key: cord-0333374-ofgv3p8y
authors: Lyngse, F. P.; Mortensen, L. H.; Denwood, M. J.; Christiansen, L. E.; Moller, C. H.; Skov, R. L.; Spiess, K.; Fomsgaard, A.; Lassauniere, R.; Rasmussen, M.; Stegger, M.; Nielsen, C.; Sieber, R. N.; Cohen, A. S.; Moller, F. T.; Overvad, M.; Molbak, K.; Krause, T. G.; Kirkeby, C. T.
title: SARS-CoV-2 Omicron VOC Transmission in Danish Households
date: 2021-12-27
journal: nan
DOI: 10.1101/2021.12.27.21268278
sha: 59980039b9b659e44e484253cc854d624e7ff178
doc_id: 333374
cord_uid: ofgv3p8y

The Omicron variant of concern (VOC) is a rapidly spreading variant of SARS-CoV-2 that is likely to overtake the previously dominant Delta VOC in many countries by the end of 2021. We estimated the transmission dynamics following the spread of Omicron VOC within Danish households during December 2021. We used data from Danish registers to estimate the household secondary attack rate (SAR). Among 11,937 households (2,225 with the Omicron VOC), we identified 6,397 secondary infections during a 1-7 day follow-up period. The SAR was 31% and 21% in households with the Omicron and Delta VOC, respectively. We found an increased transmission for unvaccinated individuals, and a reduced transmission for booster-vaccinated individuals, compared to fully vaccinated individuals. Comparing households infected with the Omicron to Delta VOC, we found an 1.17 (95%-CI: 0.99-1.38) times higher SAR for unvaccinated, 2.61 times (95%-CI: 2.34-2.90) higher for fully vaccinated and 3.66 (95%-CI: 2.65-5.05) times higher for booster-vaccinated individuals, demonstrating strong evidence of immune evasiveness of the Omicron VOC. Our findings confirm that the rapid spread of the Omicron VOC primarily can be ascribed to the immune evasiveness rather than an inherent increase in the basic transmissibility.

The SARS-CoV-2 variant B.1.1.529, which is referred to as the Omicron variant of concern (VOC), has overtaken the Delta VOC in South Africa and has spread rapidly to at least 28 countries countries in Europe (7), Asia, the Middle East and South America (9; 17). The Omicron VOC has been reported to be three to six times as infectious as previous variants A current concern worldwide is that the Omicron VOC is able to evade immunity induced by the currently used vaccines, and a preliminary meta-analysis of neutralization studies indicated that the vaccine effectiveness is reduced to around 40% against symptoms and to 80% against severe disease, but that the effect for booster vaccinations is at 86% and 98%, respectively (12). These results are supported by laboratory studies establishing a markedly reduced elimination of the Omicron VOC by neutralizing antibodies, indicating that the vaccination effectiveness with Pfizer-Biontech against infection is only at 35% for the Omicron VOC (5) . This was corroborated by another in vitro study reporting an 8.4-fold reduction in neutralization for the Omicron VOC vs. the PV-D614G reference strain, whereas there was only a 1.6-fold reduction in neutralization for the Delta VOC (27) . Therefore, the advantage of the Omicron VOC seems to be a combination of high transmissibility and increased immune evading abilities.

Studies on the transmission of the Omicron VOC are yet sparse and a critical prerequisite for effective control of this variant worldwide (3) . In particular, it is important to clarify whether the growth advantage can be ascribed to immune evasiveness, i.e., a higher proportion of vaccinated or previously infected individuals being susceptible to infection, an increased inherent transmissibility for this variant, or both. 3 The aim of the present study is to investigate the household transmission of the Omicron VOC. Specifically, we address the following questions: 1) Is the secondary attack rate higher for the Omicron VOC than for the Delta VOC? 2) Does the Omicron VOC show a higher immune evasiveness relative to the Delta VOC? 3) Is booster vaccination effective for reducing transmission?

Since July 2021, the Delta VOC has been the dominant variant in Denmark. The first Danish case infected with the Omicron VOC was detected on 22 nd November 2021 (Danish Covid-19 Genome Consortium, DCC), and community transmission was determined to be present by late November 2021. On 8 th December, Danish authorities discontinued intensive contact tracing of close contacts for cases specifically infected with the Omicron VOC. We therefore started the study period on 9 th December 2021 when cases of both variants were treated approximately equally, thus reducing bias from intensified contact tracing and active case finding of the Omicron VOC that was implemented shortly after it's discovery in Denmark (24). The end of the inclusion period for primary cases was set at 12 th December to balance the inclusion of enough cases for proper estimation and early dissemination of the results. Potential secondary cases were followed up to 7 days after, i.e., until 19 th December 2021 to allow for test results to be obtained. We obtained the last test results on 21 st December.

We used Danish register data for this study. All individuals in Denmark have a unique identification number, enabling cross linking between administrative registers. Using this, we obtained individual level information on home address, data on all antigen and RT- is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint Households were defined based on residential addresses. We included households with 2-6 members to exclude care facilities and other places, where many individuals share the same address. If two individuals tested positive on the same day, we excluded these households from the data, to ensure proper identification of the primary case within each household.

We defined a primary case as the first individual within a household to test positive with an RT-PCR test within the study period. We followed all tests of other household members in the study period. A positive secondary case was defined by either a positive RT-PCR test or a positive antigen test (10) . Almost all samples that tested positive with RT-PCR were tested with Variant PCR to determine the VOC (21) 

We defined the secondary attack rate (SAR) within households as the proportion of potential secondary cases within the same household that tested positive between 1-7 days following the positive test of the primary case within the household (15). Adjusted odds 5 . 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 December 27, 2021. ; https://doi.org/10.1101/2021.12.27.21268278 doi: medRxiv preprint ratios (OR) were estimated from multivariable logistic regression models fit to the binary outcome of test status of each potential secondary case, with the primary explanatory variable reflecting the strain type in the household (Omicron vs. Delta VOC) and potentially confounding variables of age and sex of the primary case, age and sex of the potential secondary case, and household size (2-6 individuals). To test if vaccine status conferred differential protection against the Omicron and Delta VOC, we included an interaction term between vaccination status of primary and potential secondary cases and the variant. Standard errors were adjusted to control for clustering at the household level.

We have conducted a number of supplementary analyses to support the main analysis.

We re-analysed each strata of the data separately using another set of logistic regression models (see appendix 7.4). To examine the potential mediating role of viral load of primary cases infected with the Omicron VOC relative to the Delta VOC, we plotted the distributions of Ct values for each variant (appendix Figure 3) . We also examined the extent to which the Ct value of the primary case could explain the difference in transmission between the variants (see appendix Table 8 ). Our study relies on Variant PCR testing to determine if each primary case was Delta or Omicron. We estimated the intra-household correlation of variants, i.e., the probability that the positive secondary case was infected with the same variant as the primary case. To investigate if there was bias in the selection of samples for Variant PCR, we investigated the probability of sampling for Variant PCR by sample Ct value and age. Moreover, we tested the robustness of potential secondary cases being tested and testing positive by only using RT-PCR tests, which are more sensitive.

This study was conducted using data from national registers only. According to Danish law, ethics approval is not needed for this type of research. All data management and analyses were carried out on the Danish Health Data Authority's restricted research servers with project number FSEID-00004942. The study only contains aggregated results 6 . 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 data used in this study are available under restricted access due to Danish data protection legislation. The data are available for research upon reasonable request to The Danish Health Data Authority and Statens Serum Institut and within the framework of the Danish data protection legislation and any required permission from Authorities. We performed no data collection or sequencing specifically for this study.

A total of 2,225 primary cases with the Omicron VOC and 9,712 primary cases with the Delta VOC were included ( is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint Notes: The secondary attack rate (SAR) is expressed as a percentage (%). Summary statistics based on primary cases are shown separately from summary statistics on potential secondary cases, positive secondary cases and SAR.

We found that the cumulative probability of potential secondary cases to be tested at least once increased from 33-41% to around 87-89% at 7 days after the primary case tested positive ( Figure 1 , panel a). The probability was higher when the primary case was infected with the Delta VOC compared to the Omicron VOC. The probability of potential secondary cases being tested twice increased from 9% to 69-74% at 7 days after the primary case tested positive. The probability of potential secondary cases testing positive increased from 3-4% on day 1, to 21% and 31% on day 7, when the primary case was infected with the Delta VOC and Omicron VOC, respectively ( Figure 1 is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint Notes: Panel (a) shows the probability of potential secondary cases being tested after a primary case has been identified within the household. Panel (b) shows the probability of potential secondary cases that test positive subsequently to a primary case being identified within the household. Note that the latter is not conditional on being tested, i.e. the denominator contains test negative individuals and untested individuals. The x axes shows the days since the primary case tested positive, and the y axes shows the proportion of individuals either being tested (a) or testing positive (b) with either antigen or RT-PCR tests, based on the variant of the primary case. The SAR for each day relative to the primary case can be read directly from panel (b). For example, the SAR on day 7 is 31% and 21%, whereas the SAR on day 4 is 24% and 15%. The shaded areas show the 95% confidence bands clustered on the household level. See appendix Figure 5 for the same two panels, only using RT-PCR tests.

The effect of vaccination on susceptibility and transmissibility of SARS-CoV-2 within households is shown in 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 December 27, 2021. ; https://doi.org/10.1101/2021.12.27.21268278 doi: medRxiv preprint cases, while a booster-vaccinated primary case was associated with a decreased OR of 0.72 (CI: 0.56-0.92). Notes: This table shows odds ratio estimates for susceptibility and transmissibility by vaccination status. Column 1 shows the susceptibility based on the vaccination status of the potential secondary case, conditional on being in a household infected with the Omicron VOC. Column 2 shows the susceptibility based on the vaccination status of the potential secondary case, conditional on being in a household infected with the Delta VOC. Column 3 shows the transmissibility based on the vaccination status of the primary case, unconditional on the variant in the household. Note, all estimates are from the same model, but with a different reference category across column 1-3. The estimates are adjusted for age and sex of the primary case, age and sex of the potential secondary case, and size of the household. The estimates are furthermore adjusted for vaccine status of the potential secondary case interacted with the household variant, and the vaccine status of the primary case. 95% confidence intervals are shown in parentheses. Standard errors are clustered on the household level. The odds ratio estimates for the full model are presented in Appendix Table 8 , column I.

The relative difference in transmission between the Omicron and Delta variants when comparing potential secondary cases with the same vaccine status is shown in Table 3 . is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint We compared the Ct values of primary cases with the Omicron VOC and the Delta VOC (Appendix 7, Figure 3 ). The distribution of Ct values for cases with the Omicron VOC were slightly skewed to the left compared to cases with the Delta VOC, but the median values (27.24 and 28.29, respectively) did not differ substantially. Adjustment for Ct values of the primary cases did not materially alter the findings, suggesting that the difference between the Omicron and Delta VOC transmission is not due to differences in viral load in the primary case (appendix Table 8 ). Similarly, the distribution of time since last vaccination/booster/infection among positive secondary cases were approximately the same across the two variants (appendix Figure 4) . However, these analyses are limited by the fact that vaccine roll out in Denmark is largely determined by age, which makes it difficult to estimate true associations for variables that are also correlated with age.

The probability that a sample was selected for Variant PCR was stable across Ct value, varying between 97% and 100% (Appendix 7, Figure 2 ). Selection for Variant PCR was also similar across age groups, varying between 95% and 99%. The Variant PCR was validated with high quality whole genome sequencing (WGS) data, and showed an 11 . 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 December 27, 2021. ; https://doi.org/10.1101/2021.12.27.21268278 doi: medRxiv preprint estimated false negative rate of 0.06% and a false positive rate of 1.15% (Appendix section 6.3).

We found an intra-household correlation between primary case and secondary case variants of 98.0% and 97.7% for households with primary cases infected with the Omicron VOC and the Delta VOC, respectively (Appendix section 7.2). This supports the assumption that subsequent infections within the same household are likely to be associated with the same variant as was identified for the primary case.

Finally, misclassification of primary and secondary cases is also a potential concern. It is not given that the index case is actually also the case with the first positive RT-PCR test.

However, we believe that this potential bias would impact both the Delta and Omicron VOC transmission estimates equally. Figure 1 , panel b, shows the proportion of secondary cases testing positive on each day after the primary case. One day after, 3% and 4% have tested positive, two days after, 7% and 12% have tested positive. These cases could have been misclassified as secondary cases, instead of primary cases. To investigate the sensitivity of our results for this potential misclassification, we estimated our regression model including only secondary cases that tested positive on day 2-7 and on day 3-7 (Appendix Table 9 ).

Our results show that the Omicron VOC is generally 2.7-3.7 times more infectious than the Delta VOC among vaccinated individuals (Table 3 ). This observation is in line with data from (18) is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint Surprisingly, we observed no significant difference between the SAR of Omicron versus Delta among unvaccinated individuals (Table 3 ). This indicates that the increased transmissibility of the Omicron VOC primarily can be ascribed to immune evasion rather than an inherent increase in the basic transmissibility. If this observation can be confirmed by independent studies, it has important ramifications for the understanding of the current challenges for control of the epidemic. Our data indicate that the non-pharmaceutical interventions that were used to control the previous variants of SARS-CoV-2 are also likely to be effective against the Omicron VOC. On the other hand, although we showed that booster vaccines did offer some protection against household transmission, the reduced level of protection means that vaccination is less likely to be sufficient to curb transmission within a population. Furthermore, the duration of the protective effect is currently unknown, and the rapidly waning effectiveness of the second dose against the Omicron VOC as well as data from neutralization assays (14; 8) do raise some concerns about the longevity of the booster response. This means that the current vaccines are unlikely to mitigate the spread of the Omicron VOC to the extent that has been achieved for previous variants on the long term. We therefore suggest that adapted or improved vaccines may be necessary to mitigate the spread of the Omicron VOC. However, both a primary series and a booster dose is likely to play an important role in reducing transmission on a short term and modifying the outcome of infection by reducing severity. We were unable to address this question as detailed clinical data were not available in our registries, and in any case the relatively short follow-up time precludes a thorough investigation of any subsequent hospitalisations and deaths.

(1) note that the existing circulating immunity within a country is of major importance in limiting the severity of the epidemic with the Omicron VOC. We found that booster- 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 December 27, 2021. There are some potential biases in this study. Firstly, in this initial phase, the spread of the Omicron VOC was characterized by spreading events, and has therefore not yet reached an even distribution throughout the population (3). We found that a large proportion of the Omicron VOC cases were aged 20-30 years, and therefore different from the age distribution of Delta cases (Table 1) . Furthermore, the Omicron VOC is mainly found in households with 2 members, whereas the Delta VOC is mainly found in households with 4 members. This could potentially impact the comparability of our estimates for the two the variants. However, our results are robust, when we exclude households with a primary case <10 years and when we only include 2-person households (Appendix Table 8 ).

Recently, self testing kits have become widely available for purchase in Denmark. This could influence the results, for instance if individuals that self-test at home refrain from also being tested in public testing facilities, meaning that their test results are not registered in the national databases. However, the probability of being tested was very similar between RT-PCR and any test (Figure 1 and Appendix Figure 5 ), and consequently most positive antigen tests were confirmed with a positive RT-PCR test, so we assume that has likely also been the case for positive self-tests as well. Moreover, we believe that this potential bias will impact both the Delta and the Omicron VOC transmission estimates equally.

There are also a number of other confounding variables that might lead to biases within our study. Vaccines are not randomly distributed within the population, so immunocom-14 . 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 December 27, 2021. ; https://doi.org/10.1101/2021.12.27.21268278 doi: medRxiv preprint promised and other vulnerable individuals are more likely to have had access to a booster vaccine in particular. To some extent, this has been addressed by including age as an explanatory variable in the models, but it was not possible to include information on underlying risk conditions, so this information is incomplete. Similarly, there are likely underlying behavioural drivers for an individual being unvaccinated, which are likely to confound with other risky behaviours that might be expected to increase both transmission and susceptibility to infection (e.g. poor use of face masks, reduced attention to hygiene). The use of registry data limits our inference to associations between transmission/susceptibility and vaccination status of individuals, where part of the association is due to general characteristics of the individuals themselves rather than their vaccination status. However, a key point of our study is that these biases are non-differential with respect to variant, i.e. they would be expected to affect transmission of both the Delta and Omicron VOC identically because risky behaviours and/or underlying health concerns should affect both variants equally. Consequently, although the associations between vaccination status and transmission for each variant are likely to each be susceptible to bias, we believe that any such bias is non-differential with respect to variant. Therefore, our results concerning the relatively higher transmissibility of the Omicron vs. Delta VOC for vaccinated individuals should be robust to these potential biases.

The SAR was found to be higher for the Omicron VOC than for the Delta VOC across all age groups (Table 1) . Notably, the SAR for individuals aged 70+ years was 32% for the Omicron VOC compared to only 18% for the Delta VOC. This has implications for e.g. care facilities, highlighting the need for increased protection against transmission, now that the Omicron VOC will likely overtake the Delta VOC in many countries.

To conclude, we found an increased susceptibility for unvaccinated individuals, and a reduced susceptibility for booster-vaccinated individuals, compared to fully vaccinated individuals in households infected with the Delta VOC. Additionally, we found a reduced susceptibility for booster-vaccinated individuals in households infected with the Omicron VOC. Furthermore, we found an increased transmissibility from unvaccinated individuals, and a reduced transmissibility from booster-vaccinated individuals, compared to fully 15 . 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 is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint [7] ECDC (2021). Weekly epidemiological update: Omicron variant of concern (VOC)

-week 50 (data as of 19 December 2021) https://www.ecdc.europa.eu/en/newsevents/weekly-epidemiological-update-omicron-variant-concern-voc-week-50-data-19december-2021 (Accessed 2021-12-21).

17 . 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 is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint [17] Public Health England (2021a). SARS-CoV-2 variants of concern and variants under investigation in England. technical briefing 32.

[18] Public Health England (2021b). SARS-CoV-2 variants of concern and variants under investigation in England. technical briefing 31. [23] SSI (2021b). Status for udvikling af SARS-CoV-2 Varianter der overvåges i Danmark.

[24] SST, S. (2021). Naere kontakters naere kontakter skal ikke laengere gå i selvisolation.

https://covariants.org/per-country (Accessed 2021-12-20).

. 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 is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint 

The authors declare no competing interests.

. 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 December 27, 2021. ; https://doi.org/10.1101/2021.12.27.21268278 doi: medRxiv preprint

This section provides some background information in order to better understand the circumstances underlying our study. The results build on all data available in Denmark, i.e., not restricted to our study sample.

Denmark provides free testing with both antigen and RT-PCR tests. Antigen tests provide a quick test result (positive/negative) with a median time from sample to result of less than 30 minutes (Table 4 ). All positive antigen tests are recommended to be confirmed with an RT-PCR test. RT-PCR tests are more sensitive (25), but also require a longer time before the result is known. The median time is approximately 24 hours (Table 5) .

Only samples from positive RT-PCR tests are selected for Variant PCR and whole genome sequencing. 

. 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 December 27, 2021. ; https://doi.org/10.1101/2021.12.27.21268278 doi: medRxiv preprint 

This subsection provides details on the sampling probability for variant PCR. There was a higher proportion of cases aged 15-30 years that tested positive with the Omicron VOC (green, panel b).

23 . 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 December 27, 2021. ; https://doi.org/10.1101/2021.12.27.21268278 doi: medRxiv preprint 96 97  100 98  9898  97  98  98  98  96  97  97  99  9797  98  98  100100 98  97  98  97  97  97  98  97  97  98  98  98  98  99  99  98  98 

This subsection provides additional results in order to examine the robustness of the Variant PCR used to detect the Omicron VOC.

The Variant PCR was validated using the available high quality whole-genome sequencing (WGS) data. The rate of false positives was estimated to 1.15% (18/1,567) ( is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint 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 December 27, 2021. ; https://doi.org/10.1101/2021.12.27.21268278 doi: medRxiv preprint

In this subsection, we investigate the intra-household correlation of variants, i.e., the probability that the primary and positive secondary cases are infected with the same variant.

In households where the primary case was infected with the Omicron VOC, we found 1,160 positive secondary cases that also had a Variant PCR result (Table 7) . Of these, 1,137 (98%) were also Omicron VOC and 23 (2%) were Delta VOC. Similarly, in households where the primary case was infected with the Delta VOC, we found 3,917 positive secondary cases. Of these, 3,827 (98%) were also Delta and 90 (2%) were Omicron VOC.

The overall intra-household correlation of variants was 97.8 (CI: 97.3-98.3). is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint 

This subsection provides additional results in order to validate the robustness of the results shown in the main paper. Figure 5 shows the same as Figure 1 , but only using RT-PCR tests (i.e. excluding antigen tests).

28 . 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 Notes: This figure shows the same as Figure 1 , but only including RT-PCR tests. Panel (a) shows the probability of potential secondary cases being tested after a primary case has been identified within the household. Panel (b) shows the probability of potential secondary cases that test positive subsequently to a primary case being identified within the household. Note that the latter is not conditional on being tested, i.e. the denominator contains test negative individuals and untested individuals. The x axes shows the days since the primary case tested positive, and the y axes shows the proportion of individuals either being tested (a) or testing positive (b) with an RT-PCR test, based on the variant of the primary case. The SAR for each day relative to the primary case can be read directly from panel (b). For example, the SAR on day 7 is 30% and 21%, whereas the SAR on day 4 is 23% and 14%. The shaded areas show the 95% confidence bands clustered on the household level. Tables 8 and 9 show the results of logistic regression models fit to specific strata of the data.

Column I of Table 8 shows the odds ratio estimates of our main specification, i.e., the same as used in Table 2 and 3. Column II shows the odds ratio estimates, when excluding households with a primary case <10 years. Column III shows the odds ratio estimates, when only including 2-person households. Column IV shows the odds ratio estimates, while adjusting for Ct value of the primary case.

In Table 9 we show the estimates, while restricting the inclusion criteria for positive secondary cases. Column I shows the odds ratio estimates, including positive secondary cases found on day 1-7, i.e, our main specification, also presented in Table 2 and 3 .

Column II includes positive secondary cases found on day 2-7. Column III includes positive secondary cases found on day 3-7.

. 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 Notes: Column I shows the odds ratio estimates of our main specification, i.e., the same as used in Table  2 and 3. Column II shows the odds ratio estimates, when excluding households with a primary case <10 years. Column III shows the odds ratio estimates, when only including 2-person households. Column IV shows the odds ratio estimates, while adjusting for Ct value of the primary case. 95%-Confidence interval in parentheses. Standard errors are clustered on the household level.

. 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 December 27, 2021. ; https://doi.org/10.1101/2021.12.27.21268278 doi: medRxiv preprint Notes: This table shows the sensitivity for the choice of inclusion of positive secondary cases. In Table  9 we show the estimates, while restricting the inclusion criteria for positive secondary cases. Column I shows the odds ratio estimates, including positive secondary cases found on day 1-7, i.e, our main specification, also presented in Table 2 and 3 . Column II includes positive secondary cases found on day 2-7. Column III includes positive secondary cases found on day 3-7. 95%-Confidence interval in parentheses. Standard errors are clustered on the household level.

. 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 December 27, 2021. ; https://doi.org/10.1101/2021.12.27.21268278 doi: medRxiv preprint

Potential severity, mitigation, and control of Omicron waves depending on pre-existing immunity and immune evasion

Effect of a nationwide booster vaccine rollout in Israel on SARS-CoV-2 infection and severe illness in young adults. Travel medicine and infectious disease

Outbreak caused by the SARS-CoV-2 Omicron variant in Norway

How bad is Omicron? What scientists know so far

SARS-CoV-2 Omicron has extensive but incomplete escape of Pfizer BNT162b2 elicited neutralization and requires ACE2 for infection. medRxiv

Genome Con