Observed serial intervals of SARS-CoV-2 for the Omicron and Delta variants in Belgium based on contact tracing data, 19 November to 31 December 2021

The SARS-CoV-2 Omicron BA.1 variant is rapidly spreading worldwide, possibly outcompeting the Delta strain. We investigated the empirical serial interval for both variants using contact tracing data. Overall, we observed a shorter serial interval for Omicron compared to Delta, suggesting faster transmission. Furthermore, results indicate a relation between the empirical serial interval and the vaccination status for both the Omicron and the Delta variant. Consequently, with the progression of the vaccination campaign, the reasons for and extent of dominance of Omicron over Delta may need further assessment.


Background
The WHO designated the SARS-CoV-2 Omicron BA.1 variant (B.1.1.529) as a variant of concern (VoC) on 26 November 2021 [1]. Omicron shows a fast epidemic growth and has taken over as dominant VoC from the previously dominant Delta variant (B.1.617. 2) worldwide. In Belgium, the Omicron variant was the dominant circulating strain during the period from 27 December 2021 to 9 January 2022, identified in 88.5% of the sequenced samples [2]. It has been found that the Omicron variant is more efficient at evading immunity, acquired from previous infection or vaccination [3,4], compared to the Delta variant. Another epidemiological characteristic that may contribute to the rapid spread of Omicron is increased transmissibility, possibly due to an increase in the reproduction number or a shortened serial interval, i.e. the time between symptom onset in an infector and infectee [5]. In this study, we estimate the mean and the standard deviation of the serial interval for both the Omicron and Delta variants using contact tracing data collected in Belgium and assess whether these variants are associated with a different observed serial interval. To gain more insights on the possible impact of vaccination, we also compare the observed serial intervals for different combinations of vaccination status in transmission pairs.

Data
Belgium has a contact tracing system in place, where COVID-19 confirmed cases are asked about their contacts during two days prior to symptom onset until 10 days after. Variant detection was done using genotype sequencing. If a variant was found in a transmission chain, all cases belonging to that chain were assumed to be infected by that variant. We collected transmission pairs that could be linked either to Omicron or Delta infections, where the infector reported first symptoms between 19 November and 31 December 2021. During this period Omicron started to spread in Belgium and took over dominance from the Delta variant [6]. The same non-pharmaceutical interventions were in place throughout the considered time window, with a fairly low stringency index (around 48) compared to neighboring countries.
We assumed that the first confirmed case in a reconstructed transmission pair (i.e. the 'index' case) was the infector, and hence the contact was the infectee. Transmission pairs for which symptom onset was not available for either the infector or the infectee were excluded, as well as pairs for which the observed serial interval was less than -5 days or more than 15 days in order to ensure biologically plausible serial intervals [7,8]. Vaccination status was assigned to both cases in a transmission pair as (i) unvaccinated (including partially vaccinated), (ii) . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 30, 2022. ; https://doi.org/10.1101/2022.01.28.22269756 doi: medRxiv preprint vaccinated (i.e. completed vaccination cycle), and (iii) vaccinated plus booster. Household status was assigned based on information collected during the interview with the index case or provided by the healthcare worker. A more detailed description of the data and analysis is provided in Supplement S1.

Overall serial interval
Of the 2495 included transmission pairs, 86.61% were linked to transmission of the Omicron variant ( Figure S1). In what follows we will report the mean and standard deviation (SD) of the observed serial intervals, the median was 3 days for all stratifications. All reported pvalues are based on a Mann-Whitney U test. Tables S1 and S2 show the proportion of transmission pairs by household and vaccination status.

We identified 2161 Omicron and 334 Delta transmission pairs during the period from 19
November to 31 December 2021. The empirical serial interval distribution for Omicron had a mean of 2.75 days (SD 2.53 days), compared to 3.00 days (SD 2.48 days) for Delta (p = .019; Table 1 shows the parameter estimates of the normal distribution fit to both empirical serial interval distributions ( Figure 1C). is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

Serial interval stratified by household status
There was a significant difference between the mean empirical serial interval of both variants within households, but not between households (p = .034 for within-household, and p = .686 for between-household transmission pairs; Figures 2A-D). Figure 2E represents the fit of a normal distribution to the empirical serial intervals for both variants, and in Table 1 we report the parameters of these fitted distributions.

Serial interval stratified by vaccination status
No difference in empirical serial intervals was found for pairs where both cases were unvaccinated or only partially vaccinated (2.69 vs 2.54 days, p = .931; Figure 3A

Discussion
Our estimates of the empirical serial interval for Omicron are in line with those previously reported. Lee  is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 30, 2022. ; https://doi.org/10.1101/2022.01.28.22269756 doi: medRxiv preprint asymptomatic transmission may lead to different serial and generation interval distributions [11]. Future studies estimating the generation time for both variants are needed to shed more light on this.
We found an indication that the empirical serial interval might depend on vaccination status.
Precisely, we observed a shorter empirical serial interval for Omicron (2. This work has several limitations. Date of symptom onset is self-reported by the case and may be subject to recall bias. Likewise, there might be bias in the reporting of contacts and the level of reporting may differ for each index case. We have used all reported transmission pairs, although some of them may have been wrongly assigned. We further assume that directionality of transmission was from index to contact. We also do not explicitly account for right truncation [12], but this is assumed not to affect our estimates since it is unlikely that symptomatic infectees were missed as we used the data available on 17 January 2022, but only including infectors that had symptom onset until 31 December 2021 and limiting the serial interval to at most 15 days. However, because contacts are reconstructed until two days prior to symptom onset of the index case, possible left truncation might lead to exclusion of . CC-BY 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted January 30, 2022. ; https://doi.org/10.1101/2022.01.28.22269756 doi: medRxiv preprint some transmission pairs. While the transmission risk is higher around symptom onset [8], it is possible that not all the infection events caused by an index case are detected. There is also possible selection bias due to the targeted genotype sequencing of suspected Omicron cases (such as previously infected cases or travelers), whereas genotype sequencing resulting in confirmed Delta cases might have been done on samples from a hospital setting as severe disease was an indication for sequencing during the study period. This analysis does not correct for age and previous infection and hence reinfections might be over-represented among the Omicron cases.

Ethical and privacy statement
The data analyses carried out in this work are among the legal tasks of Sciensano (artikel 4, §4 Wet tot oprichting van Sciensano; article 4, §4 Law establishment Sciensano). Article 4, §4 explicitly states that Sciensano is authorized to collect, validate, analyse, report and archive data of a personal nature concerning public health. Sciensano is further authorized to make these processed data available with approval of the qualified sectoral committees. Such approval for these data was obtained from the Information Security Committee Social Security and Health (ISC).

Acknowledgements
Data is provided by Sciensano, the Belgian institute for health. We thank Nicola Low for the insightful comments on an earlier draft of this manuscript. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint

Data availability
Transmission pair data (without vaccination status for identifiability reasons) and R code for the analysis are available on https://github.com/cecilekremer/serial_interval.