Home transmission of the SARS-CoV-2 Omicron variant in Denmark

Study design and participants

Delta VOCs have been the dominant variant in Denmark since July 2021. The first Danish case of Omicron VOC was detected on November 22, 2021.18and community transmission was in place by early December 202119. On December 8, Danish authorities stopped extensive tracing of close contacts for cases specifically infected with Omicron VOC. We therefore chose a study period beginning on December 9, 2021, when the cases of both variants were treated roughly equally, thus reducing bias from previously extensive contact tracing and active case detection for Omicron VOC20. The end of the inclusion period for initial cases was set for December 15, with household contacts followed up until 7 days after the initial case, i.e. until December 22, 2021. We chose this as the end of our study period because the Christmas holiday in Denmark often begins on December 23, and often It includes extended family visits, interrupting the usual pattern of transmission within families. For additional information on the number of new cases, the proportion to Omicron and the number of tests conducted in Denmark during December 2021, see Appendix Section 1.

We used Danish registration data for this study. All individuals in Denmark have a unique identification number, which enables cross-linking of administrative records. Using this, we obtained person-level information on all residential addresses from the person’s central registry, and complete data on all antigen and RT-PCR tests for SARS-CoV-2 from the Danish Microbiology Database (MiBa).21), and all vaccination records from the Danish Immunization Registry22.

We identified households in Denmark using their unique residential address and assigned the same household identifier to all individuals registered at that address: this was used to determine household size. We only included households of 2-6 members to exclude care facilities and other settings, where many individuals share the same address.

We defined the initial case as the first family member to test positive using RT-PCR during the study period. We followed all other family members’ tests in the study period. A secondary case was identified by either a positive RT-PCR test or a positive antigen test23. All samples that tested positive with RT-PCR were subsequently tested with Variant PCR for VOC determination.24 (Annex Table S1 and Figure S1). Based on the result of the Variant PCR test for the initial case, we classified the families as having either Omicron or Delta VOC associated. Delta VOC has been the dominant variant in Denmark since early July 2021, accounting for nearly all positive RT-PCR samples from August to November 2021.25. We excluded families with a positive RT-PCR test 60 days before the initial case and families in which the initial case was ambiguous because two people tested positive on the same day.

We classified individuals into three groups: (1) unvaccinated; (2) fully fortified; or (iii) a booster vaccination. The definition of complete vaccination includes individuals who were infected more than 14 days ago, but were defined according to the vaccine used as follows: 7 days after the second dose of Comirnaty (Pfizer/BioNTech); 15 days after the second dose of Vaxzevria (AstraZeneca); 14 days after the second dose of Spikevax (Moderna); 14 days after vaccination with Janssen (Johnson & Johnson); 14 days after the second cross-vaccination dose. The booster vaccination was defined as 7 days after the booster vaccination26, 27. As of December 22, 2021, the distribution of vaccines in Denmark was: 85% Comirnaty, 14% Spikevax, 1% Janssen, and approximately 0% AstraZeneca28. All other individuals, including 59 partially vaccinated individuals, were considered non-immunized.

Statistical analysis

The causal effect of family exposure to Omicron VOC rather than Delta VOC on SAR may be confused. This is evidenced by the difference in characteristics between families exposed to Omicron and Delta VOC, the latter of which was more widely prevalent at the beginning of the study period (Table 1 and Appendix Figure S1b). We hypothesize that these differences result from the spatio-temporal patterns of Omicron VOC transport when first introduced. The causal interpretation of our results is conditional on the assumption that all effects of non-random assignment of variables to families are intercepted by the observed family characteristics. Causal assumptions are described in Appendix Section 2.

We defined the secondary attack rate (SAR) as the proportion within the household of household contacts defined as secondary cases16. Adjusted odds ratios (OR) for infection were estimated using multivariate logistic regression models with the binary score of test results for each household contact as the response variable, and the household variable (Omicron vs Delta VOC) as the main explanatory variable of interest. Variables representing age and sex of the initial case, age and sex of the household contact, and family size (2-6 individuals) were included as additional explanatory variables to account for confounding factors. In order to test whether vaccine status confers differential protection against Omicron and Delta VOC, we included an interaction term between vaccination status for primary cases, contacts, and variant. We found no evidence of an interaction between vaccination status of the primary cases and the variant (P = 0.14). In particular, we estimate the following equation:

$${{log}}left(frac {{{Pr}}({y}_{c,p}=1)}{1 – {{Pr}}({y}_{c) , p} = 1)} right) = {{{{{rm{Constant}}}}}} + {{{{{rm {Variant}}}}}}} _{p} + {{{{{rm {VaccineStatu}}}}}} {{{{rm {s}}}}} _{c} \ + {{{{{{rm{Variant}}} }}}}} _{p}times {{{{{rm {VaccineStatu}}}}}} {{{{{rm {s}}}}}}} _{c} + {{{ {{{rm {VaccineStatu}}}}}}} {{{{rm {s}}}}}} _{p} \ + {{{{{rm{Ag}}}} }}} {{{{{rm {e}}}}}}_{p} + {{{{{rm{Sex})}}}}}}_{p} + {{{{{}} {rm{HouseholdSiz}}}}}}{{{{{rm {e}}}}}_{p} + {{{{rm{Ag}}}}}}{{{ {{{rm {e}}}}} _{c} + {{{{{rm {Sex}}}}}}}_{c}, $$


where yc, p equals one if the contact c Tested positive 1-7 days after exposure to the initial case s, and zero otherwise. a variables Determines whether the initial state s Hit by omicron or delta. Vaccine status represents the fixed effects of vaccination status (categorical variable) for the initial condition s and contact c. Age represents the fixed effects of age at 10-year intervals (categorical variable), gender represents the fixed effects of sex, and family size represents the fixed effects of family size (categorical variable). Cluster strong standard errors with family-level aggregation using Taylor series linearity were used to estimate the covariance matrix of the regression coefficients.29.

We also performed a number of complementary analyzes to support the main analysis. To test the robustness of the results, we compared different specifications of the main logistic regression model (Appendix, Section 4.4). To examine the potential mediating role of viral load of primary cases infected with Omicron VOC relative to Delta VOC, we plotted distributions of cycle threshold (Ct) values ​​for each variant (Supplementary Figure S3). We also examined to what extent the Ct value for the initial condition could explain the difference in transmission between variables (Annex Table S21). Our study is based on the assumption that we correctly distinguish between primary and secondary cases, and that household secondary cases are infected with the primary case and not from the outside community. To assess this potential misclassification of cases, we performed a series of robustness assays (Appendix Section 4.2). First, to investigate the potential role of the differential occurrence of grade III cases across variables, we compared the RSR across two-person and multi-person households, as tertiary cases are not possible in two-person households. Second, to investigate the possible role of misclassification of primary cases, we took advantage of the fact that a high proportion of contacts in our sample were tested multiple times. Because untested contacts or secondary cases that tested positive in the first test could be the true primary case, we restricted our sample to only families in which all contacts tested negative after the point at which the primary case was tested positive. Third, to investigate the possible misclassification of infected secondary cases by the outside community rather than the family, we estimated the probability of secondary cases having the same variable as the primary case. In addition, to maximize the likelihood of misclassification being identified, we focused on families with a variant different from the most prevalent in the corresponding locus.

Our study also relied on the Variant PCR assay to determine whether each primary case had Omicron or Delta VOC. To check whether there was a sample selection bias for the Variant PCR test, we investigated the sampling probability of Variant PCR by sample Ct value and age (Supplementary Figure S1). Furthermore, we validated the variant by PCR using whole-genome sequencing data (Supplement Table S5). Finally, we tested the durability of household contacts that are tested and only test positive using RT-PCR assays, which have higher sensitivity and specificity than antigen tests (Supplement Figure S7).

moral statement

This study was conducted using data from national registries only. According to Danish law, ethics approval is not required for this type of research. All data management and analyzes were performed on the Danish Health Data Authority restricted research servers with project number FSEID-00004942. The study contains only aggregated results and does not contain personal data.

Report summary

More information about research design is available in the Nature Research Reporting Summary linked to this article.