Skip to main content

COVID-19 deaths detected in a systematic post-mortem surveillance study in Africa

Our take —

This study, available as a preprint and thus not yet peer reviewed, enrolled 364 deceased individuals, whose deaths were registered at the University Teaching Hospital in Lusaka, Zambia of whom 19.2% tested positive for SARS-CoV-2. Most of these deaths occurred in the community and the median age of deaths was 48 years. While this method of testing post-mortem provides a period-prevalence estimate for SARS-CoV-2 infection among decedents between June and September 2020, it does not provide detailed information on the cause of death.

Study design


Study population and setting

Post-mortem SARS-CoV-2 testing (nasopharyngeal swab, RT-qPCR) was performed on 372 deceased individuals at the University Teaching Hospital in Lusaka, Zambia between June and September 2020. In order to be included in this study, individuals had to have died in the 48 hours before enrollment and next of kin had to provide written, informed consent (~99% consented). The study used quasi-random sampling (enrolling every 5th death in July and every 3rd death in August with daily cap of 5 deaths; and without a daily cap in September). Deaths were included regardless of age or cause. Testing results were described by month, by whether the death occurred at a facility or in the community, by age, and by sex.

Summary of Main Findings

Testing results were obtained for 364/372 of the deceased (97.8%) (n=10 samples could not be matched to enrollment, and n=2 tested positive antemortem outside of the study). A two-week period in September 2020 accounted for n=59 deaths (~16%) enrolled, while a two-week period in July accounted for n=30 deaths(~8%). Among the 364 deaths with test results, about a quarter were deaths that happened in a facility/health center (n=96) and 74.2% were deaths the occurred in the community (n=268). 19.2% of deceased in this sample tested positive for SARS-CoV-2 (70/364). The greatest proportion of deaths were documented in late July and early August. 70% of those testing positive for SARS-CoV-2 were male, while in the burial registry overall, about 60% were male. The median age among those testing positive for SARS-CoV-2 was 48 years (IQR: 36-72).

Study Strengths

Enrollment of the deceased took place from the University Teaching Hospital in Lusaka, Zambia, where close to 80% of all deaths in the city are registered, including those that occur at the hospital itself and from the community. The investigators took steps during enrollment to get a representative sample of the deceased registered at the facility.


The authors switch between using the terminology of “CV19 deaths” and “CV19+ deaths.”; this study did not ascertain whether deaths were caused by COVID-19 (COVID-19 attributable deaths) and, therefore, the use of “CV19 deaths” could be misleading. The authors cite US CDC guidance indicating that in deaths where SARS-CoV-2 is detected the virus should be assumed to be underlying cause of death or a contributor. Given competing risks of death in low-income communities in Zambia (e.g. tuberculosis, HIV, etc.), US CDC guidance may not be directly applicable and attributing deaths to COVID-19 may be incorrect. Additionally, because of the sampling strategy, the prevalence estimate will weighted more heavily towards periods in which enrollment was the highest (i.e. September 2020).

Value added

This study uses post-mortem SARS-CoV-2 testing to ascertain prevalence among recently deceased. This strategy may be particularly useful to determine burden of disease in places where antemortem testing is not done.

This review was posted on: 22 January 2021