Skip to main content

Our take —

This retrospective study, available as a preprint and thus not yet peer-reviewed, assessed whether the emergence of the Omicron SARS-CoV-2 variant of concern was associated with an increased risk of reinfection using data for all positive test results in South Africa between March 4, 2020, and November 27, 2021 (n= 2,796,982). Suspected reinfection (two positive tests >90 days apart) was identified for 35,670 participants. Two statistical models were used to assess changes in risk—one based on a constant hazard of reinfection throughout the study period, and one based on time-varying hazards of primary infection and reinfection. Both approaches demonstrated that the emergence of the Omicron variant was associated with an increased risk of reinfection. For the second model, the hazard ratio for reinfection vs. primary infection following the introduction of the Omicron variant was 2.39 (95% CI: 1.88–3.11), demonstrating population-level evidence of immune escape, a finding not observed with emergence of either Beta or Delta variants. This study demonstrates the risk that the Omicron variant poses for reinfection in the setting of high rates of prior SARS-CoV-2 infection. Additional research is urgently needed to assess the ability of prior infection and/or vaccination to protect against severe disease and death from Omicron.

Study design

Retrospective Cohort

Study population and setting

This study examined whether the emergence of the Beta, Delta, and Omicron SARS-CoV-2 variants of concern led to increased risk of reinfection in South Africa. Study data included all laboratory-confirmed SARS-CoV-2 positive test results received by the Notifiable Medical Conditions Surveillance System (NMC-SS) between March 4, 2020, and November 27, 2021 (n= 2,796,982). For the sake of consistency and data completeness, the date that each specimen was received in the lab was used as a proxy for calendar date of infection. Reinfection was defined as two positive SARS-CoV-2 tests at least 90 days apart for the same participant. Two statistical approaches were used to assess changes in risk of reinfection over time. In the first approach, observed reinfection patterns were compared to a null model that assumed risk of reinfection was proportional to overall incidence with a constant hazard coefficient. In the second approach, empirical hazard coefficients for primary infection vs. reinfection were calculated at each time point and compared over time. A sensitivity analysis was conducted to assess the impact of vaccination.

Summary of Main Findings

The study population included 35,670 participants with at least two suspected SARS-CoV-2 infections. Using the first approach (constant hazard of reinfection), the emergence of the Omicron variant was associated with an increase in observed vs. projected reinfections, which is a signature of immune escape. Similar deviations from the null model were not observed after the introduction of the Beta or Delta variants. Using the second approach (time-varying hazards of primary infection and reinfection), the emergence of the Beta and Delta variants was associated with an increased hazard of primary infection without corresponding changes to the hazard of reinfection. In contrast, the introduction of the Omicron variant was association with a decreased hazard of primary infection and an increased hazard of reinfection; the hazard ratio for reinfection vs. primary infection in November 2021 was 2.39 (CI: 1.88–3.11). Sensitivity analyses demonstrated that vaccination may be partially responsible for the observed decline in the hazard of primary infection, but vaccine coverage during the study period was quite low.

Study Strengths

This study analyzed a large, comprehensive national data set for all confirmed SARS-CoV-2 cases in South Africa. Results were consistent between both statistical approaches used.

Limitations

These analyses do not account for changes in testing practices, rates of detection, participant behavior, participant demographics (age, occupation, socioeconomic status), or healthcare access which may have occurred over the course of the pandemic. Results from rapid antigen tests may be underreported despite mandatory reporting requirements, so the study may underestimate rates of both primary infection and reinfection. Civil unrest in July 2021 in Gauteng and KwaZulu-Natal provinces may have also led to underreporting of test results and underestimation of infections during this period. Reinfection was not confirmed by sequencing and did not require negative test results between primary infection and reinfection. Symptom severity in primary infection vs. reinfection was not assessed because these data were not collected. While vaccine coverage in South Africa was low during the study period, vaccination may have impacted hazards of both primary infection and reinfection. The vaccination status of individual study participants was unknown.

Value added

This study demonstrated early population-level evidence of an increased risk of SARS-CoV-2 reinfection immediately following the emergence of the Omicron variant in South Africa.

Our take —

This preprint, which has yet to undergo peer-review, examined changes in menstrual period timing and flow after COVID-19 vaccination in 1,273 adults in the United Kingdom. About half of the surveyed individuals reported changes in period timing (earlier or later than expected) and flow (lighter or heavier than expected) in their subsequent period after COVID-19 vaccination, though the study population is self-selected and the author notes that the frequency of menstrual changes after COVID-19 vaccination is likely overreported and does not represent the prevalence of menstrual changes after vaccination. Vaccine type did not impact menstrual changes, but participants on hormonal contraceptives were more likely to report changes in menstrual flow. Prospective studies with control groups are needed to estimate the frequency of menstrual changes following COVID-19 vaccination compared to baseline changes over time, and future vaccine trials should incorporate menstrual changes as a reportable side effect.

Study design

Retrospective Cohort

Study population and setting

This study described menstrual changes in a self-selected sample of adults (18+ years) in the United Kingdom (UK) who had received vaccination against SARS-CoV-2. A total of 2,241 participants reported their age, length of their normal menstrual cycle, contraceptive use, breastfeeding status, history of gynecologic conditions, vaccine type, vaccine date during their menstrual cycle, and the timing and flow of their subsequent period after each vaccination dose (doses were administered approximately 8 weeks apart in the UK regardless of vaccine type) on an online form. Among these respondents 1,273 (57%) provided complete data and were included in the analysis. The study assessed the timing and flow of each participant’s next period after vaccination based on the following variables: vaccine brand (Pfizer, Moderna, AstraZeneca); use of hormonal contraceptives; and vaccination timing during each participant’s menstrual period using Chi squared tests. It also assessed how subsequent period changes after the first vaccine dose related to period changes after the second vaccine dose. In a post-hoc analysis, it assessed differences in period timing and flow in participants with several gynecologic conditions. All analyses were unadjusted for potential confounders.

Summary of Main Findings

The participants’ median age was 33 years old (interquartile range [IQR] 29-39), the median cycle length was 28 days (IQR: 27-30), and the majority (1,117, 87.6%) were not using hormonal contraception. Among the participants who were vaccinated with Pfizer (n = 778, 61%), AstraZeneca (346, 27.1%), or Moderna (n = 136, 10.7%) there were no significant differences between subsequent period timing (late 29-35%, on time 40-47%, early 24-27%) or flow (heavier 31-35%, the same 49-55%, lighter 13-17%). While there were no statistically significant changes in subsequent period timing by hormonal contraceptive use (23-28% early, 43-45% on time, 29-32% heavier), participants on hormonal contraceptives were more likely to report lighter (19% vs. 14%) or heavier (42% vs. 32%) flow compared to participants not on hormonal contraceptives (p-value = 0.001). Vaccination timing during the menstrual cycle (relative to predicted ovulation dates among participants not on hormonal contraception) did not have a clear effect on subsequent period timing or flow. Among the subset of 813 respondents who had completed both vaccine doses, menstrual changes following vaccine dose one were very similar to changes following the second vaccine dose. Finally, among the small subset of individuals with endometriosis (n = 60, 4.7%) and polycystic ovarian syndrome (n = 87, 6.8%), vaccination was statistically significantly associated with earlier (38% vs. 23%) and later (41% vs. 31%) periods respectively, compared to participants with no history of gynecologic conditions.

Study Strengths

This study provided thoughtful explanations for who was included (and why they were included) in each analysis. It also provided a link to a pre-specified analysis plan and a de-identified dataset.

Limitations

This study notes that it likely overestimates menstrual changes after COVID-19 vaccination given that people who experienced changes were more likely to respond to the survey. It therefore cannot be used to estimate the prevalence of menstrual changes following COVID-19 vaccination. This bias may also influence why participants using hormonal contraception reported more menstrual changes than non-users. There is also insufficient detail about participant recruitment to determine the temporal period of data collection or how representative the study population is of people who are menstruating. For example, 87.6% were not using hormonal contraception, which may reflect that this study population is actively attempting pregnancy. Despite providing reasoned explanations of which characteristics qualified participants for each analysis, this study does not provide information on how many participants were included in each analysis. Finally, each factor was evaluated without adjusting for factors which may confound their association with menstrual cycle timing and flow.

Value added

This study addresses an underreported side effect of COVID-19 vaccination, temporary changes in menstrual timing and flow.

Our take —

This preprint, which has yet to be peer reviewed, retrospectively examined 240,648 individuals in the US who were diagnosed with COVID-19 before May 31, 2021 and evaluated the effectiveness of COVID-19 vaccination both before and up to 12 weeks after diagnosis on preventing the occurrence of long COVID associated illness or symptoms as defined by COVID-19 interoperability alliance. 12-20 weeks following acute COVID-19 diagnosis. Most individuals in the study (91.6%) were unvaccinated within 12 weeks after their COVID-19 diagnosis, 7.4% (17796) were vaccinated within 12 weeks of their diagnosis and only 1% (2392) were vaccinated prior their diagnosis. Compared to unvaccinated individuals, the authors observed a protective effect of COVID-19 vaccination in reducing the likelihood of long COVID-19 continuing symptoms occurring within 12 to 20 weeks following diagnosis, with a greater reduction in odds with shorter durations of vaccination after COVID-19 diagnosis. On the other hand, those who were vaccinated prior to their diagnosis had the lowest risk of experiencing long COVID-19 12-20 weeks after diagnosis (OR: 0.22, 95%CI 0.196-0.245). Although the authors relied on sets of symptoms and conditions that have not been fully validated as associated with long COVID, this study provides additional evidence that vaccination provides some protection against the occurrence of clinical symptoms associated with long COVID 12-20 weeks following diagnosis. 

Study design

Retrospective Cohort

Study population and setting

This retrospective cohort study evaluated the role of vaccination (Pfizer, Moderna, and Johnson and Johnson) on reducing the likelihood of experiencing long COVID-19 symptoms, defined as the presence of at least one COVID-19-associated symptom (fever, loss of taste or smell, etc) between 12 and 20 weeks after initial diagnosis. The study used the Arcadia Data Research Database to identify individuals in the U.S. who had SARS-CoV-2 infection at least 20 weeks before May 21, 2021, the deadline for data extraction. Individuals had to have had at least one health care visit before and after Jan 1, 2020.  The primary exposure of interest was timing of the first dose of a COVID vaccine relative to the date of COVID-19 diagnosis: prior to diagnosis, 0-4 weeks, 4-8 weeks, 8-12 weeks, and more than 12 weeks post-diagnosis, with unvaccinated as the reference group. The outcome was defined as the occurrence of at least one or an aggregate of COVID-19-associated symptoms 12-20 weeks after diagnosis. The following additional covariates were collected for each individual: age, sex, race, ethnicity, insurance type, and comorbid conditions prior to COVID-19 diagnosis. Multiple logistic regression models were used to estimate the association between vaccination timing and the presence of any long COVID-19 symptom. Sensitivity analyses examined the association between timing of vaccination as a continuous metric with a count of the number of long-COVID symptoms 12 weeks post-diagnosis. 

Summary of Main Findings

Between Feb 2020 and May 2021, 240,648 individuals were diagnosed with COVID-19 and had at least one primary care visit both before January 1, 2020 and at least 20 weeks after their diagnosis. Of these, 91.6% (n=220,460) hadn’t received the first dose of a COVID-19 vaccine 12 weeks after their COVID-19 diagnosis, while 17.4% (n=17,796) were vaccinated within 12 weeks and 1% (n=2,392) were vaccinated prior to their COVID-19 diagnosis. Patients vaccinated before a COVID-19 diagnosis were 78% less likely to have any long COVID symptoms than unvaccinated individuals (OR: 0.22, 95%CI 0.196-0.245). Previously unvaccinated individuals who were vaccinated 0-4 weeks, 4-8 weeks, and 8-12 weeks after their COVID-19 diagnosis also had reduced odds of reporting any long COVID-19 symptoms compared to unvaccinated individuals (OR: 0.382, 95%CI 0.353-0.413; OR: 0.535, 95%CI 0.506-0.567; and OR: 0.747,95%CI 0.713-0.784 respectively). 

Study Strengths

This study used an extensive, broadly-representative database to assess the effectiveness of the timing of COVID-19 vaccination on preventing long-COVID symptoms 12-20 weeks after diagnosis. 

Limitations

Only 7.4% of the total study cohort had any vaccination within 12 weeks after their date of COVID-19 diagnosis. The authors did not consider potential confounding by other notable confounding variables such as pre-existing co-morbidities, which may contribute to the likelihood of COVID-19 symptoms and may be differentially distributed across vaccinated and unvaccinated individuals. Extrapolation, inaccurate inference, and residual confounding could have occurred if there is a big difference in the distribution of the confounding variables among these groups. Finally, no distinction was made between the 3 major COVID-19 vaccines available in the US. 

Value added

The study was one of the larger, broadly-representative study suggesting a protective effect of vaccination with either Pfizer, Moderna, or Johnson and Johnson on preventing long COVID-19 symptoms. 

Our take —

This matched case-control study from Israel, taking place between July 30, 2020 and September 23, 2021, found that a third booster dose of the Pfizer BNT162b2 vaccine was 93% (95% CI, 88-97) more effective in protecting individuals against COVID-19 hospitalizations, 92% (95% CI, 82-97) more effective against severe disease, and 81% (95% CI, 59-97) more effective against COVID-19 related death compared to two vaccine doses received at least five months prior. Severe disease was defined as having SpO2 <94% on room air at sea level, PaO2/FiO2 30 breaths/min, or lung infiltrates >50%, according to US National Institutes of Health criteria. However, these outcomes were rare among individuals in both groups, with only 53 combined hospital admissions, cases of severe disease, and deaths in the three-dose group and 432 combined in the two-dose group. Individuals at high risk for severity and death, including the elderly and those with pre-existing comorbidities, benefitted most from a third booster dose.

Study design

Retrospective Cohort

Study population and setting

Due to many countries experiencing a resurgence of COVID-19 which may be associated with loosening of restrictions, possible waning immunity over time, and the Delta variant wave, some are beginning to administer third doses of vaccine. This retrospective study, designed to emulate a target trial, studied the effectiveness of a third/booster dose of Pfizer’s BNT162b2 vaccine in preventing severe COVID-19 outcomes. Data from Clalit Health Services, the largest healthcare provider in Israel, was examined from between July 30, 2020 and September 23, 2021. Individuals who received a third dose between these dates were matched with individuals who were demographically and clinically similar and had only received two doses. Controls who later received a third dose on a future date then became eligible to be “recruited” to the third dose group. All participants had received their second dose at least five months prior to their recruitment date, had no previous documented SARS-CoV-2 infection, and no contact with the healthcare system in the three days prior to recruitment. Individuals had to be eligible to receive a third dose at least one of the days during the study period in order to be included. Immunocompromised individuals as well as people who were healthcare workers, living in long-term care facilities, or were medically confined to their homes were excluded. The primary outcomes of this study were COVID-19 related hospital admission, severe disease, and COVID-19 related death. Secondary outcomes were SARS-CoV-2 infection confirmed by PCR test and symptomatic infection. All outcomes were assessed starting seven days after administration of the third dose. Third dose effectiveness for each outcome was estimated by calculating 1 – risk ratio using a Kaplan-Meier estimator.

Summary of Main Findings

Both the third dose and control groups included 728,321 individuals after matching. In both the booster and control groups, cases of hospital admission, sever COVID-19 and death were incredibly rare. The effectiveness of the third dose compared to two doses of BNT162b2 vaccine against hospital admission was estimated to be 93% (95% CI, 88-97), 92% (95% CI, 82-97) against severe disease, and 81% (95% CI, 59-97) against COVID-19 related death. The incidence of hospitalization began to diverge between the two groups around six days following the third dose. There was also a divergence seen between the groups at 8-9 days following the third dose in the rates of severe disease and COVID-19 related deaths. The third dose effectiveness against documented infection was estimated to be 88% (95% CI, 87-90) and 91% (95% CI, 89-92) for symptomatic infection. It was also observed that the incidence trends began to decline in each age group shortly after the third dose was approved for that particular age group compared to other ineligible groups. The vast majority of the protective effect of a third dose was seen in older (>70 years) individuals and people with multiple co-morbidities, whereas efficacy in younger healthier people, while significant, was limited in magnitude.

Study Strengths

This study included patients with many common comorbidities and COVID-19 risk factors as defined by the CDC, which is reflective of the general population. The controls were also matched by a wide-ranging list of criteria, including age, sex, place of residence, number of pre-existing comorbidities/risk factors, month that they received their second vaccine dose, and the number of SARS-CoV-2 PCR tests they had performed in the nine months prior to their index date. The latter two criteria served as indicators of health-seeking behaviors.

Limitations

This study had several limitations that should be noted. First, the optimal time in order to achieve maximum protection from a third vaccine dose is not known. The estimated effectiveness here was calculated starting from seven days after receipt of the third dose, but it is possible that protection could begin earlier to some degree. Next, the secondary outcomes of symptomatic infection and PCR confirmed infection could be biased due to differing testing frequencies between groups due to differences in health seeking behaviors, although this was partially accounted for in the matching criteria. Third dose effectiveness could not be estimated in individuals younger than 40 years old due to low numbers of the outcomes of interest. Additionally, this study was performed in a smaller country with a fairly homogenous population compared to other countries. Finally, specific high-risk populations were excluded from the analyses even though they are the groups likely to be targeted to get booster doses first.

Value added

Important data on third dose effectiveness for the Pfizer BNT162b2 vaccine.

Our take —

This study from Israel describes rates of breakthrough COVID-19 cases in July of 2021 among people who received two doses of the Pfizer vaccine. Among people older than 60 years, the rate of breakthrough infection was 2.2 times higher among those who were vaccinated six months before the study period compared to those who were vaccinated two months before the study.  Similarly, among people aged 40-59 years, the infection rate was 2.1 times higher among those who were vaccinated four and five months before the study compared to those who were vaccinated two months before the study; among those aged 16-39 years, the rate was 1.6 times higher. In all age groups, the risk of infection increased with time since full vaccination, however, estimated vaccine effectiveness against severe COVID-19 disease only declined modestly in people aged 60 years and older (from 92% to 85%). Although the authors attempted to control for some possible sources of bias, such as differences in rates of PCR testing, they could not adjust for others (e.g., comorbidities, differences in health care access, and differences in risk behaviors), so results should be interpreted cautiously. This study provides further evidence that vaccine induced immunity to infection wanes over time, but protection against severe disease remained strong. 

 

Study design

Retrospective Cohort

Study population and setting

The study was conducted in Israel between July 11 and July 31, 2021 among 4,791,398 residents who had been fully vaccinated (with the Pfizer vaccine) between January 16 and May 31, 2021 with no history of prior SARS-CoV-2 infection. By the start of the study period, the Delta variant accounted for over 98% of all new SARS-CoV-2 infections in the country; July 31, 2021 was selected as the end date of the study because of the rollout of the booster dose.  The study modeled the rate of SARS-CoV-2 infections and severe COVID-19 cases using Poisson regression.  Participants were stratified into the following age categories; 16-39 years, 40-59 years, and 60 years and older. Participants in each age category were grouped into vaccination intervals (using two-week periods) starting from each group vaccination eligibility date. Full vaccination status was defined as 7 days after the second dose of Pfizer vaccine. Severe infections were defined as COVID-19 pneumonia with a respiratory rate of more than 30, oxygen saturation less than 94% on ambient air, or P arterial O2 over FiO2 of <300. An interaction term between each age group and vaccination period was included to evaluate differences in waning immunity by age.  Regressions were adjusted for the following possible confounders: week of infection, the number of PCR tests that were done for each individual before vaccination (to account for possible ascertainment bias), sex, and population groups (general Jewish, Arab, and ultra-Orthodox). In sensitivity analyses, models were adjusted for socioeconomic status, and were restricted to the general Jewish population.

Summary of Main Findings

Across all age groups of fully vaccinated individuals, 13,426 people tested positive for SARS-CoV-2 infection, and 403 had severe COVID-19. In the adjusted model among individuals above the age of 60 years, the rate of SARS-CoV-2 infection was 2.2 times (95% CI: 1.3-3.6) higher among those vaccinated more than six months prior, relative to those who had their vaccine two months before the study.  Similar trends were noted among age groups 16-39 and 40-59 years, with rates of infection 1.6 (95% CI: 1.3-2.0) and 2.1 (95% CI: 1.4-3.0) times higher, respectively, among those who were vaccinated four months or more before the study period compared to those who were vaccinated two months before the study period. Comparing rates of severe COVID-19 among those above the age of 60 years, vaccinated persons six months or more before the study period had 1.8 times (95% CI: 1.1-2.9) the risk of severe disease compared to those vaccinated four months before the study period.  Among individuals below the age of 60 years, there were no statistically significant differences in the rates of severe COVID-19 across different vaccination periods.  In an additional analysis that used an unvaccinated cohort of individuals above the age of 60 years, vaccine efficacy for severe COVID-19 disease declined from 92% to 85% during the vaccination period. and efficacy against SARS-CoV-2 infection across all age groups declined from 82% to 57%. 

Study Strengths

The study accounted for the possibility of differential detection among individuals by adjusting for the number of PCR tests for each individual before the vaccination campaign. Several sensitivity analyses were done, including adjustment for socioeconomic status, and using ten-year age groups. 

Limitations

The high risk of unmeasured confounding should result in caution when interpreting results as solely the consequence of waning immunity. No data regarding comorbid conditions were included in the study, which could have resulted in confounding.  Similarly, health-related behaviors may have differed between early and late vaccine recipients (e.g., mask-wearing, social distancing, etc.). Socioeconomic status was only included as a variable in sensitivity analysis, but this may have affected both exposure and outcome risk. The rates of severe COVID-19 disease among age groups below the age of 60 years were low, which limits the evaluation of vaccine effectiveness against severe COVID-19 disease. Also, the definition of severe COVID-19 disease included patients with oxygen saturation below 94%, which could overestimate disease severity if the prevalence of chronic lung disease was elevated in the study cohort.  

Value added

This study provides observational data that corroborate laboratory-based studies indicating a waning of immunity against SARS-CoV-2 infection after vaccination.

Our take —

This preprint, which has yet to undergo peer-review, used a large-scale network of electronic health records in the US to evaluate the risk of various post- COVID-19 symptoms within 6 months following infection among those vaccinated and not vaccinated for COVID-19. To account for potential differences between those who were and were not vaccinated, the investigators compared those vaccinated against COVID-19 with a comparison group of unvaccinated individuals but who had been vaccinated for influenza in a prior year. These data show that having at least one dose of a COVID-19 vaccine prior to infection was associated with a significantly lower risk of numerous clinical outcomes (each a composite outcome with death), including ICU admission (hazard ratio [HR] 0.75), respiratory failure (HR 0.70), and intubation/ventilation (HR 0.72), seizures (HR 0.73), hypercoagulopathy/venous thromboembolism (HR 0.81), especially among those younger than 60, but not a lower risk of various long-COVID symptoms, including mood, sleep, or anxiety disorders. Fewer benefits of one dose were observed among those 60 or older, however, which supports calls for additional vaccine doses among older individuals. These findings should be interpreted with caution given the observational nature of the study and the measurement uncertainty inherent to electronic health record data.

Study design

Retrospective Cohort

Study population and setting

This study used data from the TriNetX electronic health records network to compare the incidence of COVID-19 sequelae within 6 months following SARS-CoV-2 infection between those who received at least one dose of a COVID-19 vaccine and individuals unvaccinated for COVID-19 but who had received an influenza vaccine. TriNetX Analytics maintains a network of 59 healthcare organizations with data on approximately 81 million people in the US, both insured and uninsured. All study participants had a confirmed SARS-CoV-2 infection (either laboratory-confirmed or by clinical diagnosis) between January 1, 2021 and August 31, 2021. Among the exposed group, infection must have occurred at least 14 days after documented COVID-19 vaccination. To capture individuals who were not vaccine-averse (and thus might be considerably different from the vaccinated individuals in unknown ways), investigators included individuals without a COVID-19 vaccine but who had previously received at least one influenza vaccine in the comparison group. Outcomes consisted of numerous acute and post-acute ICD-10 codes previously associated with COVID-19, including hospitalization, ICU admission, respiratory failure, ventilation, and death, as well as several conditions associated with long-COVID. To achieve balance in baseline characteristics, 1:1 propensity score matching was employed, and Cox proportional survival methods were stratified by 2-month periods to account for potential changes in diagnostics and other temporal variations. Analyses accounted for death as a competing risk. Secondary analyses assessed potential effect modification by age at the time of infection and the impact of 1 versus 2 vaccine doses. Bonferroni corrections were utilized to account for multiple testing.

Summary of Main Findings

10,024 individuals were identified as having had a SARS-CoV-2 infection at least 2 weeks after documented COVID-19 vaccine (mean age: 57 years; 59.4% female). Of these, 9,479 were matched to individuals without a COVID-19 vaccine prior to their infection but who had had at least one previous influenza vaccine. Individuals who received a COVID-19 vaccine prior to infection were significantly less likely to experience deleterious acute outcomes, including death and respiratory failure (composite outcome; hazard ratio [HR] 0.70; Bonferonni-corrected, p<0.0001), ventilation (HR 0.72, p=0.0024), ICU admission (HR 0.75, p<0.0001), venous thromboembolism (HR 0.81, p=0.014), and an oxygen requirement (HR 0.83, p=0.011), than those who hadn’t received a COVID-19 vaccine. No differences between the study groups were observed for the following outcomes: composite of death and any long-COVID symptom, Type 2 diabetes, or mood or anxiety disorders. In secondary analyses, lower risks among those who received 2 doses of a COVID-19 vaccine were observed for additional outcomes, including myocarditis, cerebral hemorrhage, interstitial lung disease, and death, compared to the unvaccinated, although these risks were not statistically different from those who had received only 1 dose. These associations were generally stronger (i.e., vaccines were associated with noticeably lower risks) among those younger than 60 years old compared to those 60 or older. 

Study Strengths

This study examined an extensive list of outcomes, both acute and post-acute, previously reported with COVID-19 and their apparent association with SARS-CoV-2 among those who had and had not been vaccinated against COVID-19 using data from a large-scale electronic health records network. 

Limitations

As noted in the study, electronic health records may be limited by missing data, incomplete and non-standardized records, and inadequate information on contextual factors, such as socioeconomic status and key behavioral factors. Moreover, the investigators were not able to consider the role of variants of concern, compare the different vaccine types on the likelihood of breakthrough infections or their sequelae, or how booster vaccines impact these outcomes. While including death as a composite for each outcome reduced the risk of survivorship bias from competing risks, it made it difficult to assess how SARS-CoV-2 vaccination impacted rare COVID-19 sequelae, such as those related to long-COVID. To account for potential confounding by both known and unknown behavioral factors, the investigators utilized a control group that had been vaccinated for influenza in a prior year but not for COVID-19. While this likely removed some unobserved differences between vaccinated and unvaccinated individuals, given the politicization of vaccines during the COVID-19 pandemic, influenza vaccination in any previous year is an imperfect proxy of openness to vaccination. Finally, this study only included individuals who sought healthcare for their infection and may therefore not generalize to those who did not get tested or seek healthcare assistance. 

Value added

This is one of the first and largest studies to compare the likelihood of a broad array of acute and post-acute sequelae of COVID-19 six months following SARS-CoV-2 infection, between those who were vaccinated vs. those who were not vaccinated for COVID-19. 

Our take —

Since the start of the pandemic, surprisingly little data has emerged on how the risks of secondary SARS-CoV-2 transmission may depend on the masking status of infected individuals and their close contacts. This study, conducted through a contact tracing program in one Iowa county from October 2020 to February 2021, collected information from 431 infected cases and 969 of their close contacts to estimate differences in the transmission risk by masking status. The authors found that when at least one person was unmasked during the exposure, the risk of secondary transmission was double compared to when both the infected person and their close contact were masked (26% vs. 13%). The risk of transmission was higher when only the contact was unmasked, compared to when only the infected case was unmasked, but the sample sizes were too small to draw a firm conclusion about this difference. Although this observational study is subject to several possible sources of bias, it provides rare individual-level data supporting the effectiveness of masks in preventing infection by the wearer.

Study design

Retrospective Cohort

Study population and setting

This study, conducted by the public health department of Johnson County, Iowa, compared SARS-CoV-2 secondary attack rates (SARs) by the masking status of index cases and their close contacts from October 23, 2020 to February 28, 2021. Through the Iowa state contact-tracing program, the authors identified 969 close contacts of 431 cases (people who had tested positive for SARS-CoV-2 infection) who met inclusion criteria and for whom data were available on masking status during exposure and subsequent SARS-CoV-2 testing results. Close contacts were defined as: people who spent more than 15 minutes within 6 feet of a case during that case’s infectious period; who spent 2 hours or more in the same enclosed space as a case; or who experienced “substantial direct exposure” to the case (this latter criterion was evaluated on a case-by-case basis). The masking status of both case and contact during exposure was assessed via interviews with cases. Cases were also asked about dates and durations of exposure, symptoms during exposure, and the setting of the exposure (indoors vs. outdoors). Contacts were interviewed and asked for demographic information, date of symptom onset, previous COVID-19 history, and vaccination history. Contacts were excluded from the study if exposure occurred in a household, health care, or long-term care setting; or if no SARS-CoV-2 testing results from 2-14 days after the exposure were available. Secondary attack rates were calculated for combinations of masking status of the case and contact.  Multivariable logistic regression was used to estimate the association between mask score (the number of persons masked during an exposure: 0, 1, or 2) and SARS-CoV-2 transmission, adjusting for age, exposure setting, whether the case was symptomatic during exposure, and exposure duration. The study was initiated in response to a change in guidance from the Iowa Department of Health on September 29, 2020 in which close contacts were advised to perform symptom monitoring for 14 days instead of home quarantine if both the initial case and the contact were fully masked during exposure.

Summary of Main Findings

The average number of contacts per case was 2.25, and the median age of contacts was 18 years (range: 0 to 90 years). The overall secondary attack rate (SAR) was 20.5% (95% CI: 18.1 to 23.2), with the following SARs by masking status of case and contact, respectively: unmasked/unmasked 26.4% (95% CI: 22.9 to 30.7), unmasked/masked 10.0% (95% CI: 4.0 to 25.3), masked/unmasked 29.1% (95% CI: 19.3 to 43.9), masked/masked 12.5% (95% CI: 9.6 to 16.3). Among the 590 contacts (61%) in which at least one person was unmasked, the SAR was 25.6% (95% CI: 22.3 to 29.4). In multivariable logistic regression, an increase in mask score of 1 unit (representing one additional person masked, from 0-2) was associated with 30% lower odds of a secondary case (odds ratio 0.70, 95% CI: 0.57 to 0.84). Longer exposure and older age were associated with higher odds of secondary infection. Results were similar when restricted to children aged 5-18 years. Only 16 contacts had received at least one vaccine dose prior to exposure; all 16 tested negative to SARS-CoV-2. Of the 3 contacts with a prior positive SARS-CoV-2 test result, one tested positive after exposure.

Study Strengths

This study was able to measure masking status among both cases and contacts, and calculated secondary attack rates for each combination of masking status. Nesting the study within a contact tracing program allowed for standardized collection of covariates.

Limitations

The sample by definition did not include individuals who could not be contacted or who refused to participate in contact tracing investigation; the authors did not report the number of possible cases who were thus excluded from analysis. The resulting sample may not be representative of the wider population of cases and their close contacts. Additionally, an unknown number of contacts without testing results were excluded, which may have resulted in overestimation of the SARs. The sample size was too small to permit strong inference about differences in protection when the case vs. the contact was masked. Classification of mask use relied on self-report, which is subject to several possible biases including faulty recall, social desirability (i.e., individuals may have reported what they felt the interviewer wanted to hear), or a desire to protect contacts from the possibility of quarantine. Finally, this study took place before widespread vaccination; these results might or might not apply to a population with higher vaccination rates.

Value added

This is one of the very few studies to estimate and compare secondary attack rates among close contacts by the masking status of the initial case and the close contact during exposure.

Our take —

This observational study assessed myocarditis rates among Israelis 16 years and older between December 2020 to May 2021 by SARS-CoV-2 vaccination status in Israel by age and sex. Myocarditis cases were clinically adjudicated and incidences per 100,000 persons after first and second Pfizer-BioNTech mRNA vaccine doses were compared with unvaccinated individuals during the same time period and with myocarditis incidence observed before the COVID-19 pandemic. Vaccine-associated myocarditis cases were rare in all age and sex groups and over 90% of cases were mild and resolved quickly. Myocarditis was more likely after the second dose of the vaccine in all individuals, compared to both control groups. Adolescent males (aged 16-19 years) were at the highest risk of myocarditis after their second dose of the vaccine (15.07 cases per 100,000) compared to other age and sex groups. This study was not able to compare the incidence of vaccine-associated myocarditis with negative outcomes of SARS-CoV-2 infection (myocarditis, hospitalization, multisystem inflammatory syndrome, death, etc.).

Study design

Retrospective Cohort

Study population and setting

This observational study assessed myocarditis rates among individuals 16 years and older in Israel following the initiation of a national vaccination campaign (Pfizer-BioNTech vaccine) on December 20, 2020 through May 31, 2021. Myocarditis cases were initially identified via passive and active surveillance through the Israeli Ministry of Health. All cases were subject to adjudication from cardiologists and rheumatologists, and cases categorized as definitive or probable myocarditis were included in the analysis. Definitive and probable myocarditis case severity was described with available clinical data (length of stay, cardiac imaging, laboratory values, and time to symptom or imaging resolution). Myocarditis cases per 100,000 persons were compared by sex and age group among individuals after their first and second doses of the Pfizer-BioNTech SARS-CoV-2 vaccine (within 21 days of the first dose and 30 days of the second dose) with unvaccinated individuals during the same time period. Observed numbers of myocarditis cases were also compared with those seen among individuals 16 years and older in Israel from 2017-2019, before the COVID-19 pandemic. The study assessed myocarditis incidence per 100,000 persons, rate differences by group, and rate ratios between groups.

Summary of Main Findings

Approximately 9.2 million Israeli residents were included in the study period, about 5.1 million of whom received two doses of the Pfizer-BioNTech COVID-19 vaccine. They adjudicated 304 reported myocarditis cases and excluded 21 with a reasonable alternative diagnosis, 59 with insufficient data, and 4 suspected cases without sufficient information to classify them as probable. Of the remaining probable or confirmed cases (n=220), 107 occurred within 21 days of the first vaccine dose or 30 days of the second vaccine dose, 31 occurred in vaccinated individuals more than 21 days after the first dose or 30 days after the second dose (thereby classified as not vaccine-related), and 82 occurred in unvaccinated individuals (29 in individuals with diagnosed COVID-19). Of the 138 cases in vaccinated individuals, 129 (93.5%) were classified as mild with rapid resolution, 4 had severely-reduced ejection fraction (2.9%), and one died (0.7%).

Both male (RD 3.19 per 100,000 persons, 95% CI: 2.37, 4.02) and female (RD 0.39, 95% CI: 0.10, 0.68) participants were more likely to be diagnosed with myocarditis after their second dose compared to the first dose, with males aged 16-19 years experiencing the highest risk (15.07 cases per 100,000 persons). Overall, cases of myocarditis after vaccination were more common after the second dose than during the 2017-2019 reference period (standardized incidence ratio 5.34, 95% CI: 4.48, 6.40), with the highest incidence ratio in males aged 16-19 years (13.6, 95% CI: 9.3, 19.2). Overall, compared to the unvaccinated group, the second dose of the vaccine was associated with an increased rate of myocarditis (rate ratio 2.35, 95% CI: 1.1, 5.02), particularly among 16-19 year-old males (8.96, 95% CI: 4.50, 17.83).

Study Strengths

This study used population-level data with robust clinical adjudication to estimate the incidence of myocarditis after vaccination with the Pfizer-BioNTech vaccine by sex and age group. They also had access to baseline myocarditis incidence prior to the COVID-19 pandemic and incidence among unvaccinated individuals for comparison.

Limitations

While this study compared myocarditis incidence following vaccination to control individuals prior to the COVID-19 pandemic and unvaccinated individuals during the same time period, it did not include a comparison group of individuals diagnosed with SARS-CoV-2, which would have allowed readers to assess the risk of myocarditis following vaccination compared to the risk of myocarditis following COVID-19. It also did not have information on case severity among control participants, making it impossible to compare myocarditis severity following vaccination to other scenarios. Furthermore, this study did not provide information on myocarditis incidence in those younger than 16 years. Finally, given the small sample size, the analysis did not adjust for potential confounders beyond age and sex, such as underlying heart disease, immunological conditions, or medications that could cause myocarditis, all of which may be more common among vaccinated individuals.

Value added

This study provides evidence that adolescent males may be at increased risk of myocarditis following the second dose of the SARS-CoV-2 vaccine compared to baseline myocarditis incidence and incidence in similar unvaccinated individuals. However, it is unclear how clinically significant vaccine-associated myocarditis is compared to COVID-19 in this age group.

Our take —

This study, available as a preprint and thus not yet peer reviewed, used data from nationwide registries in the Netherlands to estimate vaccine effectiveness against COVID-19 hospitalization during periods when the Alpha and Delta variants were dominant. The study found that full vaccination (from the Pfizer, Moderna, AstraZeneca, and Janssen vaccines) was highly effective in both periods (94% and 95%, respectively), and found no evidence that effectiveness waned over a 5-month span after full vaccination or across different age groups. Although there are some possible sources of bias in this study, the use of nationwide registries provides strong evidence for persistently robust protection of vaccination against COVID-19-associated hospitalization, even with the dominance of the Delta strain.

Study design

Retrospective Cohort

Study population and setting

This study in the Netherlands linked a nationwide vaccination registry with a nationwide hospitalization database to estimate vaccine effectiveness (VE) against COVID-19 hospitalization and ICU admission from April 4 to August 19, 2021. VE was estimated by vaccine (Pfizer, Moderna, AstraZeneca, and Janssen), age group (15-49 years, 50-69 years, 70 years and older), time since vaccination, and prevailing SARS-CoV-2 variant (April 4 to May 29, when 95% of sequenced SARS-CoV-2 isolates were Alpha, vs. July 4 – August 29, when 99% of sequenced isolates were Delta). COVID-19 hospitalizations and ICU admissions were taken from NICE, a nationwide registry of all hospitalized individuals with a positive SARS-CoV-2 test or COVID-19 diagnosis. Vaccination status was ascertained via the nationwide vaccine registry CIMS. Upon vaccination, informed consent was sought for CIMS registration; among the 84% of vaccinations provided through the Netherlands municipal health services, 7.3% declined to be registered in CIMS. The size of the unvaccinated group was calculated by subtracting the vaccinated group from the full population. Vaccination status was considered to be “partial” 14 days after the first dose and “full” 14 days after the second dose (or 28 days after the single dose of the Janssen vaccine). The median duration from symptom onset to hospitalization, by age group (3-7 days), was applied to each admission date to classify vaccination status. Incidence rates per 10,000 person-days were calculated and incidence rate ratios (IRRs) were calculated with negative binomial regression adjusted for calendar date.

Summary of Main Findings

Of the 15,571 patients hospitalized with COVID-19 during the study period, 6% were fully vaccinated, 7% were partially vaccinated, and 87% were unvaccinated. For full vaccination, the estimated VE against hospitalization during the Alpha period was 94% overall (95% CI: 93% to 95%) and was consistently high across all three age groups. The estimated VE against hospitalization during the Delta period was 95% (94% to 95%) and also above 90% across age groups. Estimated VE against hospitalization by vaccine during the Delta period was as follows: Pfizer 96% (95% to 96%), Moderna 84% (80% to 87%), AstraZeneca 94% (92% to 95%), and Janssen 91% (88% to 94%). Estimated VE against ICU admission was generally slightly higher than that for hospitalization across categories of age, vaccine, time since vaccination, and viral variant period. There was no evidence of waning VE in any age group up to 20 weeks after full vaccination.

Study Strengths

Linkage of national hospitalization and vaccination registries allowed for a full nationwide cohort analysis of vaccine effectiveness and a comparison of vaccine effectiveness across age and vaccine cohorts.

Limitations

Vaccine allocation was nonrandom; for example, the authors noted that the Moderna vaccine was provided to medically high-risk patients, which may explain the lower apparent VE for that vaccine relative to the other vaccines. Furthermore, individuals who received the Pfizer and Moderna vaccine received the doses with a median interval of five weeks, which is longer than the intervals tested in their clinical trials (three and four weeks respectively), but may have implications for long-term effectiveness. Additionally, other than age, no information on patient characteristics or comorbidities were available, so unmeasured confounding is likely. There were few hospitalizations in several strata, limiting precision of estimates. An unknown proportion (likely around 7%) of vaccinated individuals did not consent to CIMS registration, and the resulting misclassification would result in overestimation of VE. However, a sensitivity analysis showed minimal impact on VE estimates. The population at risk included people with prior SARS-CoV-2 infection, which may have also led to bias if prior infection was associated with both vaccination status and hospitalization. Finally, COVID-19 hospitalization may have included some people who were hospitalized for other reasons but who incidentally tested positive for SARS-CoV-2, which may have led to a slight bias in an unknown direction.

Value added

This study is one of the first nationwide cohort studies to estimate vaccine effectiveness in the era of dominance of the Delta strain of SARS-CoV-2.

Our take —

This retrospective study from Israel examined the effect of a third (booster) dose of the BNT162b2 (Pfizer) vaccine. The population studied included individuals ages 60 years and older who received their second dose of vaccine at least five months earlier. It was estimated that boosters decreased the rate of confirmed COVID-19 infection by a factor of 11.3, and reduced the rate of severe illness by a factor of 19.5. However, the rates of infection and severe disease were extremely low in both groups, and so while the factors of protection may be high, the overall real-world benefit of boosters may be somewhat limited.

Study design

Retrospective Cohort

Study population and setting

Administration of a third (booster) dose of the Pfizer BNT162b2 vaccine was approved on July 30, 2021 in Israel for individuals age 60 years and older who had received their second dose at least five months prior. This retrospective analysis extracted data from July 30 to August 31, 2021 from the Israeli Ministry of Health database on September 2, 2021. The goal of the study was to examine the effect of the booster dose on the rate of COVID-19 infections and severe illness in this population. Data was pulled only for individuals in the 60 years and older group who had their second dose at least five months earlier, which yielded information from 1,137,804 individuals. The rate of COVID-19 infection and severe illness was compared between those who received a booster at least 12 days earlier and those who did not (non-booster). COVID-19 illness was confirmed by PCR, and severe illness was defined as having a resting respiratory rate of more than 30 breaths per minute, oxygen saturation of less than 94% while breathing ambient air, or a ratio of partial pressure of arterial oxygen to fraction of inspired oxygen of less than 300. For the non-booster group, days at risk began 12 days after beginning of study (August 10, 2021) and ended at either time of occurrence of a study outcome, the end of study the period, or at the time of receipt of a booster dose. Finally, the rate of infection 4-6 days after receiving a booster was compared to the rate of infection at least 12 days after receiving a booster. A Poisson regression was used to calculate rates after adjusting for possible confounding factors, including age (60-69, 70-79, 80+), sex, demographic group (general Jewish, Arab, ultra-Orthodox Jewish), and date of second vaccination in half month intervals.

Summary of Main Findings

From this retrospective cohort, there were 4,439 cases of infection out of 5,193,825 person-days at risk in the non-booster group compared to 934 cases out of 10,603,410 person-days at risk in the booster group. This amounted to a lower rate of infection in the booster group compared to non-booster group by factor of 11.3 (95% CI, 10.4 to 12.3) at least 12 days after receiving a booster. There were 294 cases of severe illness out of 4,574,439 person-days at risk in the non-booster group compare to 29 cases out of 6,265,361 person-days at risk in the booster group, which translates to a 19.5 fold lower rate of severe illness (95% CI, 12.9 to 29.5) in the booster group. Finally, the rate of confirmed infection at least 12 days after booster was lower than the rate of infection from 4-6 days after receiving a booster by a factor of 5.4 (95% CI, 4.8 to 6.1).

Study Strengths

The authors chose the cutoff of 12 days post-booster dose carefully, allowing for 7 days of immunity to build up plus 5 days of delay in detection of infection by PCR test. The date of second vaccine dose was included in the regression in order to account for waning effects of earlier vaccination, and the fact that high-risk groups received their vaccines earlier.

Limitations

Rates of infection in the 4-6 days post-booster time period could be underestimated due to booster recipients possibly undergoing less frequent PCR testing and behaving more carefully in the days following their booster dose. Next, the definition for severe illness may not be completely accurate. An oxygen saturation of 95% could be normal for some individuals, such as the elderly or people with COPD, so a saturation below 94% would not be severe in these cases. Additionally, deaths were not mentioned in this analysis. Finally, the rate ratio analysis is slightly unclear to the reader since the number of people in the booster and nonbooster groups was dynamic, and therefore one cannot calculate rates by dividing number of instances by number of people, in addition to the fact that the authors adjusted for specific factors within their model.

Value added

This is one of the first reports on the effectiveness of receiving a booster (third) dose of the Pfizer BNT162b2 vaccine in reducing rates of COVID-19 infection and severe illness in adults ages 60 years and older.