Study design.
Serial, cross-sectional Sero epidemiological surveys were carried out in urban (Karachi) and rural (Matiari) areas of Sindh, Pakistan (Fig. 1). The study spanned 16 months, from July 2020 to November 2021, with sampling conducted at four time points in the same population. We adapted the WHO 30 by 7 cluster methodology to ensure population representation. In Karachi, the data was collected from its six districts, i.e., Central, East, West, South, Korangi, and Malir. In the Matiari region, data was collected from both the Hala and Matiari sub-districts. Seven union councils were randomly selected from the available list in each district, which served as primary sampling units. We randomly selected 20 Households from each union council to achieve the required sample size. We enrolled individuals who were residents of the respective districts for at least six months, regardless of age, with no stated intent to migrate.
Data collection and blood sampling were done by six research teams, three in Karachi and three in Matiari. Each research team comprised a phlebotomist, a data collector, and a team leader. Once selected, the same households were revisited at subsequent rounds. The research staff received training on personal hygiene procedures, the use of personal protective equipment, and the safe handling of laboratory samples. After each interview, it was provided with personal protective equipment (aprons, gloves, surgical face masks, and shoe and hair covers) disposed of as hospital waste.
Data were collected from each participant using a structured questionnaire through a computer-assisted personal interview (CAPI) approach. The questionnaire included sections on demographic information, exposure assessment, underlying comorbidities, and a history of any clinical symptoms like chest pain, sore throat, and fever in the past two weeks, as well as a history of travel outside their local region in the past two weeks.
Blood sample collection
After administration of the questionnaire, trained phlebotomists collected 3 ml of whole blood from each participant through venipuncture. The blood samples were transported to the Nutrition Research Laboratory (NRL) at AKU, Karachi, under a cold chain. These samples were used for the detection of SARS-CoV-2 antibodies, measurement of hemoglobin, biochemical assessment of vitamin D, C-reactive protein (CRP), and micronutrient (zinc) concentrations.
Measurement of Anti SARS-CoV-2 antibodies
Antibodies against SARS-CoV-2 were measured in sera through Roche Cobas e411 automated analyzer using Roche Elecsys Anti-SARS-CoV-2 assay kit (Roche, Basel, Switzerland) per manufacturer’s instructions. The assay qualitatively detects total polyclonal antibodies (IgA, IgG, and IgM) against SARS-CoV-2 nucleocapsid (N) antigen. Before each batch, the assay was validated using positive and negative quality controls. Assay values above the cutoff of 1.0 were considered reactive for the presence of anti-SARS-CoV-2 antibodies. Quantitative serum 25-OH vitamin D was determined using the LIASION 25 OH Vitamin D TOTAL assay (DiaSorin, Saluggia, Italy). Zinc levels in plasma were measured using a Thermo scientific atomic absorption spectrometer. Hemoglobin levels were measured using the HemoCue Hb 301 analyzer (HemoCue, CA, USA), while CRP levels were quantified using a Roche CRP (CRPLX) kit on the Cobas c311 analyzer (Roche, Basel, Switzerland).
Detection of SARS-CoV-2 RNA using RT-PCR
Nasopharyngeal swabs were collected by inserting a sterile swab (Virus RNA collection kit, Beaver, China) into a single nostril and rotated 3–4 times against the nasopharyngeal surface. Respiratory specimens were stored at -80°C until shipped to the National SARS-CoV-2 Reference Lab of the National Institute, Islamabad (NIH-Islamabad) on dry ice. RNA was extracted from 200µl of the media using MagMax Viral/Pathogen II Nucleic Acid Isolation Kit (Thermo Fisher Scientific Inc. Waltham, Massachusetts, US). Real-time PCR was performed using TaqPath COVID-19 CE-IVD RT-PCR Kit on Applied Biosystems 7500 Fast Dx Real-Time PCR Instrument (Thermo Fisher Scientific Inc. Waltham, Massachusetts, US).
COVID-19 antigen testing of respiratory samples
According to the manufacturer's guidelines, the rapid antigen test was performed using the SARS-CoV-2 Rapid Antigen Test kit (Roche, Basel, Switzerland). Briefly, the nasopharyngeal swab was collected and inserted into extraction buffer vials and rotated a few times to transfer contents into the buffer, after which the suspension was loaded on the cartridge, and results were read after 15 minutes. A Positive test was recorded when the test line was visible on the cartridge, and the visible control line confirmed the validity for each record. Quality control for each batch run was ensured by testing the commercially available Roche SARS-CoV-2 Antigen Control kit, including positive and negative controls (Roche, Basel, Switzerland).
Sample Size Estimation and Statistical Analysis
We estimated that at least 2974 individuals (at least three individuals per household) stratified by age were needed to estimate an expected seroprevalence of 36% with 5% precision at a 95% level of confidence, with 80% power and a non-response rate of 20%. Reference calculations were based on serological data available at the time [21]. Categorical variables were reported as frequencies and proportions. Seroprevalence was reported by age, gender, socioeconomic status, clinical symptoms, type of residence (rural/ urban), and district for baseline and each follow-up round. Socioeconomic status was assessed using a wealth index derived through principal component analyses of household assets.
Analysis was performed using an intention-to-treat approach. The Pearson Chi-square test was used to establish an association between categorical variables, and trend p-values were used to compare seroprevalence over time by baseline characteristics.
Mixed effect logistic regression was used to explore the impact of time on the seropositive status of the respondent. Individuals were incorporated as a random effect to model individual heterogeneity within seropositive status. The time interaction was tested to measure the impact of change in seropositive status. Furthermore, the analysis adjusted for confounding variables like age, gender, socioeconomic status, type of residence, district, area, medical symptoms, travel history, vitamin D and zinc concentrations, and vaccination status, keeping a positive COVID-19 antibody test as the outcome.
Adjusted odds ratios (OR) with 95% confidence intervals were reported. Covariates examined included age, gender, socioeconomic status, type of residence, district, area, medical symptoms, travel history, vitamin D and zinc levels, and vaccination status. The Cox proportional regression model was used to estimate hazard ratios (HR) and the corresponding 95% confidence intervals for positive cases for all possible risk factors (including area, age, sex, mother tongue, household size, etc.). A p-value of < 0.05 was considered statistically significant. Analysis was carried out using STATA Version 17 (StataCorp LLC, College Station, Texas, USA).