidoses, Cryptoccocus, Mycotoxins	Transmissible Spongiform Encephalopathy (BSE, Kuru)	Escherichia coli, Bovine TB, Lyme disease, Cholera, Q-fever, Leptospirosis
Some hosts of infective agents
Birds, Rodents, Foxes, Fish, Cats, Dogs, Pigs, Bats, Snails, Reptiles, Insects, Cows, Horses, Chimpanzees, Sheep, Goats Humans

Bats have been identified as of particular concern for four reasons:

• they are evolutionarily ancient
• they can move horizontally and vertically in the environment (they move in 3 dimensions) and therefore come in contact with a wide variety of other creatures
• they are social creatures and therefore can transmit illnesses with other bats easily
• they are hard to study.

Humans share features with bats: they move horizontally and vertically within the global environment, assisted by international travel to enable global movement, and are social creatures.

Predicting pandemics

Prediction of the next pandemic is important, yet the monitoring of appropriate viral outbreaks may not integrate data sufficiently to enable effective prediction of pandemics or indeed any infection. This is partly due to the inherent variability of viruses, and the failure of any model to predict a potential pandemic outbreak. It also requires an understanding of how viruses evolve to understand the likely impact on humans.^{16 17} More generally, there is a lack of appropriate surveillance systems of infectious diseases to produce a useful simulation model with predictive possibilities.

Mathematical models of outbreaks have not been tested against real-world outbreak data and therefore are not simulation models that can be updated with real-time data (like weather forecasting models). Large-scale (herd) studies of vaccination or within families would provide useful data for testing models, and in particular identify whom to prioritise for vaccination when vaccines are scarce. ¹⁸ Certainly we need wider use of surveillance methods.^{19 20}

The public health preparedness paradigm

Current thinking on the spread of detected viral outbreaks of pandemic interest focuses on:

• speed to produce an appropriate vaccine
• availability of first line defences (currently for influenza neuraminidase inhibitor antiviral drugs oseltamivir and zanamivir) and access to (or existence of) stockpiles
• speed before the virus becomes drug-resistant.

Other approaches would include alternatives to vaccine development, stockpiling and prepandemic vaccination. Mathematical models can be helpful in deciding strategy, quantifying outcomes and transmission rates.²¹

The following have been identified as objectives for a public health programme of preparedness:²²⁠

• limiting the circulation of avian and other animal influenza viruses to reduce human exposure
• strengthening and expanding control activities
• improving early warning
• strengthening pandemic preparedness and response capacities.

In addition, the international travel behaviour of humans adds complexity to the origins of infections. Cultural behaviours toward the burial of those who have died from infections can continue to add risk and transmission.

These approaches may turn out to be under-powered and that a more integrated and less event-driven approach would be preferable.

Development of a vaccine against potentially pandemic strains is an essential part of the strategy to control and prevent a pandemic outbreak. The use of existing technologies for influenza vaccine production would be the most straightforward approach, because these technologies are commercially available and licensing would be relatively simple. Approaches currently being tested include subvirion inactivated vaccines and cold-adapted, live attenuated vaccines. Preliminary results have suggested that, for some pandemic antigens, particularly H5, subvirion inactivated vaccines may be poorly immunogenic, for reasons that are not clear. Additional experimental approaches are required to achieve the ability to generate durable protection in humans. ²³

Earlier and better detection of potential zoonosis

We need to detect the transmission of infections from animals to humans better and how these infections evolve that ability. Given the global scope of disease itself, more rapid detection and determination of specific animal hosts and their reservoirs is required. This includes understanding the interspecies movement of infections (e.g. from bats to horses, or birds to pigs) before transmission to humans as well as the transmission of disease from humans to animals, and the specific geography where this is happening before it happens in effect – building anticipatory capacity.

Faster methods of confirming diagnosis are needed in the context of determining whether a specific diagnosis is one which would be a signal indicator of potential for spread to other humans.

Improving decision making

Integration of information to enable better detection of infections and management of pandemics is needed, linking information from across the infection production chain. Given the disparate nature of the institutions and organisations involved when pandemics are anticipated, reliance on emergency, one-time systems may be less appropriate than a system that operates independently of the existence or not of a pandemic. Such real-time information management is realisable using modern internet/mobile technologies, coupled with real-time sensor and surveillance systems that harvest data from medical record databases. This would include data-mining for the purposes of predictive screening of at-risk individuals, linked to cartographic visualisation and epidemiological databases.

While no pandemic has been predicted in advance, evidence of their potential is real; however, mathematical and predictive/anticipatory models need to be compatible with the need for real-time support for decision-makers. Improvements here are warranted, particularly modelling technologies to better understand the balance of public health response, economic dimensions and the underlying unpredictability of pandemics.²⁴

I would add here:

• anticipatory and causal models of pandemigenesis
• emergency response capabilities, organisational effectiveness, skill mix utilisation.

1. 1 Mark Woolhouse and Eleanor Gaunt, “Ecological Origins of Novel Human Pathogens,” Crit Rev Microbiol. 2007;33(4):231-42

1. 2 John A Lednicky and Jonathan O Rayner, “Uncommon respiratory pathogens,” Current Opinion in Pulmonary Medicine 12, no. 3 (2006): 235-239⁠

1. 3 Jakob Zinsstag, et al, “Human Benefits of Animal Interventions for Zoonosis Control”, Emerg Infect Dis.  2007;13(4):527-531

1. 4 See, for instance, www.conservationmedicine.org of which UK organisations are members. Also, “Daszak_et_al_Annals_NYAcSci_CM202004.pdf,” to be found at: www.conservationmedicine.org/papers/Daszak_et_al_Annals_NYAcSci_CM%202004.pdf⁠

1. 5 The Threat of an Avian Influenza Pandemic,” January 27, 2005, /352/4/323.

1. 6 ⁠Frederick A. Murphy, “Emerging zoonoses: The challenge for public health and biodefense,” Preventive Veterinary Medicine 86, no. 3-4 (September 2008): 216-223⁠

1. 7 This appears to have happened in Mexico in late April.

1. 8 Dany Shoham, “Review: Molecular evolution and the feasibility of an avian influenza virus becoming a pandemic strain––a conceptual shift,” Virus Genes 33, no. 2 (October 1, 2006): 127-132⁠

1. 9 Jonathan A. McCullers, “Planning for an Influenza Pandemic: Thinking beyond the Virus,” The Journal of Infectious Diseases 198, no. 7 (October 1, 2008): 945-947⁠

1. 10 David M Morens, Jeffery K Taubenberger, and Anthony S Fauci, “Predominant Role of Bacterial Pneumonia as a Cause of Death in Pandemic Influenza: Implications for Pandemic Influenza Preparedness,” The Journal of Infectious Diseases 198, no. 7 (2008): 962-970⁠⁠

1. 11 Colin R. Parrish and Yoshihiro Kawaoka, “The origins of new pandemic viruses: The Acquisition of New Host Ranges by Canine Parvovirus and Influenza A Viruses,” September 9, 2005

1. 12 David B. Finkelstein et al., “Persistent Host Markers in Pandemic and H5N1 Influenza Viruses,” J. Virol. 81, no. 19 (October 1, 2007): 10292-10299.

1. 13 All prion diseases are untreatable and fatal.

1. 14 Ken Alibek and S. Handelman. Biohazard: The Chilling True Story of the Largest Covert Biological Weapons Program in the World – Told from Inside by the Man Who Ran it. 1999. Delta (1999)

1. 15 A database on agents for biological warfare is available at http://sis.nlm.nih.gov/enviro/biologicalwarfare.html

1. 16 Derek J. Smith, “Predictability and Preparedness in Influenza Control,” Science 312, no. 5772 (April 21, 2006): 392-394.

1. 17 Jeffery K. Taubenberger, David M. Morens, and Anthony S. Fauci, “The Next Influenza Pandemic: Can It Be Predicted?,” JAMA 297, no. 18 (May 9, 2007): 2025-2027.

1. 18 Derek J. Smith, “Predictability and Preparedness in Influenza Control,” Science 312, no. 5772 (April 21, 2006): 392-39.

1. 19 Harold S Monto et al., “Epidemiology of Pandemic Influenza: Use of Surveillance and Modeling for Pandemic Preparedness,” The Journal of Infectious Diseases 194, no. s2 (2006): S92-S97⁠

1. 20 Kumanan Wilson and John S. Brownstein, “Early detection of disease outbreaks using the Internet,” CMAJ 180, no. 8 (April 14, 2009): 829-831

1. 21 Derek J. Smith, “Predictability and Preparedness in Influenza Control,” Science 312, no. 5772 (April 21, 2006): 392-39.

1. 22 WHO “avian influenza and influenza pandemics.pdf,” https://www.who.int/entity/bulletin/volumes/82/4/en/242.pdf.

1. 23 Kristin L Nichol and John J Treanor, “Vaccines for Seasonal and Pandemic Influenza,” The Journal of Infectious Diseases 194, no. s2 (2006): S111-S118⁠

1. 24 Jeroen K. Medema et al., “Modeling pandemic preparedness scenarios: health economic implications of enhanced pandemic vaccine supply,” Virus Research 103, no. 1-2 (July 2004): 9-15⁠