Effectiveness of second booster compared to first booster and protection conferred by previous SARS-CoV-2 infection against symptomatic Omicron BA.2 and BA.4/5 in France

Vaccination contre le SARS-CoV-2 : efficacité du second rappel comparé au premier contre l’infection symptomatique par Omicron BA.2 et BA.4/5 en France et protection conférée par une infection antérieure à la vaccination

Publié le 23 mai 2023

Real-world evaluation of the protection provided by a vaccine for a target disease implies comparing incidence levels of the target disease according to vaccination status. This requires data on the disease and vaccination status. Santé publique France’s surveillance of the COVID-19 epidemic is based on several information systems. These include the SI-DEP system, which records all SARS-CoV-2 test results, and the VAC-SI system, which records COVID-19 vaccination status within the population (see Box 1). Crossing these two databases enabled an evaluation of the protective effect of the second booster dose, which is recommended for people aged 60 years and over. 

The Omicron variant (B.1.1.529) – including its sub-lineages (BA.1, BA.2, BA.4/5) – has been the dominant variant since December 2021. Compared to previous variants, Omicron is more transmissible and possesses immune escape properties with respect to vaccine protection. In view of the vaccine’s rapidly waning efficacy against symptomatic infection, the recommendation for a second booster dose against SARS-CoV-2 (i.e., after the initial two doses of a complete primary series and a first booster) was made in March 2022. This recommendation concerned the most vulnerable populations; principally people aged 80 and over. However, given the progression of the epidemic, the recommendation was quickly extended to people aged 60 years and over. During this period, the vaccines used for the second booster dose were monovalent mRNA vaccines, i.e. vaccines against the original strain of the virus.

The first bivalent mRNA vaccines were introduced in September 2022 with a recommendation for people aged 60 and over.

This context raised several questions of interest. Firstly, what is the gain in protection conferred by a second booster of monovalent vaccine? Furthermore, did the level of protection differ according to whether the dominant variant was Omicron BA.2 or Omicron BA.4/5? Finally, what is the additional gain in protection conferred by bivalent vaccines? Several recent studies conducted by Santé publique France give answers to these questions. As a result, two articles focussing on monovalent vaccines were published in the journal Vaccine [1, 2] and a third article addressing bivalent vaccines is available from the open archive MedRxiv [3].

3 questions for: Cynthia Tamandjou and Vincent Auvigne, Infectious Diseases Department, Santé publique France



Your study focuses on people over the age of 60 targeted by a second booster of COVID-19 vaccine. What methodology allowed you to study the increase in protection conferred by the first and second booster vaccines in this population?

Given the evolution of the COVID-19 epidemic, protecting the most vulnerable people has been a priority objective since the beginning of the vaccination campaign. In this context, it was important to evaluate the effectiveness of vaccines distributed to this population on an ongoing basis. We therefore conducted two primary studies on monovalent vaccines, which target the original SARS-CoV-2 strain: 

  • One investigating the effectiveness of the complete two-dose primary vaccination series and of the first booster against severe forms of COVID-19 linked to the Delta and Omicron BA.1 variants (Study 1)
  • Another investigating the effectiveness of the second booster against symptomatic infections with Omicron BA.2 and BA.4/5 variants (Study 2), taking into account previous infections.

These studies contended with several challenges: the emergence of new variants, natural immunity, reinfections, as well as a sharp decline in the number of “unvaccinated” people who usually constitute the reference population in studies on vaccine effectiveness. We used a test-negative design, which compares the vaccination status of people with COVID-19 symptoms according to the results (positive or negative) of a diagnostic test. If the vaccine is effective, the proportion of vaccinated people is higher among SARS-CoV-2-negative individuals than among positive individuals. This method, initially developed to evaluate the effectiveness of the seasonal flu vaccine, has been widely used to evaluate the real-world effectiveness of COVID-19 vaccines in France and elsewhere. Upon the launch of the vaccination campaign in 2021, this was the method retained for all vaccine effectiveness studies conducted by Santé publique France.

While Study 1[1] used unvaccinated people as a reference, this became very difficult in subsequent studies because most people in France were already vaccinated against COVID-19: more than 90% of people aged 50 and over have received at least a first dose of vaccine. For this reason, the reference group in Study 2[2] consisted of people who had received a booster vaccination (following two doses of a complete primary series) 6 to 7 months before the diagnostic test included in the study. The protection measured in this study – i.e. protection provided by the second booster – is therefore additional protection to that provided by the first booster taken 6 to 7 months earlier. Thus, protection estimated at 0% here does not mean no protection, but that the second booster did not provide any additional protection compared to the first booster taken 6 to 7 months before. 

In contrast to clinical trials, there is no random draw to allocate vaccines in real-world studies. Our protocol ensures that cases (people who test positive for SARS-CoV-2) and controls (people who test negative for SARS-CoV-2) are comparable, except for vaccination status, particularly in terms of their risk behaviours and use of SARS-CoV-2 screening. Thus, we compared people (cases and controls) who lived in the same area and performed the same type of test (RT-PCR or antigen test) in the same week. The statistical analysis also takes into account other factors that could influence the vaccination status of individuals, including gender, age or the presence of comorbidities. Particular consideration was given to the time interval between the last vaccination and the diagnostic test, as well as to prior Sars-CoV-2 infections, both of which are known to influence the effectiveness of the vaccination. 

Does your study quantify the gain in protection conferred by the second booster, recommended for people aged 60 and over? What are the main findings?

In Study 1, the effectiveness of a complete two-dose primary vaccination series with an mRNA vaccine against risk of hospital admission in subjects aged 50 years and older was estimated at 96% against the Delta variant (circulating between June and December 2021) and 92% against the Omicron variant (dominant since January 2022). This level of effectiveness applies to the first month (8-30 days) after taking the second dose of vaccine. The effectiveness against risk of admission to intensive care or death in hospital was estimated at 95% against the Delta variant and 98% against the Omicron variant. While these levels of protection were maintained at high levels over several months against the Delta variant, a decrease over time was observed against the Omicron variant, particularly in the 80+ age group. The booster dose restored high levels of protection against this variant. This study was able to demonstrate the major benefit of the booster dose during the Omicron epidemic, which is ongoing since January 2022.

Study 2 showed that, all else being equal, people who received a second booster 7 to 30 days before the diagnostic test included in the study were better protected against symptomatic infection than those whose last vaccination was a first booster taken 6 to 7 months earlier. The additional protection conferred was 41%. In practice, the risk of developing SARS-CoV-2 infection was almost halved in people who had received a second booster dose compared to those who had received a first booster 6 to 7 months earlier. The study therefore confirms that booster vaccines targeting the original virus are effective against infections with Omicron BA.2 and BA.4/5, which were the dominant variants in the study. However, as with the primary vaccination series and the first booster dose, we observed a decrease in protection over time. The level of added protection at 3 to 4 months after the second booster, compared to that conferred by a first booster taken 6 to 7 months earlier, was only marginal (10%). However, it should be remembered that the protection against severe illness (requiring hospitalisation and/or leading to death), which we did not measure in this study, generally lasts longer than that against milder symptomatic forms.

Study 2 also quantified the protection conferred by an old infection against a new symptomatic SARS-CoV-2 infection in vaccinated individuals. It proved to be strong and lasting. For example, vaccinated individuals who had no prior infection recorded in the SI-DEP database were twice as likely to be infected as those who were vaccinated and infected before the onset of the Delta variant, i.e. at least 1 year before our study. For those who had been infected in the preceding 5 months, the risk of infection was divided by ten.

In September 2022, new bivalent vaccines were introduced and recommended for the same 60+ age group. Are there any results on their effectiveness in terms of added protection compared to previous monovalent vaccines? Does this knowledge have implications for the near future and for controlling the epidemic?

The bivalent vaccines on the market since September 2022 contain the messenger RNA of two strains of virus: the original virus (the same as in the first monovalent vaccines) and a strain of Omicron (BA.5 in most cases). They were designed to provide better protection against the Omicron variants in circulation. 

Also using SI-DEP and VAC-SI data, we conducted a cohort study to compare the protection against symptomatic infection conferred by a booster of bivalent vaccine with that conferred by a booster of monovalent vaccine[3]. We followed over 135,000 people vaccinated in the same weeks for a median of 77 days. The bivalent vaccine provided 8% additional protection.

The results of these studies carried out by Santé publique France were considered by the French National Authority for Health (HAS, Haute Autorité de Santé) in the work undertaken to prepare its recommendation for the 2023 vaccination strategy. The HAS noted that the data confirms the protection conferred by a booster dose and recommended organising a booster campaign for autumn 2023 to target all people at risk of severe illness (people aged 65 years and over; infants aged 6 months and over; children, adolescents and adults with comorbidities making them more vulnerable to severe illness; immunocompromised persons; persons with any other comorbidity) and their families. In addition, the HAS is considering recommending a springtime booster for the most vulnerable people (people aged 80 years and over, immunocompromised people and people at very high risk of severe illness).

VAC-SI and SI-DEP: two databases used to study the effectiveness of COVID-19 vaccines

SI-DEP (Système d’Information de DEPistage populationnel) is the national information system for population screening. It was designed to obtain automatic feedback on SARS-CoV-2 diagnostic tests (RT-PCR, serology and antigen) carried out across the French territory by all medical biology laboratories, pharmacies, hospitals or authorised health professionals. This database was created through a partnership between the Ministry of Health and Prevention (with the General Directorate of Health responsible for processing) and the network of public healthcare facilities known as Assistance Publique - Hôpitaux de Paris (AP-HP, project manager). An automated system that was less time-consuming for the laboratories was required to receive results quickly. The system was developed in 2 months and launched in May 2020, at the end of the first wave. 

All vaccinations carried out in France since the beginning of the COVID-19 vaccination campaign are recorded in the VAC-SI database. This database is managed by the CNAM (Caisse Nationale d'Assurance Maladie), the national health insurance administration. Since February 2021, the health professionals that perform vaccinations have fed it with data on their activity. This database contains comprehensive information about vaccinations, including date administered, vaccine type and vaccine name, as well as indications related to priority populations for vaccination (those with comorbidities, health or social workers, care-home residents). Santé publique France analysed VAC-SI on a daily basis to estimate vaccination coverage by age group, by place of vaccination, and by specific population.

For more information (French only):

Santé publique France and vaccine logistics: the case of COVID-19 vaccines

Santé publique France operates a Pharmaceutical Establishment (PE) managed by a senior pharmacist. The missions are defined by the Code of Public Health (Code de la santé publique) in article L. 1413-4: “At the request of the Minister of Health, the Agency shall acquire, manufacture, import, store, transport, distribute and export products and services necessary to protect the population against serious health threats. It ensures, under the same conditions, their renewal and destruction if necessary.”

In the context of the pandemic, France launched a COVID-19 vaccination campaign on 27 December 2020 to counter the spread of coronavirus in France. The logistics of this campaign were entrusted to Santé publique France. The pharmaceutical company is responsible for the reception, storage and shipping of vaccines and medical devices such as injection equipment (needles, syringes) to more than 20,000 pharmacies participating in the vaccination campaign, vaccination centres and hospitals.

For more information: Logistique vaccinale (French only)

[1] Cynthia Raissa Tamandjou Tchuem, Vincent Auvigne, Sophie Vaux, Charline Montagnat, Juliette Paireau, Stéphanie Monnier Besnard, Amélie Gabet, Nabil Benhajkassen, Yann Le Strat, Isabelle Parent Du Chatelet, Daniel Levy-Bruhl Vaccine effectiveness and duration of protection of COVID-19 mRNA vaccines against Delta and Omicron BA.1 symptomatic and severe COVID-19 outcomes in adults aged 50 years and over in France, Vaccine, Volume 41, Issue 13, 2023, Pages 2280-2288, https://doi.org/10.1016/j.vaccine.2023.02.062 

[2] Cynthia Tamandjou, Vincent Auvigne, Justine Schaeffer, Sophie Vaux, Isabelle Parent du Châtelet. Effectiveness of second booster compared to first booster and protection conferred by previous SARS-CoV-2 infection against symptomatic Omicron BA.2 and BA.4/5 in France. Vaccine, 2023. 
DOI: https://doi.org/10.1016/j.vaccine.2023.03.031  

[3] Protection against symptomatic SARS-CoV-2 BA.5 infection conferred by the Pfizer-BioNTech Original/BA.4-5 bivalent vaccine compared to the mRNA Original (ancestral) monovalent vaccines – a matched cohort study in France. Vincent Auvigne, Cynthia Tamandjou, Justine Schaeffer, Sophie Vaux, Isabelle Parent du Chatelet. medRxiv 2023.03.17.23287411; DOI: https://doi.org/10.1101/2023.03.17.23287411