Serological survey for SARS-CoV-2 antibodies among tribal communities of Odisha post-second wave

The COVID-19 pandemic has spread to almost all countries of the world affecting over 258 million persons and causing over 5.16 million deaths1. While India accounts for around 34.54 million of these cases and 0.46 million deaths, the State of Odisha in Eastern India has reported around 1.04 million cases, largely spread over the two waves2. India introduced vaccines against COVID-19 in a phased manner in January 20213. The vaccination coverage has increased steadily, with over a billion doses across India and over 42 million in Odisha completed at present2. Vaccination has since added another layer to the existing dynamics of population-level immunity against SARS-CoV-2.

While there have been four rounds of national serological surveys undertaken in India since December 2020, similar surveys have also been undertaken in the State of Odisha in targeted urban and rural regions4-6. The National Serosurvey fourth round has shown a seroprevalence of SARS-CoV-2 antibodies in two tribal-dominated districts of Odisha in Koraput (32.6%) and Rayagada (37.3%) but not exclusively in the tribal community7. Such surveys would also help inform the levels and trends of immunity among specific vulnerable groups including the tribal population with an expanding vaccination programme in the backdrop.

Scheduled tribes (STs) constitute around eight per cent of the population in India and are generally residents in well-recognized pockets8. Odisha is unique among Indian States in that it has a significant (22.85% of ST population) and diverse [62 STs and 13 Primitive Vulnerable Tribal Groups (PVTGs)] tribal population9. These indigenous communities are generally considered socially and economically marginalized and at a disproportionately higher risk in case of public health emergencies10,11. While there is some evidence that tribal communities in India report a lower caseload and higher recovery rates, the mechanisms are not well understood12,13.

A serological survey was hence carried out to estimate and compare the age-specific prevalence of SARS-CoV-2 antibodies in tribal populations across Odisha, India. Since the study population in this survey exclusively involved the tribes of Odisha, which makes this first of its kind and will help plan and implement various upcoming tribal-level policy-making decisions of the State.

Material & Methods

A population-based, age-stratified, cross-sectional study design was adopted for the present survey based on the survey design and sampling framework adopted for national serosurveys by the Indian Council of Medical Research (ICMR)4,5. A written informed consent, parents/guardians informed consent and assent were obtained for study participants. Approval from the State Ethics Committee and the Institutional Human Ethical Committee on human research was taken from ICMR-Regional Medical Research Center (RMRC), Bhubaneswar. The study was conducted from August 30 to September 16, 2021 in seven districts of Odshia State (tribal sub-plan areas) where the tribal population is 50 per cent or more14. The profile of study districts is provided in Supplementary Table I. The total population of the seven districts is 20.1 per cent of the total population of Odisha. Overall, tribal population in the seven districts constitutes 61.8 per cent of the total population of the study districts.

Multistage random sampling method was used where clusters within the districts were selected through probability proportionate to size, and households in clusters were selected by systematic random sampling. The sample size for each district was calculated to be 395 (rounded off to 400) with an assumption of an expected prevalence of 50 per cent (reported in previous surveys), relative precision of 16 per cent, design effect of 2.5 and non-response rate of 10 per cent for a 95 per cent level of confidence15. Ten clusters in each district (a total of 70 clusters) were selected using the probability proportionate to size (PPS) sampling methodology using household population size from the census16. At least 40 consenting individuals (4 aged 6-9 yr, 8 aged 10-17 yr and 28 aged 18 yr and above) were enrolled from each cluster.

Households whose adults agreed to provide written informed consent for themselves and their children were included in the study. Pregnant women, bedridden patients and those with recognizable cognitive impairment were excluded from participating. Data were collected in an electronic data collection tool Open Data Kit (Open Source). Venous blood samples (3 ml) were collected in a gel vacutainer and were transferred to a local healthcare facility in less than 6 h (maintaining cold chain) for centrifugation. Separated serum from whole blood was tested in the ICMR-RMRC, Bhubaneswar, to determine the presence of total antibodies, including IgG antibodies against SARS-CoV-2 nucleocapsid (N) protein. IgG antibodies against N-protein was tested for all participants in an automated analyzer Cobas e411 (Roche Diagnostics, GmbH, Mannheim, Germany) using Elecsys anti-SARS-CoV-2 kit insert. Cut-off index (CoI) values of <1.0 were considered as nonreactive and CoI of ≥ 1.0 as reactive.

The seroprevalence of COVID-19 for each study district was estimated. The design weights in our analyses were accounted for and calculated as the product of the inverse of the sampling fraction for the selection of individuals, households, and clusters within each district17 (Supplementary Annexure). Persons who had received two doses of any approved COVID-19 vaccine and where at least 14 days had passed since the 2^nd dose were considered ‘fully immunized’. Those who had not received any dose of vaccine were categorized as ‘unimmunized’ and the remaining as ‘partially immunized’. Analysis was carried out in the statistical software R version 4.0.3 (R Foundation for Statistical Computing, Vienna, Austria)18. The prevalence was calculated for seven districts using the presence of antibodies against SARS-CoV-2 with 95 per cent confidence interval (CI).

Results & Discussion

A total of 3127 households were approached and 2855 participants were included in the final analysis with an overall non-response rate of 6.14 per cent across all districts. Figure 1 provides the flow diagram of sample collection process. The overall weighted seroprevalence was 72.8 per cent (95% CI: 70.1-75.3) across all districts in Odisha. National serosurvey reports have shown a steady increase in the population level seroprevalence, from 0.73 per cent (95% CI: 0.34-1.13) in May-June 2020 to 24.1 per cent (95% CI: 23.0-25.3%) in December 2020-January 20215,19. Similarly, in the region, surveys in Delhi, Mumbai and Bhubaneswar in the urban areas have reported a seroprevalence between 24.71 per cent (95% CI: 24.01-25.42), 17.3 per cent (95% CI: 16 to 18.7) and 49.04 per cent (95% CI 46.39–51.68), respectively20-22. The fact that seroprevalence in a population is a function of time and the progression of the pandemic is well recognized, and repeated cross-sectional surveys from Indian cities such as Delhi, Mumbai and Bhubaneswar have shown such increasing trends20-22, the findings suggest that the pandemic has affected these groups similar to the general rural population in this region. This indicates a resumption of existing interconnections in the rural areas and forests after the inevitable relaxation of the restrictions imposed during the lockdowns. Furthermore, unawareness regarding the disease and protocols to prevent contamination were the other reasons for the spread.

Close

Fig. 1

Study flow diagram of the sample collection process.

Export to PPT

The sociodemographic characteristics of the survey participants are depicted in the Table. The seroprevalence was found higher among males (74.2%, 95% CI: 71.5-76.8) than females (71.7%, 95% CI: 68.0-75.2); there was no significant difference across them. However, the higher prevalence in males may be explained by the relatively better vaccination coverage23.

Serological prevalence was the highest among 18-44 yr (74.4%, 95% CI: 71.3-77.3). The seroprevalence among children aged 6-9 yr was 68.6 per cent and among those aged 10-17 yr was 73.8 per cent. This was similar to other age groups of study participants, indicating a uniform spread of the pandemic across age groups. The third national serosurvey have reported similar trends among children and adolescents, with no significant difference found across the age groups5. Seroprevalence was higher among those with multimorbidity (79.65%, 95% CI: 72.85-85.40%) as compared to single chronic disease (68.38%, 95% CI: 59.86-76.08%) or no chronic diseases (73.07%, 95% CI:71.30-74.78%). Serological prevalence or presence of COVID-19-specific antibodies across the districts was between 69 and 76 per cent respectively, as shown in Figure 2. The seroprevalence was higher as compared to two tribal-dominated districts, Koraput (32.6%) and Rayagada (37.3%), reported in the fourth round of the national serosurvey7; however, the study participants enrolled in the present study exclusively belonged to the tribal community. Furthermore, the national serosurvey did not present disaggregated data on tribal groups7.

Close

Fig. 2

Seroprevalence and vaccination coverage among the tribal population across districts.

Export to PPT

Among the participants, 41.93 per cent had received at least one dose of any COVID-19 vaccine. Only 6.17 per cent of participants were hesitant about accepting the COVID-19 vaccines. The relative isolation of habitat and unique sociobehavioural characteristics necessitates a specific effort to generate evidence on the extent of the COVID-19 pandemic in these groups and identify the challenges related to vaccination coverage. We also found a high seroprevalence in PVTG groups (68%), although this was relatively lower than the other listed STs (74%), it appears that the pandemic did spread significantly among the most isolated groups in mainland India.

The proportion of unimmunized persons in the younger age groups was significantly lower (12.2%) than in the middle (29.8%) and older (29.2%) ages implying a better reach and acceptance in this age. Seroprevalence was higher in fully immunized persons (75.27%), but there was no significant difference between the two groups.

Vaccination coverage across the tribal population was in line with the general population, where around 42 per cent had received at least one dose of any of the COVID-19 vaccines. This is an outcome of the rapid pace of expansion of the world’s largest vaccination programme where 1.19 billion doses were administered along with good acceptance across the diverse groups in the region and will increase further with time2. The common reason for not availing vaccination was the unavailability of vaccination slot (36.19%), unawareness regarding COVID-19 vaccines (35.43%) and lack of knowledge on where vaccines are available (8.20%). History of respiratory illness in the past three months and seroprevalence is provided in the Supplementary Table II.

The vaccination coverage among the PVTGs was higher than the other tribes, with over 89 per cent having received at least a single dose of the COVID-19 vaccine. This implies excellent reach of the region’s vaccination programme and little to no vaccine hesitancy in the PVTGs.

It is encouraging to note that we found minimal levels of vaccine hesitancy (6%) among the tribal communities of Odisha. This is in contrast to the higher levels of vaccine hesitancy seen in other parts of the world24. The major causes of not receiving the vaccine to date were supply-side bottlenecks and lack of knowledge on the availability of such vaccines. Due to their genetic makeup, the severity of COVID-19 was not much among them and was well managed with limited healthcare.

The study draws its strength from the large sample size and robust sampling design that improves the generalizability of the results to the target population. However, this survey study had one major limitation. As the study was conducted during an expanding vaccination drive, it was not possible to differentiate whether the presence of antibodies against SARS-CoV-2 was due to prior natural infection or vaccination. An additional IgG estimation against the spike (S) protein of SARS-CoV-2 could, however, substantiate this differentiation.

Our findings show a high seroprevalence against SARS-CoV-2 among the tribal population and PVTGs. This is evident across sociodemographic subgroups such as age and gender. The vaccination coverage is at par with the general population, and efforts may be required to address some knowledge gaps to improve coverage during the coming days.