Product-specific COVID-19 vaccine effectiveness against secondary infection in close contacts, Navarre, Spain, April to August 2021

COVID-19 vaccine effectiveness by product (two doses Comirnaty, Spikevax or Vaxzevria and one of Janssen), against infection ranged from 50% (95% CI: 42 to 57) for Janssen to 86% (70 to 93) for Vaxzevria-Comirnaty combination; among ≥ 60 year-olds, from 17% (−26 to 45) for Janssen to 68% (48 to 80) for Spikevax; and against hospitalisation from 74% (43 to 88) for Janssen to > 90% for other products. Two doses of vaccine were highly effective against hospitalisation, but suboptimal for infection control.


Identification of close contacts and information sources
In Spain, to control COVID-19 transmission, all laboratory-confirmed COVID-19 cases are interviewed to detect their close contacts [5]. Close contacts are defined as any person who either had high-risk exposure to a confirmed COVID-19 case within a timeframe ranging from 2 days before the onset of symptoms in the case to 10 days after the onset of symptoms, or, for asymptomatic cases, in the 2 days before the sample which led to confirmation was taken, to 10 days after the sample was taken [6]. Close contacts are encouraged to be tested immediately and 7 to 10 days after the last exposure by RT-PCR. In symptomatic contacts, a positive antigen test within 5 days from the symptom onset is also considered a confirmed infection [7].
In our study, we excluded close contacts with a prior positive test, nursing home residents and those who did not complete the testing protocol. Age, sex, chronic conditions and contact setting were obtained from the enhanced epidemiological surveillance of COVID-19 [5]. Additional information about methods is described in the Supplementary Material.

Vaccine products administered
In Spain, the COVID-19 vaccination campaign targeted adults starting with older age groups and gradually moving towards younger ones [8]. Although the product administered mainly depended on the availability at the particular moment, people aged ≥ 70 years were mainly vaccinated with Comirnaty or Spikevax, and those aged 60-69 years were mainly vaccinated with Vaxzevria. All products were used in people from 18 to 59 years of age; however, those who received the first dose of Vaxzevria could choose the second dose between Vaxzevria or Comirnaty. With this exception, all people received the same product for the second dose.
Information about COVID-19 vaccine doses, product and date of administration were obtained from the regional vaccination register. We considered a person fully vaccinated ≥ 14 days after receiving one dose of Janssen or the second dose of other vaccines, and partially vaccinated ≥ 14 days after receiving only the first dose of Spikevax, Comirnaty or Vaxzevria.

Statistical analysis of product-specific vaccine effectiveness
We compared the incidence of SARS-CoV-2 infection, of symptomatic cases and of COVID-19 hospitalisation by product-specific COVID-19 vaccination status as a variable with nine categories, with unvaccinated people as reference category. The same risk period was assigned to everyone in the cohort; therefore, the Cox regression provided estimates of the crude and adjusted relative risks (RR) with 95% confidence intervals (CI). Adjusted models included age groups, sex, chronic conditions, contact setting, month and COVID-19 vaccination status of the index case.

Estimates of product-specific vaccine effectiveness
The study cohort included 30,240 close contacts of 12,263 index cases. In the cohort of close contacts, 7,177 (23.7%) SARS-CoV-2 infections were confirmed and 263 (3.7%) of them led to hospitalisation. Characteristics of the study population by outcomes and product-specific vaccination status are shown in Table 1 and Supplementary Tables S1 and S2.
The effectiveness of full vaccination in preventing SARS-CoV-2 infection was higher in people aged 18-59 years than in those aged ≥ 60 years for all vaccine products. Among people aged ≥ 60 years, full vaccination estimates ranged from 17% (95% CI: − 26 to 45) for Janssen to 68% (95% CI: 48 to 80) for Spikevax ( Table  3).
The VE estimates did not change substantially when close contacts of vaccinated index cases were excluded or only household contacts were considered ( Table 4). The time since the last dose was < 90 days for most vaccinated close contacts. The VE declined 90 days after the second dose of Spikevax (85% to 67%; p comparison = 0.003) and Comirnaty (70% to 63%; p comparison = 0.035), but not 90 days after the first dose of Vaxzevria (40% to 52%; p comparison = 0.746) (Supplementary Table S3).
Several VE estimates seemed slightly lower against the Delta variant than against the Alpha variant; however, all products maintained a notable protective effect against both variants (Supplementary Table S4).

Ethical statement
This study was approved by the Navarre's Ethical Committee for Clinical Research (PI2020/45), which waived the requirement of obtaining informed consent.  and vaccination status of index case. These analyses estimate the vaccine effectiveness by product, but product comparison may not be valid.

Discussion
The results show a clear effectiveness of the full COVID-19 vaccination with each of the four evaluated vaccine products and highlight some improvement points. Our findings are consistent with previous studies that reported high effectiveness of full COVID-19 vaccination in preventing hospitalisations [9][10][11] and moderate VE in preventing symptomatic and asymptomatic SARS-CoV-2 infection [12,13]. However, few studies evaluated VE for specific products [11,14].
Two doses of mRNA vaccine or the heterologous vector/mRNA vaccination provided high protection against infection in the population < 60 years of age. However, a decrease in the VE was observed from 90 days after the last dose that raises doubts about the duration of this effect.
The COVID-19 VE for the same product was lower in the elderly people. Therefore, full vaccination effect seemed suboptimal to prevent infection in this age group and to achieve transmission control in locations with a high proportion of older population as has been shown in other studies [10,12]. Younger age groups vaccinated with one dose of Janssen or two doses of Vaxzevria had also a suboptimal protection against SARS-CoV-2 infection [13].
Our results are consistent with the stronger immunological response observed in individuals who received the Vaxzevria-Comirnaty combination than those who received two doses of Vaxzevria [15,16]. This supports the recommendation of this heterologous vaccination in Spain and several other countries [8,17].
Although the COVID-19 VE may be slightly lower against the Alpha and Delta variants, full vaccination remains notably protective against these variants of concern as also shown by Lopez Bernal et al. [18].
Regardless of the product, two-dose vaccination remained highly effective in preventing hospitalisation; however, the VE of one dose of Janssen seemed slightly lower. The suboptimal VE in preventing infection in fully vaccinated people, because of older age, vaccine type or time from the last dose, may be insufficient for infection control. A booster dose of vaccine in some cases has been suggested as a possible solution [19,20].
The present study has advantages of the cohort design while all cases were laboratory-confirmed and non-cases were those who tested negative for SARS-CoV-2 as in the test-negative design and the analysis of close contacts ensures similar exposure to infection [12,13,21]. However, this study also has some limitations. Vaccination coverage varied with age, comorbidity and month; nevertheless, analyses were adjusted for these variables. Since characteristics of people who received each vaccine product were different, specific estimates can be considered separately and care must be taken in comparisons. The correlation among close contacts of the same index case was not controlled for. Epidemiological and vaccination conditions of this study may not be similar to other sites. Estimates refer to the first 8 months since vaccination and VE may wane over time.

Conclusions
Regardless of the product, two vaccine doses were highly effective against hospitalisation, and mRNA or heterologous vaccination provided high protection within 90 days against SARS-CoV-2 infection in those younger than 60 years of age. However, protection against infection was suboptimal for vector-based vaccines in adults and for all products among people ≥ 60 years old. Additional preventive measures should be maintained while SARS-CoV-2 is circulating to interrupt transmission in the population and product-specific VE should be monitored in the long term.