EBOLA HAS MUCH TO TEACH US ABOUT THE MANAGEMENT OF COVID-19

The author is an independent disaster management consultant and was the UK Government’s Humanitarian Health Adviser in Kinshasa during the Ebola epidemics of 2018 in the Democratic Republic of Congo. He was also special adviser to the Minister of Health in Sierra Leone during the Cholera epidemic of 2012 and Ebola epidemic of 2014.

It’s never too late to learn from a virus, especially a new zoonotic RNA virus to which we are all susceptible. These are some of the relevant non-clinical lessons learned from the various Ebola (EVD) outbreaks of 2014 in Sierra Leone and the Democratic Republic of Congo (DRC) in 2018.

Before we start, though, just a few words to put things in perspective. It’s not over until it’s over. At the moment, there are no vaccines to prevent us contracting C-19 and no therapeutic drugs to cure us once we have. Until such time there is a vaccine, the population reaches herd immunity, or the bug mutates to become benign, we will eventually all contract the disease; it’s just a matter of when[2]. Since lockdown cannot be sustained forever, we should expect at least three waves of infection between now and the end of 2021. And, as I was once told by an exhausted Congolese doctor who had not been paid for six months, “You can recover from being broke; the only thing you can’t recover from is being dead.”

The list and the recommendations that follow is not exhaustive. It also focuses on health systems management rather than clinical aspects

INTERNATIONAL HEALTH: Treat epidemics as natural disasters from the outset.

Covid-19 is a man-made ‘natural’ disaster, not just a health crisis. It needs to be treated as such, with the full range of inter-sectoral and inter-disciplinary skillsets and coordination processes brought to bear. The enemy is not so much the pathogen, but fear. This makes anthropology as important as medicine and sociology as important as epidemiology. One of the main reasons the West Africa Ebola epidemic got out of control was that for too long it was seen as a ‘health emergency’ “beyond the wit of the lay person and which could therefore only be dealt with by doctors[3].” This medico-centric approach was then exacerbated by a ten-week delay in response. The UK Government’s Department for International Development (DFID) was warned of what was likely to occur on 5^th July 2014 yet did not respond at the scale required until mid-September, by which time the outbreak had spiralled out of control.

SCIENCE: Epidemiology is only part of the story. Every other discipline must be represented, including Behavioural and Health Economics.

Science can be described as ‘The testing of truth until a better truth comes along.’ It is not exact in other words. It can inform but not decide. Medicine, Public Health and Epidemiology are ‘Health Sciences’ but are completely different disciplines.

Epidemiologists are data analysts, not disaster managers and are not used to making rapid decisions based on uncertain and flaky data. They tend to spend hours trying to understand minute changes in the data in forensic detail in case this unearths some previously unnoticed insight into how the bug is behaving. This is understandable but slows down the response. Chasing the dragon of exactitude is not helpful during a disaster response; it diverts attention from getting the job done and fosters a mindset where perfection gets in the way of good enough.

Medical doctors, even those with Masters’ degrees in public health and business administration, do not automatically make good disaster managers. This is because their education and work experience is not only extremely hierarchical, but based on saving individual lives.

Public health is about managing public goods and taking decisions that do least harm to the most people in situations where resources are inadequate.

The ‘epi-curve’ is not a definitive tally of morbidity and mortality, but represents the best approximation to truth. The actual curve is normally layered over the ‘most likely’ predicted curve to show whether the trajectory of the disease is as expected.

RECOVERY STRATEGY: Whack-a-Mole

Once an infectious disease becomes endemic – as C-19 seems to be in the process of doing – only a combination of vaccination, therapy and isolation will keep it under control[4]. In the absence of the first two control measures, the only realistic option left (apart from allowing herd immunity) is to knock it on the head by isolating any localised outbreak as soon as it is discovered. This means not letting anyone out of the infected area unless tested as negative and not letting anyone in unless immune for twice the infectious period i.e 28 days. This is only doable in rural settings, but even then will only be successful with the consent of the community concerned. Supplies have to be delivered to the periphery and then collected by the community. Roadblocks don’t work.

If a vaccine becomes available, there won’t be enough for a mass vaccination programme for some time. In such circumstances, a ‘ring vaccination’ strategy will have to be adopted. The first ring includes all known and likely contacts of the index case, with the second ring being their contacts. This was used successfully during the Ebola outbreak in North Kivu, DRC in late 2018 as part of clinical vaccine trials.

LEARN LESSONS: Conduct a mid-term and after-action review.

John Meynard Keynes always maintained that “When the facts change, I change my mind.” For all large-scale ‘natural’ disasters, the UN is obliged to conduct a 30-Day Review, a Mid-Term Review, and an After Action Review. Guidelines for the facilitation of these already exist. Governments should not be afraid to use them and advertise that they are being planned.

DISEASE SURVEILLANCE: Test and Trace.

You cannot manage what you cannot measure. This business-school mantra is especially true of disease outbreaks. This means laboratory antigen testing – preferably with rapid diagnostic tests of high sensitivity and specificity – and contact tracing. Until such time as rapid diagnostic tests are available, sentinel, syndrome surveillance will suffice. This means observing symptoms within a sample of the population rather than universal coverage.

As with other coordination efforts, technical advice should be provided from the centre but operational management should be carried out by local public health authorities. Contact tracing teams in particular can be recruited, trained and deployed within one week if volunteers have already been identified.

LEADERSHIP: Appoint a single authority figure with clear limits of authority and responsibility. In a national emergency this is likely to be the President or Prime Minister (or designate).

Politicians lead; Scientists inform; Advisers advise. Politicians must make the decisions, preferably cross-party. If these are informed by the evidence, people will go along with those decisions however painful they might be. Authority can be delegated but responsibility cannot. For purposes of public perception if not actual day-to-day management, it is critical that there is only one focal point – a Czar, if you like – through whom everything flows. This is not just for operational oversight but for accountability. Everything communicated by this individual must be scripted; one word out of place can be catastrophic. Trust is born of honesty and transparency. It is easily squandered if legitimate questions go unanswered or data is selectively applied.

INFORMATION MANAGEMENT: Share KPI updates.

It’s all about choosing the correct Key Performance Indicators (KPIs) and tracking them over time. It’s not just a matter of applying Key Health Performance Indicators as per WHO guidelines. To be included, they must be useful for managing individual and collective aspects of the response. Datasets for each pillar of response can be complicated but summaries can be displayed graphically and shared with the public via mass or social media once the senior management team has drawn its conclusions from what the data is telling them. Some graphic displays – or dashboards, as they are sometimes called – are more useful than others. Their design is critical. The dashboards put together by the World Health Organisation, the US Centres for Disease Control and the Congolese government for the Ebola responses in 2018 are the gold standard. There is no need to re-invent the wheel.

CRITICAL THINKING: It’s about trends and patterns, not just numbers and statistics.

It’s not just about the data; it’s about the analysis of the data and the assumptions behind the interpretation of this analysis. It’s not just about epidemiological curves and rates and ratios, either; it’s about spotting patterns and trends and taking informed decisions quickly based on these patterns and trends. Do not let the perfect become the enemy of the good.

Evidence-based decision making is not always driven by data. Yes, you cannot manage what you cannot measure, but anecdote, experience and the vagaries of behavioural economics is also needed if the response is to be properly framed.

STRATEGIC COMMUNICATIONS: Be totally transparent. Apply the 4-C’s … inform the public carefully, concisely, consistenty and coherently without patronising or dissembling.  

Communicate with communities. Use the Red Cross for this as they alone have the ‘auxiliary status’. Leaving issues to be teased out by the media is reactive and automatically puts government in a defensive posture, especially when the media turns from being sceptical to being cynical. In a crisis, the fastest way for political leaders to lose people’s trust is to dissemble and obfuscate about what is going on while over-promising and under-delivering. The fastest way to gain people’s trust is to level with them; to tell them exactly what is going on, what assumptions are being made, what the current challenges are, and what is being done to resolve them. This sort of transparency does not come easily to politicians who think they cannot afford to be seen as ignorant or indecisive, either in cabinet or in front of their electorate. This is understandable during an election but is the wrong approach in a crisis. As any ministerial spin-doctor will tell you, communicating essential messages in this way needs to be careful, concise, consistent and coherent.

People quickly become numb to large numbers. Most absolute numbers such as xxx died yesterday actually tell the public very little. What they need is comparative data which tells them whether the response is on track or not. Averages can also be misleading, with means, medians and modes often giving different skews to the analysis.

The public must be educated on the science involved, including basic statistics and epidemiology. Without this basic knowledge it becomes difficult to maintain trust within a ‘blame culture’ fuelled by a 24/7 media panting for the next scoop and more difficult to manage public expectations. This needs to be made available through, for example, The Open University and mainstreamed as a key ‘stratcomms’ pillar.  

The law of diminishing returns kicks in once political mantras are repeated too often and ends up being socially and politically counter-productive. ‘Stay Home’ and ‘Do the Five’ are public health messages and are about communicating risk; ‘Save the NHS’ is a political message.

Key messages must be memorable and have to be repeated often. However, contrary to WHO advice, underpinning them with more complicated explanation does not dilute the message; it makes them more credible. The better educated the audience, the more complicated the messages can be.

Conveying the impression of being forward leaning and ahead of the curve means proactively addressing issues before they are raised by others.

Communications technology must be state of the art. “Next slide please” is so 1980’s and gives the impression of backwardness.

People need to know there is light at the end of the tunnel. They need to know at the earliest opportunity what an exit strategy might look like; what the options are; and what criteria need to be met. Their expectations need to be managed, and managed proactively.

VACCINES AND THERAPEUTICS: Discuss the ethics of profiting from public goods.

Pharmaceutical companies are not charities. Vaccine and drug development by the private sector is a highly competitive business, with both reputation and profits at risk. Commercial companies only collaborate when government’s guarantee losses in research and development. Should a viable vaccine be developed, it should be assumed that vaccination programmes will be conducted ‘at cost’ in developing countries, but only for a limited period. Generics will be resisted during the patent period.

While nowadays the genome sequence of a new zoonotic pathogen is quickly shared, the race is on to develop a vaccine. When it comes to efficacy trials, pharmaceutical companies jealously guard their findings in their bid to be the first to win a global contract. It is not unknown for data to be withheld or even faked during this phase in order to mislead the competition.

COMMUNITY ENGAGEMENT: Integrate community groups and national NGOs into the response.

In principle, outbreak control measures should be centrally coordinated but locally managed. This means making use of local knowledge and resources which take into account cultural and linguistic variables. Monocultural approaches to public-health messaging lead to misunderstandings and, eventually, disengagement.

HEALTH ECONOMICS: Justify the financial cost in terms of the cost to society of doing nothing.

It’s not just about the financial per-capita cost of the intervention but how this stacks up against the cost to society of doing nothing. Costs are not just financial, but social and political. Cost-Benefit Analysis (CBA) necessarily uses some form of financial metric to measure the cost per avoidable death averted and quality adjusted life years (QALYs) gained.

Such analysis has to factor in the considerable secondary (indirect) impacts in terms of reduced economic output, increased poverty, reduced care-seeking for other ailments, declines in vaccination coverage, and mental health impacts. In DRC, there were ten times more excess deaths from cholera, measles and malaria than there were from Ebola in the same place over the same time period. Something similar is likely to be the same for C-19. Ultimately, the key metric becomes ‘cost per net excess death or disability averted’[5].

BEHAVIOURAL ECONOMICS: Nudge more.

People are herd animals; they will do what the herd does. The trouble is, humans are exceptionally bad at assessing risk, especially risks they cannot see, don’t understand, or are not immediate. They need help in assessing what levels of residual risk are acceptable

Communities don’t wait for official guidance once they know the routes of transmission and how to interrupt them. Buckets of chlorinated water – sometimes with soap – or hand-sanitiser dispensers spring up in front of stores, on street corners and people’s homes. Handshaking gives way overnight to fist bumping or elbow knocking. Cash is shunned. There will come a moment when hand hygiene and social distancing behaviours will become normalised i.e are adopted widely enough to the point of acceptable risk where R-o remains sustainably below one. Such behaviour may have to be nudged i.e incentivised.

LOGISTICS: Assume re-use.

There is never enough Personal Protective Equipment (PPE) for health and care providers wherever they are in the care pathway, from intensive care nurses to mortuary technicians. Some of it inevitably ends up being re-used. This means supplying extra detergent for washing gowns and aprons. Health and Care Workers (HCW) use whatever barrier methods they can, including ski goggles and washing-up gloves.

There are never enough body bags for burial teams, either. Disposing of the bodies of victims of infectious disease with dignity and safety is psychologically demanding and physically dangerous work, especially where mortuaries are overwhelmed. Until arrangements can be made for burial or cremation, sealable body bags may be required. These must have windows so next of kin can identify the body.

RECOMMENDATIONS

1. Build trust by being brutally frank with the public about what the challenges are and how they are being addressed. It is patronising to dumb down complicated messages and repeat slogans.  Use the ‘key performance indicator’ dashboard to illustrate the assumptions and evidence behind each point being made.
2. Engage in a public education campaign on the basics of disaster epidemiology. Involve academics and the media.
3. Focus on trends, not absolute numbers.
4. Consider a weekly ‘Any Questions’-style debate with a live audience in conjunction with a daily briefing to which selected media may be invited.
5. Appoint an independent ‘international health adviser’ to work alongside the Chief Medical Officer within SAGE.
6. Plan for and publicise a Mid-Term and After-Action Review using existing UN (WHO) guidelines.
7. Embed the Red Cross in community mobilisation and communications efforts.
8. Incentivise (nudge) hand hygiene and social distancing behaviour. Make it socially unacceptable not to wash hands at every opportunity or wear a mask on public transport.
9. Incentivise self-isolation by providing some form of universal basic income per family member for the extent of the isolation.
10. Provide hazard pay along military lines to front-line health and care workers.
11. Ensure contact tracing teams are locally recruited and highly visible.
12. Invite opposition politicians to join COBR discussions.

James Shepherd-Barron, 24 April 2020

---

[1] Two times the incubation period (i.e 42 days) without a laboratory-confirmed case

[2] How badly we get the disease is apparently down to a number of risk factors, including the infective dose, the viral load, age, gender, and extent of any pre-existing conditions. This is more or less the same for any communicable disease.

[3] Absolute Disasters by James Shepherd-Barron, KissyFish Books 2015

[4] BY keeping Ro under 1

[5] See ‘Calculus of Calamity’, a book about disaster epidemiology by the author of this article. Not yet published. Available on request.