Between March and December, 2020, more than 20 000 laboratory-confirmed cases of SARS-CoV-2 infection were reported in Zambia. However, the number of SARS-CoV-2 infections is likely to be higher than the confirmed case counts because many infected people have mild or no symptoms, and limitations exist with regard to testing capacity and surveillance systems in Zambia. We aimed to estimate SARS-CoV-2 prevalence in six districts of Zambia in July, 2020, using a population-based household survey.
Between July 4 and July 27, 2020, we did a cross-sectional cluster-sample survey of households in six districts of Zambia. Within each district, 16 standardised enumeration areas were randomly selected as primary sampling units using probability proportional to size. 20 households from each standardised enumeration area were selected using simple random sampling. All members of selected households were eligible to participate. Consenting participants completed a questionnaire and were tested for SARS-CoV-2 infection using real-time PCR (rtPCR) and anti-SARS-CoV-2 antibodies using ELISA. Prevalence estimates, adjusted for the survey design, were calculated for each diagnostic test separately, and combined. We applied the prevalence estimates to census population projections for each district to derive the estimated number of SARS-CoV-2 infections.
Overall, 4258 people from 1866 households participated in the study. The median age of participants was 18·2 years (IQR 7·7–31·4) and 50·6% of participants were female. SARS-CoV-2 prevalence for the combined measure was 10·6% (95% CI 7·3–13·9). The rtPCR-positive prevalence was 7·6% (4·7–10·6) and ELISA-positive prevalence was 2·1% (1·1–3·1). An estimated 454 708 SARS-CoV-2 infections (95% CI 312 705–596 713) occurred in the six districts between March and July, 2020, compared with 4917 laboratory-confirmed cases reported in official statistics from the Zambia National Public Health Institute.
The estimated number of SARS-CoV-2 infections was much higher than the number of reported cases in six districts in Zambia. The high rtPCR-positive SARS-CoV-2 prevalence was consistent with observed community transmission during the study period. The low ELISA-positive SARS-CoV-2 prevalence might be associated with mitigation measures instituted after initial cases were reported in March, 2020. Zambia should monitor patterns of SARS-CoV-2 prevalence and promote measures that can reduce transmission.
US Centers for Disease Control and Prevention.
The Zambian Government acted swiftly to control the spread of SARS-CoV-2, initiating a whole-of-government response, restricting travel into the country, closing public gathering spaces (eg, restaurants, bars, churches), and invoking the Public Health Act to expand authority of the Zambian Government agencies.
From the outset, contact tracing teams rapidly responded to newly reported cases. With the exception of a localised outbreak in Nakonde District in May, 2020, the number of positive cases remained sporadic until June, 2020 (appendix p 2). The number of laboratory-confirmed cases rapidly increased in July, 2020, coinciding with a gradual relaxation of physical distancing measures in May and June, 2020. According to the Zambia National Public Health Institute (ZNPHI), as of Feb 18, 2021, 72 467 confirmed COVID-19 cases had been identified from 1 038 573 tests in Zambia.
COVID-19 symptoms overlap with those of other common upper respiratory tract infections that are usually self-limited.
Furthermore, limited testing capacity and surveillance system gaps are likely to have contributed to under-ascertainment of SARS-CoV-2 infections in Zambia. Although testing criteria were rapidly expanded in the country to capture cases without an international travel history,
this strategy was implemented incompletely throughout the country, partly due to low rates of testing as a result of poor availability of testing supplies and reagents (approximately 0·25 tests per 1000 people per week between March and July, 2020).
This situation is similar to other parts of the world; serological studies from the USA, Spain, and Brazil identified an order of magnitude or more difference between laboratory-confirmed case counts and community infections.
Evidence before this study
Since the start of the SARS-CoV-2 outbreak in early 2020, many national and subnational prevalence estimates have been reported around the world. However, few prevalence estimates have been reported for Africa. Although testing has been less widely available in Africa than other parts of the world, the disease burden predicted early in the pandemic on the basis of the experience of other countries has not been observed in the continent. We searched PubMed from database inception to Jan 3, 2021, for peer-reviewed and preprints using the search terms “COVID-19” AND “prevalence” AND “Africa”. Additionally, we searched bibliographies of identified studies, a database of seroprevalence studies maintained by WHO, and the Google search engine for manuscripts and unpublished reports. We identified 11 studies reporting prevalence estimates from seven African countries: Congo (Brazzaville; n=1), Ethiopia (n=2), Kenya (n=1), Malawi (n=1), Nigeria (n=3), South Africa (n=2), and Togo (n=1). All studies were cross-sectional and included varying populations (eg, blood donors, antenatal clinic attendees, health workers, people with HIV), and most had small sample sizes. Most studies used only antibody tests to estimate seroprevalence and none were population-based studies. SARS-CoV-2 prevalence estimates ranged from 1·6% to 45·1%.
Added value of this study
To our knowledge, this is the first population-based SARS-CoV-2 prevalence study done in Africa. The findings showed high prevalence of rtPCR-positive SARS-CoV-2 infections in Zambia in July, 2020, which was a period of community transmission in the country. Transmission might have been minimal before the first wave in July and August, 2020. Few people who tested positive for SARS-CoV-2 were symptomatic.
Implications of all the available evidence
This study demonstrates that laboratory-confirmed case counts might be underestimated by an order of magnitude or more in Zambia. The inclusion of rtPCR testing in the study design identified a substantial proportion of people who tested positive for SARS-CoV-2, compared with serological testing alone. The low proportion of symptomatic infections could partly explain the discrepancy between the number of estimated infections and the lower than expected impact of COVID-19 in Zambia. Reasons for the apparent lower COVID-19 severity observed in Africa warrant further study.
In Niger State, Nigeria, seroprevalence among a small sample of randomly selected individuals was 25·4% in late June, 2020.
In Cape Town, South Africa, seroprevalence among several selected groups was 44·6% during the downslope of the first wave.
In May and June, 2020, SARS-CoV-2 seroprevalence was 12·3% among health-care workers in Blantyre, Malawi.
Among blood donors in Kenya, SARS-CoV-2 seroprevalence was 5·2% from April to June, 2020.
Modelled estimates from Kenya suggest more widespread disease in the country, with lower severity than that observed in other regions of the world.
Differences in population demographics (ie, young age structure of populations) and disease epidemiology (ie, high prevalence of infectious diseases such as HIV, tuberculosis, and malaria) in Africa compared with other heavily affected areas might affect SARS-CoV-2 epidemiology. Representative studies are needed to understand the epidemiology of SARS-CoV-2 in Africa to inform national public health responses. We aimed to estimate SARS-CoV-2 prevalence in six districts of Zambia in July, 2020, using a population-based household survey.
Study design and study population
Within each district, 16 standardised enumeration areas were randomly selected as primary sampling units using probability proportional to size. All households within each standardised enumeration area were listed and 20 households from each standardised enumeration area were selected using simple random sampling. All individuals (of any age) who had slept in the house the night before the survey was done were eligible for participation in the survey.
Participants were administered a questionnaire that included information about demographics, medical history, SARS-CoV-2 exposures, and history of recent illness on a tablet using the research electronic data capture (REDCap) application hosted by the Zambia Ministry of Health (Lusaka, Zambia). SARS-CoV-2 exposures included known contact with a laboratory-confirmed case, travel (domestic or international), usual means of transportation, health facility use in the past month, in-person attendance to work or school, and the number of visits to markets or grocery stores. Recent illness was assessed by asking if the participant had experienced any illnesses since February, 2020 (ie, before the first reported case in Zambia); if they responded affirmatively, symptomology was ascertained. All responses to the questionnaire were self-reported.
Participants were tested for SARS-CoV-2 infection by real-time PCR (rtPCR) using nasopharyngeal specimens, and for anti-SARS-CoV-2 antibodies by ELISA using plasma specimens at the University Teaching Hospital (Lusaka, Zambia) and the Centre for Infectious Disease Research in Zambia, (Lusaka, Zambia). Nasopharyngeal specimens were collected using scored swabs (Citoswab; Citotest Labware, Haimen, China). With a participant tilting their head back slightly, the swab was inserted until encountering physical resistance, rotated briefly, and withdrawn and placed into a 5 mm specimen bottle containing a viral transport medium. Blood specimens for antibody testing were collected in 500 μL edetic acid cryovial microtainer tubes using finger-prick or heel-prick (for children aged <6 months); venepuncture for blood was used as an alternative procedure in the event that finger-prick or heel-prick was unsuccessful, or according to the participant’s preference. All study specimens were transported in cooler boxes on ice to a local laboratory in each district on the same day. Blood specimens were centrifuged to separate plasma, which was transferred into a separate cryovial and stored at −20°C or below pending testing.
to resolve or confirm any non-negative results and the result of the CDC assay was considered final. Primers and probes for the CDC assay were obtained from Inqaba (Johannesburg, South Africa).
The Euroimmun ELISA (PerkinElmer, Waltham, MA, USA) for anti-spike protein IgG was done in single replicate according to manufacturer’s instructions. Positive or negative results were considered final. Borderline results were re-run in duplicate and considered positive or negative if both results from the duplicate run were positive or negative; the final result was deemed borderline if both results from the duplicate were borderline or if either duplicated result was discrepant.
Positive rtPCR results were communicated to district teams for case investigation and contact tracing per national guidelines. Negative rtPCR and all ELISA results were returned to participants by study staff.
SARS-CoV-2 prevalence and 95% CIs were calculated as the number of positive test results divided by the total number of tests done overall and per district and overall during the survey period. Estimates were calculated for rtPCR and ELISA separately and for the combined measure (rtPCR and ELISA). We calculated prevalence ratios (PRs) for the combined measure using Poisson regression to assess for associations between demographic and behavioural factors and for SARS-CoV-2 prevalence. The χ2 test was used to assess differences in prevalence across districts. Sampling weights were calculated based on the sampling frame and non-response weights (for questionnaire and each laboratory test) were calculated at the household, standardised enumeration area, and district levels. Additionally, each set of weights was calibrated to the population estimates at the district level by age and sex. Estimates were weighted, thus raw participant numbers were not reported in the analysis. Variance estimation accounted for clustering at districts and standardised enumeration areas when calculating 95% CIs and during hypothesis testing. An intracluster correlation coefficient of 0·12 (95% CI 0·06–0·18) was calculated using ANOVA to assess the degree of household clustering of SARS-CoV-2. Analyses were done using SAS (version 9.4) and the svy package in R (version 4.0.3).
These numbers were compared with the total number of reported cases in each district at the end of the study (July 31, 2020) to estimate the ratio of reported cases to total SARS-CoV-2 infections in each district. Additionally, the proportion of people who reported knowing their positive SARS-CoV-2 status before testing was reported for rtPCR and ELISA tests separately. We also did a sensitivity analysis excluding 333 participants for whom epidemiological data were disassociated from laboratory results during the study.
Role of the funding source
The funder of the study was involved in the study design, data analysis, and data interpretation, and writing of the report.
Table 1Participant demographics
Data are % (95% CI). Estimates were weighted, thus raw participant numbers were not reported.
* To maintain the 333 test results that were dissociated from the epidemiological data of the participants, prevalence estimates were weighted using the standardised enumeration area testing response rate instead of household testing response rate and age and sex were calibrated at the district level instead of the individual level; the results of a sensitivity analysis excluding these dissociated test results were not significantly different from the main study findings.
Table 4Associations between demographic and behavioural factors and SARS-CoV-prevalence (combined measure) in six districts of Zambia in July, 2020 (n=1952)
Estimates were weighted, thus raw participant numbers were not reported.
In France, an estimated 14% of symptomatic infections were detected by the public health system between May and June, 2020.
Paradoxically, people reporting contact with a confirmed COVID-19 case had lower SARS-CoV-2 prevalence in this study. It is possible that individuals with known exposure to people with COVID-19 took additional individual preventive measures to avoid becoming infected. Furthermore, many people were likely being unknowingly exposed within the community considering the widespread transmission in July, 2020, in Zambia.
whereas studies from the USA and Europe have suggested similar severity in people with and without HIV infection.
In Cape Town, people with HIV had higher seroprevalence than women attending antenatal clinics.
In an HIV clinic in Barcelona, Spain, the incidence of COVID-19 among patients was lower than that among the general population of the city.
Our study was done in a country experiencing a generalised HIV epidemic, and thus a large proportion of the study population had HIV. Although the prevalence of SARS-CoV-2 was higher among people with HIV than those who were HIV negative in this study, the difference was not significant; however, this study was not powered to detect such a difference. It is unclear what effect ART, which nearly all people with HIV reported taking, had on this finding. Further studies are needed to understand the effect of HIV on SARS-CoV-2 infection severity and prevalence considering the burden of HIV in Zambia and Africa overall.
the proportion of asymptomatic infections in this study is higher than reported elsewhere.
Recall bias could have reduced symptom reporting; however, a high proportion of asymptomatic SARS-CoV-2 infections could help explain the paradox between the large number of SARS-CoV-2 infections estimated in this study and the relatively mild strain on hospital services observed during the first epidemic peak in Zambia compared with experiences in Europe and North America. The lower apparent severity observed might be a result of the young population in Zambia, since younger individuals are less likely to have symptoms and develop severe illness than older individuals.
Participants voluntarily participated in each aspect of the study (ie, interview and nasopharyngeal and blood specimen collection), and the response rate for participants who had both rtPCR and ELISA tests was low (46% of all participants), which could have biased estimates; therefore, rtPCR and ELISA prevalence estimates were also reported separately. The pooled estimate should be interpreted with caution because the districts in this study were purposefully selected. Past medical history (including HIV status) and potential exposure to SARS-CoV-2 might have been misreported. Data collection occurred during a dynamic period in the COVID-19 outbreak in Zambia, complicating interpretation of the estimates. Some individuals shed SARS-CoV-2 genetic material for weeks and some can quickly mount an antibody response; however, rtPCR positivity is likely to reflect SARS-CoV-2 infection in the past 2–3 weeks, whereas ELISA for IgG antibodies is likely to reflect past infection. The ELISA used has a reported sensitivity of about 90%, and serological cross-reactivity is an emerging area of investigation in Africa.
Since this is an observational study, causality between reported associations cannot be determined.
will help rapidly detect SARS-CoV-2 infections, allowing for early isolation of infected people and timely identification of contacts, while helping to curb SARS-CoV-2 transmission in Zambia. Serial prevalence surveys will provide the Zambian Government insight with regard to the true extent of disease transmission over time and can inform vaccine strategy. Depending on the stage of the epidemic at the time of subsequent prevalence studies, the use of both rtPCR and ELISA testing should be considered because—as shown in this study—the relatively large proportion of people with rtPCR-positive SARS-CoV-2 infection would have been missed if participants had only been tested by ELISA for anti-SARS-CoV-2 antibodies. While SARS-CoV-2 continues to spread in the community, the Zambian Government should continue to aggressively promote community mitigation measures, including rapid detection and isolation of people with confirmed SARS-CoV-2 infection, identification and quarantine of people who have been in close contact with confirmed cases, universal mask wearing in public, and physical distancing measures, which have been shown to reduce SARS-CoV-2 transmission.
LBM, JZH, SF, SY, AW, DTB, JF, JEZ, DB, NB, BC, MK, NS, NK, PMZ, DC, FM, CC, VM, SA, and KM designed the study. SF, LC, MS, JEZ, DB, KIN, DK, AS, CM, SS, KDZ, AMa, and AMw collected data. JZH and JF accessed and verified the underlying data. JZH did the literature search. JZH, DTB, JF, JEZ, and DB analysed data. LBM, JZH, SF, SY, AW, DTB, JF, JEZ, DB, SA, and KM did data analysis. LBM, JZH, SF, LC, MS, SY, AW, DTB, JF, JEZ, DB, KIN, DK, NB, BC, BH, TLSJ, AS, CM, SS, KDZ, AMa, MK, AMw, NS, NK, PMZ, DC, FM, CC, VM, SA, and KM wrote the manuscript. JZH, DTB, AW, and SA produced the figures. All authors had full access to all the data in the study and had final responsibility for the decision to submit for publication.
Declaration of interests
We declare no competing interests.
Deidentified participant data used for this analysis can be requested from the Zambian Ministry of Health after July 31, 2021. Interested researchers must submit a research proposal for consideration by the study investigators. If approved, the requestor must sign a data use agreement. Additionally, the study protocol is available for request. All data requests should be directed to the corresponding author.
This work was funded by the US President’s Emergency Plan for AIDS Relief through the US Centers for Disease Control and Prevention (CDC) and the CDC Emergency Response to the COVID-19 pandemic. The findings and conclusions in this Article are those of the authors and do not necessarily represent the official position of the US Centers for Disease Control and Prevention.
First 100 persons with COVID-19— Zambia, March 18–April 28, 2020.
MMWR Morb Mortal Wkly Rep. 2020; 69: 1547-1548
Press briefing on COVID-19 and additional preventive and control measures.
Evidence supporting transmission of severe acute respiratory syndrome coronavirus 2 while presymptomatic or asymptomatic.
Emerg Infect Dis. 2020; 26: E1-E6
Coronavirus disease 2019 case surveillance—United States, January 22–May 30, 2020.
MMWR Morb Mortal Wkly Rep. 2020; 69: 759-765
Signs and symptoms to determine if a patient presenting in primary care or hospital outpatient settings has COVID-19 disease.
Cochrane Database Syst Rev. 2020; 7CD013665
Let Africa into the market for COVID-19 diagnostics.
Nature. 2020; 580: 565
Prevalence of SARS-CoV-2 antibodies in a large nationwide sample of patients on dialysis in the USA: a cross-sectional study.
Lancet. 2020; 396: 1335-1344
Estimated SARS-CoV-2 seroprevalence in the US as of September 2020.
JAMA Intern Med. 2020; 30329: 1-11
Prevalence of SARS-CoV-2 in Spain (ENE-COVID): a nationwide, population-based seroepidemiological study.
Lancet. 2020; 396: 535-544
Remarkable variability in SARS-CoV-2 antibodies across Brazilian regions: report on two successive nationwide serological household surveys.
SSRN. 2020; ()
Bicycle taxi operators with greater exposure to Coronavirus in Quelimane.
Most people exposed to COVID-19 have no symptoms 2020.
Sero-prevalence of anti-SARS-CoV-2 antibodies in Addis Ababa, Ethiopia.
bioRxiv. 2020; ()
Seroprevalence of COVID-19 in Niger State.
medRxiv. 2020; ()
SARS-CoV-2 seroprevalence in the Cape Town Metropolitan Subdistricts after the peak of infections.
Higher SARS-CoV-2 seroprevalence in workers with lower socioeconomic status in Cape Town, South Africa.
Research Square. 2020; ()
High SARS-CoV-2 seroprevalence in health care workers but relatively low numbers of deaths in urban Malawi.
medRxiv. 2020; ()
Seroprevalence of anti-SARS-CoV-2 IgG antibodies in Kenyan blood donors.
Science. 2020; 371: 79-82
Revealing the extent of the COVID-19 pandemic in Kenya based on serological and PCR-test data.
medRxiv. 2020; ()
2010 census of population and housing. Population and demographic projections 2011–2035.
Coronavirus disease (COVID-19) technical guidance: the Unity Studies: early investigation protocols.
World Health Organization,
CDC 2019-novel coronavirus (2019-nCoV) real-time RT-PCR diagnostic panel.
Underdetection of cases of COVID-19 in France threatens epidemic control.
Nature. 2021; 590: 134-139
Detection of B.1.351 SARS-CoV-2 variant strain—Zambia, December 2020.
MMWR Morb Mortal Wkly Rep. 2021; ()
Risk factors for COVID-19 death in a population cohort study from the Western Cape Province, South Africa.
Clin Infect Dis. 2020; ()
Description of COVID-19 in HIV-infected individuals: a single-centre, prospective cohort.
Lancet HIV. 2020; 7: e554-e564
SARS-CoV-2 and HIV coinfection: clinical experience from Rhode Island, United States.
J Int AIDS Soc. 2020; 23e25573
Outcomes among HIV-positive patients hospitalized with COVID-19.
J Acquir Immune Defic Syndr. 2020; 85: 6-10
Clinical characteristics, risk factors, and incidence of symptomatic coronavirus disease 2019 in a large cohort of adults living with HIV: a single-center, prospective observational study.
AIDS. 2020; 34: 1775-1780
Estimating the extent of asymptomatic COVID-19 and its potential for community transmission: systematic review and meta-analysis.
Off J Assoc Med Microbiol Infect Dis Canada. 2020; 5: 223-234
Age-dependent effects in the transmission and control of COVID-19 epidemics.
Nat Med. 2020; 26: 1205-1211
Serology test evaluation report for “SARS-COV-2 ELISA (IgG)” from Euroimmun.
High prevalence of pre-existing serological cross-reactivity against SARS-CoV-2 in sub-Sahara Africa.
Int J Infect Dis. 2020; 102: 557-583
Antigen-detection in the diagnosis of SARS-CoV-2 infection using rapid immunoassays: interim guideance.
World Health Organization,
Geneva11 September 2020
Published: March 09, 2021
Published by Elsevier Ltd.
For non-commercial purposes:
- Read, print & download
- Redistribute or republish the final article
- Text & data mine
- Translate the article (private use only, not for distribution)
- Reuse portions or extracts from the article in other works
- Sell or re-use for commercial purposes
- Distribute translations or adaptations of the article