eAppendix. Supplemental Methods
Customize your JAMA Network experience by selecting one or more topics from the list below.
Botton J, Semenzato L, Jabagi M, et al. Effectiveness of Ad26.COV2.S Vaccine vs BNT162b2 Vaccine for COVID-19 Hospitalizations. JAMA Netw Open. 2022;5(3):e220868. doi:10.1001/jamanetworkopen.2022.0868
Although the Ad26.COV2.S vaccine (Janssen) showed an efficacy of 85.4% against severe and critical COVID-19 in the pivotal trial,1 its effectiveness in the general population against COVID-19 hospitalization was estimated to be approximately 68%,2,3 compared with approximately 90% for mRNA vaccines.4 However, to date, the effectiveness of Ad26.COV2.S has not been compared with that of other COVID-19 vaccines. In France, the Ad26.COV2.S vaccine was used from April 24, 2021, in people aged 55 years or older, whereas the BNT162b2 mRNA vaccine (Pfizer-BioNTech) was the most widely administered (78% of first doses). As of the end of July 2021, 19 million people aged 55 years or older (84% of the population in that age group) were partially or fully vaccinated. In this comparative effectiveness research study, we compare the effectiveness of full vaccination with Ad26.COV2.S vs BNT162b2 against COVID-19–related hospitalization.
The research group has permanent regulatory access to the anonymized data from the French National Health Data System (French decree No. 2016-1871, French law articles Art. R. 1461-13/14, French data protection authority decision CNIL-2016-316). Thus, no informed consent or specific approval by an ethics committee was required. This report follows the International Society for Pharmacoeconomics and Outcomes Research (ISPOR) reporting guideline for comparative effectiveness research.
On the basis of the French National Health Data System5 (see the eAppendix in the Supplement), we constructed a matched cohort of participants aged 55 years or older vaccinated with either Ad26.COV2.S or BNT162b2 between April 24, 2021, and July 31, 2021 (ie, 99 days). Participants of the 2 groups were individually matched according to age, sex, area of residence (100 areas) and date of full vaccination (first dose for Ad26.COV2.S, second dose for BNT162b2). Each participant was followed from the date of injection (day 0) or day 14 or day 28 (time of full effectiveness) until hospitalization for COVID-19 (outcome), death, or the end of follow-up on August 31, 2021, whichever occurred first.
The COVID-19 hospitalization rate was compared between the 2 groups using inverse probability of treatment weighted Cox models. Data were analyzed using SAS statistical software version 9.4 (SAS Institute).
The cohort included 689 275 participants vaccinated with Ad26.COV2.S (94% of all individuals of this age category vaccinated with Ad26.COV2.S) and 689 275 participants vaccinated with BNT162b2. The mean (SD) age was 65.8 (9.0) years, and 341 490 participants in each group (49.5%) were women. The 2 groups were similar in terms of socioeconomic and health characteristics (Table 1). During a median (IQR) follow-up of 54 (22-74) days from day 28 after injection, 129 COVID-19–related hospitalizations occurred in participants vaccinated with Ad26.COV2.S, and 23 hospitalizations occurred in those vaccinated with BNT162b2. The risk of hospitalization for COVID-19 from day 28 after injection was 5.2 times higher in individuals vaccinated with Ad26.COV2.S compared with those vaccinated with BNT162b2 (adjusted hazard ratio, 5.2; 95% CI, 3.4-7.9) (Table 2). On the basis of these results and according to an effectiveness of BNT162b2 of 92% (95% CI, 90%-94%) estimated from the same data set,6 we obtained an absolute effectiveness of Ad26.COV2.S of 59% (95% CI, 33%-75%).
This comparative effectiveness research study, which, to our knowledge, is the largest estimating effectiveness of Ad26.COV2.S in the general population, included almost the entire population aged 55 years or older vaccinated with Ad26.COV2.S in France. Considering the high rate of vaccination uptake in this population, using an active comparator was more relevant than considering unvaccinated individuals as controls. The risk of severe COVID–19 related hospitalization after vaccination was approximately 5 times higher with Ad26.COV2.S than with BNT162b2.
On the basis of these results and an effectiveness of BNT162b2 of 92% (95% CI, 90%-94%) estimated from the same data set,6 we obtained an absolute effectiveness of Ad26.COV2.S of 59% (95% CI, 33%-75%). This finding is consistent with previous estimates of smaller populations and using test-negative or case-control designs.2,3 A limitation of this study is that, although the 2 vaccine groups were matched on vaccination day, age, sex, and area of residence and the associations were adjusted for a large number of covariables, we cannot completely exclude residual confounding. Using the active BNT162b2 vaccine comparator likely lowered this potential bias compared with using an unvaccinated group. The slightly higher risk estimates from day 0 or 14 are likely associated with partial protection by the first dose of BNT162b2 combined with delayed protection of Ad26.COV2.S immediately in the days after injection.
This study found that the Ad26.COV2.S vaccine is less effective against COVID-19–related hospitalization than the BNT162b2 vaccine. These results strengthen the evidence supporting a second dose in people who received the Ad26.COV2.S vaccine by an mRNA vaccine as recommended in both France and the US.
Accepted for Publication: January 13, 2022.
Published: March 2, 2022. doi:10.1001/jamanetworkopen.2022.0868
Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2022 Botton J et al. JAMA Network Open.
Corresponding Author: Jérémie Botton, PharmD, PhD, EPI-PHARE Scientific Interest Group in Epidemiology of Health Products, French National Agency for the Safety of Medicines and Health Products, French National Health Insurance, 143-147 Boulevard Anatole France, F-93285 Saint-Denis Cedex, France (firstname.lastname@example.org).
Author Contributions: Dr Botton and Ms Semenzato had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Botton, Semenzato, Jabagi, Weill, Dray-Spira, Zureik.
Acquisition, analysis, or interpretation of data: Botton, Semenzato, Baricault, Weill, Dray-Spira, Zureik.
Drafting of the manuscript: Botton, Semenzato, Zureik.
Critical revision of the manuscript for important intellectual content: Jabagi, Baricault, Weill, Dray-Spira, Zureik.
Statistical analysis: Botton, Semenzato, Baricault.
Administrative, technical, or material support: Jabagi, Zureik.
Supervision: Botton, Weill, Dray-Spira, Zureik.
Conflict of Interest Disclosures: None reported.
Additional Contributions: Stephane Le Vu, PharmD, PhD, and Kim Bouillon, MD, PhD reviewed the manuscript, and Marion Bertrand, MSc, and Jérôme Drouin, MSc, provided technical support related to data management (all from EPI-PHARE, Scientific Interest Group in Epidemiology of Health Products, French National Agency for the Safety of Medicines and Health Products, French National Health Insurance). These individuals were not compensated for their contributions beyond their normal salaries.