ABSTRACT
The study was conducted to estimate the protective effectiveness (PE) of complete primary or booster dose regimens of COVID-19 vaccines deployed in Bangladesh. The study was conducted in four hospitals in Dhaka between December 30, 2021, and August 31, 2022 following a test-negative design. Patients aged ≥18 years attended with COVID-like symptoms were enrolled and tested for RT-PCR. Test-negative controls were matched to confirmed cases at a 1:1 ratio considering site, date, and age groups. Conditional logistic regression was used to estimate the PE considering the association between receipt of complete primary with or without a booster regimen and development of COVID-19 disease symptoms. Whole genome sequencing (WGS) was carried out to confirm the variants. RT-PCR positive 847 cases were matched to 847 controls. WGS of strains revealed 6% to be the Delta variant and 94% was Omicron variant. The PE conferred by receipt of complete primary regimen with or without booster dose of any vaccine revealed no significant protection (15%, 95% CI: −11 to 36, p = .23) against any COVID-19 disease or severe disease (14%, 95%CI: −23 to 39, p = .42). However, there was a protective association between receipt of complete primary regimen with or without booster dose of one mRNA vaccine (Pfizer-BioNTech) against any COVID-19 disease (88% (95% CI: 26 to 98, p = .023)) for the first 90 days. The analysis suggested little vaccine effectiveness during Omicron surge, with the possible exception of one mRNA-vaccine 90 days after dosing.
KEYWORDS: Bangladesh, COVID-19 vaccines, omicron variant, test-negative design (TND) vaccine effectiveness
Introduction
The World Health Organization (WHO) first reported Omicron (B.1.1.529) as a circulating variant of SARS-CoV-2 in South Africa on November 24, 2021 which became the predominant circulating variant across the globe in just 1–2 months.1 Many subvariants including BA.1, BA.2, BA.3, BA.4, BA.5, BQ.1 and BQ.1.1 emerged following the original Omicron variant over a period of approximately one year.2,3 Omicron has been highly transmissible and spreads more quickly than any preceding variants.4 Due to the high transmission rate of Omicron, Delta has been replaced by Omicron as the dominant variant in several regions of the world.5 It has been noted that the Omicron variant’s symptoms were less severe,6 despite the fact that it was able to evade the immune system, making the vaccines less effective against it. The Omicron variant was first reported on December 6, 2021 in Bangladesh and gradually replaced the Delta variant.7
Vaccination has been one of the most pressing global issues as vaccines play a critical role in preventing deaths, and hospitalization caused by COVID-19 diseases and contribute to control the spread of the disease.8 Studies showed that two doses of Pfizer-BioNTech or AstraZeneca vaccines had modest and short-term protection against infection with the Omicron variant.9 Evidence suggested that boosting with Pfizer-BioNTech or Moderna vaccines had a substantial increase in protection against mild disease.10 The vaccine effectiveness waned rapidly after 15–20 weeks of two doses and waning also occurred 5–9 weeks even after the booster doses.11,12
Data from other studies suggest that booster dose of the vaccine was effective to prevent severe diseases including hospitalization, or death caused by Omicron variant. Therefore, a recommendation was made for booster vaccination in different parts of the world.13 Various vaccines such as Serum Institute of India (ChAdOx1 nCoV-19), Sinopharm (Vero Cell-Inactivated), Pfizer-BioNTech (BNT162b2), Moderna (mRNA-1273), and Sinovac (Vero Cell-Inactivated) vaccines were used for the complete primary regimen among Bangladeshi population.14 In addition, Bangladesh started administering the booster dose of COVID-19 vaccines since December 28, 2021.15 Initially, the Pfizer-BioNTech (BNT162b2) vaccine was used and other vaccines like Moderna (mRNA-1273), Serum Institute of India (ChAdOx1 nCoV-19), and Sinopharm (Vero Cell- Inactivated) were gradually included for the booster regimens.
In a previous analysis we have shown that Moderna (mRNA-1273) vaccine exhibits 64%; (95% CI: 10 to 86, p = .029) protection among Bangladeshi population during the surge of Delta Variant of SARS-CoV-2. The protection by the receipt of any vaccine against severe disease was 85%.14 Here we evaluated the protective effectiveness of a complete primary regimen and/or a booster regimen of homologous (same as the primary vaccine) or heterologous (different from the primary vaccine) vaccines deployed against symptomatic and severe episodes of SARS-CoV-2 diseases during the predominance of Delta and Omicron variants following a test-negative design in Bangladesh.
Materials and methods
Study site, and enrollment of patients
The study was carried out in four tertiary level Hospitals in Dhaka city, Bangladesh between 30 December 2021 and 31 August 2022.14 Patients who presented with COVID-19-like symptoms14 were enrolled in the study from the outpatient isolation ward or inpatient unit of the hospitals. Patients having performed RT-PCR testing for SARS-CoV-2 infection were approached for enrollment. Adult patients who attended the hospitals with symptoms for ≤10 days and provided informed written consent were enrolled in the study.
After enrollment, clinical information, including history and physical examination of the patient was entered electronically using a structured case report form. The socioeconomic, clinical, and demographic variables that have previously been linked to COVID-19 risk, were also obtained from them. Based on the WHO criteria, the severity of disease at the time of presentation was further classified and comorbidity information based on patients’ statements was recorded and further characterized by a modified Charlson index, as described elsewhere.16 For this analysis, we included mortality and hospitalization data of the follow-up period (30 days since enrollment). Ultimate severity included severity at presentation or death or hospitalization within 30 days following the presentation.
As the results from RT-PCR testing were available on the next day of enrollment, the assessment of criteria for enrolling patients was done without having any knowledge of patient’s status of the test results. The data was also collected in a blinded manner.
Ascertainment of vaccination status
The vaccination status of each patient was recorded at the time of enrollment. Information about the date and name of each vaccine dose was obtained by inspecting the vaccination cards. The individuals who did not bring the vaccination cards, but provided verbal confirmation, vaccination status was checked with the centralized database of vaccination records of the management information system (MIS) of the GoB by linking the national identification number of the patient.14 All aforementioned processes were performed in a manner blinded to the patient’s RT-PCR results.
Specimen collection, transportation, and laboratory testing
A swab was used to collect nasopharyngeal swab (NPS) specimen which was then placed in viral transport medium (VTM) in a cold box to maintain the temperature of 2 to 8°C. The specimens collected from enrollment sites were transported to the Virology Laboratory of the International Centre for Diarrhoeal Disease Research, Bangladesh (icddr,b) and the Institute of Epidemiology and Disease Control Research (IEDCR) laboratory for RT-PCR testing.
Following the manufacturer’s instructions, viral RNA was extracted from NPS using the Qiagen Viral RNA mini-kit (Qiagen, Germany). Real-time PCR (RT-PCR) was performed following the procedures described earlier.14 The cycle threshold (CT) value of ≤ 40 was considered as positive result.14 The test result was entered electronically. The whole-genome sequencing using the MinION sequencing platform was carried out for the RT-PCR-positive samples with a ≤30 CT value.14 Lineages were assigned by the SARS-CoV-2 pangolin (github.com/cov-lineages/pangolin; accessed on February 27, 2023) classification system.
Selection of cases and controls
The RT-PCR-negative controls against each RT-PCR-positive cases at a 1:1 ratio was selected by a statistician who was not aware of the vaccination status of enrolled participants. The selection of controls was done by enrollment site, date of presentation and age-groups of cases. The sampling frame was gradually enlarged till a frame of ±7 days was reached if one matching control was not available for a case on the date of presentation. The incidence density sampling was used where chronological order was maintained to select the cases and resampling was never considered after selecting one case. However, resampling was considered for controls to select as cases or controls later on.
Sample size estimation and statistical analysis
Before analyzing the data, we developed a statistical analytic plan (SAP) and our primary aim was to assess the effectiveness of COVID-19 vaccines offered by the Government of Bangladesh. The sample-size was calculated considering 60% protection by the receipt of two-dose (complete primary regimen) of any vaccines with at least 80% power (two-tailed, p < .05). We assumed that 90% controls would be vaccinated. Our required sample size was 576 considering the aforementioned assumptions at 1:1 case-control ratio. Complete primary vaccination was defined as the receipt of two doses of the homologous or heterologous vaccines and booster vaccination was defined as the receipt of a third dose after receiving the complete primary regimen. Our primary analysis examined the protection given by a complete primary regimen and/or a booster regimen of homologous (same as the primary vaccine) or heterologous (different from the primary vaccine) vaccines administered at least 14 days before the development of COVID-19 disease symptoms of any level of severity.
In secondary analyses, we assessed the efficacy of complete primary regimen and booster regimens against all symptomatic episodes. A sub-analysis was done to measure the effectiveness by time after receiving the second or booster doses of vaccines, we stratified the period of enrollment into ≤90 days versus >90 days. Matched controls were selected for the same periods. The majority of vaccinated cases and controls were vaccinated more than 90 days prior to presentation with either the Sinopharm (Vero Cell-inactivated), Moderna (mRNA-1273), or Serum Institute of India (ChAdOx1 nCoV-19) vaccines. Almost similar numbers of cases and controls, vaccinated with Pfizer-BioNTech (BNT162b2) were enrolled for both periods (≤90 and >90 days) before presentation (Figure 1).
Figure 2.
Distribution of vaccinees* in the study since receipt of the final dose of a complete primary or booster regimen.
*Vaccinees with complete primary or booster regimens (date of presentation at least 14 days of the complete primary or booster dose).
In addition, we evaluated the effectiveness of vaccines against severe disease where WHO criteria were used to define the severity of the disease at presentation. The “ultimate severe disease” included patients who required hospitalization or died within 30 days of enrollment. Although we were unable to get adequate statistical power to measure the many sub-group protections including individual types of vaccines used in Bangladesh, we conducted exploratory analyses of protection by each vaccine as other studies revealed different levels of protection for different vaccines.
Bivariate conditional logistic regression models were used to carry out the matched bivariate analyses between relevant covariates such as behavioral, demographic, and clinical variables. Multiple conditional logistic regression was used to evaluate the vaccine protection. The matched case-control status was considered as the outcome where independent variables included vaccination status, and selected covariates. The non-receipt of vaccine was considered as the reference category.
The odds ratios for the associations between vaccination and disease status were estimated by exponentiation of the coefficient for vaccination from the fitted models, and the standard error of the coefficient was used to calculate 95% confidence intervals for the estimated effect. Protective effectiveness was calculated as ((1 − odds ratio) × 100%)). For the evaluation of vaccine protection against all disease episodes, potentially confounding variables were introduced as independent variables in the models if the bivariate associations had p-values ≤0.05 (two-tailed).
The restriction of cases and controls to those presenting severe disease was employed to help further safeguard against residual healthcare utilization bias, as both the uptake of vaccination and the use of healthcare facilities may differ in those presenting with severe versus non-severe disease. However, because of the paucity of severely ill SARS-CoV-2 patients enrolled in the study, we also conducted an exploratory analysis comparing verified cases presenting severe disease versus controls presenting severe disease, disregarding the matched selection.
In this analysis, unconditional logistic regression models were fitted, and the matching variables for the site of presentation, and age at presentation were forced into the models as covariates.
Data freezing was done prior to the analysis according to the SAP. The R statistical software (version 4.10) was used for data analyses. For estimation of the conditional logistic regression coefficients by maximizing the conditional likelihood, the “clogit” package of R was used.
Results
Enrollment of patients and selection of cases and controls for the analysis
We approached 22,278 adult patients with COVID-19 like symptoms attending study hospitals. After obtaining informed consent, 5,333 patients were enrolled in the study based on the eligibility criteria, and RT-PCR testing was done for all enrolled patients. In this analysis, 847 cases and 3,905 test-negative controls were included after excluding 566 participants who either had received a single dose of vaccine or enrolled before 14 days of receiving the second or third dose. Among 847 cases, we found 578 and 135 vaccinees with complete primary and booster regimens respectively. Out of 847 matched controls, there were 585 vaccinees with complete primary dose series and 141 vaccinees with a booster dose (Figure 2). Whole genome sequencing was done for 525 SARS-CoV-2 isolates among which 32 (6.1%) and 493 (93.9%) were Delta and Omicron variants. All the Delta variants were detected in January 2022 whereas Omicron variants were detected throughout the study period from January to August 2022.
Figure 1.
Enrollment of patients and selection of cases and controls for the analysis.
Comparability of baseline characteristics among cases and controls
When we compared the baseline characteristics of cases and controls, we found that wealthier households, and a higher body mass index were significantly (p < 0.05) associated with SARS-CoV-2 disease. Other demographic and clinical characteristics including comorbid conditions ascertained by history did not show any association with the disease (Table 1). The distribution of the baseline features of cases and controls with severe disease at presentation and the ultimate severity 30 days after presentation did not differently significant from one another. Moreover, we found 13 (1.53%) cases and 6 (0.6%) controls who died within 30 days of enrollment.
Table 1.
Comparability of baseline characteristics among cases and controls.
| Label | Cases (n = 847) | Matched controls (n = 847) | *p-value | |
|---|---|---|---|---|
| Age | In years | 38.3 ± 14 | 38.7 ± 13.6 | .240 |
| Age groups | 18–30,yrs | 291(34.4) | 291(34.4) | |
| 31–60,yrs | 489(57.7) | 489(57.7) | ||
| 61+,yrs | 67(7.9) | 67(7.9) | NA | |
| Gender | Female | 296(34.9) | 277(32.7) | |
| Male | 551(65.1) | 570(67.3) | .322 | |
| Religion | Non-muslim | 56(6.6) | 53(6.3) | |
| Muslim | 791(93.4) | 794(93.7) | .770 | |
| Body mass index (BMI) | KG/M^2 | 24.7 ± 4.1 | 24.2 ± 4.1 | .005 |
| †HH members | Count | 4.1 ± 2 | 4.3 ± 1.8 | .144 |
| †HH Income | BD Taka | 45170.6 ± 47992.1 | 40280.4 ± 38497.5 | .023 |
| ‡Smoker | No | 706(83.4) | 679(80.2) | |
| Yes | 141(16.6) | 168(19.8) | .079 | |
| ‡Ultimate Severity | No | 823(97.2) | 831(98.1) | |
| Yes | 24(2.8) | 16(1.9) | .162 | |
| Severity at presentation | No | 831(98.1) | 834(98.5) | |
| Yes | 16(1.9) | 13(1.5) | .533 | |
| §Heart disease | No | 828(97.8) | 829(97.9) | |
| Yes | 19(2.2) | 18(2.1) | .862 | |
| §Hypertension | No | 738(87.1) | 755(89.1) | |
| Yes | 109(12.9) | 92(10.9) | .192 | |
| §Lung disease | No | 843(99.5) | 836(98.7) | |
| Yes | 4(0.5) | 11(1.3) | .083 | |
| §Diabetes | No | 788(93) | 772(91.1) | |
| Yes | 59(7) | 75(8.9) | .135 | |
| §Stomach disease | No | 846(99.9) | 845(99.8) | |
| Yes | 1(0.1) | 2(0.2) | .571 | |
| §Kidney disease | No | 844(99.6) | 846(99.9) | |
| Yes | 3(0.4) | 1(0.1) | .341 | |
| ¶Modified Charlson Comorbidity Index | Score | 0.5 ± 1.3 | 0.5 ± 1.3 | .950 |
*p-values are generated using conditional logistic regression.
†HH means household.
‡Ultimate severity includes severity at presentation and death within 30 days of enrollment.
‡A person who smokes tobacco regularly.
§Ascertained by history.
¶Modified Charlson Comorbidity Index.16
Vaccine protection of a complete primary or booster regimen against any symptomatic COVID-19 disease among the Bangladeshi population
The adjusted protective effectiveness (PE) against all episodes of COVID-19 disease conferred by receipt of a complete primary and/or booster regimen of any vaccine (Serum Institute of India (ChAdOx1 nCoV- 19), Sinopharm (Vero Cell- Inactivated), Moderna (mRNA-1273), Pfizer-BioNTech (BNT162b2), and Sinovac (Vero Cell-Inactivated) was 15% (95% CI: −11 to 36, p = .228). The complete primary and booster regimens of any vaccine demonstrated 14% (95% CI: −13 to 35, p = .277) and 24% (95% CI: −12 to 48, p = .163) protection respectively.
An exploratory analysis was done to show the protection for individual vaccines; the Pfizer-BioNTech (BNT162b2) vaccine demonstrated 52% (95% CI: −14 to 80, p = .095), 61% (95% CI: −4 to 85, p = .061), 29% (95% CI: −137 to 79, p = .578) protection for the complete primary and/or booster, complete primary, and booster regimens respectively. Analyses of other individual vaccines such as Serum Institute of India (ChAdOx1 nCoV- 19), Sinopharm (Vero Cell- Inactivated), Moderna (mRNA-1273), and Sinovac (Vero Cell-Inactivated) vaccines did not show any significant protection for any regimens. When we estimated the protection by combining two mRNA vaccines (Moderna (mRNA-1273), and Pfizer-BioNTech (BNT162b2) and two Vero Cell-inactivated vaccines Sinopharm (Vero Cell- Inactivated), Sinovac (Vero Cell- Inactivated)) for the aforementioned regimens, we did not find any evidence of protection against symptomatic COVID-19 disease (Table 2).
Table 2.
Vaccine protection of a complete primary or booster regimen against any symptomatic COVID-19 disease among the Bangladeshi population.
| Cases |
Matched control |
Adjusted protective effectiveness (PE) (95% CI) |
||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Non-vaccinees | Complete primary regimen | Booster regimen | Non-vaccinees | Complete primary regimen | Booster regimen | Complete primary and/or booster regimen* | p-value | Complete primary regimen† | p-value | Complete primary with booster regimen‡ | p-value | |
| Any vaccine§ | 134(0.16) | 578(0.68) | 135(0.16) | 121(0.14) | 585(0.69) | 141(0.17) | 15(−11,36) | .228 | 14(−13,35) | .277 | 24(−12,48) | .163 |
| Serum Institute of India (ChAdOx1 nCoV-19) | 48(0.33) | 81(0.55) | 17(0.12) | 46(0.32) | 90(0.62) | 10(0.07) | 26(−33,59) | .320 | 31(−25,62) | .215 | −59(−398,49) | .429 |
| Moderna (mRNA 1273) | 40(0.43) | 32(0.34) | 21(0.23) | 38(0.41) | 26(0.28) | 29(0.31) | 18(−62,58) | .577 | −9(−131,48) | .817 | 58(−12,84) | .084 |
| Pfizer-BioNTech (BNT162b2) | 39(0.63) | 11(0.18) | 12(0.19) | 33(0.53) | 20(0.32) | 9(0.15) | 52(−14,80) | .095 | 61(−4,85) | .061 | 29(−137,79) | .578 |
| mRNA | 58(0.35) | 54(0.33) | 52(0.32) | 50(0.3) | 59(0.36) | 55(0.34) | 24(−25,54) | .274 | 22(−33,55) | .358 | 28(−38,62) | .327 |
| Sinopharm (Vero Cell- Inactivated) | 68(0.33) | 137(0.67) | – | 67(0.33) | 138(0.67) | – | 5(−44,37) | .820 | 5(−44,37) | .820 | – | – |
| Vero Cell, inactivated (Sinopharm and Sinovac) | 70(0.33) | 142(0.67) | – | 67(0.32) | 145(0.68) | – | 9(−37,39) | .661 | 9(−37,39) | .661 | – | – |
*Protective effectiveness of complete primary 2-doses of vaccine (with or without booster/3rd dose); †Protective effectiveness of complete primary 2-doses of vaccine only.
‡Protective effectiveness of complete primary 2-doses with 3rd dose of vaccine.
§Any vaccine (Serum Institute of India (ChAdOx1 nCoV- 19), Sinopharm (Vero Cell- Inactivated), Moderna (mRNA-1273), Pfizer-BioNTech (BNT162b2), and Sinovac (Vero Cell-Inactivated)).
Adjustment was done by body mass index, and household income.
CI = Confidence Interval.
Vaccine protection of a complete primary or booster regimen against any symptomatic COVID-19 disease by time since the final dose
When we evaluated the vaccine protection for the complete primary and/or booster regimen by duration, there was 25% (95% CI: −23 to 54, p = .254) protection conferred by any vaccine against symptomatic COVID-19 disease in participants who were enrolled within 90 days of the last vaccination. Beyond 90 days of this regimen, the protection of any vaccine was 11% (95% CI: −24 to 36, p = .481). In addition, the protection by a complete primary regimen of any vaccine for the first 90 days and after 90 days after dosing was 26% (95% CI: −24 to 56, p = .251), and 12% (95% CI: −23 to 37, p = .451) respectively. For the booster regimen of any vaccine, the protection was 19% (95% CI: −94 to 66, p = .641) and 3% (95% CI: −64 to 42, p = .917) for the first 90 days and after 90 days respectively.
Exploratory analyses of the vaccine types, only Pfizer-BioNTech (BNT162b2) vaccine exhibited substantially greater 88% (95% CI: 26 to 98, p = .023) protection within 90 days of complete primary and/or booster regimen. For the booster dose of the Pfizer-BioNTech (BNT162b2) vaccine, the protection was 82% (95% CI: −61 to 98, p = .124) within 90 days of dosing (Table 3). However, no protection was evident for other individual vaccines for the duration within and beyond 90 days since the last vaccination.
Table 3.
Vaccine protection of a complete primary or booster regimen against any symptomatic COVID-19 disease by time since the final dose.
| Vaccine |
Duration |
Cases |
Matched control |
Adjusted protective effectiveness (PE) (95% CI) |
|||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Non–vaccinees | Complete primary regimen | Booster regimen | Non–vaccinees | Complete primary regimen | Booster regimen | Complete primary and/or booster regimen* | p–value | Complete primary regimen† | p–value | Complete primary with booster regimen‡ | p–value | ||
| Any vaccine§ | ≤90, days | 60(0.42) | 68(0.47) | 16(0.11) | 52(0.36) | 77(0.53) | 15(0.1) | 25(−23,54) | .254 | 26(−24,56) | .251 | 19(−94,66) | .641 |
| >90, days | 95(0.18) | 346(0.67) | 73(0.14) | 90(0.18) | 356(0.69) | 68(0.13) | 11(−24,36) | .481 | 12(−23,37) | .451 | 3(−64,42) | .917 | |
| Serum Institute of India (ChAdOx1 nCoV–19) | ≤90, days | 23(0.88) | 1(0.04) | 2(0.08) | 24(0.92) | 2(0.08) | 0(0) | 8(−568,87) | .931 | 63(−409,97) | .457 | inf | – |
| >90, days | 46(0.37) | 72(0.57) | 8(0.06) | 43(0.34) | 80(0.63) | 3(0.02) | 30(−30,62) | .265 | 35(−21,66) | .174 | −146(−1163,52) | .280 | |
| Moderna (mRNA 1273) | ≤90, days | 24(0.8) | 2(0.07) | 4(0.13) | 26(0.87) | 0(0) | 4(0.13) | −86(−764,60) | .43 | inf | – | 39(−338,92) | .620 |
| >90, days | 37(0.49) | 26(0.35) | 12(0.16) | 33(0.44) | 24(0.32) | 18(0.24) | 35(−43,71) | .284 | 26(−68,67) | .472 | 70(−6,92) | .062 | |
| Pfizer–BioNTech(BNT162b2) | ≤90, days | 31(0.84) | 3(0.08) | 3(0.08) | 23(0.62) | 10(0.27) | 4(0.11) | 88(26,98) | .023 | 92(18,99) | .034 | 82(−61,98) | .124 |
| >90, days | 29(0.69) | 7(0.17) | 6(0.14) | 31(0.74) | 7(0.17) | 4(0.1) | −15(−256,63) | .804 | 7(−236,74) | .914 | −75(−827,67) | .511 | |
| mRNA | ≤90, days | 34(0.68) | 7(0.14) | 9(0.18) | 28(0.56) | 11(0.22) | 11(0.22) | 46(−37,79) | .194 | 50(−81,86) | .291 | 43(−78,82) | .330 |
| >90, days | 45(0.41) | 37(0.33) | 29(0.26) | 43(0.39) | 40(0.36) | 28(0.25) | 12(−61,52) | .675 | 14(−62,54) | .64 | 6(−119,60) | .886 | |
| Sinopharm (Vero Cell– Inactivated) | ≤90, days | 45(0.51) | 43(0.49) | – | 42(0.48) | 46(0.52) | – | 17(−54,55) | .554 | 17(−54,55) | .554 | – | – |
| >90, days | 44(0.42) | 60(0.58) | – | 46(0.44) | 58(0.56) | – | −3(−85,42) | .908 | −3(−85,42) | .908 | – | – | |
| Vero Cell, inactivated (Sinopharm and Sinovac) | ≤90, days | 45(0.51) | 44(0.49) | – | 42(0.47) | 47(0.53) | – | 17(−54,55) | .556 | 17(−54,55) | .556 | – | – |
| >90, days | 46(0.42) | 63(0.58) | – | 46(0.42) | 63(0.58) | – | 5(−67,46) | .848 | 5(−67,46) | .848 | – | – | |
*Protective effectiveness of complete primary 2-doses of vaccine (with or without booster/3rd dose); †Protective effectiveness of complete primary 2-doses of vaccine only.
‡Protective effectiveness of complete primary 2-doses with 3rd dose of vaccine; Effects are adjusted by body mass index, and household income.
§Any vaccine (Serum Institute of India (ChAdOx1 nCoV- 19), Sinopharm (Vero Cell- Inactivated), Moderna (mRNA-1273), Pfizer-BioNTech (BNT162b2) and Sinovac (Vero Cell-Inactivated)).
CI = Confidence Interval; Duration = Time since receipt of a complete primary or booster regimen.
Vaccine protection of a complete primary or a booster regimen against ultimate severe COVID-19 disease by time since the final dose
We found 40 (2.4%) patients including 24 cases and 16 controls who were classified as ultimate severe. When we evaluated the protective effectiveness against ultimate severity considering the severe cases and severe controls, there was no significant protection conferred by any vaccines (Table 4).
Table 4.
Vaccine protection of a complete primary or a booster regimen against ultimate severe COVID-19 disease by time since the final dose*.
| Vaccine types | Ultimate severe cases |
Ultimate severe controls |
Protective effectiveness (PE)‡ (95% CI) |
|||||
|---|---|---|---|---|---|---|---|---|
| Vaccinees | Non-vaccinees | Vaccinees | Non-vaccinees | Crude PE | p-value | Adjusted§ PE | p-value | |
| Any vaccine† | 15(0.62) | 9(0.38) | 12(0.75) | 4(0.25) | 13(−21,37) | .421 | 14(−23,39) | .424 |
| ≤90 days | 2(0.18) | 9(0.82) | 2(0.33) | 4(0.67) | 17(−45,53) | .512 | 14(−65,55) | .662 |
| >90 days | 13(0.59) | 9(0.41) | 10(0.71) | 4(0.29) | 12(−24,37) | .467 | 10(−30,37) | .595 |
*Ultimate severity included severity at presentation or death or hospitalization within 30 days following presentation.
†Any vaccine (Serum Institute of India (ChAdOx1 nCoV- 19), Sinopharm (Vero Cell- Inactivated), Moderna (mRNA-1273), Pfizer-BioNTech (BNT162b2), and Sinovac (Vero Cell-Inactivated)).
‡Unconditional logistic regression performed to get PE.
§Adjusted by age groups and site of presentation based on the rule of thumb.
Discussion
Our study gave little evidence of protection during a period of Omicron predominance against any COVID-19 disease or severe disease by receipt of a complete primary regimen or booster dose of the vaccines used in Bangladesh- Serum Institute of India (ChAdOx1 nCoV- 19), Sinopharm (Vero Cell-Inactivated), Moderna (mRNA-1273), Pfizer-BioNTech (BNT162b2), and Sinovac (Vero Cell-Inactivated).
In this analysis, a point estimate of 52% (95% CI: −14 to 80; p = .095) revealed that the complete primary and/or booster dose of Pfizer-BioNTech (BNT162b2) vaccine was suggestively protective against COVID-19 disease. Within 90 days following the final dose of the complete primary regimen, the Pfizer-BioNTech (BNT162b2) vaccination demonstrated a significant (92% (95% CI: 18 to 99; p = .034)) level of protection in this study. Another trial revealed that a two-dose primary regimen of Pfizer-BioNTech (BNT162b2) demonstrated 65.5% protection at 2 to 4 weeks against symptomatic infection by the Omicron variant, which declined to 8.8% at 25 weeks.10
In our previous analysis, we did not find any evidence of protection by the Serum Institute of India (ChAdOx1 nCoV- 19) and Sinopharm (Vero Cell- Inactivated) vaccine against the Delta variant.14 Similarly, in this analysis, no protection of these vaccines was detected when Omicron was the main circulating variant in Bangladesh.
Our study has a number of possible limitations. Firstly, the study sample size was not calculated for secondary analyses to assess the protection of individual vaccines, as well as to assess the protection of vaccines against severe SARS-CoV-2 disease. As a result, these analyses were inherently underpowered, which may have contributed to the wide confidence intervals observed. This imprecision limits the reliability of the effect estimates and makes it difficult to draw firm conclusions, particularly when comparing the effectiveness across vaccine types. Secondly, since the hospitalization rate of patients with ventilation and ICU support was considerably lower when Omicron was the predominant circulating variant in Bangladesh and globally, we found low numbers of severe patients for this study. Because of the scarcity of events, analysis of protection against ultimate severe disease including mortality yielded results with wide confidence intervals. As a consequence, even if there were true differences in vaccine effectiveness against severe disease, our study may not have been adequately powered to detect them. Thirdly, the analysis was not purely against Omicron, with Delta constituting a small minority of isolates at the beginning of the study interval. However, the restriction of the analysis to the period in which only Omicron was isolated yielded virtually identical results to the overall analysis. Fourthly, the study was conducted in four hospitals in Dhaka, which may limit the generalizability of the findings to other regions of Bangladesh or countries with different healthcare access, disease epidemiology, and vaccination coverage. Therefore, applying these results to other populations should be done with caution. Fifthly, we used RT-PCR, which is considered the gold standard for COVID-19 diagnosis due to its higher sensitivity. However, we acknowledge the risk that some individuals classified as controls may actually be false negatives. Finally, we restricted our analysis to cases and control without earlier episodes of COVID-19 infections and we acknowledge that prior COVID-19 infection was not systematically assessed in our study. Patients with a history of previous COVID-19 were excluded based on verbal self-report at enrollment; however, this method may have resulted in incomplete exclusion of individuals with prior exposure due to recall bias or asymptomatic infections. We did not perform serological testing to confirm prior infection status, which limits our ability to fully assess the impact of hybrid immunity due to the combination of natural and vaccine-induced immunity. We also acknowledge that our study was limited in its ability to evaluate vaccine effectiveness within clinically relevant subgroups due to the relatively small sample sizes available for stratified analyses. For example, the number of older adults, individuals with immunocompromising conditions, and those with specific comorbidities was insufficient to generate statistically meaningful subgroup estimates. Despite these limitations, there are many strengths of this study. The major strength was that we enrolled patients with COVID-19 disease symptoms in real-time, and also carried out the laboratory testing to confirm SARS-CoV-2 infection. Ascertainment of eligibility criteria and collection of clinical information from patients was done in a manner blinded to the test results of cases and controls. In addition, the vaccination status was thoroughly determined by combining patient interviews with systematic searches of the vaccine registries created by the government of Bangladesh without being aware of the participant’s status as cases or controls. This made it possible to categorize each patient’s vaccination status precisely, including whether or not they had booster doses, which most likely guaranteed a high level of internal study validity.
In conclusion, the study was unable to demonstrate that the vaccines deployed in Bangladesh in aggregate provided significant protection while Omicron was the predominant variant. Exploratory analysis showed that the Pfizer-BioNTech (BNT162b2) vaccine may have conferred significant protection for a brief length of time. For future analyses, it will be critical to evaluate the vaccine effectiveness in preventing COVID-19 disease among the Bangladeshi population regardless of prior exposure to the disease so that the impact of hybrid immunity can be assessed. It will also be important to assess whether the failure of the vaccine to protect against COVID-19 disease was due to poor immune responses to the vaccination.
Acknowledgments
The icddr,b is grateful to the governments of Bangladesh, Canada, Sweden, and the UK for providing core/unrestricted support. We also thank the study participants and study staff. J.D.C. and F.Q. were responsible for designing the study. J.D.C., F.Q., M.T.I. and F.K. were responsible for the drafting of this report. J.D.C. and F.A. led the data analysis and interpretation. M.T.I. F.K., and P.K.B. verified the data presented in the manuscript. M.T.I., F.K., M.N.H.R., M.I.H., S.H., O.H.B.M. and M.H.A. led the data collection at the sites represented within this report. F.K., M.T.I. and F.Q. were responsible for the supervision and monitoring of field, laboratory, clinical and data management activities. J.D.C., F.Q., M.T.I., F.K., F.A., M.N.H.R., M.I.H., S.H., P.K.B., S.A.A.A., A.N.A., M.M.B., M., M.Z.R., O.H.B.M., M.T.H., M.H.A., S.M.S., A.A.S., M.A.S., A.R., and T.S. revised the report critically. All authors have full access to all the data in the study. All authors have read and agreed to the published version of the manuscript.
Biography
Firdausi Qadri, Senior Scientist, Infectious Diseases Division and Head, Mucosal Immunology and Vaccinology Unit, at icddr,b, Dhaka, Bangladesh. She is also the founder and leads to Institute for Developing Science and Health Initiatives (ideSHi). Her work includes basic and applied immunology of infectious diseases but also clinical trials and large field-based studies on vaccines. Dr. Qadri has more than 550 publications in peer-reviewed journals of high impact. She has been honored with numerous prestigious awards for her contribution to research. Such as the Gold Medal in Biological Sciences from the Bangladesh Academy of Science, Moselio Schaechter Award by the American Society for Microbiology (ASM), the Christophe & Rodolphe Mérieux Foundation Prize, the French Academy of Sciences, Prof. C.N.R. Rao Prize from TWAS. L’Oréal-UNESCO- For Women in Science Award, Ramon Magsaysay Award 2021- known as Asia’s Nobel Prize. Dr. Qadri received the country’s highest award for outstanding contributions to research and training, the Independence Award (Swadhinata Padak) in 2023. It is a unique milestone for women scientists, In 2024, Dr. Qadri received the prestigious Clara Southmayd Ludlow Medal from the ASTMH and VinFuture Prize from Viet Nam for her pioneering and inspirational contributions to the global health.
Funding Statement
The study was funded by the Bill & Melinda Gates Foundation [INV-034303]. The Gates Foundation had no role in the study design, data collection, data analysis, data interpretation, or writing of the report. https://www.gatesfoundation.org/about/committed-grants/2021/06/inv034303.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Data availability statement
Anonymous participant data and a data dictionary for each variable analyzed in this article, as well as the study protocol, the statistical analysis plan, and the informed consent form will be made available when the study is complete, upon requests directed to the corresponding author (fqadri@icddrb.org). Data can be shared after the approval of a proposal through a secure online platform. The Research Review Committee and Ethical Review Committee at the icddr,b will review the proposal and will then give their approval. Additionally, data sharing will depend on the published data access rules of the icddr,b. Petitioners will need to sign a standard data access agreement issued by the icddr,b.
Informed consent statement
Informed consent was obtained from all subjects involved in the study.
Institutional review board statement
The study was conducted in accordance with the Declaration of Helsinki, and approved by the Institutional Review Board (or Ethics Committee) of the icddr,b (PR 21,075 and 27 July 2021).
References
- 1.Wawina-Bokalanga T, Logist A-S, Sinnesael R, Van Holm B, Delforge M-L, Struyven P, Cuypers L, André E, Baele G, Maes P, et al. Genomic evidence of co-identification with Omicron and Delta SARS-CoV-2 variants: a report of two cases. Int J Infect Dis. 2022;122:212–9. doi: 10.1016/j.ijid.2022.05.043. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Arora P, Zhang L, Rocha C, Sidarovich A, Kempf A, Schulz S, Cossmann A, Manger B, Baier E, Tampe B, et al. Comparable neutralisation evasion of SARS-CoV-2 omicron subvariants BA.1, BA.2, and BA.3. Lancet Infect Dis. 2022;22(6):766–767. doi: 10.1016/S1473-3099(22)00224-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Qu P, Evans, JP, Faraone, J, Zheng, YM, Carlin, C, Anghelina, M, Stevens, P, Fernandez, S, Jones, D, Lozanski, G.. Distinct neutralizing antibody escape of SARS-CoV-2 Omicron subvariants BQ.1, BQ.1.1, BA.4.6, BF.7 and BA.2.75.2. bioRxiv; 2022. https://pubmed.ncbi.nlm.nih.gov/36299423/. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Ren SY, Wang W-B, Gao R-D, Zhou A-M. Omicron variant (B.1.1.529) of SARS-CoV-2: mutation, infectivity, transmission, and vaccine resistance. World J Clin Cases. 2022;10(1):1–11. doi: 10.12998/wjcc.v10.i1.1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Fan Y, Li X, Zhang L, Wan S, Zhang L, Zhou F. SARS-CoV-2 Omicron variant: recent progress and future perspectives. Sig Transduct Target Ther. 2022;7(1):141. doi: 10.1038/s41392-022-00997-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Das S, Samanta S, Banerjee J, Pal A, Giri B, Kar SS, Dash SK. Is Omicron the end of pandemic or start of a new innings? Travel Med Infect Dis. 2022;48:102332. doi: 10.1016/j.tmaid.2022.102332. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Releifweb . SARS-CoV-2 variants in Dhaka city: emergence of Omicron BA.5 subvariant. 2022. [cited 2023 May 2]: https://reliefweb.int/report/bangladesh/sars-cov-2-variants-dhaka-city-emergence-omicron-ba5-subvariant.
- 8.Izda V, Jeffries MA, Sawalha AH. COVID-19: a review of therapeutic strategies and vaccine candidates. Clin Immunol. 2021;222:108634. doi: 10.1016/j.clim.2020.108634. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Buchan SA, Chung H, Brown KA, Austin PC, Fell DB, Gubbay JB, Nasreen S, Schwartz KL, Sundaram ME, Tadrous M, et al. Estimated effectiveness of COVID-19 vaccines against Omicron or Delta symptomatic infection and severe outcomes. JAMA Netw Open. 2022;5(9):e2232760. doi: 10.1001/jamanetworkopen.2022.32760. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Andrews N, Stowe J, Kirsebom F, Toffa S, Rickeard T, Gallagher E, Gower C, Kall M, Groves N, O’Connell A-M, et al. Covid-19 vaccine effectiveness against the Omicron (B.1.1.529) variant. N Engl J Med. 2022;386(16):1532–1546. doi: 10.1056/NEJMoa2119451. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Ferdinands JM, Rao S, Dixon BE, Mitchell PK, DeSilva MB, Irving SA, Lewis N, Natarajan K, Stenehjem E, Grannis SJ, et al. Waning of vaccine effectiveness against moderate and severe COVID-19 among adults in the US from the VISION network: test negative, case-control study. BMJ. 2022;379:e072141. doi: 10.1136/bmj-2022-072141. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Adams K, Rhoads JP, Surie D, Gaglani M, Ginde AA, McNeal T, Talbot HK, Casey JD, Zepeski A, Shapiro NI, et al. Vaccine effectiveness of primary series and booster doses against COVID-19 associated hospital admissions in the United States: living test negative design study. BMJ. 2022;379:e072065. doi: 10.1136/bmj-2022-072065. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Au WY, Cheung PP. Effectiveness of heterologous and homologous COVID-19 vaccine regimens: living systematic review with network meta-analysis. BMJ. 2022;377:e069989. doi: 10.1136/bmj-2022-069989. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Khanam F, Islam MT, Ahmmed F, Ahmed SU, Hossen MI, Rajib MH, Haque S, Biswas PK, Tauheed I, Zaman K, et al. Measuring the effectiveness of COVID-19 vaccines used during a surge of the Delta variant of SARS-CoV-2 in Bangladesh: a test-negative design evaluation. Vaccines. 2022;10(12):2069. doi: 10.3390/vaccines10122069. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Roy DN, Azam MS, Islam E, Lounis M. Multi-dimensional potential factors influencing COVID-19 vaccine booster acceptance and hesitancy among university academic community in Bangladesh: a cross-sectional comparative study. PLOS ONE. 2023;18(4):e0281395. doi: 10.1371/journal.pone.0281395. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Sangha O, Stucki G, Liang MH, Fossel AH, Katz JN. The self-administered comorbidity questionnaire: a new method to assess comorbidity for clinical and health services research. Arthritis Rheum. 2003;49(2):156–163. doi: 10.1002/art.10993. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Anonymous participant data and a data dictionary for each variable analyzed in this article, as well as the study protocol, the statistical analysis plan, and the informed consent form will be made available when the study is complete, upon requests directed to the corresponding author (fqadri@icddrb.org). Data can be shared after the approval of a proposal through a secure online platform. The Research Review Committee and Ethical Review Committee at the icddr,b will review the proposal and will then give their approval. Additionally, data sharing will depend on the published data access rules of the icddr,b. Petitioners will need to sign a standard data access agreement issued by the icddr,b.