Effectiveness of COVID-19 vaccine booster in the general population and in subjects with comorbidities. A population-based study in Spain

Environmental Research 215 (2022) 114252 Available online 10 September 2022 0013-9351/© 2022 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/bync-nd/4.0/). Effectiveness of COVID-19 vaccine booster in the general population and in subjects with comorbidities. A population-based study in Spain Narmeen Mallah a , b , c , Jacobo Pardo-Seco a , b , c , Luis-Ricardo L´ opez-P´ erez d , Juan-Manuel Gonz´ alez-P´ erez d , Benigno Ros´ on d , María-Teresa Otero-Barr´ os e , Carmen Dur´ an-Parrondo e , Victoria Nartallo-Penas f , Susana Mir´ as-Carballal g , Carmen Rodríguez-Tenreiro a , b , c , Irene Rivero-Calle a , b , c , h , Alberto G´ omez-Carballa a , b , c , Antonio Salas a , b , c , i , Federico Martin´ on-Torres a , b , c , h , * a Genetics, Vaccines and Pediatric Infectious Diseases Research Group (GENVIP), Instituto de Investigaci´ on Sanitaria de Santiago and Universidad de Santiago de Compostela (USC), Galicia, Spain b WHO Collaborating Centre for Vaccine Safety, Santiago de Compostela, Spain c Centro de Investigaci´ on Biom´ edica en Red de Enfermedades Respiratorias (CIBERES), Instituto de Salud Carlos III, Madrid, Spain d Subdirecci´ on de Sistemas y Tecnologías de La Informaci´ on, Servizo Galego de Saude, Galicia, Spain e Direcci´ on Xeral de Saude Pública, Consellería de Sanidade, Xunta de Galicia, Galicia, Spain f Programa Gallego de Vacunaci´ on, Consellería de Sanidade, Xunta de Galicia, Galicia, Spain g Servicio de Enfermedades Transmisibles, Consellería de Sanidade, Xunta de Galicia, Galicia, Spain h Translational Pediatrics and Infectious Diseases, Hospital Clínico Universitario and Universidad de Santiago de Compostela (USC), Galicia, Spain i Unidade de Xen´ etica, Instituto de Ciencias Forenses, Facultade de Medicina, Universidade de Santiago de Compostela, And GenPoB Research Group, Instituto de Investigaci´ on Sanitaria (IDIS), Hospital Clìnico Universitario de Santiago, Servizo Galego de Saúde, Galicia, Spain ARTICLE INFO Keywords: COVID-19 booster vaccine effectiveness Comorbidities Population-based study SARS-CoV-2 Spain ABSTRACT Background: Research on the effectiveness of COVID-19 booster-based vaccine schedule is ongoing and real-world data on vaccine effectiveness (VE) in comorbid patients are limited. We aimed to estimate booster dose VE against SARS-CoV-2 infection and COVID-19 severity in the general population and in comorbid patients. Method: A retrospective test-negative control study was undertaken in Galicia-Spain (December 2020–November 2021). VE and 95% confidence interval (CI) were estimated using multivariate logistic regression models. Results: 1,512,415 (94.13%) negative and 94,334 (5.87%) positive SARS-CoV-2 test results were included. A booster dose of COVID-19 vaccine is associated with substantially higher protection against SARS-CoV-2 infection than vaccination without a booster [VE boosted =87% (95%CI: 83%; 89%); VE non-boosted =66% (95%CI: 65%; 67%)]. The high VE was observed in all ages, but was more pronounced in subjects older than 65 years. VE against COVID-19 severity was analyzed in a mixed population of boosted and non-boosted individuals and considerable protection was obtained [VE: hospitalization =72% (95%CI: 68%; 75%); intensive care unit administration =83% (95%CI: 78%; 88%), in-hospital mortality =66% (95%CI: 53%; 75%)]. Boosted comorbid patients are more protected against SARS-CoV-2 infection than those who were non-boosted. This was observed in a wide range of major diseases including cancer (81% versus 54%), chronic obstructive pulmonary disease (84% versus 61%), diabetes (84% versus 65%), hypertension (82% versus 65%) and obesity (91% versus 67%), among others. Conclusions: A booster dose of COVID-19 vaccine increases the protection against SARS-CoV-2 infection and COVID-19 severity in the general population and in comorbid patients. 1. Introduction Vaccination against Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) remains the fundamental preventive measure against spreading the virus and developing severe Coronavirus Disease 2019 (COVID-19) disease, but whether a booster dose is necessary remains a * Corresponding author. Hospital Clínico Universitario de Santiago de Compostela, A Choupana s.n., 15701, Santiago de Compostela, Spain. E-mail address: [email protected] (F. Martin´ on-Torres). Contents lists available at ScienceDirect Environmental Research journal homepage: www.elsevier.com/locate/envres https://doi.org/10.1016/j.envres.2022.114252 Received 18 June 2022; Received in revised form 22 August 2022; Accepted 30 August 2022 Environmental Research 215 (2022) 114252 2 point of debate (Shekhar et al., 2021). Studies on COVID-19 vaccine effectiveness (VE) suggested declining protection over time, yet no global consensus has been reached on the speed of this decrease (Feikin et al., 2022). A study in Israel reported that antibody titer reached the climax after one month of the second dose of BNT162b2, and then declined rapidly (Khoury et al., 2021). In their study, Naaber and colleagues documented that the antibody levels declined six months after the second dose of BNT162b2, revealing a decrease in the immune response over time (Naaber et al., 2021). A study in Northern California showed that mRNA-1273 VE against SARS-CoV-2 infection moderately decreased from 88% to 76% at 6–8 months after vaccination (Florea et al., 2022). The impact of booster dose administration on COVID-19 is ongoing research. The World Health Organization (WHO), continues to review the emerging evidence on the need for and timing of a booster dose for the currently available COVID-19 vaccines (World Health Organization, 2021), but studies are still limited. Besides, the external validity of findings on VE is influenced by setting-specific parameters such as adherence to vaccine doses and time intervals for dose administration, access to healthcare, number of undertaken SARS-CoV-2 tests, the threshold for COVID-19 hospitalization and intensive care admission, as well as COVID-19 management strategy and the applied non-pharmaceutical interventions at different time points of the pandemic (e.g., lockdowns, facemasks, social distancing, etc.). Studies in different settings and populations would therefore aid in assessing the impact of booster administration. Moreover, although it has been widely reported that comorbid patients are very likely to develop serious COVID-19 outcomes (CDC: Center for Disease Control and Prevention, 2019; He et al., 2022; Chiner-Vives et al., 2022; Gonzalez-Barcala et al., 2022; Adab et al., 2022; Ejaz et al., 2020), several vulnerable groups were not sufficiently included in clinical trials on COVID-19 vaccines. Therefore, the degree of protection offered by recall COVID-19 vaccine doses in individuals with medical conditions requires further investigation. Spain was among the countries most affected by SARS-CoV-2 pandemic with 5,111,842 cases and 87,904 deaths as of November 24, 2021 (Ministerio de Sanidad Consumo y Bienestar Social [Ministry of Health Consumption and Social Welfare], 2021). It prioritized vaccination of inmates, social and healthcare workers and residents of long-term care facilities, front-line healthcare staff and elderly individuals to reduce the risk of COVID-19 related morbidity and mortality in these populations, and thereafter decrease the burden on public health facilities (Gobierno de Espa˜ na [Spanish Government], 2022). Vaccination of the other groups of the population was introduced successively following a priority order according to age, medical status, and occupation (Gobierno de Espa˜ na [Spanish Government], 2022). Monge and colleagues reported a moderate VE of booster mRNA vaccine dose against SARS-CoV-2 infections; however, the study was limited to individuals older than 40 years and excluded individuals at risk of infection (Monge et al., 2022). Using real-world data of around 3, 000,000 Polymerase Chain Reaction (PCR)-based SARS-CoV-2 tests, we extend the study of Monge by investigating the impact of a booster dose administration against SARS-CoV-2 infection as well as COVID-19 severe illness resulting in hospitalization, admission to intensive care unit (ICU), and death. We undertook a population-based study in Galicia, Northwest Spain that involved individuals older than 11 years and explored the impact of a booster dose on a large variety of comorbidities. 2. Methods 2.1. Settings This study was initiated within the framework of a project on COVID19 VE in Galicia, a region located in Northwest Spain (Pardo-Seco et al., 2022). Galicia is an autonomous community with a total inhabitants of 2,694,245 (Instituto Galego de Estatística, 2021a), for 29,576 km 2 (Instituto Galego de Estatística, 2019) i.e,., its population density is similar to that of the European population. The gender distribution in Galicia is 48% (N =1,296,602) males and 52% (N =1,397,643) females. Almost 9% of the Galician population are aged 80 and above (N =236, 788), 17% are in the age range of 65–79 (N =457,245), 22% are aged between 50 and 64 (N =596,034), while the rest are younger adults and children (Instituto Galego de Estatística, 2021b). The Galician Healthcare Service (SERGAS) is a public health system with universal access to healthcare at low or no cost. Galicia has 6,571 public hospital beds distributed in 35 hospitals (data of 2020) (Instituto Galego de Estatística, 2020). COVID-19 vaccine has been made freely available for all the population, and it was administered by priority order according to occupation, age, and medical status (Gobierno de Espa˜ na [Spanish Government], 2022). During the study period (December 26, 2020–November 23, 2021), the following four vaccines were administered in Spain: Pfizer-BioNTech vaccine (BNT162b2), Vaxzevria (ChAdOx1 nCoV-19), Spikevax (mRNA-1273), and Janssen (Ad26. COV2–S) (European Medicines Agency, 2022). A full vaccine course of BNT162b2, mRNA-1273, or ChAdOx1 nCoV-19 consisted of two injections with a predefined time separation between the first and the second injection, while Ad26. COV2–S was given as a single injection. Until the start of the study, the vaccination campaign encompassed individuals 11 years or above, hence, those younger than that age were excluded from the study. 2.2. Ethics The study did not involve an intervention or the use of human biological samples. SERGAS provided the authors with anonymized data, waiving the need for written informed consent. SERGAS did not take part in data analysis. The study protocol was approved by the clinical research ethics committee of Galicia (CEIC, protocol number: 2022–175). 2.3. Study design A retrospective test-negative case-control study was conducted to determine the effectiveness of COVID-19 vaccines administered in Galicia-Spain, against SARS-CoV-2 infection as well as against severe COVID-19 causing hospitalization, ICU admission, or death. To estimate VE in fully susceptible people, we excluded from the analysis those individuals who had a previous positive SARS-CoV-2 PCR, antigen or antibody test result at any time before the enrolment date in the study. Antibody response to the vaccine in infection-naive individuals takes more time than in those who had been infected with the virus in the past (Tut et al., 2021). A positive SARS-CoV-2 PCR test result represented a case, while a negative test result was deemed a control. An individual could contribute to the study by one or more negative tests; however, only one positive SARS-CoV-2 PCR test was considered per patient. During the study period, SARS-CoV-2 PCR test results were performed at no cost by SERGAS health centers for clinical motives such as presenting COVID19-related symptoms or being in close contact with a SARS-CoV-2 infected individual in the last two weeks. 2.4. Exposure definition In our settings, the exposure was defined as receiving any injection of COVID-19 vaccine. Individuals were categorized into unvaccinated, partially vaccinated, non-boosted and boosted according to their exposure status, in order to focus on SARS-CoV-2 infection acquired since vaccination after a sufficient interval for biological protection. Unvaccinated individuals are those who did not receive any COVID-19 vaccine injection during the study period. Partially vaccinated individuals are those who received a single dose of BNT162b2, mRNA-1273 or ChAdOx1 nCoV-19, as well as those who were given the two injections of N. Mallah et al. Environmental Research 215 (2022) 114252 3 these vaccines or the single injection of Ad26. COV2–S, but with less than seven days after the last injection. Non-boosted individuals are those who received the complete course of the vaccine counting since the 7th day after the last injection. Boosted individuals are those who received an additional dose of the vaccine counting at least seven days after the injection. 2.5. Outcome definition The primary outcome of the study consisted of SARS-CoV-2 infection confirmed by a PCR-based test. The date of SARS-CoV-2 infection was considered the date of the first positive PCR test. Secondary outcomes included hospitalization, ICU admission, or mortality attributed to severe COVID-19. Only in-hospital death events were considered for the mortality analysis. SARS-CoV-2 was deemed a cause of hospitalization if a patient had a positive PCR test result in the 30 days preceding the hospital admission or within three days after hospitalization (Yeo et al., 2021; Mehta et al., 2021). To account for nosocomial infections, patients with a positive SARS-CoV-2 PCR test after three days of hospitalization were not included in the analysis. ICU admissions and in-hospital mortality were ascertained among the population of COVID-19 hospitalized patients. 2.6. Statistical analysis Adjusted odds ratios (ORs) and their 95% confidence intervals (CIs) were estimated using multivariate logistic regression models. Unvaccinated individuals were used as a reference. The models were directly adjusted for age and sex due to their biological plausibility, and other variables were added to control confounding. To account for socioeconomic differences the following three variables were used as a proxy of this indicator: 1) sanitary area which represents the area of residence; 2) pharmaceutical cost contribution defined as the percentage of medicine cost paid by the participants according to their income tax; and 3) receiving social support. Data on comorbidities associated with each individual were also collected. The comorbidities included: atrial fibrillation, cancer, chronic obstructive pulmonary disease (COPD), dementia, depression, diabetes, epilepsy, heart failure, human immunodeficiency virus (HIV), hypertension, ischemic cardio-pathology, kidney failure, obesity, stroke, and Parkinson disease. Information on receiving any of pneumococcal conjugate vaccine, pneumococcal polysaccharide vaccine, or flu vaccine was also retrieved. The evolution of the epidemiological situation of the pandemic was considered by computing the time period between the PCR test date and the COVID-19 vaccination date. We examined the effect of each potentially confounding variable using the change-in-estimate method. For this purpose, a univariate analysis was performed and the covariables with a pvalue <0.2 were consecutively added to the model (Greenland, 1989). A covariable was kept in the model if it changed the originally estimated OR by at least 10%. The effectiveness of COVID-19 vaccination status was then calculated as follows: VE =(1 - OR adjusted ) ×100. When the number of observations to estimate the effectiveness of the booster against a certain outcome was not sufficient, a mixed population of boosted and non-boosted individuals was used. The analysis was stratified by age group and comorbidities. All analyses were undertaken using STATA v.12 (Stata Statistical Software: Release 12; StataCorp LP; College Station, TX, USA). 3. Results Fig. 1 summarizes how the sample size of the analysis was reached. A total of 909,636 individuals fulfilled the inclusion criteria and contributed 1,606,749 SARS-CoV-2 test results to the analysis. Among the included individuals, 94,334 were infected with SARS-CoV-2, whereas the remaining 815,302 individuals were negative for SARS-CoV-2. The number of SARS-CoV-2 infected individuals was the same as that of the positive SARS-CoV-2 test results (94,334; 5.87% of total PCR tests) as only one SARS-CoV-2 PCR positive test was considered per individual. The 815,302 SARS-CoV-2 negative people, provided 1,512,415 (94.13% of total PCR tests) test results to the analysis since each individual could contribute one or more negative SARS-CoV-2 PCR test results to the study. The analysis unit in our study is the test, not the person. Fig. 2 represents the distribution of the negative tests across the study population. Of the negatively tested individuals, 58.2% contributed only one Fig. 1. Flow diagram of participants and SARS-COV-2 PCR test results entry the study. N. Mallah et al. Environmental Research 215 (2022) 114252 4 negative test result, 25.4% provided two negative test results, 16.4% shared more than two negative test results, and 2.5% gave more than 5 negative test results (Fig. 2). In 1,155 infected individuals, COVID-19 was not deemed the cause of hospitalization as SARS-CoV-2 infection was either nosocomial or the hospitalization took place more than 30 days after infection. Accordingly, 93,179 infected individuals were included in the analysis of VE against COVID-19 hospitalization, ICU admission and in-hospital mortality. Of the 93,179 infected individuals, 5,871 (6.30%) were hospitalized for COVID-19, and 925 (0.99%) were further admitted to ICU. One thousand and eight (1.08%) in-hospital deaths for COVID-19 were also registered. The general and clinical characteristics of the study population are presented in Table 1. Impact of COVID-19 booster-based vaccination on SARS-CoV-2 infection and COVID-19 severity. Receiving a booster dose of COVID-19 vaccine increased the protection against infection by SARS-CoV-2 from 66% in non-boosted individuals to 87% [VE non-boosted =66% (95%CI: 65%; 67%); VE boosted = 87% (95%CI: 83%; 89%)] (Table 2). Only four (0.07%) boosted individuals who had been tested positive for SARS-CoV-2 were hospitalized for COVID-19. VE against COVID-19 hospitalization was then estimated in the mixed subpopulation of nonboosted and boosted individuals and a protection exceeding 70% was observed [VE =72% (95%CI: 68%; 75%)] (Table 3). None of those individuals who tested positive for SARS-CoV-2 and had received a booster dose of the vaccine was admitted to ICU or died for COVID-19, thus VE against these outcomes was estimated in the mixed subpopulation of non-boosted and boosted individuals. VE against ICU admission was 83% (95%CI: 78%; 88%), and that against inhospital mortality was 66% (95%CI: 53%; 75%). In our study population, only mRNA type vaccines were administered as a booster. Restricted analysis of individuals who had received mRNA-type vaccines exclusively, showed 88% VE against SARS-CoV-2 infections [VE =88% (95%CI: 85%; 90%)]. Among individuals who had received the mRNA booster dose, only four were hospitalized and none was admitted to ICU or died of COVID-19. In a mixed subpopulation of non-boosted and boosted individuals, VE against hospitalization for COVID-19 was 74% (95%CI: 70%; 77%); ICU admission 86% (95% CI: 80%; 90%) and in-hospital mortality 71% (95%CI: 59%; 79%) (Table S2). Impact of COVID-19 booster-based vaccination on SARS-CoV-2 infection and COVID-19 severity stratified by age. When stratifying the study population by age, sufficient observations of boosted individuals were only obtained for those older than 65 years. Receiving a booster dose of the COVID-19 vaccine offers more than 80% protection against SARS-CoV-2 infection in the elderly population [66–80 years: VE =85% (95%CI: 77%; 91%); ≥81 years: VE =82% (95%CI: 75%; 87%)] (Table 2). For the younger age categories, VE was estimated in the mixed subpopulation of non-boosted and boosted individuals due to insufficient number of observations, and ranged between 51% and 76% (Table 2). Likewise, the effectiveness of receiving a booster dose of COVID-19 vaccine against hospitalization was more than 70% in people aged over 65 years [66–80 years: VE =71% (95%CI: 62%; 78%); ≥81 years: VE =73% (95%CI: 64%; 80%)] (Table 3). VE against hospitalization in the younger age groups (≥18 years) was estimated in the mixed subpopulation of non-boosted and boosted individuals and ranged between 69% and 84% (Table 2). Comparing the mixed subpopulation of non-boosted and boosted individuals to unvaccinated individuals showed a substantial VE against ICU admission for COVID-19 in individuals aged between 46 and 65 years [VE =86% (95%CI: 77%; 91%)] and in those older than 65 years [VE =83% (95%CI: 73%; 89%)]. The data also revealed considerable protection against death for COVID-19 in individuals older than 65 years [VE =65% (95%CI: 50%; 75%)]. Impact of COVID-19 booster-based vaccination on SARS-CoV-2 infection and COVID-19 severity stratified by comorbidity type. Individuals with any of the following comorbidities and who received a booster dose of COVID-19 vaccine are more protected against SARS-CoV-2 infection than those who were non-boosted: cancer [VE nonboosted =54% (95%CI: 49%; 59%); VE boosted =81% (95%CI: 68%; 88%)], COPD [VE non-boosted =61% (95%CI: 55%; 67%); VE boosted =84% (95%CI: 66%; 93%)], depression [VE non-boosted =62% (95%CI: 59%; 65%); VE boosted =83% (95%CI: 74%; 89%)], diabetes [VE non-boosted = 65% (95%CI: 62%; 68%); VE boosted =84% (95%CI: 75%; 90%)], hypertension [VE non-boosted =65% (95%CI: 63%; 68%); VE boosted =82% (95%CI: 76%; 86%)], ischemic heart disease [VE non-boosted =56% (95% CI: 49%; 63%); VE boosted =79% (95%CI: 60%; 90%)], kidney failure [VE non-boosted =67% (95%CI: 61%; 72%); VE boosted =85% (95%CI: 70%; 92%)], obesity [VE non-boosted =67% (95%CI: 65%; 69%); VE boosted = 91% (95%CI: 83%; 95%)], and stroke [VE non-boosted =66% (95%CI: 60%; 71%); VE boosted =85% (95%CI: 68%; 93%)] (Table 4). Few individuals with epilepsy or Parkinson who had received the booster dose of COVID-19 vaccine tested positive for SARS-CoV-2, therefore VE against infection in these subgroups was estimated in the mixed population of non-boosted and boosted individuals. Substantial protection against SARS-CoV-2 infection was observed for epilepsy [VE =66% (95%CI: 58%; 73%)] and Parkinson patients [VE =74% (95%CI: 64%; 81%)] (Table 4). Few patients with comorbidity who had received the booster dose of the COVID-19 vaccine were hospitalized for COVID-19, hence VE against hospitalization was estimated by comparing the odds of hospitalization in the mixed population to that of unvaccinated individuals. Vaccinated comorbid patients showed lower odds of COVID-19 related hospitalization than unvaccinated ones. VE ranged between 49% and 79% (Table 5). VE against ICU admission for COVID-19 was estimated in the mixed population of non-boosted and boosted individuals relative to unvaccinated people. Significant protection against COVID-19-related ICU admission was observed in patients with cancer [VE =87% (95%CI: 71%; 94%)], depression [VE =87% (95%CI: 73%; 94%)], diabetes [VE =83% (95%CI: 70%; 91%)], hypertension [VE =83% (95%CI: 72%; 89%)], and obesity [VE =84% (95%CI: 73%; 91%)]. VE against COVID19-related ICU admission could not be estimated for other comorbid groups due to the limited number of observations. 0 10 20 30 40 50 60 Number of SARS−CoV−2 negative PCR test results Individuals (%) 0 10 20 30 40 Fig. 2. Distribution of SARS-CoV-2 PCR negative test results per individual. N. Mallah et al. Environmental Research 215 (2022) 114252 5 Table 1 Sociodemographic and clinical characteristics of the study population per each of the outcomes: infection with SARS-CoV-2; hospitalization for COVID-19; intensive care unit (ICU) admission for COVID-19; and in-hospital death for COVID-19. Characteristic Infection N (%) Hospitalization N (%) ICU Admission N (%) In-hospital Death N (%) No Yes Total No Yes Total No Yes Total No Yes Total N (%) 1,512,415 (94.13) 94,334 (5.87) 1,606,749 (100) 87,308 (93.70) 5871 (6.30) 93,179 (100) 92,254 (99.01) 925 (0.99) 93,179 (100) 92,340 (98.92) 1008 (1.08) 93,348 (100) Age (years) Mean (SD) 49.40 (21.68) 42.63 (21.10) 45.0 (21.70) 40.82 (20.20) 64.28 (19.24) 42.30 (20.93) 42.12 (20.91) 60.05 (14.0) 42.30 (20.93) 41.93 (20.65) 81.75 (10.61) 42.36 (20.98) Range 11–111 11–107 11–111 11–107 11–102 11–107 11–107 13–94 11–107 11–107 29–104 11–107 Age quartiles y 1st quartile: 11 - 32 375,231 (24.81) 35,850 (38.00) 411,081 (25.58) 35,354 (40.49) 438 (7.46) 35,792 (38.41) 35,750 (38.75) 42 (4.54) 35,792 (38.41) 35,791 (38.76) 1 (0.10) 35,792 (38.34) 2nd quartile: 33 - 48 392,178 (25.93) 23,334 (24.74) 415,512 (25.86) 22,397 (25.65) 844 (14.38) 23,241 (24.94) 23,093 (25.03) 148 (16.00) 23,241 (24.94) 23,234 (25.16) 7 (0.69) 23,241 (24.90) 3rd quartile: 49 - 65 373,152 (24.67) 19,874 (21.07) 393,026 (24.46) 18,096 (20.73) 1524 (25.96) 19,620 (21.06) 19,266 (20.88) 354 (38.27) 19,620 (21.06) 19,556 (21.18) 84 (8.33) 19,640 (21.04) 4th quartile: 66 - 111 371,854 (24.59) 15,276 (16.19) 387,130 (24.09) 11,461 (13.13) 3065 (52.21) 14,526 (15.59) 14,145 (15.33) 381 (41.19) 14,526 (15.59) 13,759 (14.90) 916 (90.87) 14,675 (15.72) Gender* y Male 672,291 (44.45) 45,229 (47.95) 717,520 (55.34) 41,398 (47.42) 3194 (54.40) 44,592 (47.86) 43,969 (47.66) 623 (67.35) 44,592 (47.86) 44,139 (47.80) 554 (54.96) 44,693 (47.88) Female 840,124 (55.55) 49,105 (52.05) 889,229 (55.34) 45,910 (52.58) 2677 (45.60) 48,587 (52.14) 48,285 (52.34) 302 (32.65) 48,587 (52.14) 48,201 (52.20) 454 (45.04) 48,655 (52.12) COVID-19 vaccine* ‡ Unvaccinated 917,498 (60.66) 69,816 (74.01) 987,314 (61.45) 64,251 (73.59) 4609 (78.50) 68,860 (73.90) 68,067 (73.78) 793 (85.73) 68,860 (73.90) 68,279 (73.94) 730 (72.42) 69,009 (73.93) Partially vaccinated 158,103 (10.45) 8937 (9.47) 167,040 (10.40) 8459 (9.69) 431 (7.34) 8890 (9.54) 8843 (9.98) 47 (5.08) 8890 (9.54) 8770 (9.50) 124 (12.30) 8894 (9.53) Fully vaccinated 428,756 (28.35) 15,498 (16.43) 444,254 (27.65) 14,522 (16.63) 827 (14.09) 15,349 (16.47) 15,264 (16.55) 85 (9.19) 15,349 (16.47) 15,211 (16.47) 154 (15.28) 15,365 (16.46) Fully vaccinated and booster 8058 (0.53) 83 (0.09) 8141 (0.51) 76 (0.09) 4 (0.07) 80 (0.09) 80 (0.09) 0 (0.00) 80 (0.09) 80 (0.09) 0 (0.00) 80 (0.09) Flu vaccine* y Vaccinated 661,912 (43.77) 29,005 (30.75) 690,917 (43.00) 24,830 (28.44) 3396 (57.84) 28,226 (30.29) 27,740 (30.07) 486 (52.54) 28,226 (30.29) 27,251 (29.80) 836 (82.94) 28,357 (30.38) Unvaccinated 850,503 (56.23) 65,329 (69.25) 915,832 (57.00) 62,478 (71.56) 2475 (42.16) 64,953 (69.71) 64,514 (69.93) 439 (47.46) 64,953 (69.71) 64,819 (70.20) 172 (17.06) 64,991 (69.62) Pneumococcal conjugate vaccine**, y Vaccinated 165,174 (10.92) 9207 (9.76) 174,381 (10.85) 8416 (9.64) 631 (10.75) 9047 (9.71) 8933 (9.68) 114 (12.32) 9047 (9.71) 8868 (9.60) 200 (19.84) 9068 (9.71) Unvaccinated 1,347,241 (89.08) 85,127 (90.24) 1,432,368 (89.15) 78,892 (90.36) 5240 (89.25) 84,132 (90.29) 83,321 (90.32) 811 (87.68) 84,132 (90.29) 83,472 (90.40) 808 (80.16) 84,280 (90.29) Pneumococcal polysaccharide vaccine* y Vaccinated 178,026 (11.77) 7008 (7.43) 185,034 (11.52) 5404 (6.19) 1286 (21.90) 6690 (7.18) 6484 (7.03) 206 (22.27) 6690 (7.18) 6419 (6.95) 321 (31.85) 6740 (7.22) Unvaccinated 1,334,389 (88.23) 87,326 (92.57) 1,421,715 (88.48) 81,904 (93.81) 4585 (78.10) 86,489 (92.82) 85,770 (92.97) 719 (77.73) 86,489 (92.82) 85,921 (93.05) 687 (68.15) 86,608 (92.78) Presence of comorbidities (yes) Obesity* y 236,286 (15.62) 12,737 (13.50) 249,023 (15.50) 10,515 (12.04) 1829 (31.15) 12,344 (13.25) 11,960 (12.96) 384 (41.51) 12,344 (13.25) 12,108 (13.11) 287 (28.47) 12,395 (13.28) Depression* y 204,577 (13.53) 8270 (8.77) 212,847 (13.25) 6839 (7.83) 1156 (19.69) 7995 (8.58) 7849 (8.51) 146 (15.78) 7995 (8.58) 7826 (8.48) 204 (20.24) 8030 (8.60) Diabetes* y 168,868 (11.17) 7115 (7.54) 175,983 (10.95) 5274 (6.04) 1488 (25.34) 6762 (7.26) 6510 (7.06) 252 (27.24) 6762 (7.26) 6490 (7.03) 344 (34.13) 6834 (7.32) COPD* y 75,539 (4.99) 2366 (2.51) 77,905 (4.85) 1562 (1.79) 637 (10.85) 2199 (2.36) 2125 (2.30) 74 (8.00) 2199 (2.36) 2064 (2.24) 161 (15.97) 2225 (2.38) Atrial fibrillation* † 89,838 (5.94) 2906 (3.08) 92,744 (5.77) 1825 (2.09) 827 (14.09) 2652 (2.85) 2559 (2.77) 93 (10.05) 2652 (2.85) 2397 (2.60) 317 (31.45) 2714 (2.91) Hypertension* y 390,812 (25.84) 16,672 (17.67) 407,484 (25.36) 12,959 (14.84) 2980 (50.76) 15,939 (17.11) 15,479 (16.78) 460 (49.73) 15,939 (17.11) 15,342 (16.61) 734 (72.82) 16,076 (17.22) Heart failure* y 81,349 (5.38) 2191 (2.32) 83,540 (5.20) 1173 (1.34) 756 (12.88) 1929 (2.07) 1878 (2.04) 51 (5.51) 1929 (2.07) 1659 (1.80) 332 (32.94) 1991 (2.13) Kidney failure* y 79,955 (5.29) 2185 (2.32) 82,140 (5.11) 1247 (1.43) 728 (12.40) 1975 (2.12) 1896 (2.06) 79 (8.54) 1975 (2.12) 1739 (1.88) 293 (29.07) 2032 (2.18) Cancer* y 159,940 (10.58) 5289 (5.61) 165,229 (10.28) 4120 (4.72) 895 (15.24) 5015 (5.38) 4120 (4.72) 895 (15.24) 5015 (5.38) 4799 (5.20) 270 (26.79) 5069 (5.43) Ischemic cardiopathology* y 73,100 (4.83) 2585 (2.74) 75,685 (4.71) 1785 (2.04) 617 (10.51) 2402 (2.58) 2310 (2.50) 92 (9.95) 2402 (2.58) 2244 (2.43) 203 (20.14) 2447 (2.62) Stroke* y 65,833 (4.35) 2075 (2.20) 67,908 (4.23) 1431 (1.64) 493 (8.40) 1924 (2.06) 1866 (2.02) 58 (6.27) 1924 (2.06) 1775 (1.92) 198 (19.64) 1973 (2.11) Epilepsy** y 23,544 (1.56) 963 (1.02) 24,507 (1.53) 778 (0.89) 141 (2.40) 919 (0.99) 901 (0.98) 18 (1.95) 919 (0.99) 889 (0.96) 39 (3.87) 928 (0.99) Dementia** y 55,383 (3.66) 1859 (1.97) 57,242 (3.56) 1250 (1.43) 490 (8.35) 1740 (1.87) 1736 (1.88) 4 (0.43) 1740 (1.87) 1517 (1.64) 250 (24.80) 1767 (1.89) 8 (0.86) (continued on next page) N. Mallah et al. Environmental Research 215 (2022) 114252 6 4. Discussion The present population-based study showed high effectiveness of a third-dose booster schedule in preventing SARS-CoV-2 infection in Galicia-Spain. We reported that people who received a booster dose of COVID-19 vaccine are substantially more protected against SARS-CoV-2 infection than non-boosted individuals who had received any of the authorized vaccines but not the booster dose (VE =87% versus VE = 66%). Our data support the utility of a booster dose of COVID-19 vaccine in developed country settings such as that of Spain in both age and comorbidity-based indications. During our study period, most individuals who had received the booster dose were older than 65 years as this age range represents a priority group for vaccination in Spain. Upon stratification by age, elevated VE (>80%) from a booster dose against SARS-CoV-2 infection was maintained in people aged 66–80 years, as well as those 81 years and above. As for the younger age categories, VE was estimated in a mixed subpopulation of non-boosted and boosted individuals due to insufficient number of observations among individuals with a booster dose, and a protection against infection ranging between 51% and 76% was observed. We also showed that administering a booster dose is also associated with an important decrease in the likelihood of hospitalization for COVID-19 among people over 65 years (VE >70%). VE against ICU admission and in-hospital mortality was evaluated in the mixed vaccinated population of non-boosted and boosted individuals due to lack of observations among the boosted group. Vaccination against COVID-19 contributed to lowering the odds of ICU admission for COVID19 by more than 80% in people aged between 46 and 65 years as well as in those older than 65 years. It also revealed 65% protection against death for COVID-19 in individuals older than 65 years. Studies on the impact of administering a booster dose are emerging and our findings are comparable to reports available so far. A study in Chile showed that a three-dose schedule prevents infection with SARSCoV-2 between 79% and 97%, depending on the vaccine type (Jara et al., 2022). The study also reported considerable VE that ranged between 86% and 99% against hospitalization, ICU admission and death in individuals who received a booster dose (Jara et al., 2022). A preliminary study in Israel, reported 86% reduction in the odds of testing positive for SARS-CoV-2 in individuals who received a booster dose of BNT162b2 relative to those who received only two doses of the vaccine (Patalon et al., 2022). A second study in Israel also demonstrated that the rates of SARS-CoV-2 infection and COVID-19 related hospitalization was lower in the boosted group than those in the non-boosted group by factors of 11.3 and 19.5, respectively (Bar-On et al., 2021). Arbel and colleagues showed that the COVID19-related mortality rate is 90% lower in Israeli people with a booster dose of BNT162b2 (Arbel et al., 2021). Barda et al. evaluated the effectiveness of a booster dose of BNT162b2 in Israel and estimated VE of 93% against COVID-19-related hospitalization and 81% against COVID-19-related death (Barda et al., 2021). A study in England reported that the effectiveness of a booster dose against symptomatic COVID-19 ranged from 94% to 97% and was similar in all age groups, and that against hospitalization or death oscillated between 97% and 99% in all age groups (Andrews et al., 2022). In the United States, VE against COVID-19–associated hospitalization was at least 90% among persons who had received a third dose of mRNA vaccine ≥14 days earlier (Thompson et al., 2022). In Spain, to the best of our knowledge, this is the first report on the effectiveness of the boosted dose of COVID-19 vaccine. As several particularly vulnerable groups were not included in sufficient numbers in clinical trials on COVID-19 vaccines, quantifying realworld VE, including both biological and behavioral effects is essential. In the present study, we found that administering a booster dose in Table 1 (continued) Characteristic Infection N (%) Hospitalization N (%) ICU Admission N (%) In-hospital Death N (%) No Yes Total No Yes Total No Yes Total No Yes Total Parkinson disease y 15,017 (0.99) 463 (0.49) 15,480 (0.96) 289 (0.33) 132 (2.25) 421 (0.45) 413 (0.45) 421 (0.45) 366 (0.40) 64 (6.35) 430 (0.46) HIV 4339 (0.29) 143 (0.15) 4482 (0.28) 127 (0.15) 14 (0.24) 141 (0.15) 140 (0.15) 1 (0.11) 141 (0.15) 141 (0.15) 0 (0.00) 141 (0.15) Multi-medicated (yes)* y 57,037 (3.77) 1741 (1.85) 58,788 (3.66) 1161 (1.33) 458 (7.80) 1619 (1.74) 1569 (1.70) 50 (5.41) 1619 (1.74) 1495 (1.62) 150 (14.88) 1645 (1.76) Proxy of socioeconomic status Receiving social protection* y No 1,465,683 (96.91) 92,883 (98.46) 1,558,566 (97.00) 86,303 (98.85) 5521 (94.04) 91,824 (98.55) 90,933 (98.57) 891 (96.32) 91,824 (98.55) 91,058 (98.61) 921 (91.37) 91,979 (98.53) Yes 46,732 (3.09) 1451 (1.54) 48,183 (3.00) 1005 (1.15) 350 (5.96) 1355 (1.45) 1321 (1.43) 34 (3.68) 1355 (1.45) 1282 (1.39) 87 (8.63) 1396 (1.47) Contribution to paying medicine costs according to income* ‡ 50%–60% of medicine cost 275,993 (19.21) 20,257 (22.25) 296,250 (19.39) 19,649 (23.02) 550 (11.36) 20,199 (22.39) 20,082 (22.45) 117 (15.92) 20,199 (22.39) 20,195 (22.39) 4 (33.33) 20,199 (22.39) 30%–40% of medicine cost 634,009 (44.13) 45,661 (50.15) 679,670 (44.49) 44,081 (51.64) 1410 (29.12) 45,491 (50.43) 25,274 (50.60) 217 (29.52) 45,491 (50.43) 45,490 (50.44) 1 (8.33) 45,491 (50.43) 10% of medicine cost 247,385 (17.22) 10,997 (12.08) 258,382 (16.91) 9138 (10.71) 1507 (31.12) 10,645 (11.80) 10,414 (11.64) 231 (31.43) 10,645 (11.80) 10,640 (11.80) 5 (41.67) 10,645 (11.80) 0% of medicine cost 279,206 (19.44) 14,140 (15.53) 293,346 (19.20) 12,493 (14.64) 1375 (28.40) 13,868 (15.37) 13,698 (15.31) 170 (23.13) 13,868 (15.37) 13,866 (15.37) 2 (16.67) 13,868 (15.37) Unknown or missing 75,822 (5.00) 3279 (3.50) 79,101 (4.90) 1947 (2.20) 1029 (17.5) 2976 (3.20) 2786 (3.00) 190 (20.50) 2976 (3.20) 2149 (2.30) 12 (1.20) 3145 (3.40) N: the number of PCR tests; ICU: intensive care unit; Hospitalization, ICU, and in-hospital death were determined in the subpopulation of SARS-CoV-2 positive PCR tests. All variables listed in Table 1 showed a statistically significant association with SARS-CoV-2 infection (Х 2 p-value <0.001). They also showed a statistically significant association with hospitalization for COVID-19 (Х 2 p-value of Pneumococcal conjugate vaccine =0.005; Х 2 p-value of the other variables <0.001), except HIV (p-value =0.076). Х 2 p-value of the association with ICU admission for COVID-19 >0.05. Х 2 p-value of the association with ICU admission for COVID-19 >0.05. * : Х 2 p-value of the association with ICU admission for COVID-19 < 0.001. ** : Х 2 p-value of the association with ICU admission for COVID-19 < 0.005. Variables without any asterisk showed. † Х 2 p-value of the association with in-hospital death for COVID-19 < 0.001. ‡ Х 2 p-value of the association with in-hospital death for COVID-19 < 0.005. Variables without any dagger showed. N. Mallah et al. Environmental Research 215 (2022) 114252 7 comorbid patients importantly increments the protection against SARSCoV-2 infection. The increase in protection was the most noticeable in boosted patients with major health problems like cancer (from 54% to 81%) and obesity (from 67% to 91%) as compared to non-boosted patients. VE against severe COVID-19 (hospitalization, ICU admission or death) was assessed in the mixed population of non-boosted and boosted people. The odds of COVID-19 related hospitalization decreased between 49% and 79% after vaccinating comorbid patients. Important VE against COVID-19-related ICU admission (>80%) was seen in patients with cancer, depression, diabetes, hypertension, or obesity. Considering the deficit of studies that evaluated COVID-19 VE in subpopulations of comorbid patients, the findings of the present study could prove useful for future systematic reviews and meta-analyses. The main strength of our study lies in its population-based nature where all individuals vaccinated in Galicia-Spain during the study period were assessed for their inclusion in the study; thus, selection bias is only remotely probable. Exposure misclassification is also improbable to have occurred as in Galicia-Spain, all population had access to free of charge COVID-19 vaccine. Vaccination was managed by the regional health care organism and corresponding data were electronically Table 2 Vaccine effectiveness (VE) against SARS-CoV-2 infection. Covid-19 vaccination Infection No N (%) Yes N (%) VE* (95%CI); p-value ≥ 11 years Unvaccinated 917,498 (60.66) 69,816 (74.01) Reference Partially vaccinated 158,103 (10.45) 8937 (9.47) 36% (35%; 38%); p-value <0.0001 Non-boosted 428,756 (28.35) 15,498 (16.43) 66% (65%; 67%); p-value <0.0001 Boosted 8058 (0.53) 83 (0.09) 87% (83%; 89%); p-value <0.0001 11–17 years Unvaccinated 99,133 (91.70) 9997 (92.68) Reference Partially vaccinated 4516 (4.18) 705 (6.54) 1% (−9%, 9%); p-value = 0.849 Non-boosted/ boosted 4454 (4.12) 85 (0.79) 76% (69%; 82%); p-value <0.0001 18–30 years Unvaccinated 180,126 (78.11) 18,785 (83.98) Reference Partially vaccinated 21,694 (9.41) 2503 (11.19) 23% (19%; 26%); p-value <0.0001 Non-boosted/ boosted 28,781 (12.48) 1080 (4.83) 69% (67%; 71%); p-value <0.0001 31–45 years Unvaccinated 243,684 (69.21) 15,082 (70.59) Reference Partially vaccinated 35,834 (10.18) 2924 (13.69) 8% (3%; 12%); p-value <0.0001 Non-boosted/ boosted 72,552 (20.61) 3359 (15.72) 52% (50%; 54%); p-value <0.0001 46–55 years Unvaccinated 145,184 (59.23) 9432 (66.46) Reference Partially vaccinated 25,117 (10.25) 718 (5.06) 64% (61%; 67%); p-value <0.0001 Non-boosted/ boosted 74,801 (30.52) 4043 (28.49) 51% (48%; 55%); p-value <0.0001 56–65 years Unvaccinated 111,703 (54.57) 6805 (65.78) Reference Partially vaccinated 30,547 (14.92) 867 (8.38) 53% (49%; 57%); p-value <0.0001 Non-boosted/ boosted 62,435 (30.50) 2673 (25.84) 51% (47%; 53%); p-value <0.0001 66–80 years Unvaccinated 94,054 (43.75) 6474 (65.82) Reference Partially vaccinated 21,520 (10.01) 606 (6.16) 41% (35%; 46%); p-value <0.0001 Non-boosted 97,155 (45.20) 2735 (27.81) 67% (64%; 70%); p-value <0.0001 Boosted 2228 (1.04) 21 (0.21) 85% (77%; 91%); p-value <0.0001 ≥81 years Unvaccinated 43,614 (27.80) 3241 (59.58) Reference Partially vaccinated 18,875 (12.03) 614 (11.29) 48% (43%; 53%); p-value <0.0001 Non-boosted 90,037 (57.39) 1535 (28.22) 79% (77%; 81%); p-value <0.0001 Boosted 4371 (2.79) 50 (0.92) 82% (75%; 87%); p-value <0.0001 * VE was adjusted for age, sex, and time between outcome occurrence and pandemic initiation. Table 3 Vaccine effectiveness (VE) against COVID-19-related hospitalization. Covid-19 vaccination Hospitalization No N (%) Yes N (%) VE* (95%CI); p-value ≥ 11 years Unvaccinated 64,251 (73.59) 4609 (78.50) Reference Partially vaccinated 8459 (9.69) 431 (7.34) 42% (35%; 48%); p-value <0.0001 Non-boosted/ boosted 14,598 (16.72) 831 (14.15) 72% (68%; 75%); p-value <0.0001 11–17 years Unvaccinated 9945 (92.65) 45 (97.83) Partially vaccinated 704 (6.56) 1 (2.17) NA Non-boosted/ boosted 85 (0.79) 0 (0.00) NA 18–30 years Unvaccinated 18,462 (83.91) 283 (88.44) Reference Partially vaccinated 2468 (11.22) 30 (9.38) 29% (−5%; 51%); p-value = 0.086 Non-boosted/ boosted 1072 (4.87) 7 (2.19) 69% (29%; 87%); p-value = 0.006 31–45 years Unvaccinated 14,424 (70.08) 601 (84.53) Reference Partially vaccinated 2848 (13.84) 69 (9.70) 58% (45%; 68%); p-value <0.0001 Non-boosted/ boosted 3311 (16.09) 41 (5.77) 76% (66%; 83%); p-value <0.0001 46–55 years Unvaccinated 8666 (65.13) 670 (85.68) Reference Partially vaccinated 681 (5.12) 35 (4.48) 46% (22%; 63%); p-value = 0.001 Non-boosted/ boosted 3959 (29.75) 77 (9.85) 84% (78%; 88%); p-value <0.0001 56–65 years Unvaccinated 5888 (63.85) 769 (81.20) Reference Partially vaccinated 788 (8.54) 69 (7.29) 45% (27%; 59%); p-value <0.0001 Non-boosted/ boosted 2546 (27.61) 109 (11.51) 76% (68%; 83%); p-value <0.0001 66–80 years Unvaccinated 4828 (62.26) 1298 (78.24) Reference Partially vaccinated 492 (6.34) 105 (6.33) 29% (11%; 44%); p-value = 0.004 Non-boosted/ boosted 2435 (31.40) 256 (15.43) 71% (62%; 78%); p-value <0.0001 ≥81 years Unvaccinated 2038 (54.99) 943 (67.07) Reference Partially vaccinated 478 (12.90) 122 (8.68) 48% (35%; 59%); p-value <0.0001 Non-boosted/ boosted 1190 (32.11) 341 (24.25) 73% (64%; 80%); p-value <0.0001 * : VE was adjusted for age, sex, and time between outcome occurrence and pandemic initiation. NA: VE estimation is not applicable due to a lack of observations. N. Mallah et al. Environmental Research 215 (2022) 114252 8 registered. The estimated VE was adjusted for a wide range of possible confounding variables including socio-demographic, socioeconomic, and clinical variables. The extent of exposure to SARS-CoV-2 varies across settings which if present would bias VE estimates, and residual confounding from unmeasured factors such as health seeking behavior, application of different non-pharmaceutical measures during the study period and adherence to the use of facemasks might have still been present. The outcomes, SARS-CoV-2 infection and disease severity, in our study were ascertained through clinical and medical based records, yet the possibility of outcome misclassification from a false positive or a false negative test result or from inaccurate diagnosis codes cannot be Table 4 Vaccine effectiveness (VE) against SARS-CoV-2 infection and COVID-19 hospitalization stratified by comorbidity. Covid-19 vaccination Infection No N (%) Yes N (%) VE* (95%CI); p-value Obesity Unvaccinated 127,335 (53.89) 9223 (72.41) Reference Partially vaccinated 25,207 (10.67) 1077 (8.46) 41% (37%; 45%); p-value <0.0001 Non-boosted 81,711 (34.58) 2426 (19.05) 67% (65%; 69%); p-value <0.0001 Boosted 2033 (0.86) 11 (0.09) 91% (83%; 95%); p-value <0.0001 Depression Unvaccinated 99,231 (48.55) 5383 (65.09) Reference Partially vaccinated 22,995 (11.24) 811 (9.81) 36% (30%; 41%); p-value <0.0001 Non-boosted 79,917 (39.06) 2056 (24.86) 62% (59%; 65%); p-value <0.0001 Boosted 2344 (1.15) 20 (0.24) 83% (74%; 89%); p-value <0.0001 Diabetes Unvaccinated 73,600 (43.58) 4681 (65.79) Reference Partially vaccinated 19,514 (11.56) 597 (8.39) 44% (38%; 49%); p-value <0.0001 Non-boosted 73,354 (43.44) 1819 (25.57) 65% (62%; 68%); p-value <0.0001 Boosted 2400 (1.42) 18 (0.25) 84% (75%; 90%); p-value <0.0001 COPD Unvaccinated 31,329 (41.47) 1532 (64.75) Reference Partially vaccinated 8278 (10.96) 161 (6.80) 48% (38%; 56%); p-value <0.0001 Non-boosted 34,691 (45.92) 666 (28.15) 61% (55%; 67%); p-value <0.0001 Boosted 1241 (1.64) 7 (0.30) 84% (66%; 93%); p-value <0.0001 Atrial fibrillation Unvaccinated 33,498 (37.29) 1882 (64.76) Reference Partially vaccinated 10,004 (11.14) 228 (7.85) 43% (35%; 51%); p-value <0.0001 Non-boosted 44,643 (49.69) 781 (26.88) 64% (58%; 69%); p-value <0.0001 Boosted 1693 (1.88) 15 (0.52) 73% (54%; 85%); p-value <0.0001 Hypertension Unvaccinated 169,915 (43.48) 10,930 (65.56) Reference Partially vaccinated 44,592 (11.41) 1341 (8.04) 45% (41%; 48%); p-value <0.0001 Non-boosted 170,803 (43.70) 4348 (26.08) 65% (63%; 68%); p-value <0.0001 Boosted 5502 (1.41) 53 (0.32) 82% (76%; 86%); p-value <0.0001 Cardiac failure Unvaccinated 30,688 (37.72) 1441 (65.77) Reference Partially vaccinated 9201 (11.31) 210 (9.58) 33% (22%; 43%); p-value <0.0001 Non-boosted 39,902 (49.05) 530 (24.19) 70% (64%; 75%); p-value <0.0001 Boosted 1558 (1.92) 10 (0.46) 80% (61%; 90%); p-value <0.0001 Kidney failure Unvaccinated 30,755 (38.47) 1411 (64.58) Reference Partially vaccinated 9021 (11.28) 183 (8.38) 39% (28%; 48%); p-value <0.0001 Non-boosted 38,152 (47.72) 581 (26.59) 67% (61%; 72%); p-value <0.0001 Boosted 2027 (2.54) 10 (0.46) 85% (70%; 92%); p-value <0.0001 Table 4 (continued) Covid-19 vaccination Infection No N (%) Yes N (%) VE* (95%CI); p-value Cancer Unvaccinated 73,865 (46.18) 3445 (65.14) Reference Partially vaccinated 18,558 (11.60) 426 (8.05) 35% (28%; 48%); p-value <0.0001 Non-boosted 65,200 (40.77) 1403 (26.53) 54% (49%; 59%); p-value <0.0001 Boosted 2317 (1.45) 15 (0.28) 81% (68%; 88%); p-value <0.0001 Ischemic heart disease Unvaccinated 30,813 (42.15) 1686 (65.22) Reference Partially vaccinated 8202 (11.22) 193 (7.47) 38% (27%; 47%); p-value <0.0001 Non-boosted 32,960 (45.09) 697 (26.96) 56% (49%; 63%); p-value <0.0001 Boosted 1125 (1.54) 9 (0.35) 79% (60%; 90%); p-value <0.0001 Stroke Unvaccinated 25,764 (39.14) 1328 (64.00) Reference Partially vaccinated 7782 (11.82) 243 (11.71) 29% (18%; 38%); p-value <0.0001 Non-boosted 31,093 (47.23) 496 (23.90) 66% (60%; 71%); p-value <0.0001 Boosted 1194 (1.81) 8 (0.39) 85% (68%; 93%); p-value <0.0001 Epilepsy Unvaccinated 11,238 (47.73) 653 (67.81) Reference Partially vaccinated 2794 (11.87) 123 (12.77) 26% (9%; 40%); p-value = 0.004 Non-boosted/ boosted 9512 (40.40) 187 (19.42) 66% (58%; 73%); p-value <0.0001 Dementia Unvaccinated 15,244 (27.52) 1047 (56.32) Reference Partially vaccinated 7621 (13.76) 320 (17.21) 38% (29%; 45%); p-value <0.0001 Non-boosted 31,054 (56.07) 477 (25.66) 81% (77%; 84%); p-value <0.0001 Boosted 1464 (2.64) 15 (0.81) 88% (78%; 90%); p-value <0.0001 Parkinson Unvaccinated 4790 (31.90) 275 (59.40) Reference Partially vaccinated 2011 (13.39) 70 (15.12) 35% (15%; 50%); p-value =0.002 Non-boosted/ boosted 8216 (54.71) 118 (25.49) 74% (64%; 81%); p-value <0.0001 HIV Unvaccinated 2350 (54.16) 90 (62.94) Reference Partially vaccinated 531 (12.24) 10 (6.99) 35% (−30%; 67%); p-value =0.229 Non-boosted/ boosted 1458 (33.60) 43 (30.07) 30% (−16%; 57%); p-value =0.166 COPD: chronic obstructive pulmonary disease. HIV: human immunodeficiency virus. * : VE was adjusted for sex, age, and time between outcome occurrence and pandemic initiation. N. Mallah et al. Environmental Research 215 (2022) 114252 9 ruled out. If it occurred, this non-differential misclassification could have underestimated the VE. The imperfect sensitivity of PCR testing could cause misclassification, which could attenuate VE estimates (Fonseca et al., 2021). We excluded from the analysis individuals who were infected with SARS-CoV-2 before the study initiation in order to account for acquired immunity from a past infection. Nonetheless, asymptomatic patients are likely not to be tested for SARS-CoV-2 which if present could overestimate our findings on VE. The time passed since booster dose administration until outcome development could influence VE estimates. We did not stratify for this variable due to insufficient observations, however, we controlled the analysis for the time between COVID-19 pandemic initiation and outcome occurrence. Other studies are encouraged to control for time since booster dose administration. Our data lacks information on SARS-CoV-2 variant in positively tested individuals, accordingly VE stratified by SARS-CoV-2 variant is lacking. SARS-CoV-2 variants might affect COVID-19 VE, consequently, a continuous assessment of vaccine performance is needed. Future studies are required to estimate the effectiveness of booster-based schedule on COVID-19 severity in young adults. 5. Conclusions The need for booster dose(s) of COVID-19 vaccines remains an issue of debate. Vaccine performance varies across settings and populations, hence determining the extent and duration of VE using real-world data is crucial to inform related authorities and design specific prevention programs that take into account the vaccination calendar, the need for booster dose/s and the vulnerable populations. Our findings suggest that, in settings like Spain, booster-based vaccine schedule increments the protection against to SARS-CoV-2 infection and COVID-19 severity. Importantly, COVID-19 vaccine booster administration considerably protects patients with major comorbidities. Table 5 Vaccine effectiveness (VE) against COVID-19 hospitalization stratified by comorbidity. Covid-19 vaccination Hospitalization No N (%) Yes N (%) VE* (95%CI); p-value Obesity Unvaccinated 7436 (70.72) 1459 (79.77) Reference Partially vaccinated 929 (8.83) 135 (7.38) 43% (30%; 54%); p-value <0.0001 Non-boosted/ boosted 2150 (20.45) 235 (12.85) 78% (73%; 83%); p-value <0.0001 Depression Unvaccinated 4303 (62.92) 856 (74.05) Reference Partially vaccinated 713 (10.43) 88 (7.61) 52% (38%; 63%); p-value <0.0001 Non-boosted/ boosted 1823 (26.66) 212 (18.34) 69% (59%; 77%); p-value <0.0001 Diabetes Unvaccinated 3267 (61.95) 1124 (75.54) Reference Partially vaccinated 476 (9.03) 108 (7.26) 48% (34%; 59%); p-value <0.0001 Non-boosted/ boosted 1531 (29.03) 256 (17.20) 74% (65%; 80%); p-value <0.0001 COPD Unvaccinated 931 (59.60) 461 (72.37) Reference Partially vaccinated 122 (7.81) 35 (5.49) 43% (14%; 63%); p-value = 0.008 Non-boosted/ boosted 509 (32.59) 141 (22.14) 71% (51%; 82%); p-value <0.0001 Atrial fibrillation Unvaccinated 1076 (58.96) 592 (71.58) Reference Partially vaccinated 161 (8.82) 57 (6.89) 42% (19%; 58%); p-value = 0.001 Non-boosted/ boosted 588 (32.22) 178 (21.52) 68% (51%; 79%); p-value <0.0001 Hypertension Unvaccinated 8074 (62.30) 2247 (75.40) Reference Partially vaccinated 1110 (8.57) 209 (7.01) 47% (37%; 55%); p-value <0.0001 Non-boosted/ boosted 3775 (29.13) 524 (17.58) 73% (67%; 78%); p-value <0.0001 Cardiac failure Unvaccinated 682 (58.14) 543 (71.83) Reference Partially vaccinated 140 (11.94) 58 (7.67) 52% (33%; 66%); p-value <0.0001 Non-boosted/ boosted 351 (29.92) 155 (20.50) 72% (54%; 53%); p-value <0.0001 Kidney failure Unvaccinated 716 (57.42) 519 (71.29) Reference Partially vaccinated 137 (10.99) 39 (5.36) 66% (50%; 77%); p-value <0.0001 Non-boosted/ boosted 394 (31.60) 170 (23.35) 78% (60%; 88%); p-value <0.0001 Cancer Unvaccinated 2545 (61.77) 666 (74.41) Reference Partially vaccinated 364 (8.83) 56 (6.26) 45% (25%; 60%); p-value <0.0001 Non-boosted/ boosted 1211 (29.39) 173 (19.33) 66% (50%; 56%); p-value <0.0001 Ischemic heart disease Unvaccinated 1094 (61.29) 443 (71.80) Reference Partially vaccinated 145 (8.12) 40 (6.48) 39% (10%; 58%); p-value = 0.014 Non-boosted/ boosted 546 (30.59) 134 (21.72) 49% (19%; 67%); p-value = 0.004 Stroke Unvaccinated 848 (59.26) 357 (72.41) Reference Partially vaccinated 191 (13.35) 42 (8.52) 55% (35%; 69%); p-value <0.0001 392 (27.39) 94 (19.07) Table 5 (continued) Covid-19 vaccination Hospitalization No N (%) Yes N (%) VE* (95%CI); p-value Non-boosted/ boosted 70% (49%; 83%); p-value <0.0001 Epilepsy Unvaccinated 518 (66.58) 103 (73.05) Reference Partially vaccinated 100 (12.85) 20 (14.18) 8% (−62%; 47%); p-value = 0.785 Non-boosted/ boosted 160 (20.57) 18 (12.77) 32% (−55%; 71%); p-value = 0.355 Dementia Unvaccinated 634 (50.72) 322 (65.71) Reference Partially vaccinated 251 (20.08) 64 (13.06) 48% (29%; 62%); p-value <0.0001 Non-boosted/ boosted 365 (29.20) 104 (21.22) 79% (62%; 89%); p-value <0.0001 Parkinson Unvaccinated 149 (51.56) 96 (72.73) Reference Partially vaccinated 61 (21.11) 8 (6.06) 80% (60%; 91%); p-value <0.0001 Non-boosted/ boosted 79 (27.34) 28 (21.21) 73% (−4%; 93%); p-value = 0.058 HIV Unvaccinated 79 (62.20) 9 (64.29) Reference Partially vaccinated 9 (7.09) 1 (7.14) NA Non-boosted/ boosted 39 (30.71) 4 (28.57) NA COPD: chronic obstructive pulmonary disease. HIV: human immunodeficiency virus. NA: VE estimation is not applicable due to a lack of observations. * : VE was adjusted for sex, age, and time between outcome occurrence and pandemic initiation. N. Mallah et al.