Sudan’s capital, Khartoum, is a tripartite metropolis comprising Khartoum, Bahri, and Omdurman; it has >8 million inhabitants (6). We chose Omdurman, the largest of the 3 cities, as the study site for 2 surveys conducted in March–July 2020 (Appendix). One, a retrospective mortality survey, was conducted using a 2-stage cluster sampling methodology based on random geopoints with 2 recall periods, the prepandemic (January 1, 2019–February 29, 2020) and the pandemic period (March 1, 2020–date of survey); an adult representative of the household answered a standardized questionnaire. The second was a nested SARS-CoV-2 antibody prevalence survey; all the members of a subset of the household, regardless of age, were invited to participate in the seroprevalence study.

Capillary blood was collected on dried blood spot cards and directly tested with the STANDARD Q COVID-19 IgM/IgG Combo rapid diagnostic test (RDT) (SD–Biosensor, https://www.sdbiosensor.com). All participants who tested positive for any isotype were considered seropositive. Dried blood spot cards (Euroimmun, https://www.euroimmun.com) were transferred to the National Public Health Laboratory (NPHL; Khartoum, Sudan) for further analysis by ELISA (Anti–SARS-CoV-2 ELISA [IgG, S1 domain]; Euroimmun) to compare with the rapid test results (7,8). To adjust our seroprevalence estimates using published validation data for both ELISA and RDT tests, we conducted a meta-analysis with random effects and a Bayesian latent class model (Appendix).

Figure 1

Figure 1. Survey flow for cross-sectional study of SARS-CoV-2 prevalence and population-based death rates, Omdurman, Sudan, 2021. DBS, dry blood spot; RDT, rapid diagnostic test.

During March 1–April 10, 2021, a total of 2,374 (62.3%) participants from 555 households (Figure 1) agreed to provide blood; 34.3% (95% CI 32.4%–36.2%; Table 1) of them had detectable SARS-CoV-2 antibodies (IgM, IgG, or both). After adjusting for immunoassay performance for detecting previous infections, we estimated a seroprevalence of 54.6% (95% CI 51.4%–57.8%), noting a clear increase of seroprevalence risk with age (Table 1). We found the highest seroprevalence of 80.7% (95% CI 71.7%–89.7%) among participants >50 years of age. Assuming a population size of 3,040,604 for Omdurman on the basis of the data collected in the survey and the data provided by the Ministry of Planning, we estimate that 1,660,170 (95% CI 1,458,225–1,863,936) persons had been infected by SARS-CoV-2 at the time of the survey.

Figure 2

Figure 2. Comparison of estimated and reported deaths from coronavirus disease, Sudan, January 2019–April 2021. A) Distribution of all deaths as reported in a population-based cross-sectional survey in the city of Omdurman,...

We found evidence of significant clustering of seropositivity within households; 364 households (65.6%) had >1 positive household member. Living with a person who was seropositive led to a 1.68-fold (odds ratio [OR] 95% CI 1.35–2.08; p<0.001) increase in the odds of being seropositive (Appendix). Among the 4,086 households visited (Figure 1), we enumerated 27,315 persons who had been a household member at some time after January 1, 2019. Among them, 319 deaths were reported, including 206 (64.6%) among persons >50 years of age and 30 (9.4%) among children <5 years of age. The deaths increased in 2020 during the pandemic period, consistent with the reported countrywide confirmed COVID-19 deaths (Figure 2).

The overall death rate for the whole recall period was 0.16 (95% CI 0.13–0.18) deaths/10,000 population/day (Table 2). The crude death rate significantly increased by 67% (95% CI 32%–110%) from 0.12 (0.10–0.14) deaths/10,000 population/day for the prepandemic period to 0.20 (0.16–0.23) deaths/10,000 population/day for the pandemic period. This difference was even more pronounced among those >50 years of age; deaths increased 74% (95% CI 30%–133%; p<0.001) between the 2 periods. (Table 2). On the basis of on our estimates of the population size of Omdurman and the death rates, we estimated 7,113 excess deaths (95% CI 5,015–9,505) during the pandemic period and that 5,125 (95% CI 4,165–6,226) of these occurred in persons >50 years of age.

Our findings indicate that mortality rates in the overall population of Omdurman increased by 67% during the first pandemic year; the highest increase (74%) was among the population >50 years of age. We estimated an excess of 7,113 all-cause deaths during the pandemic period, compared with 287 COVID-19–related deaths officially reported for Omdurman; these data were obtained from the Khartoum Ministry of Health. We have considered the potential limitation of having a recall period >2 years for mortality estimates, which could introduce bias for deaths occurring at the beginning of the recall period. Surveyors were trained to be aware of this factor to mitigate those bias (Appendix).

The crude seroprevalence estimate shows how widespread SARS-CoV-2 infection was, affecting all age groups, especially persons >50 years of age. However, the estimates based on RDT results might have underestimated the seroprevalence as a result of several limitations. First, we conducted our survey 1 year after the earliest SARS-CoV-2 infection was detected in Sudan, so a varying degree of antibody decay over time could be expected (9,10). Second, when antibodies remain present in the blood, their detection is limited by the performance of the RDT (11). To overcome those limitations, we adjusted the crude results; we observed a 20% increase in the overall seroprevalence. With that estimation we calculated that the number of infections was 50 times higher than the number of COVID-19 cases recorded by the end of the survey, which was consistent with other case-to-infection ratios in low-income settings in Africa and Asia (12,13). Despite this high seroprevalence, another wave of infection occurred right after the survey (May–June 2021); comparing it with the previous wave, we saw that fewer cases but more deaths per case were reported. Three more waves occurred during September 2021–January 2022, the latest one reporting a record number of weekly cases (14). No sequencing data was available as of January 2022; therefore, it was impossible to discuss the emergence of new variants and their impact on the new waves of infections given the prior seroprevalence we estimated in this survey.

In summary, this population-based cross-sectional survey in Omdurman, Sudan, demonstrated significantly higher death rates during the COVID-19 pandemic compared with those of the prepandemic period, particularly affecting persons >50 years of age. We also found elevated SARS-CoV-2 seropositivity, affecting older populations the most. Our results suggest that Omdurman, one of the largest population centers in Africa, was severely affected by the COVID-19 pandemic and that excess mortality rates were much higher than reported COVID-19 deaths.