COVID-19 Surveillance in Madagascar and Urban Burkina Faso: Addressing Underreporting of Disease Burden Through Integrative Analysis of Diverse Data Streams

Abstract

Background: Coronavirus disease 2019 (COVID-19) caused substantial disease and death worldwide since December 2019, but the burden was lower in Africa than in high-income countries. To address potential underreporting, we modeled severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection and disease burden in Burkina Faso and Madagascar.

Methods: Prospectively enrolled patients who presented with fever at sentinel healthcare facilities were assessed for active SARS-CoV-2 infection. Household members of SARS-CoV-2-infected patients were prospectively followed for confirmed SARS-CoV-2 infection. Archived serum specimens that spanned the pandemic onset in Madagascar to the start of prospective surveillance were tested for anti-SARS-CoV-2 immunoglobulins. Data from these multiple sources contributed to an integrated analysis to calibrate an epidemiologic mass action model.

Results: COVID-19 accounted for a substantial fraction of healthcare-ascertained febrile illness in both Burkina Faso and Madagascar, with symptom profiles consistent with those previously reported. SARS-CoV-2 vaccination coverage was very low in Burkina Faso and unavailable in Madagascar. The household secondary attack rate was 28% (95% confidence intervals [CI], 22%-35%] in Madagascar and 31% (95% CI: 9%-68%) in Burkina Faso, indicating substantial transmission of the disease within households in both locations. Model simulations estimated that the actual number of SARS-CoV-2 infections was at least nine times higher than the reported number of febrile COVID-19 cases.

Conclusions: Africa has faced persistent challenges due to underinvestment in vaccination programs and disease surveillance programs. There was substantial underreporting of COVID-19 cases during the pandemic in both countries. Our findings call for improving systems and resources in disease surveillance during epidemic and interepidemic periods in these countries.

Keywords: Burkina Faso; COVID-19; Madagascar; serosurveillance.

Figure 1.

Observed and susceptible, infectious, and recovered (SIR)–modeled weekly frequency of reverse-transcription polymerase chain reaction–or rapid antigen test–positive symptomatic SARS-CoV-2 cases for the period of prospective surveillance in Burkina Faso (A) and Madagascar (B). Observed case counts are shown as solid black points. The red line plots the weekly average for 1000 synthetic epidemics, a randomly selected 10% of which are shown in gray. For Madagascar, the period seroprevalence of anti–SARS-CoV-2 IgG positivity among retrospectively analyzed serologic specimens collected from febrile illness cases (blue solid points) is compared to the model predicted proportion of individuals with a history of at least 1 SARS-CoV-2 infection (mean of 1000 epidemics, red; 100 sample epidemics, green). Dashed vertical gray lines denote the 3-week period during which prospective surveillance in Madagascar was being initiated. SIR model–derived seroprevalence curve for SARS-CoV-2 IgG seropositivity (mean, red; 95% confidence bands, gray) in Burkina Faso (C) and Madagascar (D) starting in early March 2020. The data collection times are indicated as ticks on the x-axis, with green indicating the start of each collection period for retrospectively analyzed archived serum specimens. Abbreviations: COVID-19, coronavirus disease 2019; Ig, immunoglobulin; SARS-CoV-2, severe acute respiratory syndrome coronavirus 2.

Figure 2.

Diagram of dynamic epidemic model describing the transmission of SARS-CoV-2 at surveillance SETA/COVIA surveillance sites in Madagascar and Burkina Faso. Abbreviations: FOI, force of infection; Ig, immunoglobulin; SAR, secondary attack rate; SARS-CoV-2, severe acute respiratory syndrome coronavirus 2.