We obtained information on COVID-19 patients and household contacts reported in the jurisdiction of the Itako PHC in Ibaraki Prefecture, Japan, during April 21–30, 2022. The epidemiologic investigation and data collection procedure on COVID-19 index patients was almost the same as that described in our previous study (5). The study design was approved by the Ibaraki Prefecture Epidemiologic Research Joint Ethics Review Committee (protocol no. R3-10). We performed statistical analyses by using R version 4.4.1 (The R Foundation for Statistical Computing, https://www.r-project.org).

Genomic sequencing surveillance showed that 90% of 5,630 variants of concern sampled in Japan during April 18–May 1 were BA.2 (6). In Ibaraki, sequencing of 69 samples collected during April 23–May 13 demonstrated that 65 (94%) samples were BA.2 and 4 (6%) were BA.1.

Figure

Figure. Flowchart of enrollment of index patients and household contacts in study of vaccine effectiveness against SARS-CoV-2 among household contacts during Omicron BA.2–dominant period, Japan.

Itako PHC undertook or requested SARS-CoV-2 testing of household contacts who had any symptoms during their 5-day observational and quarantine period of a COVID-19 patient (5,7,8). Asymptomatic household contacts were not tested. During the study period, the PHC jurisdiction identified 729 confirmed COVID-19 cases (4) (Figure). After excluding 112 index cases without household members, we identified 617 infected household members, and a total of 1,575 household members, including cases and contacts. We deemed patients with the earliest symptom onset date as index cases in the household, which comprised 388 patients. Of 1,187 household contacts of the 388 index patients, we included 952 who had available data and no history of previous infection in the study. Of the 235 excluded household contacts, we excluded 7 because of a history of infection before the study period. Of the 952 contacts, 395 were infected; the overall HCAR was 41.5%.

Excluding health professionals, 82.5% of residents in Ibaraki had completed 2 COVID-19 vaccine doses, and 56.3% had received a booster dose by April 20, 2022 (9). Among initial vaccinations in Ibaraki, 81% were BNT162b2 vaccine (Pfizer-BioNTech, https://www.pfizer.com) and 19% mRNA-1273 vaccine (Moderna, https://www.modernatx.com) (5).

We stratified vaccination status on the basis of vaccination doses or recent vaccination, which we defined as receiving the last vaccine dose within 3 months. HCAR was 30.9% for contacts with 3-dose vaccination and 32.4% for contacts with recent vaccination, substantially lower than the 49.6% for contacts without complete vaccination (0–1 doses) and 46.0%, without recent vaccination (Table 1). Of the 307 contacts with 3-dose vaccination status, 272 had been vaccinated within the previous 3 months.

The HCAR for contacts of an index patient who had a 2-dose vaccination status was 37.2% and HCAR was 33.1% for contacts of index cases with a 3-dose status. Those results were substantially lower than the 46.2% for contacts of an index patient without complete vaccination status. However, the HCAR for contacts of an index patient with recent vaccination was not substantially lower than that for contacts of an index patient without recent vaccination (Table 1).

Among household contacts who had a 2-dose vaccination status, HCAR was 34.8% for contacts of index patients with complete vaccination but was 49.5% for contacts of index patients without complete vaccination status. However, among household contacts with 3-dose vaccination and without complete vaccination, the HCAR was not much lower for contacts of index patients with complete vaccination than for contacts of index patients without complete vaccination status (Table 2). The HCAR among household contacts without recent vaccination was 37.9% for contacts of index patients with complete vaccination, compared with 52.1% for contacts of index patients without complete vaccination (relative risk 0.73 [95% CI 0.61–0.87]); VE 0.27 [95% CI 0.13–0.39]) (Table 2). We obtained similar results in multivariate analyses (Appendix Table). Among household contacts with recent vaccination, HCAR was not substantially different for contacts of index patients with complete vaccination status and contacts of patients without complete vaccination status (Table 2).

Our study found the HCAR among household contacts of index SARS-CoV-2 Omicron BA.2 patients with complete vaccination was substantially lower than that among contacts of index patients without complete vaccination. In our previous study on the Omicron BA.1 variant–dominant period, the VE was 0.43 against infectiousness (5). Several other previous studies also reported VE against Omicron infectiousness (10–12); however, the vaccination doses of index patients in those studies might have been associated with the vaccination doses of their household contacts.

Among household contacts without recent vaccination in our study, HCAR was lower for contacts of index patients with complete vaccination status than for contacts of index patients without complete vaccination, a VE of 0.27. However, among household contacts with recent vaccination, we noted no major effect of vaccination completeness for index patients.

VE was greatly affected by vaccination completeness for index patients among household contacts with 2-dose vaccination but not among household contacts with 3-dose vaccination. Another study in Japan also showed relatively low VE against infectiousness among contacts with 3-dose vaccination compared with VE against infectiousness among other contacts (13).

In this study, the HCAR for contacts of an index patient with recent vaccination was not much lower than for contacts of an index patient without recent vaccination. Previous studies have reported relatively stable VE against infectiousness compared with VE against susceptibility (12,14).

The first limitation of this study is that we did not adjust for confounding factors, such as demographic data, for analyses on all household contacts. The second limitation is that we did not confirm specific variant types with genomic sequencing and different variants might have had different infection rates. Finally, our assessment of risk factors and potential confounders might be limited to the data reported to the PHCs.

In summary, our results indicate that the SARS-CoV-2 Omicron BA.2 attack rate among household contacts without recent vaccination was lower for household contacts of vaccinated index patients than for contacts of unvaccinated index patients. These results strengthen data on indirect effectiveness of COVID-19 vaccines and the value of continued vaccine campaigns to prevent severe illness among index cases and household contacts.