Re: Covid-19 deaths in Africa: prospective systematic postmortem surveillance study
Dear Editor
Rapid response to “Covid-19 deaths in Africa: prospective systematic postmortem surveillance study”.
The British Medical Journal (BMJ), on 17th February 2021, published a paper titled “Covid-19 deaths in Africa: prospective systematic postmortem surveillance study,” which concluded that cases of covid-19 were under-reported because testing was rarely done and that if their data were generalizable, the impact of covid-19 in Africa had been vastly underestimated. These conclusions are highly questionable because all the study participants in the paper were enrolled within a period where all deaths were swabbed and tested at the University Teaching Hospital (UTH) mortuary [appendixes 1, 6-13].
As background, in April 2020, the Zambia National Public Health Institute initiated the Covid-19 Facility Active Screening and Mortality Surveillance in community deaths. This was first conducted at the UTH mortuary in Lusaka [appendix 1] and later scaled-up to include selected mortuaries around the country [appendix 2-4]. Data in the UTH mortuary show that between April 2020 and February 2021, 3562 deaths were tested for covid-19. Thus mortality and disease surveillance has remained an important component of the covid-19 fight in Zambia [1].
We observe that the paper title is misleading and inaccurate because the study subjects were drawn from Lusaka, Zambia, a country in sub-Saharan Africa. We are concerned as to why the authors conclude that a study sampling three hundred and seventy covid deaths in Lusaka would represent a general picture of all African cities. Sub-Saharan cities are different from North African ones, and their diversity is difficult to conceptualise [2]. This title is sensationalist and alarmist to the scientific community in general.
We submit in retrospect that the title should have read "Covid-19 deaths in Lusaka: prospective systematic postmortem surveillance study".
We observe that the introduction of the paper does not provide the reader with background information or a theoretical context on the fatal impact of covid-19 in an urban African population. The rationale of the study is challenging to deduce, and neither is the objective stated. We note that the main purpose of an introduction is to describe the study's importance by providing a rationale behind the design used to accomplish the stated objective [3].
The objective set out was to directly measure the fatal impact of coronavirus disease 2019 (Covid-19) in an urban African population. We observe that to achieve this objective, this study should have tabulated the number of deaths in the previous years before the covid-19 pandemic and the number of deaths during the pandemic. The authors further assume that not testing people in the community before they died was undoubtedly an important factor in underestimating the impact of Covid-19 in Lusaka. We argue that the fatal impact of covid-19 in Lusaka was not underestimated as community deaths were all tested by the surveillance team. So the policy-makers had a real picture of the burden of covid-19 at the time [4].
The methodology does not spell out how the study would measure the fatal impact of the covid-19, which is the object of the study. We categorically state that the authors have not measured the fatal impact of covid-19 in this study based on their methods as this study does not answer its objective. The authors, instead, enumerated deaths that had a positive test on PCR.
The authors got carried away writing about Zambia's economy. We are at pains to understand how this strengthens their objective. This piece of information may be irrelevant and misplaced to the study's title and object.
The authors mislead the scientific community by stating that health facilities may issue death certificates legally required to inter a body. To inter a body, one would require either a Medical Certification of the Cause of Death, Brought-in-Dead Certificate or a Coroners' Authority for Burial [appendix 14-17]. Issuance of a DC is the responsibility of the Registrar-General at the Ministry of Home Affairs [5, 6, 7]. We are concerned that this misrepresentation of the process may be cited as the correct process of obtaining a burial permit, given the journal's authoritative nature (The BMJ). We, therefore, urge the authors to correct this anomaly.
The authors state that the study enrolled community deaths, but we observe no violent deaths (homicides, suicides, accidents) in the results. We note that sampling natural deaths from the community is not representative of all community deaths, and hence positivity rate in this study is biased towards the natural causes.
The authors should have declared clearly that they are testing deaths that seemingly had a natural cause. Of note, the ZPRIME protocol is biased towards respiratory symptoms. The authors observed a high prevalence of mortality related to covid-19 because the study subjects were pooled from natural causes only. This in itself may denote a confirmation bias.
The authors used an abbreviated verbal autopsy tool to identify underlying risk factor comorbidities. We are concerned that this is mere hearsay with no scientific proof based on testing or examination. There was no effort to collaborate what was said by the informant and what was present in the deceased person. Were tuberculosis, hypertension, HIV/AIDS, alcohol misuse, and diabetes tested for postmortem? We do not observe this in the methodology.
The authors state that they were surprised to observe covid-19 deaths in children and further assume that it is a distinct feature of the presentation of covid-19 in Africa. There was no effort to validate the cause of deaths in these children. The study lost potential key information about the real mechanisms underlying deaths as verbal autopsies and PCR testing do not allow an extensive evaluation of the disease. This may even be overestimated given that the precise cause of death remains elusive [1, 8, 9].
The study has failed based on its methods to measure the fatal impact of covid-19 in Africa. We are confident that our concerns and observations will be taken into account by your highly revered editorial team.
Competing interests: No competing Interests
Reference
[1] Luchenga Mucheleng’anga, Cordilia Himwaze, The role of forensic pathology in the COVID-19 pandemic in Zambia, Forensic Science International: Reports, Volume 2,
2020, 100147, ISSN 2665-9107, https://doi.org/10.1016/j.fsir.2020.100147.
[2] Open edition journals, African Cities and the Development Conundrum
Actors and Agency in the Urban Grey Zone, https://journals.openedition.org/poldev/262. Retrieved on 25/02/2021 at 22:19 hours.
[3] Koehler, Steven A. Forensic Epidemiology (International Forensic Science and Investigation) (p. 249). Taylor and Francis CRC ebook account. Kindle Edition.
[4] Ministry of Health Zambia is at Ministry Of Health - Ndeke House,Lusaka, https://web.facebook.com/mohzambia. Retrieved on 25/02/2021 at 20:01 hours.
[5] The births and deaths registration act, chapter 51 of the laws of Zambia, http://www.parliament.gov.zm/sites/default/files/documents/acts/Births%2.... Retrieved on 27/02/2021 at 11:04 hours.
[6] Zambia-National-Strategic-Action-Plan-for-CRVS, http://www.crvs-dgb.org/wp-content/uploads/2015/11/Zambia-National-Strat.... Retrieved on 27/02/2021 at 11:10 hours.
[7] The inquests act chapter 36 of the laws of Zambia, http://www.parliament.gov.zm/sites/default/files/documents/acts/Inquests.... Retrieved on 27/02/2021 at 11:13 hours.
[8] Murray, C.J., Lopez, A.D., Black, R. et al. Population Health Metrics Research Consortium gold standard verbal autopsy validation study: design, implementation, and development of analysis datasets. Popul Health Metrics 9, 27 (2011) doi:10.1186/1478-7954-9-27
[9] Salerno M, Sessa F, Piscopo A et al (2020) No autopsies on COVID-19 deaths: a missed opportunity and the lockdown of science. J Clin Med 9. https://doi.org/10.3390/jcm9051472
Rapid Response:
Re: Covid-19 deaths in Africa: prospective systematic postmortem surveillance study
Dear Editor
Rapid response to “Covid-19 deaths in Africa: prospective systematic postmortem surveillance study”.
The British Medical Journal (BMJ), on 17th February 2021, published a paper titled “Covid-19 deaths in Africa: prospective systematic postmortem surveillance study,” which concluded that cases of covid-19 were under-reported because testing was rarely done and that if their data were generalizable, the impact of covid-19 in Africa had been vastly underestimated. These conclusions are highly questionable because all the study participants in the paper were enrolled within a period where all deaths were swabbed and tested at the University Teaching Hospital (UTH) mortuary [appendixes 1, 6-13].
As background, in April 2020, the Zambia National Public Health Institute initiated the Covid-19 Facility Active Screening and Mortality Surveillance in community deaths. This was first conducted at the UTH mortuary in Lusaka [appendix 1] and later scaled-up to include selected mortuaries around the country [appendix 2-4]. Data in the UTH mortuary show that between April 2020 and February 2021, 3562 deaths were tested for covid-19. Thus mortality and disease surveillance has remained an important component of the covid-19 fight in Zambia [1].
We observe that the paper title is misleading and inaccurate because the study subjects were drawn from Lusaka, Zambia, a country in sub-Saharan Africa. We are concerned as to why the authors conclude that a study sampling three hundred and seventy covid deaths in Lusaka would represent a general picture of all African cities. Sub-Saharan cities are different from North African ones, and their diversity is difficult to conceptualise [2]. This title is sensationalist and alarmist to the scientific community in general.
We submit in retrospect that the title should have read "Covid-19 deaths in Lusaka: prospective systematic postmortem surveillance study".
We observe that the introduction of the paper does not provide the reader with background information or a theoretical context on the fatal impact of covid-19 in an urban African population. The rationale of the study is challenging to deduce, and neither is the objective stated. We note that the main purpose of an introduction is to describe the study's importance by providing a rationale behind the design used to accomplish the stated objective [3].
The objective set out was to directly measure the fatal impact of coronavirus disease 2019 (Covid-19) in an urban African population. We observe that to achieve this objective, this study should have tabulated the number of deaths in the previous years before the covid-19 pandemic and the number of deaths during the pandemic. The authors further assume that not testing people in the community before they died was undoubtedly an important factor in underestimating the impact of Covid-19 in Lusaka. We argue that the fatal impact of covid-19 in Lusaka was not underestimated as community deaths were all tested by the surveillance team. So the policy-makers had a real picture of the burden of covid-19 at the time [4].
The methodology does not spell out how the study would measure the fatal impact of the covid-19, which is the object of the study. We categorically state that the authors have not measured the fatal impact of covid-19 in this study based on their methods as this study does not answer its objective. The authors, instead, enumerated deaths that had a positive test on PCR.
The authors got carried away writing about Zambia's economy. We are at pains to understand how this strengthens their objective. This piece of information may be irrelevant and misplaced to the study's title and object.
The authors mislead the scientific community by stating that health facilities may issue death certificates legally required to inter a body. To inter a body, one would require either a Medical Certification of the Cause of Death, Brought-in-Dead Certificate or a Coroners' Authority for Burial [appendix 14-17]. Issuance of a DC is the responsibility of the Registrar-General at the Ministry of Home Affairs [5, 6, 7]. We are concerned that this misrepresentation of the process may be cited as the correct process of obtaining a burial permit, given the journal's authoritative nature (The BMJ). We, therefore, urge the authors to correct this anomaly.
The authors state that the study enrolled community deaths, but we observe no violent deaths (homicides, suicides, accidents) in the results. We note that sampling natural deaths from the community is not representative of all community deaths, and hence positivity rate in this study is biased towards the natural causes.
The authors should have declared clearly that they are testing deaths that seemingly had a natural cause. Of note, the ZPRIME protocol is biased towards respiratory symptoms. The authors observed a high prevalence of mortality related to covid-19 because the study subjects were pooled from natural causes only. This in itself may denote a confirmation bias.
The authors used an abbreviated verbal autopsy tool to identify underlying risk factor comorbidities. We are concerned that this is mere hearsay with no scientific proof based on testing or examination. There was no effort to collaborate what was said by the informant and what was present in the deceased person. Were tuberculosis, hypertension, HIV/AIDS, alcohol misuse, and diabetes tested for postmortem? We do not observe this in the methodology.
The authors state that they were surprised to observe covid-19 deaths in children and further assume that it is a distinct feature of the presentation of covid-19 in Africa. There was no effort to validate the cause of deaths in these children. The study lost potential key information about the real mechanisms underlying deaths as verbal autopsies and PCR testing do not allow an extensive evaluation of the disease. This may even be overestimated given that the precise cause of death remains elusive [1, 8, 9].
The study has failed based on its methods to measure the fatal impact of covid-19 in Africa. We are confident that our concerns and observations will be taken into account by your highly revered editorial team.
Competing interests: No competing Interests
Reference
[1] Luchenga Mucheleng’anga, Cordilia Himwaze, The role of forensic pathology in the COVID-19 pandemic in Zambia, Forensic Science International: Reports, Volume 2,
2020, 100147, ISSN 2665-9107, https://doi.org/10.1016/j.fsir.2020.100147.
[2] Open edition journals, African Cities and the Development Conundrum
Actors and Agency in the Urban Grey Zone, https://journals.openedition.org/poldev/262. Retrieved on 25/02/2021 at 22:19 hours.
[3] Koehler, Steven A. Forensic Epidemiology (International Forensic Science and Investigation) (p. 249). Taylor and Francis CRC ebook account. Kindle Edition.
[4] Ministry of Health Zambia is at Ministry Of Health - Ndeke House,Lusaka, https://web.facebook.com/mohzambia. Retrieved on 25/02/2021 at 20:01 hours.
[5] The births and deaths registration act, chapter 51 of the laws of Zambia, http://www.parliament.gov.zm/sites/default/files/documents/acts/Births%2.... Retrieved on 27/02/2021 at 11:04 hours.
[6] Zambia-National-Strategic-Action-Plan-for-CRVS, http://www.crvs-dgb.org/wp-content/uploads/2015/11/Zambia-National-Strat.... Retrieved on 27/02/2021 at 11:10 hours.
[7] The inquests act chapter 36 of the laws of Zambia, http://www.parliament.gov.zm/sites/default/files/documents/acts/Inquests.... Retrieved on 27/02/2021 at 11:13 hours.
[8] Murray, C.J., Lopez, A.D., Black, R. et al. Population Health Metrics Research Consortium gold standard verbal autopsy validation study: design, implementation, and development of analysis datasets. Popul Health Metrics 9, 27 (2011) doi:10.1186/1478-7954-9-27
[9] Salerno M, Sessa F, Piscopo A et al (2020) No autopsies on COVID-19 deaths: a missed opportunity and the lockdown of science. J Clin Med 9. https://doi.org/10.3390/jcm9051472
APPENDICES (https://drive.google.com/drive/folders/1ENXRLJ8qR6t6HfRpm8RXBnU_CtCn1vCu...)
Competing interests: No competing interests