Can symptoms from the #CoronaTracker predict the daily increase in COVID 19 cases in India?

Team C-Voter’s #CoronaTracker asked respondents to report ILI (Influenza Like Illness) symptoms in their household (including themselves) and neighbourhood. In today’s blog we breakdown trends in self-reported ILI symptoms and whether they can be used to predict overall number of Coronavirus cases in India.

Read more about ILI symptoms and their classification in our blog post from yesterday.

Trust in police at an all-time high

The C-Voter survey carried out between 23–30th April revealed that while 69.9% of respondents said they have a lot of trust in the police 20.6% said they have some trust in the police. This is a drastic increase from 2018 when a mere 29.9% of the respondents said they had a lot of trust in the police, and 28.4% said they had no trust in the police. In fact, from 2010 to 2020, the trust in the police has increased by 60.3% and the rank of police forces (out of the 18 institutions surveyed about) has increased from 18 in 2010 to 4 in 2020.

Uncovering trends in self-reported ILI symptoms

In order to collect data on ILI symptoms, the #CoronaTracker asks respondents to report all 7 major symptoms as reported by the World Health Organization and the Indian Council of Medical Research (ICMR) including high fever, tiredness, dry cough, difficulty in breathing, cold/nasal congestion/runny nose, body pain and sore throat pain. Starting from March 23rd, this data is being collected every day, among 18+ adults statewide, including every major demographic.

We observe that in the initial weeks of the survey, reports of self-infection were much higher than reports of infection in households or in their neighbourhood. However, after mid-April this pattern is reversed with more reports of infection in households and neighbourhood than reports of self-infection. Both ILI symptoms and COVID-19 mainly spread from person to person, including when an infected person coughs or sneezes, through airborne transmission, surface transmission, amongst others. Hence, this is likely to be the reason why individuals who reported having any ILI symptoms in the initial weeks of the survey no longer exhibit them and have instead passed it on to someone in their household or neighbourhood.

Out of all reports of ILI symptoms (self, household and neighbourhood) the highest proportion is that of reports of symptoms in the neighbourhood. However, this proportion reduces drastically, stabilizing at around 0.4% in towards the end of April.

The ILI symptoms collected by the tracker can be representative of “self-reported cases” in a thoroughly random population. Since the severity of symptoms takes a few days to appear, the information from the daily tracker could show the number of positive cases ahead of time (as reported by Govt and Hospitals).

So can these symptoms predict the daily increase in COVID 19 cases in India?

In addition to reporting symptoms, respondents from April 1st are also being asked to report which of the typical symptoms associated with COVID19 are they seeing in their household or neighbourhood. There seems to be around a one week gap in specific COVIID 19 symptoms reported in the tracker and the actual number of confirmed cases in the country.

It is visible that both graphs are following similar bumps in their trajectory, and if they continue to go in parallel, the tracker data can be useful in predicting the number of cases in India. Meanwhile, the total number of positive Coronavirus cases in India is at 35,043 (May 1st) with a recovery rate of 25% and a mortality rate of 3.2%. The nation-wide lockdown is scheduled to end on May 3rd. Some states have however extended the lockdown till mid-May.

Read the infographic here in Hindi and English

-Shreya Maskara/New Delhi
From Polstrat, a non-partisan political consultancy which aims to shift the narrative of political discourse in the country from a problem-centric to a solutions-oriented approach. All survey findings and projections are based on the Team CVoter “Trust in Institutions” survey carried out in 2010, 2018 and 2020 among 18+ adults statewide, including every major demographic. Team CVoter (Center For Voting Opinions and Trends in Election Research) is a leading international stakeholder research organization with its roots in India.

Read more about Polstrat here. Follow us on Medium to keep up to date with Indian politics.