On March 24th, 2020, India went into a 3 weeks-long lockdown to curb the spread of the coronavirus. As the current Indian lockdown enters in its final week, let us introspect from the data on what has changed / will change in our lives and livelihood due to this lockdown. India is home to almost 1.4 Bn people in this world (1/5th of the world population) with a very unique demographic distribution across the urban and rural population. Imposing a nation-wide lockdown is not only challenging from an administrative point-of-view, but also from the economic viability perspective. Few of us still debating in our minds that whether a country like India should ever go into a lockdown state like now. At Eugenie, our mission is to remove bias from decision making via discovering insights using AI from the data. Hence, we felt the need to bring the data-driven-insight point-of-view from the current epidemiology data fig.1. shows state wise statistical data for COVID-19.
To analyze and forecast with an epidemic data like COVID-19 we have introduced newer models in our AI workbench. We have included compartmental models to analyse and forecast with the data. In our model, we have included disease spread dynamics (susceptible, exposed, asymptomatic, waiting for the test result, confirmed, recovered, and deceased) and population dynamics (India population across different age groups and their corresponding vulnerability basis current statistics and the probability of getting infected via community transmission). Details of our modeling fundamentals can be found https://eugenie.ai/paper/covid-19/
Using our model, we have analyzed the data so far and also have forecasted the number for the future. Using these results, we have analyzed the lockdown scenario and its impact.
In the absence of the lockdown, the following grave effects would have transpired. (Based on our forecast before the lockdown on the 24th of March):
-By April 14th, 2020, an exponential increase in the infected cases would have taken place resulting in more than 57,000 infected cases in India.
-According to the report , India has 7,13,986 total government beds, out of which 5-8% are ICU beds (35,699 to 57,119 ICU beds). Even in the best-case scenario where all ICU beds are equipped with ventilators, we have a maximum of ~57k ventilators to cater to a growing number of COVID-19 patients. According to this we would have exhausted our hospital’s ICU capacity by 14th of April, 2020.
-We would have needed to test at least 140K people. This would have put the hospitals in tremendous pressure in procuring the test kits as well as performing extensive testing. Also, treating the infected would have demanded additional resources for the hospitals, which would have resulted in a gigantic health crisis for a country as populous as India.
-Approximately fatal cases would have reached to a number 1108.
As we are standing on the 7th of April, let us look into these numbers. According to India Ministry to health and family welfare website (https://www.mohfw.gov.in/), the statistics as reported as follows.
Eugenie predicted that the number would have reached close to 25K active infected cases, with lockdown we are close to 4K.
Can we safely assume that we have successfully flattened the curve? Are we in a better situation now?
As per Eugenie's, data-based insights it is very evident that we are in a better position than what we would have been without lockdown. However, with the lockdown, we could only delay the process of reaching the maximum havoc.
The lockdown has given a couple of weeks’ time to the government, authorities, and hospitals to prepare for the infrastructure such as ICU beds, ventilators and other necessary equipment to fight against this pandemic. This might result in a comparatively lesser burden on the country's healthcare infrastructure and lesser fatal cases in the future. However, from the data, it is very distinct that the fight is not yet over.
Authorities are working round the clock in devising a more sustainable strategy for executing rapid testing facilities at scale. They are also working towards how to increase the health capacity understanding future demands etc. At the same time to reduce the burden on the economy, the Government is also identifying strategies on how to release the lockdown scenario in a more optimal way to save our lives and livelihood. As a data/AI-first company, we feel that it is our duty to help Organizations and Government in helping with data analysis and using AI to devise operation strategy. In the following blogs, we will talk about how Eugenie’s point of view on these topics.
Assuming that the people will return to the normal lives after the lockdown ends on the 14th of April, in all likelihood, the infections may show a spike, however, the lockdown will certainly leave its imprint in making people more responsible in terms of social distancing and hygiene.
Despite the conflicting views regarding the implementation of the lockdown, the data clearly shows its impact in slowing the spread of the virus in India, which in turn is resulting in saving plentiful human lives and giving much-needed preparatory time to the healthcare system.
Going forward, the situation might turn more challenging after the lockdown ends as there will be a real possibility of community spread of the infection. To avoid the drastic spike in the infections after the lockdown ends, social distancing and discretionary measures relating to safety and health still need to be practiced.
Rigorous testing, contact tracing, clustering of the high-risk regions will be necessary to avoid community transmission.
We’d like to maintain the stance that our analysis is derived purely from the epidemiological and statistical models. However, decision-makers also need to consider the societal, economic, and ethical factors to make the most suitable decisions. The lockdowns tend to be difficult for the countries with the resulting panic and agony. Such decisions are daunting in nature as they present a conflict between economical and public health measures. As stated in the article by the Harvard business review, dealing with COVID-19 involves quick learnings, immediate actions, and responding to the data with swift decision-making and no partial solution.
The mathematical forecasting models cannot capture the human suffering involved; however, data-adaptability combined with agility can surely save many human lives and ease pressure on the already overstretched healthcare systems.
1. Yeolekar, M. E., and S. Mehta. “ICU care in India-status and challenges.” JOURNAL-ASSOCIATION OF PHYSICIANS OF INDIA 56, no. R (2008): 221.