Incubation Period and Reproduction Number for novel coronavirus (COVID-19) infections in India

Introduction

While the novel coronavirus (COVID-19) spread rapidly from China to other developed countries, India saw a steady flow of patients mod early March and by 10 May 2020, it had gripped the country with 63,420 confirmed cases and 2109 deaths^1-2. The novel coronavirus (SARS-CoV-2) though related is distinct from severe acute respiratory syndrome (SARS) coronavirus and Middle East respiratory syndrome (MERS) coronavirus³. Many researchers struggled to estimate the magnitude of the epidemic wherein the epidemiological parameters remained uncertain. Knowledge of key epidemiological parameters including incubation period and reproduction number is important in controlling any epidemic. The distribution of these parameters helps estimate the epidemic size and transmission potential^3-7 of the disease.

The incubation period is defined as the time from infection to the onset of illness⁸ and is crucial for epidemiological modelling in predicting the transmission dynamics, infectiousness and quarantine period⁹. It is also important for several important public health activities like length of active monitoring, surveillance and control.

The reproduction number (R₀) is the most fundamental parameter in infectious disease dynamics describing the contagiousness or transmissibility of infectious agents and is defined as the average number of secondary cases caused by a single infectious individual in a entirely susceptible population¹⁰. An outbreak is expected to continue if R₀ has a value >1 and to end if R₀ is <1. R₀ fluctuates if the rate of human–to-human or human and vector interactions varies over time or space. There exists scant evidence which supports the applicability of R₀ outside the region for which the value was calculated¹¹. Estimation of changes in transmission over time can provide insights into the epidemiological situation and identify whether outbreak control measures are adequate and are having the desired measurable effect, and help in undertaking midcourse corrections.

Herd immunity is defined as the resistance to the spread of a contagious disease within a population that results if a sufficiently high proportion of individuals are immune to the disease, especially through vaccination or immunity post natural infection¹². When a high proportion of the population is immune, it is difficult for infectious diseases to spread, because there are not many people who can be infected and the transmission chain gets broken.

Our current understanding of these epidemiological parameters for India is limited. Hence this study was undertaken to address above issue and estimate incubation period and reproduction number of COVID-19 for India utilizing data reported by Ministry of Health and Family Welfare (MoHFW), Government of India (GOI) and the data available in public domain.

Materials and Methods

Results

Data on 268 lab confirmed cases were extracted, the range of age was from 1.5 years to 89 years with a mean age of 36.45 years (SD=±17.27). Ratio of male to female was 1.5:1 with 60.3% males and 39.7% females.

Incubation Period

Using date of exposure and date of confirmation of disease status/illness onset, the estimates of incubation period were determined. Table 1 gives the incubation period estimates for various distributions. We fitted five distributions to the data: Normal, Weibull, Log normal, Gamma and Erlang. The normal distribution provided best fit for data with median and mean incubation period of 6.93 (SD=±5.87, 95% confidence interval CI: 6.11-7.75). The incubation period ranged from 1 to 19.26 days (5^th percentile to 97^th percentile) for this best fit. The Weibull distribution was the best next fit with mean incubation period to be higher than normal distribution with mean of 8.17 (SD=8.10). Figure 1 show cumulative distribution functions for best fit. The median incubation period for other distributions ranged 3.46 to 6.06. The probability and cumulative distribution functions for various distributions in order of best fit is provided in the supplementary.

View this table:

• View inline
• View popup
• Download powerpoint

Table 1: Incubation Period distribution

• Download figure
• Open in new tab

Figure 1:

Cumulative distribution function for best fit Normal distribution for Incubation period for India

Discussion

This data driven paper to the best of our knowledge presents the estimates of incubation period and reproduction number of COVID-19 in India for the first time. We characterised the distribution of incubation period and reproduction number for COVID-19 for India. We estimated that the incubation period follows the best fit of normal distribution with around 7 days ranging from 1 to 22 day with 95^th percentile of the distribution at 17.8 days. The incubation period seems to be longer for India as compared to 2 to 10 days by WHO¹³ and 2-12 days by ECDC¹⁷. The United States’ CDC¹⁸ estimates the incubation period for COVID-19 to be between 2 and 14 days. Most of existing studies are on data originating from Wuhan, China. A study on travellers from Wuhan on 88 cases estimated the mean incubation period to be 6.4 days with a range from 2.1 to 11.1 days. The upper limit of 11.1 days was considered conservative¹⁹. Two other studies from Wuhan, China on early transmission dynamics of COVID-19 estimated the mean incubation period as 5.2 days (95% CI, 4.1 to 7.0), with the 95th percentile of the distribution at 12.5 days²⁰ and other study with 2-14 days with Log Normal as best fit²¹. A study by John Hopkins Bloomberg School of Public Health on regions and countries outside Hubei²² gave median incubation period as 5.1 days (95% CI, 4.5 to 5.8 days), and also stated that 97.5% of those who develop symptoms will do so within 11.5 days (CI, 8.2 to 15.6 days) of infection. Understanding incubation period is crucial so as to introduce and implement effective quarantine duration and preventing the spread of the virus.

Weibull distribution seems to be the best fit for reproduction number distribution in India. The reproduction number for India is estimated to be 2.6 in pre lock down phase and reduced to around 1.6 post lockdown phase with a herd immunity threshold of 61.5%. Though the value of R₀ has reduced from 2.6 in pre lock down period to 1.57 in post lock down period. i.e. 60.38% reduction it is still greater than 1. COVID-19 epidemic will increase as long as R₀ is greater than 1 and control efforts to bring R₀ below 1 needs to be implemented aggressively.

A report based on the impact of the interventions across 11 European countries estimated a posterior mean of 0.97 [0.14-2.14] for Norway and 2.64 [1.40-4.18] for Sweden, with an average of 1.43 across the 11 country posterior means, a 64% reduction compared to the pre-intervention values²³. A study from Wuhan, Hubei Province, China²⁰ estimated an R₀ of approximately 2.2, however this estimate was limited to early epidemic period up to January 4. The median R0 from Diamond Princess cruise ship²⁴ was estimated to be 2.28 (95% CI=2.06-2.52). Switzerland²⁵ estimated the basic reproduction number R₀ of COVID-19 at 2.78 (95% confidence interval, CI: 2.51 - 3.11). Transmission decreased with the strengthening of social distancing measures by 89% (95% CI: 83%-94%) which brought the R₀ to below 1 at 0.32. Study²⁵ by Wu et. al. estimated the reproductive number for COVID-19 as 2·68 (95% CI: 2·47–2·86). The wide range of R₀ from different studies indicates the challenges in estimating R₀. Additionally, R₀ is dynamic in the sense that any factor having the potential to influence the contact rate, including population density, social distancing, and seasonality will ultimately affect R₀.

Our estimate of herd immunity threshold as 61% (with R₀ =2.6) is staggering as 826,873,494 people need to be infected to achieve herd immunity in India which may result in high death rate. Many countries have implemented aggressive lock downs however some countries like Sweden had a different approach to tackle COVID-19 by taking individual responsibility for social distancing and keeping society functioning with no official lock down by the Sweden government and still successful in keeping the reproduction number in control²⁶. Long term lockdowns are not sustainable as its ill effects impacts many other health programmes. As per WHO²⁷ the weekly detection of new tuberculosis cases in India has gone down by nearly 75 per cent during the COVID-19 lockdown. The malaria modelling analysis by the Global Malaria Programme at WHO²⁸ and the HIV modelling study, convened by the WHO and UNAID²⁹, predicts the collateral damage of COVID will be as devastating as pandemic, with millions succumbing to preventable, treatable illness and disease. However, in absence of vaccine and other uncertainties regarding the virus, herd immunity is debatable issue as a preventive measure and achieving herd immunity is likely to be a long-drawn process.

Conclusions

This study provides the estimates for key epidemiological charachteristics of COVID-19 in India. The mean incubation period seems to be larger at 6.93 and 95^th percentile estimate for best fit normal distribution to be 17.8 days. Best fit for reproduction number follows weibull distribution. Our estimates for pre lockdown reproduction number was 2.6. and post lockdown the reproduction number reduced to 1.57. This implies and shows that the epidemic response strategies adopted by India are effective. However the herd immunity is estimated in the range of 36-61% for R₀ of 1.57 and 2.6 respectively.