Modelling Related QnA
Below we address questions related to models and modelling approaches for the epidemic spread of COVID-19.
Question: For India, the graph is shown to be climbing. From this, is there any way to predict when the peak will occur and how quickly it will fall ?
Short Answer: Yes, this can be done using models, but it depends on how the physical distancing and other restrictions are implemented after the lockdown, and incorporated within the models. For example, lockdown followed by a strict implementation of testing, tracing and quarantining may lead to a relatively low peak after about a few months.
Yes, it is possible, but the answer depends on the measures that are taken post-lockdown – whether the lockdown is extended or not, what activities are allowed post-lockdown, how much transport is allowed between cities and states, etc. The INDSCI-SIM model developed here, can simulate the effect of these measures independently in different states in India. If we assume that the post-lockdown period will see a return of all activities under a strict quarantine testing protocol where all symptomatic and several asymptomatic infected individuals will be quarantined, then we will see a very gradual increase in the graph. The peak will occur much later in time and will have a much lower value. This is beneficial since it will ensure that the cases are growing slowly and there is not too much stress on the health system at any time. However, it will also mean that the disease will be around for a long time.
Question: If there are only a small number of cases for a few weeks, can the cases surge to a large number ?
Short Answer: Yes, this is possible over a long period of time. Also, it can happen if there are changes in how physical distancing policies are implemented . For example, a strict lockdown together with identifying those infected can reduce cases to a small number. This number can grow again to a large number if the lockdown is lifted without adequate precautions such as testing, tracing and quarantine.
Yes, it is possible, but for it to happen, there has to be some policy change. Consider a case where a strict lockdown and adequate identification of those infected is enforced for a long duration and causes the cases to reduce to a small number. In this case, most of the population has not yet been infected with the disease and do not have immunity. If the lockdown is lifted and the physical distancing measures are relaxed, then the cases can surge to a large number once again since there are a large number of susceptible people. If the lockdown measures are lifted but testing, tracing and quarantining procedures are in place, then the rate of increase of cases will be smaller. The cases may still increase to a large number but the increase will be over a much longer duration giving time to get health systems in place. The only scenario when there will be no surge in cases after lifting the lockdown is when we have reached herd-immunity, where a significant population (about 60-70%) is immune to infection either due to a vaccine or as a consequence of having been infected before.
Question: Are there models predicting resurgence of the Novel Coronavirus (SARS-CoV-2) even if we get over it now?
Short Answer: Yes, most models predict recurrent waves of infection until enough people develop immunity to the virus, either through recovery or via a vaccine. This is because the disease could be present in low rates to surge again, or may get into the system through travellers from infected areas and spread again through the people who are not yet immune.
Reaching zero new cases in a country of India’s size is difficult to achieve in the short run with physical distancing measures alone. Even if that is achieved in some states or the entire country, it is highly likely that the disease will re-enter through travellers. As long as a substantial portion of the population remains susceptible to the virus and as long as there are infected individuals who can spread the infection, models predict that there will be a resurgence if physical distancing measures are relaxed. This has in fact already happened in some countries such as Singapore. It is therefore imperative to keep the number of infections at a manageable level while the lockdown measures are removed. A possible strategy is referred to as periodic or lightswitch lockdown strategy where the lockdown measures are reinforced after a fixed period, or when the number of infections exceeds a certain manageable threshold. This strategy along with quarantine testing can help manage the spread of the epidemic.
Question: What will happen when Indians stranded abroad start returning to India?
Short Answer: Since there is a risk of the returnees carrying infection, they will likely be isolated on arrival and quarantined and/or tested for the infection. Every traveller, whether symptomatic or not, should have to undergo testing or some form of quarantine.
It is possible that those stranded abroad could have acquired the infection while there or in-transit. Once they arrive in India, they would need to be quarantined for an appropriate period of time and tested to ensure they are not infected, before they return to their homes. This will be done to ensure that they do not bring the virus into the community. It is therefore essential that a mechanism for thorough screening of such individuals is in place before they are allowed to return. During the early stages of COVID-19 in India, several states asked asymptomatic foreign travellers to self-quarantine in their homes. The measures adopted the next time will have to be a lot more stringent. All travellers including asymptomatic ones will have to undergo testing/quarantine to ensure that the disease does not come into the community again.
Question: If a reasonably good drug to treat severe Covid-19 patients is found tomorrow, by when will the world return “to normal”?
Short Answer: Even if a good drug to treat severe COVID-19 patients is developed, the infections will grow exponentially. The world can return to normal only when countries have adequate hospital facilities to treat the surge of severe cases post normalcy, and that they also procure sufficient stocks of drugs.
This depends on the hospital facilities to handle severe cases, the amount of drugs that can be produced, and how many people are able to receive it. If there are sufficient stocks of drugs in India, even then the physical distancing measures cannot be removed since our public health system will not be able to cope with testing as well as hospitalisation of severe cases. and things will not return to normal rapidly. The same holds for other countries as well. However, there are additional complications due to logistics of drug production, economic inequalities across nations and within a nation, because of which not all governments would be able to afford the drug for each and every infected citizen. Therefore, even if a drug is discovered, it will take some time to build up adequate quantities of drugs in each region and we will need to follow physical distancing measures to avoid rapid exponential growth of hospitalisations.
Question: If a COVID-19 vaccine is found tomorrow, by when will the world return “to normal”?
Short Answer: The world will return to normal when countries are confident that they have administered the vaccine to a substantial portion of the population, about 60%-70%, so that herd immunity is built in the community.
According to model predictions, if a vaccine for COVID-19 is discovered, it would be enough to vaccinate 60-70% of the population to achieve herd-immunity. So ideally, if the vaccine is administered to around 60%-70% of the population, then the world can return to normal in less than a month (considering the 14 days incubation plus possibility of a second infection). However, due to economic inequalities across nations and within a nation as well as logistics of production and supplies of the vaccine, not all governments/individuals may obtain the vaccine immediately after a vaccine is discovered. Moreover they will need to prioritise vulnerable sections of society with available resources Therefore, even if a vaccine is discovered, it would take several months for the world to return to normal.
Question: How many new cases per day is socially acceptable as risk?
Short Answer: The number of cases per day has to be sufficiently small, so that contact tracing and quarantine measures can control the spread of the virus. For a city like Chennai with 6 million population, this number can be as low as 10 in the absence of total lockdown.
If the number of new cases each day increases, it means that the population is at risk. However, if the number is sufficiently small, so that it is possible to identify all infected individuals and trace their contacts, then, the disease can be controlled through quarantine measures. The value of this number depends on the social contact network and the efficiency of contact tracing – which in turns depends on capacity of the public health sector. Let us assume that an individual, on average, comes in physical contact with 5 new people every day. If this person is infected, it is likely that he/she has been potentially exposing people to virus for about 5 days, then the total number of primary and secondary contacts over 5 days is 6X6X6X6X6, or 7776 people. Thus, if 10 people are infected, then tracing their contacts for 5 days will involve 77760 people. For a city like Chennai with 6 million population, this is about 1% of the population and is a crude estimate for the maximum number of people who can be tracked in a day, with sufficient available manpower.
Question: What is meant by “active cases”? Why is this number compared to the number of deaths and the number of recoveries ?
Short Answer: Active cases refer to the individuals who are infected and either undergoing treatment or under quarantine. The number of active cases is simply the total number of infections less the number of recoveries and the number of deaths.
An individual who is infected by the virus eventually recovers or succumbs to the disease. Until this happens, the individual is infectious and is said to be an “active case”. Thus the total number of infected people is simply the sum of the number of active cases, the number of recoveries and the number of deaths. The number of deaths due to the infection compared to the number of infected people gives the fatality of the disease. Note that often the official numbers capture only a fraction of the infected and therefore the active cases within a zone/city/state.
Question: Does one model work for the whole world; don’t we need specific models for each country, each region in a country ?
Short Answer: Models that account for epidemiology of the disease can capture broad trends across all countries. However, it is essential to specialize the models for each country and region in the country. Different regions have different age structures in their population and different social contact networks. Further, the travel and migration trends vary across regions and affect the spread of the disease.
Mathematical models have several parameters — clinical parameters and demographic parameters. While the clinical parameters are likely to be similar across countries, the latter will have to be specific to different countries and regions. For example, India has a much younger population than western Europe. Europe has a more prominent cafe culture than India. In India, it is often common to have joint families where children live with their parents and grandparents. Thus, a model specific to India will have different parameters than those specific to Europe. The INDSCI-SIM model developed in the group here, is an India-specific model which can further be specialized to each state. In this model, the migration trends in each state are also used as input to calculate the rate at which infections spread.
Question: If the lockdown is lifted on May 4, but people continue to wear masks and maintain physical distance, will we flatten the curve ?
Short Answer: This can happen, but is highly dependent on an efficient contact tracing of emerging cases , testing and quarantine policy after the lockdown. It is more likely that the cases will increase if the lockdown is lifted, and a series of more lockdowns might be required to flatten the curve.
Even if the lockdown is lifted and certain precautions such as wearing masks and maintaining physical distance are observed, it will still require a lot more effort to flatten the curve. A very stringent testing, tracing and quarantine policy will have to be in place. Even so, it might not be sufficient since some physical contacts are inevitable in daily life and many contacts may not be traced. The INDSCI-SIM model predicts that periodic and lightswitch lockdowns, where the lockdown is again enforced when the number of cases increases beyond a threshold along with quarantine measures can flatten the curve.