Responding to the Pandemic Modeling

Ratna Megawangi’s article, “Coronavirus and Panic Statistics” (Kompas, 20/6/2020), is interesting to discuss. The article comments on the prediction of the spread of the coronavirus by mathematical modeling, as widely presented by experts.


In particular, Ratna reviewed predictions made by Dr. Neil Ferguson of Imperial College London, an epidemiologist who from the start was keen to suggest that a lockdown be implemented in England. With the mathematical modeling he made, in mid-March Ferguson predicted the death toll due to the coronavirus in the UK would reach 510,000 people if no lockdown was tightly applied. This prediction is what makes many parties panic. According to Ratna, Prime Minister Boris Johnson, who initially believed in the herd immunity scenario, eventually shifted to trusting Ferguson to apply the regional quarantine alias total lockdown. And, the fact was that, as of Wednesday (24/6/2020), Worldometer data recorded the number of positive cases in the UK at 306,210, of whom 42,927 had died. It is very low compared to the 510,000 infections according to Ferguson’s prediction if the lockdown was not implemented.


This case is considered a prominent example of predictions with the modeling of many experts. It turns out that many are inaccurate and are judged to only cast panic.


In Indonesia, our experts also made models. In mid-March, almost simultaneously with the release of Ferguson’s prediction, the Center for Mathematical Modeling and Simulation of the Bandung Institute of Technology (ITB) also made a prediction. Dr. Nuning Nuraini and her team used Richard’s Curve modeling, because this model successfully predicted the beginning, peak and end of the SARS outbreak in Hong Kong in 2003.


Nuning from the start frankly explained the limitations of the modeling. “It should be noted, this is the result of modeling which I feel is quite simple, and it does not include other factors with high complexity,” she said as quoted by the ITB news website. According to this modeling, it is estimated that Indonesia will experience a peak in the number of Covid-19 daily cases in mid-April with the largest number of new daily cases at around 600 cases.


Data on 20 June 2020, the addition of new cases reached 1,226 people raising the total to 45,029 cases.


Let’s look at the note-taking on the field. Data in mid-April (16/4/2020) showed that there was an addition of 380 new cases, bringing the total to 5,516 cases. The death toll increased by 27 to a total of 496 people. It turned out, by mid-April the pandemic had not yet reached its peak. The peak occurred in June, and nationally it has not subsided. Data on 20 June 2020, the addition of new cases reached 1,226 people raising the total to 45,029 cases. The number of fatalities that day increased by 56 people to 2,429 people. It means the prediction of Nuning and her team was lower than the real numbers.


In the above context, Nuning’s prediction did not spread panic as Ratna blamed on the Ferguson case. Ferguson’s prediction was indeed incredibly wrong, where his number of cases jumped to 1,200 percent compared to the facts (official records). However, there may be reasons, that the official record number was low, because it followed the lockdown recommendation.


Another modeling was also made by the University of Indonesia School of Public Health (FKM UI) team, comprising Dr. Pandu Riono, Iwan Ariawan, Muhammad N Farid, and Fafizah Jusril. In the second week of April, the team made a prediction that in early July there would be 250,000 positive patients in Jabodetabek (Greater Jakarta) if mudik (exodus during Idul Fitri) was not prohibited. However, if mudik was prohibited or prevented, that number could be reduced by around 200,000 people, to only 50,000 people entering the hospital. The peak of the case was predicted to occur in May, and if efforts were made to prevent and limit movement properly, starting in June there would be a decrease in cases. The number was quite daunting.


The team explained the effect of the intervention on the addition of daily cases. If the government implements low intervention, for example without imposing limitation on social movements, the positive case would jump to 1.5 million people. However, if the government implements a high-scale and strict intervention, the addition of positive cases can be prevented from spiking. The modeling of the FKM UI team also did not produce predictions that spread panic. The results were even softer than official government records. Indeed, in June the addition of the number of cases was still very high, but the increase was not as high as in the modeling. In fact, prevention and restriction efforts have not been maximized as predicted by the FKM UI.


There are still many teams from research institutions or universities that are moved to compile mathematical modeling with the building of theories and diverse methods, to make predictions of the spread of the coronavirus. The method differs with the results which are also not exact, but at least there is no contradiction. All is done in good faith, namely to help predict the duration and massive spread of the coronavirus. The calculation was done with various variables that affect, especially government policy variables and citizens’ behavior in implementing health protocols.


A form of scientific care


Compiling the mathematical modeling to predict the spread of the coronavirus is clearly a form of application of the scientific method. This is the same as the political survey method, which in the past 15 years has been very popular and has become an important instrument for strategies to win the regional elections, general elections and presidential elections. There are theories and methods for building assumptions, taking samples, and limiting them with complex variables.


There are simulations and stages to test. As long as the researchers are obedient and strict in using their methods and procedures, whatever the results are part of the application of scientific principles. As is always the case with the scientific method, space is always available for constant criticism, improvement and revision. What we think is true today is not necessarily true in the future. Let alone, the situation also continues to change and develop.


There is always a process of searching, researching, testing, applying in massive scale, monitoring, evaluating, and revising.


Two months ago, the use of hydroxychloroquine sulfate was believed to be effective as a drug for sufferers of Covid-19. Now, the FDA (Food and Drug Administration) of the United States, which is very powerful, does not recommend the use of this drug. There is always a process of searching, researching, testing, applying in massive scale, monitoring, evaluating, and revising. And so on so that science will be more accurate and advanced, including in this case the use of mathematical modeling to develop predictions of the spread of the coronavirus.


The assumption that predictions resulting from the modeling spread panic is only one perception. Ferguson’s predictions can be considered anomalous, because the conclusions are very extreme and jump far beyond the number produced by other modeling compilers. If this is the case, that is just the case with Ferguson. Experts can question the methods, data and procedures used by Ferguson. It can even track the track record and credibility of the person in question. If the predictions widely miss, of course they have lost in the reputation bet.


Such cases have also occurred in several political survey institutions in Indonesia. The institutions announced the results of a survey that sparked controversy, because it was very much different from the survey results of most similar institutions, as well as with factual results. The methodology and procedures were questioned by many parties, so that finally the association of political survey institutions acted. Their reputation is at stake.


However the modeling still helps. So are political trends surveys or other surveys. There is an approximate picture of the direction of a phenomenon. The more that carries it out, the better. Moreover, renowned institutions or universities also do it. That, some are panicking because of reading the modeling, maybe. It depends on perception and critical power. However, in the case of Indonesia, panic does not occur in response to the modeling by many institutions. The government even welcomes positively by inviting the modeling teams to give presentations at agencies dealing with Covid-19, such as the Coordinating Human Development and Culture Ministry, the Covid-19 Task Force, and in local governments.


This mathematical modeling can be one of the scientific materials for decision making with moderation and adjustment. We remember, the Jakarta Government’s policy to impose restrictions on social movements was directly bombarded by the slurs from netizens and politicians. The governor was accused of panicking. In reality the health experts actually recommended the need for social restrictions with strict supervision. For the medical community, social restrictions are an important step in preventing transmission of the outbreak.


Thus, the central and regional governments begin gradually opening restrictions for factories, offices and trade centers little by little, still accompanied by health protocols.


Even so, the emergence of criticism over restrictions on mobility is also important. What is loud is criticism from the economic sector. So, to prevent the economy from experiencing prolonged hibernation, the voice of the business community is accommodated. However the economy works like pedaling a bicycle. If the pedaling stops for too long, it can collapse. Thus, the central and regional governments begin gradually opening restrictions for factories, offices and trade centers little by little, still accompanied by health protocols.


In the midst of the pandemic which continues to haunt our minds, what is needed is synergy from all sectors. It should be appreciated that there are still experts who want to make various mathematical modeling to make predictions. Its large function is as a warning bell for the government and citizens to get ready. It is like the scream of a train whistle coming from afar. The warning is very useful for our mutual safety. Along with other experts who analyze and sort the SARS-CoV-2 alias Covid-19 genome, they struggle to find antivirals, or concoct prospective drugs. Experiment after experiment, including mistakes and failures in its process, remains useful. This is the virtue of the get-together to deal with the super virus. No need to panic.


Djoko Santoso, Professor at the Medical School of Airlangga University