Effectiveness of BBV152 vaccine against SARS-CoV-2 infections, hospitalizations and deaths among healthcare workers in the setting of high delta variant transmission in New Delhi, India

Study design

This was an observational (retrospective cohort) study which collected data from different HCWs employed at the All India Institute of Medical Sciences (AIIMS), New Delhi, India, a public-funded, teaching, and multi-speciality tertiary care institute. The contact details of all HCWs (22 723) were obtained from the administrative division, which maintains service records of all HCWs within the Institute. The study period in this report was from 10 April to 24 June 2021.

Study participants

All employees (regular/contractual) and students at AIIMS were eligible to participate. Throughout the pandemic, AIIMS ran a special employees’ health clinic for testing and management of Covid-19. Only the BBV152 vaccine was offered at AIIMS from 16 January 2021. HCWs who received Covishield vaccine were excluded from the current report.

Data collection

All employees and students were contacted through e-mail, social media platforms like WhatsApp, and telephone from 12 May to 24 June 2021. The data were collected either through a self-administered quality assurance measure in-built and web-based form (Google LLC, Mountain View, CA) or through telephone interviews by trained personnel. The questionnaire included demographics (age/gender/type of HCW), comorbid condition, self-reported height/weight, diagnosis of Covid-19 (yes/no), history of vaccination (none/single dose/two doses and dates of receipt) and among those with a positive diagnosis of Covid-19, information on the date of infection, number of episodes (single or more), mode of diagnosis (RTPCR [Reverse Transcription Polymerase Chain Reaction], rapid antigen test [RAT], CBNAAT [cartridge based nucleic acid amplification test]), type of symptoms, severity including hospitalization details, and the outcome (recovery/persistence of symptoms/death). Additionally, family members of the deceased HCWs were interviewed telephonically to enquire about deaths during the study period. All HCWs who reported infection and 20% of those who were not infected were contacted again to validate the reported findings through vaccination cards and test reports. We followed the STROBE guidelines for reporting the study findings.¹⁷

Definitions

1. SARS-CoV-2 infection positivity detected by either molecular tests or antigen tests (RT-PCR/CBNAAT/RAT).¹⁰

2. Severity of disease was graded as asymptomatic/symptomatic, with details of symptomatic persons along with their severity (based on the WHO ordinal scale for clinical improvement), including hospitalization and death.¹⁸

3. Symptomatic SARS-CoV-2 infection: Participant with any of the following symptoms was considered symptomatic: fever, rhinorrhoea, sore throat, cough, chest pain, wheezing, difficulty in breathing, shortness of breath, anosmia, dysgeusia, fatigue, myalgia, headache, abdominal pain, nausea, and diarrhoea.

4. Comorbid conditions considered were diabetes, hyper-tension, heart disease, chronic lung disease, chronic kidney disease, cancer, hypothyroidism, and any other self-reported chronic condition.

Statistical analyses

Data management and statistical analysis were done using Stata/SE 15.0 (StataCorp LLC, College Station, TX, USA). Categorical variables were presented as numbers (%), and continuous variables were presented as mean (SD)/median (IQR).

Demographic and clinical variables such as age (<25, 25–44 and >45 years), sex (male and female), HCW category (student/administrative/clerical staff, faculty/scientist/research staff, nursing staff, junior/senior resident, paramedical/support staff), body mass index (BMI<18.5, 18.5–24.9 and >25 kg/m²) and comorbid condition status (none and any comorbid condition) with respect to SARS-CoV-2 infection and vaccination status were compared using Pearson’s chi-square test. The Covid-19 test positivity rate was very low from mid-January to the beginning of April 2021. For assessing the vaccine effectiveness in this retrospective cohort study, we determined the exposure status as of 10 April 2021. On this date, the Covid-19 test positivity crossed 10% for the first time in Delhi.³ Also, this cutoff would provide an adequate number of subjects who would have received both doses of BBV152. Hereafter, this date is referred to as the “start of the second wave”.

Based on the status of the vaccine received, we had the following categories of subjects: A–unvaccinated throughout the study; B–unvaccinated at the start of the second wave, received dose one before the end of the study and infected before getting dose 2; C–unvaccinated at the start of the second wave and received two doses before end of the study; D–partially vaccinated at the start of the second wave and received dose two before end of the study; E–partially vaccinated throughout; and F–fully vaccinated at the start of the second wave.

The vaccine effectiveness was estimated for three outcomes: (i) any SARS-CoV-2 infection; (ii) symptomatic SARS-CoV-2 or hospitalization or death; and (iii) hospitalization or death due to SARS-CoV-2. The follow-up time was calculated from the start of the second wave to the date of diagnosis of infection, to the date of death, and to the date of interview for infected, died, and not infected due to SARSCoV-2, respectively. To avoid bias, all the HCWs with deaths and prior infection of SARS-CoV-2 on or before the second wave were excluded from the analysis. The primary group of interest for determining vaccine effectiveness (VE) was ‘F’.

The probability of infection was estimated with respect to vaccination status using the Kaplan-Meier curve, and a log-rank test was used to compare the probability of infection among the groups. Next, we used the Poisson regression model between vaccination status and outcome, controlling for demographic and clinical characteristics that could confound the association between vaccination and outcomes, including age, sex, HCW type, BMI, and comorbid status. From the estimated adjusted incidence rate ratio, we calculated the VE by subtracting the incidence rate ratio from one (VE=1–incidence rate ratio). p value <0.05 was considered statistically significant.

Ethics

The study was approved by the Institute Ethics Committee of All India Institute of Medical Sciences, New Delhi (Ref. IEC321/07.05.2021). The informed consent was obtained remotely through electronic/telephonic mode, conforming to consent procedures as per national guidelines for ethics committees reviewing biochemical and health research during the Covid-19 pandemic.¹⁹

Study flow and participants

A total of 22 723 contact numbers of employees/students were obtained from the administration and academic departments of AIIMS. These were contacted telephonically. Of these, 7479 were excluded on account of ineligibility (1924): retired/left the institute, no response (3926), non-consent (672), and duplicate/wrong entries (975). Finally, 15 244 healthcare workers (HCWs) participated in the study (Fig. 1). Further, for assessing VE, 2398 infections and 18 deaths were excluded as these had occurred before the second wave. HCWs (519) who received Covishield (483) or did not know their vaccine type (36) were also excluded. Thus, in this current report, data from 12 327 HCWs were analysed. The mean (SD) age of HCWs was 36 (11) years, and 66% were men. The different categories of HCW included were: paramedical/support staff (6871, 55.7%), nursing staff (2410, 19.5%), students/administrative/clerical staff (1293, 10.5%), and faculty/scientists/research staff/physicians (1753, 14.2%). Obesity (BMI >30 kg/m²) and any comorbid condition were reported by 8% and 16% of HCWs, respectively. The median follow-up of HCWs in this report was 48 days (IQR 35 to 69 days).

Close

FIG 1.

Study flow

Export to PPT

Vaccination exposures

HCWs were divided into six categories (A to F) according to their vaccination status at the start of the second wave. There were 3989 (32%) HCWs who were unvaccinated (A) and 3288 HCWs (27%) who were fully vaccinated (F). Among fully vaccinated HCWs, there were 2208 HCWs with intervals between the second dose and 10 April >14 days. 704 HCWs were partially unvaccinated (E) throughout the study period. The remaining HCWs (4346) had different statuses of vaccination between 10 April 2021 and the end of the study period and were distributed in the following three categories: Category B comprised 1503 HCWs. Category C had 903 HCWs, while Category D included 1940 HCWs.

Baseline characteristics and SARS-CoV-2 infection

The distribution of HCWs with respect to baseline demographic and clinical characteristics as per vaccination status is shown in Table 1, and there were significant differences observed. Among the study participants, 2484 (20%) reported having any SARS-CoV-2 infection. The incidence of SARS-CoV-2 infection among unvaccinated (A) and fully vaccinated (F) HCWs was 26.5% (1056/3989) and 18.1% (595/3288), respectively. The median (IQR) interval between receipt of the second dose and infection among category F HCWs was 47 (34–68) days.

VE

The total person days contributed by unvaccinated and fully vaccinated HCWs were 167 973 and 139 020, respectively. The incidence density for any SARS-CoV-2 infection events among unvaccinated and fully vaccinated HCWs was 62.9 (95% CI 59.2 to 66.8) and 42.8 (95% CI 39.5 to 46.4) per 10 000 person-days, respectively. The adjusted VE for preventing any SARS-CoV-2 infection was 43% (95% CI: 36 to 48, p<0.001). The incidence density for symptomatic SARSCoV-2 infection or hospitalization or death among unvaccinated and fully vaccinated HCWs was 57.5 (95% CI: 54.0 to 61.3) and 38.5 (95% CI: 35.4 to 41.9) per 10 000 person-days, respectively. The adjusted VE against the same was 44% (95% CI: 38 to 50, p<0.001). The incidence density for hospitalization or death due to SARS-CoV-2 in unvaccinated and fully vaccinated HCWs was 4.3 (95% CI 3.5 to 5.5) and 2.2 (95% CI 1.6 to 3.2) per 10 000 person-days, respectively. The VE against the same was 65% (95% CI 46 to 78, p<0.001).

We also examined the VE for fully vaccinated HCWs and segregated them as per the interval between the receipt of the second dose and 10 April 2021. The VE was almost similar for all the outcomes for HCWs with intervals of <14 days and ≥14 days (Table 2). The adjusted VE for hospitalization or death due to SARS-CoV-2 in fully vaccinated HCWs who had elapsed >14 days since receiving the second dose was 61% (95% CI 37 to 76, p<0.001). A varying significant effect was found for HCWs in categories B to D, as shown in Supplementary Table 1. Category C and D HCWs, who received nil and partial doses at the onset of the second wave, had higher effectiveness rates of 99% and 80%, respectively, against both outcomes—any SARS-CoV-2 infection and symptomatic SARS-CoV-2/hospitalization/death. VE was not found among HCWs who were partially vaccinated throughout (category E of HCWs), as shown in Supplementary Table 1. The Kaplan-Meier curves and estimates are shown in Fig. 2.

Close

FIG 2.

Kaplan-Meier curves showing vaccine effectiveness among the groups for various outcomes: (a) Any SARS-CoV-2 infection, (b) Symptomatic SARS-CoV-2 or hospitalization or death, and (c) Hospitalization or death due to SARS-CoV-2

Export to PPT

Conclusion

The BBV152 vaccine (two doses) offers modest protection against SARS-CoV-2 infection and symptomatic infection, including hospitalization and death, among a cohort of HCWs in real-world scenarios. Our estimates for VE represent the setting when the delta variant was predominant. In the absence of a protective response with the partial receipt of BBV152, efforts for maximizing coverage with a full dose schedule are vital.