Vaccine effectiveness after 1^st and 2^nd dose of the BNT162b2 mRNA Covid-19 Vaccine in long-term care facility residents and healthcare workers – a Danish cohort study

Introduction

In Denmark, the national Covid-19 vaccination strategy aims at protecting those at risk of severe outcomes of a SARS-CoV-2 infection resulting in hospitalization or death, and secure key societal functions such as the healthcare system ¹. Hence, residents in long-term care facilities (LTCF) and front-line healthcare workers (HCW) were target groups prioritized for vaccination, when the vaccination program started on 27 December 2020.

On 16 December 2020 (week 51), a nationwide partial lock down was initiated following widespread community transmission of SARS-CoV-2. In week 53 when the immunization program started, there were several reports of Covid-19 outbreaks in LTCF and the positive percentage for tested LTCF residents was 4.21% (Figure 1a) while it was 2.21% for HCW (Figure 1b).

• Download figure
• Open in new tab

• Download figure
• Open in new tab

Figure 1a and 1b: Weekly number of laboratory confirmed SARS-CoV2, number of tests and proportion of positive tests distributed on test weeks for LTCF residents and HCW in Denmark from week 49, 2020 to week 7, 2021

Number of tests include a maximum of one test per day per person until and including the first positive test. Tests of LTCF residents have been identified through The Danish Health and Medicines Authority. Tests of Healthcare workers have been identified through The Danish Agency for Labour Market and Recruitment, in which the current work affiliation is used (updated December 2020).

The aim was to estimate VE for first and second dose of the BNT162b2 mRNA vaccine in two separate cohorts: 1) all LTCF residents, and 2) all HCW, both with no history of SARS-CoV-2, using data from Danish registries.

Methods

Results

A total of 39,040 persons (median age at time of first dose: 84 years, IQR (77-90), 63.6% female) were included in the LTCF residents cohort (Table 1). From 27 December 2020 through 18 February 2021, 37,172 LTCF residents received the first dose of the vaccine and 33,567 received the second dose. The HCW cohort included 331,039 persons (median age at time of first dose: 47 years, IQR (36-57), 82.2% female) (Table 1). During the study period, 91,865 HCW received first dose and 80,839 received second dose of the BNT162b2 mRNA vaccine (Table 1). Days between first and second dose for LTCF residents and HCW were 24 (IQR; 20-52) and 25 (IQR 20-51), respectively. More than 99% of the vaccines administered to LTCF residents were BNT162b2 mRNA, which was the case for 89% in HCW (on 18 February 2021).

No significant VE was observed in LTCF residents at any time point between first and second dose after adjusting for calendar time (Table 2). From day 0-7 after second dose, the adjusted VE (VE_adj) was 52% (95% CI; 27-69) in LTCF residents and after 7 days after second dose it increased to 64% (95% CI; 14-84) for this group (Table 2). A moderate VE_adj of 17% (95% CI; 4-28) was observed in HCW from 14 days after first vaccination until second dose (Table 2). VE_adj for HCW was 46% (95% CI; 28-59) in day 0-7 after second dose and increased to 90% (95% CI; 82-95) from 7 days after second dose. (Table 2). Adjusting for age, sex and comorbidities did not affect the VE estimates significantly (data not shown). The unadjusted VE was markedly higher than VE adjusted for calendar time, both among LTCF residents and HCW (Table 2).

Discussion

For LTCF residents, we found an adjusted VE of 52% (95% CI; 27-69) within 0-7 days after second dose, which increased to 64% (95% CI; 14-84) from 7 days after second dose. In HCW, we observed the same pattern; an adjusted VE of 46% (95% CI; 28-59) in day 0-7 that increased to 90% (95% CI; 82-95) after 7 days following the second dose.

It is very reassuring that we find high VEs in a real-world setting where one cohort had a median age of 84 years at time of first vaccination The findings for the LTCF cohort are encouraging as this group comprises the most vulnerable part of the population and immunosenescence in the elderly population is generally a challenge for vaccines ¹².

When the vaccination program was initiated, the national Covid-19 incidence was at the highest level since the beginning of the pandemic and the proportion of LTCF residents tested positive for Covid-19 was increasing (Figure 1a). However, it is worth noting that a partial lockdown had been initiated to reduce circulation of Covid-19 a few weeks before the immunization program started. These interventions introduced consecutively might explain the higher VE observed in the unadjusted analysis. The calendar time adjusted analysis prevents bias related to the comparison of unvaccinated Covid-19 cases observed at the start of the vaccination campaign (characterized by high Covid-19) with Covid-19 cases observed later on where the Covid-19 incidence had decreased.

When vaccination started, some LTCF residences were dealing with Covid-19 outbreaks, why vaccination was postponed in these facilities. This selection for vaccination based on Covid-19 outbreaks may potentially inflate the observed VE. The effects of the national lockdown implemented before Christmas became visual in the beginning of January 2021, where positive test rates decreased among LTCF residents and HCW (Figure 1a and 1b). Therefore, the rate of infection decreased during the study period. The high vaccination coverage achieved in LTCF may have decreased risk of infection within the facilities, as they are relatively closed settings.

The main strengths of this study was that the two population-based cohorts of LTCF residents and HCW were followed using national, high quality information on vaccination status and laboratory SARS-CoV-2 positive samples. Due to the situation at nursing homes, there was increased focus on testing in LTCF. Furthermore, there was a national testing strategy, where HCW were offered weekly PCR-testing. Therefore, we believe that our study includes close to all cases of SARS-CoV-2. Further, our cases included both asymptomatic and symptomatic disease based on all conducted laboratory PCR SARS-CoV-2 tests, nationally. Test results from rapid antigen tests and PCR tests at private test centers were not available. However, they constitute a small fraction of test conducted in the study cohorts. Antigen tests were only offered in limited extent to some HCWs.

The adjusted VE for the HCW cohort presented here, is similar to the vaccine efficacy of 95% (95% CI; 90-98) reported by Polack et al. ¹⁰ in a phase III trial including 43,448 participants with a median age of 52 at time of vaccination. Another interesting finding that aligned with the phase III trial was the unaffected VE estimates after adjustment for age, sex and comorbid conditions, which may be explained by these populations assumingly being relatively young and healthy at time of vaccination ¹⁰. Expectedly, the estimated VE among LTCF residents was lower (VE; 64%, 95% CI; 14-84) than the VE reported by Polack and colleagues. This could be explained by the higher vulnerability and age distribution in our cohort, a median age of 84 years in LTCF residents compared to 52 years among the trial participants. It is also evident from studies of influenza vaccines, that vaccines are less effective in the elderly ¹². We observed a moderate, statistically significant, VE for healthcare workers of 16% for the period from 14 days after first dose until second dose. A similar, though higher, VE was found in a study from Israel at day 13-24 after first dose (VE; 51%, 95% CI; 7.2-78). The results from the two studies suggest a possible moderate effect after first dose. Hunter et al. (UK) reanalyzed the data presented by Chodick et al. using a different approach ¹³, and they found an increasing VE from day 14 after first dose, peaking at day 21 at 90%, suggesting a very high level of immunity even after first dose of the vaccine ¹³. Recently, another Israeli study have presented VEs for HCW on day 1-14 after first dose compared to day 15-24 in vaccinated compared to unvaccinated persons. This study observed a VE of 75% in the latter time period and 35% in the first period. The authors note that 55% of the study population had received second dose on day 21-22, but still interpret the findings as related to first dose.

We observed no or low VE after just one dose. However, it is important to note that there was a short interval (median 24 and 25 days) between first and second dose, and therefore it is unlikely that the full impact of a single dose can be extrapolated from our findings. Our findings compared to findings from studies including only first dose corroborate that a two-dose schedule to obtain maximum benefit of the BNT162b2 mRNA vaccine is important. Nonetheless, studies with a longer follow-up between first and second dose are warranted in order to explore potential advantages of a strategy with a postponed second dose.

The cohorts included in this study are two different segments of the Danish population. The findings on VE in LTCF residents may apply to other vulnerable elderly populations with similar conditions. This is the first study to report VE in LTCF residents, an age group not included in previous studies, where the populations were relatively young. The estimated VEs for HCW might be more generalizable as this population in many ways is comparable with other subpopulations. HCW have special knowledge on importance of protective equipment, why the VE in this group may not extrapolate to all populations in the same age groups. On the other hand, HCW have an increased risk of infection, through patient contacts, that may be higher than in the general population.

All in all, the present study confirms, from the field, that two doses of BNT162b2 mRNA vaccine offers protection against SARS-CoV-2 infections in two critical groups, albeit with a better VE among HCW with a median age of 47 years than among a vulnerable population of long-term-care facility residents with a median age of 84 years.

Conflicts of interest

Authors declare no conflicts of interest.