A comparative analysis of epidemiological characteristics of MERS-CoV and SARS-CoV-2 in Saudi Arabia

Abstract

In this study, we determine and compare the incubation duration, serial interval, pre-symptomatic transmission, and case fatality rate of MERS-CoV and COVID-19 in Saudi Arabia based on contact tracing data we acquired in Saudi Arabia. The date of infection and infector-infectee pairings are deduced from travel history to Saudi Arabia or exposure to confirmed cases. The incubation times and serial intervals are estimated using parametric models accounting for exposure interval censoring. Our estimations show that MERS-CoV has a mean incubation time of 7.21 (95% CI: 6.59-7.85) days, whereas COVID-19 (for the circulating strain in the study period) has a mean incubation period of 5.43(95% CI: 4.81-6.11) days. MERS-CoV has an estimated serial interval of 14.13(95% CI: 13.9-14.7) days, while COVID-19 has an estimated serial interval of 5.1(95% CI: 5.0-5.5) days. The COVID-19 serial interval is found to be shorter than the incubation time, indicating that pre-symptomatic transmission may occur in a significant fraction of transmission events. We conclude that during the COVID-19 wave studied, at least 75% of transmission happened prior to the onset of symptoms. The CFR for MERS-CoV is estimated to be 38.1% (95% CI: 36.8-39.5), while the CFR for COVID-19 1.67% (95% CI: 1.63-1.71). This work is expected to help design future surveillance and intervention program targeted at specific respiratory virus outbreaks, and have implications for contingency planning for future coronavirus outbreaks.

Keywords: Case fatality rate; Incubation period; MERS-CoV; Pre-symptomatic transmission; SARS-CoV-2; Serial interval.

Fig. 1

Comparison of nonparametric and parametric estimations of the distribution of MERS-CoV and COVID-19 infection incubation times in Saudi Arabia. Using data from Saudi Arabia, the panels (a: COVID-19, b: MERS-CoV) compare the nonparametric Kaplan-Meier estimate of the incubation period distribution with the parametric lognormal, Weibull, and gamma distributions.

Fig. 2

Comparison of early and late nonparametric and parametric estimates of the distribution of the incubation time for MERS-CoV and COVID-19 infections in Saudi Arabia. Using data from Saudi Arabia, panels (a: COVID-19, b: MERS-CoV) contrast the Kaplan-Meier nonparametric estimate of the incubation duration distribution with the parametric fitted lognormal, Weibull, and gamma, distributions.

Fig. 3

Incubation period of MERS-CoV and COVID-19 infection outbreak in Saudi Arabia. Panel (a) shows the density estimation of COVID-19 incubation periods with gamma, Weibull and log normal distributions. Panel (b) shows the density estimation of MERS-CoV incubation periods with gamma, Weibull and log normal distributions.

Fig. 4

Network diagram for (a) COVID-19 and (b) MERS-CoV infection in Saudi Arabia, where the source and transmission chain could be determined.

Fig. 5

Fitted serial interval distribution for (a) COVID-19 and (b) MERS-CoV based on reported transmission pairs in Saudi Arabia. We fitted three commonly used distribution, lognormal, gamma, and Weibull distributions, respectively.

Fig. 6

Serial intervals of possible case pairs in (a) COVID-19 and (b) MERSCoV. Pairs represent a presumed infector and their presumed infectee plotted by date of symptom onset.

Fig. 7

Pre-symptomatic infection as estimated by samples of (serial interval - incubation period), accounting for covariation. grey vertical line: 0. Samples below zero indicate pre-symptomatic transmission.

Fig. 8

Using the distributions of onset to death, the updated case fatality rates for MERS-CoV and COVID-19 in Saudi Arabia were derived. The panels a and b show the results of a gamma distribution fit to outbreak data from Saudi Arabia. Mean and median distribution durations for COVID-19 and MERSCoV were (13.4 and 12.7) and (16.6 and 13.8) days, respectively (shown by the red and blue lines).