Excess mortality associated with the COVID-19 pandemic during the 2020 and 2021 waves in Antananarivo, Madagascar

Abstract

Introduction: COVID-19-associated mortality remains difficult to estimate in sub-Saharan Africa because of the lack of comprehensive systems of death registration. Based on death registers referring to the capital city of Madagascar, we sought to estimate the excess mortality during the COVID-19 pandemic and calculate the loss of life expectancy.

Methods: Death records between 2016 and 2021 were used to estimate weekly excess mortality during the pandemic period. To infer its synchrony with circulation of SARS-CoV-2, a cross-wavelet analysis was performed. Life expectancy loss due to the COVID-19 pandemic was calculated by projecting mortality rates using the Lee and Carter model and extrapolating the prepandemic trends (1990-2019). Differences in life expectancy at birth were disaggregated by cause of death.

Results: Peaks of excess mortality in 2020-21 were associated with waves of COVID-19. Estimates of all-cause excess mortality were 38.5 and 64.9 per 100 000 inhabitants in 2020 and 2021, respectively, with excess mortality reaching ≥50% over 6 weeks. In 2021, we quantified a drop of 0.8 and 1.0 years in the life expectancy for men and women, respectively attributable to increased risks of death beyond the age of 60 years.

Conclusion: We observed high excess mortality during the pandemic period, in particular around the peaks of SARS-CoV-2 circulation in Antananarivo. Our study highlights the need to implement death registration systems in low-income countries to document true toll of a pandemic.

Keywords: COVID-19; epidemiology; public health; respiratory infections.

Figure 1

Temporal trends of all-cause crude mortality from 2016 to 2021 in the five central districts of Antananarivo (n=45 959).

Figure 2

Weekly temporal trends, in five districts of Antananarivo from 2020 to 2021. (A) All-cause excess mortality using 2016–19 crude deaths as baseline years. (B) All-cause excess mortality by age group using 2016–19 crude deaths as baseline years (no death record was observed for age group 5–14 years during 1 week in 2020 and 2 weeks in 2021). (C) SARS-CoV-2 test results and positivity rates (data from the Institut Pasteur de Madagascar (IPM) for Antananarivo). Supplementary materials representing weekly all-cause deaths by age group (online supplemental table S1) and all-cause excess mortality by sex (online supplemental figure S2) support this figure.

Figure 3

Cross-wavelet coherence and phase-difference of the SARS-CoV-2 positivity rate and all-cause death rate in Antananarivo. These two time series were in phase from June 2020 to September 2021 (phase-differences between $0$ and $\frac{\pi }{2}$). During the first wave of COVID-19 (June 2020), the death rate seemed to lead the SARS-CoV-2 positivity rate (phase-differences $-\frac{\pi }{2}$ and $0$). In the second wave (February 2021–May 2021), SARS-CoV-2 positivity rate significantly led the death rate (phase-differences between $0$ and $\frac{\pi }{2}$). The significant coherence level is ranked from blue (0.0) to red (1.0) colour scale, and statistical significance is marked by the white contour plots, when p<0.05. Supplementary materials representing all-causes deaths (online supplemental figure S1B) and SARS-CoV-2 tests positivity (online supplemental figure S1D), also online supplemental method S3 support the interpretation of this figure.

Figure 4

Observed and predicted life expectancy at birth in Antananarivo (1976–2021) for males (green) and females (blue). We projected to 2020–24, the observed life expectancy at birth and the predicted (bold) for both sexes based on the rates measured from 1990 to 2019.