By Joseph Hanlon
"Rapid, decisive and collective action could save millions of lives," says the Imperial College (London) Covid-19 Response Team, in a report published yesterday. If nothing is done, 7 billion people will be infected and 41 million will die, according to their model. If governments impose tight restrictions on social contact, this falls dramatically to 2 billion infected and 2 million dead.
The team headed by Professor Neil Ferguson conducts sophisticated modelling on how the Covid-19 pandemic will progress. The projections in its 16 March report on Britain were so shocking that they forced Prime Minister Boris Johnson to totally reverse UK policy and move to near lockdown. (It was not soon enough. Johnson and Health Secretary Matt Hancock said today they were both infected - probably from a contact after Ferguson published his earlier report.)
Ferguson's report yesterday (https://bit.ly/Covid-Imperial-12) builds a model based on reducing social contact. "Social contact" is defined as physical contact, such as shaking hands, or non-physical contact and exchanging more than three words. The most common recommendation for reducing contact is keeping more than 2 metres away from other people. Strategies that reduce social contact by 40% would halve the global deaths, but 20 million would die and the health systems of all countries in the world would be overwhelmed. To get the numbers down means supressing transmission and to do that social contact must be cut by 75% and done soon.
Based on data from Europe, each person with Covid-19 infects three other people (often before they have symptoms) and the daily death rate doubles every three days, in what is know as exponential growth. The only possibility is to stop people with the disease infecting others. That means quickly reducing social contacts - and testing, isolating those with the disease, and tracing contacts.
Implications for Mozambique
Ferguson and his team warn that they may be underestimating the impact on low income countries which have poor health services "which will be quickly overwhelmed". They admit that for poor countries the wider social and economic costs of suppression of transmission will be very high. But the only way to avert health system failure and massive deaths is "rapid, decisive and collection action."
The new report lists model results for each country, including Mozambique. It takes into account World Bank and other data on income, the health service, age profiles, family size, etc. It also recognises that social contact varies between countries, and so for Mozambique it uses data from a recent Zimbabwe study (https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0170459). The model looks at the disease over the next 250 days. Computer models are not precise predictions and always contain assumptions; in this model they are based on experience so far from China and Europe, so might be different for Africa.
The model is then tested with various assumptions. The first is based on how quickly action is taken. They model taking actions when the death rate is 2 people per million population per week (63 per week in Mozambique) and 16 deaths per million per week (500 per week in Mozambique) and show acting earlier makes a huge difference. They also look at different rates of infection.
The numbers are frightening. If government does nothing, between 61,000 and 65,000 Mozambicans will die; 94% of the population will get the disease and at the peak 190,000 will need hospital beds of which 32,000 will need critical care, far beyond the capacity of the health services.
Reducing social contacts by 45% cuts the infection rate to 63% of the population, but the deaths only fall to 48,000.
If the government moves to suppression to cut social contacts by 75% (isolation or lockdown for most families) but waits too long (until there are 500 deaths per week) then 37% of the population will be infected, 30,000 Mozambicans will die and peak hospital demand will be 84,000.
A successful response only comes if government moves fast for suppression, acting when there are only 63 death per week. The model suggests that only 11% of the population would be infected, there would be 9000 deaths (compared to over 60,000 if nothing is done) and peak hospital demand would be 19,000 patients, of whom 3500 would need critical care.
South Africa, with 927 cases and 2 deaths, imposed a three week lockdown starting at midnight last night. Everyone must stay home. Shopping is restricted to food and other essentials (sale of alcohol and cigarettes prohibited) and people are not even allow out for exercise or to walk the dog. At least 23,000 Mozambicans including many mineworkers have returned to Mozambique from South Africa in the past few days to avoid the lockdown, and some are likely to be infected and pass on the disease to their friends and family.
A South Africa style rapid lockdown would create huge problems for Mozambique. Most people are poor and have little money, and do not even have soap and running water. But Frelimo has a history of organising such a massive operation. Both the 1980 money change (to the metical) and the urban food ration system in the early 1980s in response to drought and war were done with huge collaboration of the population - and with no corruption and extra payments. If senior officials were seen to be sharing in the burden and no one gained special benefit, would it be possible to get basic food to everyone, get the ill to hospital, and do essential contact tracing? It could prevent more than 50,000 deaths.