Content uploaded by Amobi Onovo Ph.d
Author content
All content in this area was uploaded by Amobi Onovo Ph.d on Aug 12, 2020
Content may be subject to copyright.
Estimates of the COVID-19 Infection Fatality Rate for 48
African Countries: a model-based analysis
Amobi Andrew Onovo,1,3 Akinyemi Atobatele,1 Abiye Kalaiwo,1 Christopher Obanubi,1 Gertrude
Odezugo,2 Ezekiel James,1 Dolapo Ogundehin,1 Pamela Gado,1 Doreen Magaji,1 Babatunji Odelola,1
Temitayo Odusote,1 Michele Russell,1 Janne Estill,3,4* Olivia Keiser,5*
1Office of HIV/AIDS and TB, U.S Agency for International Development, Nigeria
2Office of Health, Population and Nutrition, U.S Agency for International Development, Nigeria
3Institute of Global Health, University of Geneva, Switzerland
4Institute of Mathematical Statistics and Actuarial Science, University of Bern, Switzerland
5Professor, Head of Unit Infectious Disease and Modelling, Institute of Global Health, University of
Geneva
* equally contributed last authors
Word count: Abstract: 342; Main text: 4,073; Number of tables: 2; Number of figures: 3;
Appendix: 4
Corresponding author:
Amobi Andrew Onovo, Ph.D. (Graduand), MPH, Epidemiology
Global Health Institute, University of Geneva, Switzerland.
Project Management Specialist, Performance Management, USAID Nigeria
+2347030538954
amobiandrewonovo@gmail.com
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=3657607
Abstract
Introduction
The infection fatality rate (IFR) is key to determining the effect of the pandemic at population level, as well
as the effects of public policies and regulations. We examine global data from 48 African countries to
estimate the SARS-CoV-2 IFR.
Methods
We analyzed time series data on the confirmed cases and deaths from COVID-19 disease outbreak across
Africa. We define IFR as the ratio of the number of deaths caused by COVID-19 (numerator) and the total
number of people in the population who were infected by the virus (denominator). We controlled for the
upward bias associated with the denominator, to accommodate for the untested individuals by adjusting for
population density, population aged 65 years and older, population with basic handwashing facilities,
extreme poverty, diabetes prevalence, and death rate from cardiovascular disease in a Bayesian prediction
model based on the technique of Monte Carlo.
Results
We analyzed data on the 135,126 confirmed cases and 3,922 deaths from COVID-19 disease outbreak in
Africa through May 30, 2020. After adjusting for potential risk factors in the Bayesian model, we predicted
a total of 1,686,879 COVID-19 infections, corresponding to 13 infections per confirmed case. The IFR in
Africa was estimated to be 0.23% (95%CI: 0.14% to 0.33). Country-specific rates varied from 0.004% in
Botswana and Central African Republic, to 1.53% in Nigeria, respectively. The estimated IFR is twelvefold
higher than WHO reported estimate (0.02%) from the 2009 H1N1 influenza pandemic. The inverse distance
weighted (IDW) interpolation map shows concentrations of extreme IFR in four countries: Morocco,
Nigeria, Cameroon, and South Africa.
Conclusion
The infection fatality rate of COVID-19 can vary substantially across different locations, and this may
reflect differences in demographics, underlying health issues in the population, capacity of the healthcare
system, positive health seeking behavior, as well as other factors. Variability in testing and cause-of-death
data across African countries might have impacted on the results. Our model and our estimates can help
disease and policy modelers to obtain more accurate predictions for the epidemiology of the disease and
the effect of alternative policy strategies to contain this pandemic.
Keywords: COVID-19, Infection Fatality Rate, Bayesian prediction, Monte Carlo method, Influenza,
Africa.
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=3657607
Introduction
As of July 18, 2020, 13,876,441 confirmed cases and 593,087 deaths due to coronavirus disease 2019
(COVID-19), caused by the novel severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), had
been reported worldwide [1]. The epidemic began in mainland China, with a geographical focus in the city
of Wuhan, Hubei. From February 26, 2020 onwards, the daily number of cases became greater in the rest
of the world than inside China [2]. According to global data on COVID-19 disease compiled on May 30,
2020 by the European Center for Disease Prevention and Control (ECDC), the spread had reached 54
countries in Africa with a total of 135,126 laboratory-confirmed COVID-19 cases and 3,922 COVID-19
related deaths. At the time of this analysis, six out of the 54 countries in Africa [i.e. Eritrea, Lesotho,
Namibia, Rwanda, Seychelles and Uganda] had not yet reported any COVID-19 deaths.
To understand the severity of infection during an outbreak, i.e., the virulence of the causative agent, the
common epidemiological practice is to estimate the case fatality ratio (CFR) as the risk of death among
cases. However, crude CFR obtained simply by dividing the number of deaths by the number of reported
laboratory-confirmed cases, such as those compiled daily by the World Health Organization during the
SARS epidemic [3] and those presented on the COVID-19 map dashboard by John Hopkins University,
can be misleading [4, 5]. During an outbreak of a pandemic or emerging infectious disease such as SARS-
CoV-2, the infection fatality rate (IFR) is a more reliable metric to estimate the fatality rate in all the affected
countries. The IFR is key to determining the effect of the pandemic at population level, as well as the effects
of public policies and regulations, such as social distancing measures and the effects of potential future
shortfalls in health care services.
Knowledge of the IFR of SARS-CoV-2 is necessary to tackle the COVID-19 pandemic [6, 7]. The IFR is
the ratio of two numbers—the number of deaths caused by COVID-19 (numerator) and the total number of
people in the population who were genuinely infected by the virus (denominator). However, for many
reasons, both the numerator and the denominator of the IFR are measured with error. For example, errors
in the denominator arise because patients remain asymptomatic during the first few days of the infection,
testing is not universal and selective at best, and longitudinal data on COVID-19 patients are unavailable
at the national level [8]. The IFR can be biased upwards because we do not know the actual number of
individuals who are infected. It can be biased downward because some of those who are currently infected
could die in the future, or deaths are underreported (errors in numerator). The upward bias is likely to be
much larger during the early phase of testing. The vast difference in the reported CFR by country is
primarily a product of testing availability [9,10]. In the early phases of the outbreak and in countries where
testing was limited, only hospitalized patients with advanced COVID-19 symptoms were tested. Therefore,
the CFR is an inflated estimate of the IFR because many infections in the population are unidentified [11,
12].
Representative seroprevalence studies provide an important opportunity to estimate the number of
infections in a community, and when combined with death counts can lead to robust estimates of the IFR.
Recent serosurvey for the canton of Geneva, Switzerland using a Bayesian framework estimated a
population-wide IFR of 0.64% (0.38–0.98) [13]. For Spain as a whole, the infection fatality rate was 1.15%
and ranged from 0.13% to 3.25% across 19 Spanish regions [14]. Another study examined seroprevalence
of antibodies to SARS-CoV-2 in a community sample drawn from Santa Clara County, California. The
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=3657607
study reported an IFR of 0.17% [15]. Such findings show that the IFR for SARS-CoV-2 varies across
countries and regions. At the time of this analysis, there were no reports of seroprevalence studies conducted
in Africa. In this paper, to effectively estimate the IFR of COVID-19 for 48 African countries, we provide
a new statistical approach for eliminating measurement errors in the denominator. We attempted to account
for people with undetected COVID-19 disease in the denominator, i.e. untested individuals by adjusting for
the underlying socio-demographic, economic, and potential biological risk-factors in a Bayesian statistical
model using laboratory confirmed reported cases of COVID-19 as the response variable in our model.
Because of its potential to use prior information or experimental evidence (e.g. risk factors correlated with
COVID-19) in a data model, we used Bayesian statistical modeling to produce more realistic outcomes (i.e.
estimated number of people infected with COVID-19). Consequently, we calculated the IFR by dividing
the total number of reported deaths by the adjusted or predicted denominator.
Methods
Setting and Data sources
Africa accounts for about 16 percent of the world’s human population with 1.3 billion people as of 2018
[16]. We gathered and analyzed data on the 135,126 confirmed cases and 3,922 deaths from COVID-19
disease outbreak across Africa between February 15, 2020 through May 30, 2020. We used publicly
documented COVID-19 datasets created by Our World in Data and utilized time-series aggregate data
compiled by ECDC on the total number of confirmed cases, total number of deaths and other variables of
potential interest. The following other variables were included: population density (number of people
divided by land area, measured in square kilometers), proportion of people aged 65 and above, access to
handwashing facilities (proportion of the population with basic handwashing facilities on premises), socio-
economic situation (proportion of people living in extreme poverty), diabetes prevalence (among the
population aged 20 to 79 years), and death rate from cardiovascular disease. We excluded six of the 54
African countries from this study, which at the time of this analysis had not reported any COVID-19 deaths
yet. See the link to the datasets from Our World in Data (https://covid.ourworldindata.org/data/owid-covid-
data.csv).
Assumptions
We made three assumptions for our analysis: (1.) Errors in the numerator and the denominator lead to
underreporting of true SARS-CoV-2 deaths and infections, respectively; however, error is smaller for
deaths than for infections. (2.) The predicted total number of SARS-CoV-2 infections represents the IFR
calculation denominator. In our study, the predicted total number of COVID-19 infections were
summarized using the Bayesian posterior summary statistics. To obtain deeper insights into the uncertainty
around our estimates, we examined the total predicted COVID-19 infections over a range of mean and
maximum posterior summary statistics [75%, 90% and 95%] through a sensitivity analysis. The calculated
IFR using each summary statistic was compared with the IFR from recent seroprevalence surveys (Annex
1: Sensitivity analysis). (3.) We assume that the severity of COVID-19 depends on the covariates in our
model (figure 1). Assumption #1 is self-evident; both the deaths and the actual infections are undercounted
during the initial phase of the epidemic [17,18]. Because deaths are much more visible events than
infections, we posit that, at any point in time, the errors in the denominator are larger than the errors in the
numerator. Assumption #2 is our central assumption that adjusts for biases in the denominator. Here we
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=3657607
posit that the denominator, i.e. the predicted total number of individuals infected with COVID-19,
represents the estimated upper limit of the 95% credible interval that we use as proxy for the actual total
number of infections. Assumption #3 is our constructed conceptual framework that acts as a bridge between
model adjustments for the denominator and empirical observations.
Statistical Model
We used the Bayesian parametric model to predict the total number of individuals infected with COVID-
19 as of May 30, 2020. Firstly, we fitted a Bayesian normal regression model using Gibbs sampling based
on the technique of the Markov Chain Monte-Carlo (MCMC) to specify the posterior model.
Posterior ∝ Likelihood × Prior
The posterior model combines a “likelihood function”, which includes information about model parameters
based on the observed data, and a “prior”, which includes prior information (before observing the data)
about model parameters. The model parameters included the response variable “confirmed reported cases
of COVID-19 across Africa” and the independent covariates of interest; population density, age 65 or
above, cardiovascular death rate (CVD), diabetes prevalence, handwashing facilities and extreme poverty.
Secondly, we computed Bayesian predictions for the outcome variable. Based on results from the fitted
posterior model, we predicted the total number of individuals infected with COVID-19 using the
“bayespredict” model in StataMP v.16. Here we simulated 1,000 MCMC samples of outcome values for
each of the 48 countries and calculated the posterior means and predicted p-values for each observation.
We used a random-number seed to ensure reproducibility. Finally, we conducted posterior predicted checks
by comparing the observed data with the MCMC replicates (simulated data from the posterior predictive
distribution). Unlike classical prediction, which produces a single value for each observation, Bayesian
prediction produces an MCMC sample of values for each observation.
Spatial Analysis
We georeferenced the estimated IFR across 48 countries and performed two analyses. We used inverse
distance weighted interpolation (IDW) technique in the Geostatistical Analyst tool of ArcGIS 10.8 software
to create a raster showing the spatial distribution of COVID-19 IFR. Second, we constructed a thematic
map contrasting the predicted COVID-19 IFR and Influenza IFR 2018-2019 to obtain clear insight into the
magnitude of COVID-19 outbreak.
This dataset and code for analyses are available on the following website: https:/github.com/onovo007.
Results
Descriptive analysis
Our analysis was based on 48 of the 54 African countries that reported confirmed cases and deaths from
coronavirus between February 15, 2020 through May 30, 2020. At the time of this analysis, nine countries
(Egypt, Algeria, South Africa, Nigeria, Sudan, Morocco, Cameroon, Mali, and Somalia) accounted for 80%
of Africa’s COVID-19 deaths (figure 2.a). The scatter plot shows an uphill pattern from left to right; this
indicates a strong positive correlation between total deaths due to COVID-19 and total confirmed cases of
COVID-19, (rs (47) = 0.92, p = 0.001) (figure 2.b). The scatter plot illustrates variability in pattern of
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=3657607
reported deaths and confirmed cases throughout the nine countries. The result is suggestive that the risk of
death among cases varies by location and is typically changing over time. Among these countries the
calculated crude CFR was the highest in Algeria (7.0%) followed by Mali (6.0%) and the lowest in South
Africa (2.1%). The lower crude CFR for COVID-19 in South Africa compared with the other eight countries
may be caused by differences in demographics, socio-economic and biological characteristics. This could
also suggest variability in the testing capacity of COVID-19 across these countries.
Bayesian regression model
The summary of the fitted Bayesian multiple linear regression model provided a summary (Table 1). The
response variable was log-transformed to control for skewness and ensure effective linear relationship with
the explanatory variables. The model summary shows the default priors used for the model parameters:
normal (mean 0, standard deviation 10000) for the regression coefficients and inverse gamma (shape 0.01,
scale 0.01) for the variance parameter. The first two columns of the Bayesian normal regression report the
posterior means and standard deviations of the model parameters. The posterior means and standard
deviations of the regression coefficients were very similar to the least-square estimates. The posterior mean
estimate for the variance, 2.20, is close to the residual mean squared estimate, 2.15. The minimum efficiency
in the model is 0.76, and the mean efficiency 0.97. Our acceptance rate (AR) is good and efficiencies are
high. We do not have a reason to suspect nonconvergence. Nevertheless, we explored convergence by
computing graphical diagnostic plots for all models to confirm this. Overall, graphical diagnostic plots
show that MCMC converged and mixes well for all parameters in the model (see graphical diagnostic plots
in Annex 2).
Bayesian prediction model
We calculated the posterior summary statistics for all simulated outcome observations (Annex 3) and
computed the posterior predictive summaries to test our prediction. The simulated outcomes values were
saved in a prediction dataset (see Appendix 5). We accessed the prediction results to compare the agreement
for the mean, minimum, and maximum test statistics between the predicted data and observed data (Annex
4). The posterior predictive p-value was 0.46 for the mean statistic, 0.38 for the minimum, and 0.69 for the
maximum. When this probability is close to 0.5, the replicated and observed data agree with respect to the
test statistic. However, Gelman et al., [19] indicated that values between 0.05 and 0.95 are often considered
acceptable. According to our results, the mean, minimum and maximum statistics appear to agree in the
observed and replicated data.
Estimated Infection Fatality Rate by Country
We predicted a total of 1,686,879 COVID-19 infections, which is the denominator used in our study to
calculate the IFR. In order to measure the IFR, we divided the total number of deaths reported (3,922) by
the estimated denominator. Overall, the 48 African countries are estimated to have an IFR of 0.23% (Std.
Dev: 0.04%) with a 95 percent confidence interval of 0.14 % to 0.33% as of May 30, 2020. The confidence
intervals around the IFR was calculated using the normal-based confidence intervals in StataMP v.16. The
rates varied from 0.004% in Botswana and Central African Republic to 1.53% in Nigeria, respectively.
Fourteen African countries: Nigeria, South Africa, Cameroon, Morocco, Niger, Burkina Faso, Sierra Leone,
Democratic Republic of Congo, Egypt, Kenya, Somalia, Chad, Sudan and Senegal represent 80% of the
deaths among the infected population with Nigeria presenting the highest IFR (Table 2). Interpolated map
displaying high, moderate and low rates of COVID-19 IFR spread across Africa, using IDW technique, is
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=3657607
shown in Figure 3 (Left panel). The map shows concentrations of high IFR in four countries: Morocco,
Nigeria, Cameroon, and South Africa. IFR is moderately high in specific regions in Northern and Western
Africa, whereas predominantly low in the southern, central, and eastern regions, except for countries such
as Democratic Republic of Congo, Kenya, and Somalia (Horn of Africa) with comparable rates to regions
with high IFR. Figure 3 (Right panel) compares estimated COVID-19 IFR and Influenza IFR 2018-2019
(0.1%) [20] by country. Countries shaded black have COVID-19 IFR above 0.1% (which is the 2018-2019
Influenza IFR) and countries shaded blue have COVID-19 IFR below 0.1%.
Discussion
From an extensive analysis of data from different regions of Africa, our best estimate at the current time
for the IFR of COVID-19 in 48 African countries is 0.23% (95% CI: 0.14% – 0.33%). Although this value
was lower than China's overall IFR calculation of 0.66 percent (95% CI: 0.39% - 1.33%) as reported by
PCR testing of foreign residents of Wuhan returning on repatriation flights [21], it is twelvefold higher than
WHO reported estimate (0.02%) from the 2009 H1N1 influenza pandemic [22, 23]. Our findings seem to
align with estimates observed in a recent seroprevalence study showing IFR estimates ranging from 0.02%
to 0.40% [24]. In Iceland, the country with highest number of tests per capita, the IFR lies somewhere
between 0.03% and 0.28% [25]. IFR varied disproportionality across African countries. Fifteen countries
had IFRs higher than the overall estimate of 0.23%, ranging from 0.31% to 1.53%. For these 15 countries,
the average IFR was comparable with the overall IFR from China (0.61% vs. 0.66%). Two countries,
Liberia and Algeria with IFR 0.24% mirror the continent-wide average, and 31 countries have IFRs ranging
from 0.004% to 0.17% and lower than the country's wide average. By region, twenty-three (48%) of the
countries presented IFRs above 0.1% for influenza IFR in 2018-2019. Comparison of the two maps in
Figure 4 (interpolated map on the left and the thematic map on the right) display a near glove fit or mirror
image. The different IFRs across Africa probably indicate that countries are at different stages of the
pandemic, but various other factors may also be important. These include for example underlying health
issues in the population, and differences in demographics (e.g. detailed population age structure), in health
care systems, in testing practices (including testing practices among diseased persons) and differences in
the capacity in responding to the pandemic.
The estimated IFR for each summary statistics was compared with the IFR and the total number of
infections per confirmed case from recent seroprevalence surveys conducted in Geneva, Spain and Santa
Clara County, California. The calculated IFR of the mean statistics was roughly 8.2%, 6.9% and 6.2% at
the 75%, 90% and 95% range respectively. The calculated IFR was 0.31%, 0.26% and 0.23% in the different
range for the maximum statistics. The mean IFR was an 11-fold-increase compared to the reported
population-wide IFR of 0.64% in Geneva on average, and a 6-fold increase compared to the estimated IFR
of 1.15% in Spain. This was 11 times higher than the IFR of 0.17% reported in Santa Clara County,
California. On average, the calculated IFR for the maximum statistics was moderately lower than the IFR
of 0.64% reported in Geneva. While the maximum IFR was five times smaller than the estimated IFR of
1.15% in Spain, the maximum IFR aligned markedly with the 0.17% IFR reported in Santa Clara County,
California. The estimated total number of infections of the maximum statistics were 1,265,159, 1,518,191
and 1,686,879 over the three-separate range, approximately corresponding to 9, 11 and 13 total number of
infections per confirmed case. In terms of the mean statistics, estimated total number of infections were
47,366, 56,839 and 63,154, corresponding to less than one infection per confirmed case. Across the three-
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=3657607
separate range of the mean statistics, the total number of infections per confirmed case was substantially
lower. That is a much smaller share of unreported infections compared to the maximum statistics. The
reported overall number of infections in a seroprevalence study performed in Stockholm, Sweden was
74,089, corresponding to 44 infections per confirmed case [26]. We estimated a total of 11 infections per
confirmed case in the Geneva serosurvey and 54 infections per confirmed case in the Clara County,
California survey. Such findings indicate geographical variation in the distribution of SARS-CoV-2
infections across the globe. This is also an indicator that certain countries are experiencing far more rapid
and broader transmission of SARS-CoV-2 infections than others. Since COVID‐19 is extremely contagious
and a single case will infect dozens of people, we present in this analysis the overall total COVID-19
infections predicted as 1,686,879, corresponding to a total of 13 infections per confirmed case. Our result
is slightly higher than the total number of infections per confirmed case presented in the Geneva study, and
it resonates with the larger number of infections per confirmed case as seen in the other serosurveys.
Our approach was to control for the upward bias. We did not control for the downward bias that may arise
because some of the detected cases may die in the future. Since the first reported case in Egypt on February
14, 2020 [27], several African countries have intensified efforts to reduce infections and prevent
coronavirus deaths by restriction in movements through curfews and lockdowns, deploying trained public
health forces, expanding public health surveillance activities to identify all suspected cases, setting up
facilities to isolate and treat patients, and ramping up testing capacity. In this study, we attempted to account
for a proportion of the untested individuals in the denominator by adjusting for population density,
proportion of the population that is 65 years and older, proportion of population with basic handwashing
facilities, extreme poverty, prevalence of diabetes, and death rate from cardiovascular disease in a Bayesian
prediction model. These variables were included in our analyses because a multitude of demographic, social
and economic characteristics have been attributed as potential determinants for the observed variety in the
outcomes of coronavirus. Several experts say the countries that were hardest affected have had an ageing
population [28, 29], or an underdeveloped healthcare system [30, 31]. In an epidemiological study, David
et al found that among 282 patients needing mechanical ventilation, 97.2 percent of the patients aged 65 or
above died [32]. According to the WHO, inadequate housing and overcrowding are major factors in disease
transmission, and disease outbreaks become more frequent and extreme when there is a high population
density [33]. Studies suggest that elderly adults with clinical comorbid illness, such as diabetes and
cardiovascular disease, are at higher risk of hospitalization and COVID-19 death [34, 35]. To avoid
infectious diseases like COVID-19 it is essential to wash hands regularly and correctly with soap and water.
The latest global estimates show that 3 billion people lacked soap and water at home, 900 million children
lacked soap and water at school, and 40% of health care facilities were not equipped to practice hand
hygiene at care points [36]. On average, according to our analysis, the proportion of the population in Africa
with access to water and soap hand-washing facilities is 35% (95% CI: 26-44).
Of the 46 member states of the World Health Organization (WHO) African regions, only one can provide
high-quality cause-of-death data (Mauritius), with another three able to provide low or medium-quality data
(Seychelles, South Africa, and Zimbabwe). In addition, Egypt and Morocco can provide low to medium-
quality cause-of-death data [37]. In our study, we note that there might have been underreporting of
coronavirus related deaths across certain African countries. However, given the early surveillance systems
and measures instituted by countries to stymie the outbreak, we believe registration of deaths from COVID-
19 would have improved over time. Governments across Africa are taking a wide range of testing measures
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=3657607
in response to the COVID-19 outbreak, according to University of Oxford COVID-19 Government
response tracker [38]. The data show that testing policy is not standardized and varies substantially across
African countries: 27 (50%) countries are testing only those who have symptoms and meet specific criteria
(i.e. key workers, persons admitted to hospital, encountered a known case or returned from overseas); 14
(26%) countries are testing anyone showing COVID-19 symptoms; nine (17%) countries are implementing
Open public testing (e.g. “drive through” testing available to asymptomatic people); and four (7%) countries
have no testing policy [39]. There is a high possibility that infections are far higher than reported [40].
Publicly available testing data suggests enormous differences in testing capacity and case identification
across Africa. As of June 29, 2020, the positivity rate for COVID-19 testing in Africa varied from 0.4% in
Uganda to 25.7% in Nigeria [41]. Some countries, like Australia, South Korea and Uruguay have a positive
rate of less than 1%. This implies that it takes hundreds, or even thousands of tests to find one case in these
countries. Countries such as Mexico and Nigeria [42], on the other hand, have positivity levels of 20%-
50%, or even more. According to WHO report, countries with high positivity rate are unlikely to be testing
widely enough to find all cases. WHO recommends a positive rate for around 3% to 12% as a general
benchmark of adequate testing [43].
There are however a number of limitations to this analysis. Serosurveys indicate strong association between
age and IFR. At the time of our analysis, age disaggregated data across all age groups was not available.
Other important risk-factors, particularly cancer, chronic kidney disease, obesity, and sickle cell disease
were not available for analysis. The variability in testing and cause-of-death data across African countries
might have impacted on the results.
Conclusions
Assessing the infection fatality rate of COVID-19 is crucial to determine the appropriateness of mitigation
strategies and to enable planning for healthcare needs as epidemics unfold. Without population-based
serologic studies in Africa, it is not yet possible to know what proportion of the population has been infected
with COVID-19. Our study shows that Bayesian modeling is a helpful tool that can account for missed
cases, such as those untested due to country’s low testing capacity, and the mild cases which are potentially
missed in current surveillance activities. Using an estimated number of total infections, the IFR can be
calculated. In many countries in Africa, owing to weaker health-care systems, informal settlements,
overcrowded cities and public transportation and a lack of clean water and sanitation, the current approaches
to self-protection, social distancing and containment as measures to control the outbreak may not be viable.
Scaling up surveillance efforts and growing COVID-19 research and testing capability across Africa may
help to provide a deeper understanding of how the pandemic is advancing, and to define hotspots for
targeted and pooled testing, case isolation, and early treatment. Our estimates of the underlying infection
fatality rate of this virus will inform assessments of health effects likely to be experienced in different
countries, and thus decisions around appropriate mitigation policies and strategies that should be adopted.
Contributors
AAO conceived of the study including design and method. He was the principal in data management,
analyzed the data and drafted the article. AA, AK, CO, GE EJ, DO, PG, DM, TO, BO, MR contributed to
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=3657607
reviews of the first draft. JE and OK are joint last authors and performed critical reviews of the second and
final version of the manuscript.
Declaration of interests and funding
We declare no competing interests. O Keiser was funded by grants from the Swiss National Science
Foundation (no 163878 and 196270).
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=3657607
Figure 1: Categorization of SARS-CoV-2 cases according to severity and variables included in the
Bayesian model
At the top of the pyramid, those meeting the WHO case criteria for severe or critical cases are likely to be
identified in the hospital setting. Many more cases are likely to be symptomatic (i.e., with fever, cough, or
myalgia), but might not require hospitalization. These cases will have been identified through links to
international travel to high-risk areas and through contact-tracing of contacts of confirmed cases. They
might also be identified through population surveillance. The bottom part of the pyramid represents mild
(and possibly asymptomatic) cases. These cases might be identified through contact tracing and
subsequently via serological testing. In a Bayesian prediction model, potential demographic, socio-
economic, and biological variables were adjusted to account for the fraction of untested people.
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=3657607
Figure 2: (2a.) Reported COVID-19 deaths in each African country (2b.) Total confirmed COVID-19
deaths vs. cases for the nine countries that represent 80% of reported deaths, May 30, 2020
a.
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=3657607
Table 1: Bayesian normal regression using Gibbs sampling
MCMC iterations = 3,500
Burn-in = 2,500
MCMC sample size = 1,000
Number of obs = 48
Acceptance rate = 1
Efficiency: min = 0.7658
avg = 0.9707
max = 1
Mean
Std.Dev
MCSE
Median
Equal-tailed
95% Cred. Interval
Confirmed_Cases
Population density
-0.0052613
0.0021246
0.000067
-0.0051941
-0.009677
-0.0011185
Aged 65 older
0.1831922
0.1896953
0.005703
0.1855481
-0.174675
0.550733
CVD death rate
-0.0011224
0.0036282
0.000113
-0.0011888
-0.008117
0.0061924
Diabetes prevalence
0.062837
0.0799469
0.002528
0.0618234
-0.093096
0.2195652
Handwashing facilities
0.0200225
0.0102218
0.000319
0.0200439
0.0006078
0.0405081
Extreme poverty
-0.0028848
0.010585
0.000335
-0.0029234
-0.024479
0.017851
_cons
6.22023
1.139261
0.036027
6.19992
3.77586
8.390462
var
2.209552
0.5010807
0.018107
2.147209
1.413217
3.442549
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=3657607
Table 2: Total Estimated Infection Fatality Rate for 48 African Countries as of May 30, 2020.
Country
Total Cases
reported
Total Deaths
reported
Total Predicted
infections
Estimated IFR
Crude CFR
Algeria
9,134
638
272,017
0.24%
7.0%
Angola
77
4
19,187
0.02%
5.2%
Benin
224
3
8,884
0.03%
1.3%
Botswana
35
1
22,819
0.004%
2.9%
Burkina Faso
847
53
10,257
0.52%
6.3%
Burundi
42
1
1,959
0.05%
2.4%
Cameroon
5,436
177
17,603
1.01%
3.3%
Cape Verde
405
4
12,464
0.03%
1.0%
Central African Republic
874
1
23,043
0.004%
0.1%
Chad
759
65
17,517
0.37%
8.6%
Comoros
87
2
4,894
0.04%
2.3%
Congo
587
19
40,170
0.05%
3.2%
Cote d'Ivoire
2,750
32
10,386
0.31%
1.2%
Democratic Republic of Congo
2,833
69
14,499
0.48%
2.4%
Djibouti
2,914
20
23,332
0.09%
0.7%
Egypt
22,082
879
205,083
0.43%
4.0%
Equatorial Guinea
1043
12
39,917
0.03%
1.2%
Ethiopia
968
8
19,147
0.04%
0.8%
Gabon
2,613
15
27,858
0.05%
0.6%
Gambia
25
1
5,357
0.02%
4.0%
Ghana
7,616
34
21,248
0.16%
0.4%
Guinea
3,656
22
16,782
0.13%
0.6%
Guinea-Bissau
1,256
8
8,235
0.10%
0.6%
Kenya
1,745
62
15,440
0.40%
3.6%
Liberia
273
27
11,455
0.24%
9.9%
Libya
118
5
25,774
0.02%
4.2%
Madagascar
698
5
30,097
0.02%
0.7%
Malawi
273
4
5,695
0.07%
1.5%
Mali
1,226
73
42,636
0.17%
6.0%
Mauritania
423
20
18,496
0.11%
4.7%
Mauritius
335
10
19,972
0.05%
3.0%
Morocco
7,714
202
25,380
0.80%
2.6%
Mozambique
234
2
11,812
0.02%
0.9%
Niger
955
64
12,248
0.52%
6.7%
Nigeria
9,302
261
17,052
1.53%
2.8%
Sao Tome and Principe
463
12
10,292
0.12%
2.6%
Senegal
3,429
41
13,239
0.31%
1.2%
Sierra Leone
829
45
8,729
0.52%
5.4%
Somalia
1,828
72
19,366
0.37%
3.9%
South Africa
29,240
611
47,859
1.28%
2.1%
South Sudan
994
10
26,287
0.04%
1.0%
Sudan
4,521
233
71,606
0.33%
5.2%
Swaziland
279
2
22,122
0.01%
0.7%
Tanzania
509
21
41,536
0.05%
4.1%
Togo
428
13
9,720
0.13%
3.0%
Tunisia
1,071
48
302,601
0.02%
4.5%
Zambia
1,057
7
14,677
0.05%
0.7%
Zimbabwe
160
4
20,130
0.02%
2.5%
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=3657607
Figure 3: (Left panel) - Interpolated map displaying high, moderate and low rates of COVID-19 Infection
Fatality Rate spread across Africa, using inverse distance weighted interpolation technique. (Right panel)
– COVID-19 IFR by country vs. Influenza IFR 2018-2019 (0.1%); Black color represents countries with
COVID-19 IFR above 0.1% (which is the 2018-2019 Influenza IFR) and blue color represents countries
with COVID-19 IFR below 0.1%.
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=3657607
Annex 1: Sensitivity Analysis of the Infection Fatality Rates of the Posterior Summary Statistics, May 30,
2020
Posterior Summary
Statistics
Mean
Maximum
75%
Cred. Interval
90%
Cred. Interval
95%
Cred. Interval
75%
Cred. Interval
90%
Cred. Interval
95%
Cred. Interval
Total COVID-19
infections predicted
(as of May 30, 2020)
47,366
56,839
63,154
1,265,159
1,518,191
1,686,879
Calculated IFR
(as of May 30, 2020)
8.28%
6.90%
6.21%
0.31%
0.26%
0.23%
Total number of infections
per confirmed case
(as of May 30, 2020)
0.35
0.42
0.47
9.36
11.24
12.48
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=3657607
Annex 2: Posterior Predictive Checks for Convergence across all model parameters
The mean and variance parameters for the trace plots mix very well. Autocorrelation is essentially
negligible for all positive lags. The kernel density estimates based on the first and second halves of the
sample are very similar to each other and are close to the overall density estimate. The histogram and kernel
density plots resemble the shape of an expected inverse-gamma distribution.
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=3657607
Annex 3: Posterior Summary Statistics for the predicted total number of people infected with COVID-19
disease, May 30, 2020.
Country
Mean
Std. Dev.
MCSE
Median
Equal-tailed
[95% Cred. Interval]
Algeria
9.035484
1.731037
0.050338
8.976333
5.695366
12.51362
Angola
6.800663
1.528316
0.046247
6.871423
3.739788
9.861989
Benin
6.127642
1.515889
0.054591
6.212643
3.262433
9.091983
Botswana
6.908195
1.617089
0.050982
6.94061
3.765597
10.03533
Burkina Faso
6.249863
1.573747
0.048913
6.286927
3.102572
9.235723
Burundi
4.380423
1.696422
0.056442
4.36187
1.069757
7.580187
Cameroon
6.772041
1.557751
0.04753
6.839904
3.706919
9.77584
Cape Verde
6.301612
1.601749
0.047189
6.322357
3.223055
9.43059
Central African Republic
6.901788
1.617589
0.051153
6.895058
3.785971
10.04513
Chad
6.687963
1.521467
0.051301
6.693711
3.657518
9.770955
Comoros
5.165101
1.668911
0.05344
5.223573
1.979436
8.495773
Congo
7.664578
1.524615
0.048366
7.618533
4.751974
10.60087
Cote d’Ivoire
6.482437
1.509022
0.050418
6.511117
3.584275
9.248247
Democratic Republic of Congo
6.499718
1.577042
0.049542
6.496328
3.364091
9.581859
Djibouti
6.767163
1.572878
0.05048
6.798374
3.814581
10.05758
Egypt
8.850436
1.737491
0.054985
8.710103
5.592242
12.23117
Equatorial Guinea
7.262961
1.609625
0.050183
7.264424
4.167636
10.59455
Ethiopia
6.686207
1.566906
0.04707
6.773801
3.572496
9.859878
Gabon
7.102963
1.583957
0.051103
7.119374
4.006258
10.23486
Gambia
5.418166
1.652224
0.053492
5.454941
2.216676
8.586081
Ghana
6.925883
1.510375
0.047762
6.951968
3.920189
9.964019
Guinea
6.550571
1.567091
0.045
6.579614
3.706756
9.728083
Guinea-Bissau
5.957774
1.56904
0.049617
5.917164
2.863463
9.016197
Kenya
6.524509
1.541617
0.04875
6.479735
3.500848
9.644734
Liberia
6.29074
1.501581
0.04511
6.321856
3.403189
9.346176
Libya
7.177052
1.537261
0.0444
7.168366
4.070487
10.15712
Madagascar
7.060992
1.649434
0.049418
7.00947
3.908797
10.31217
Malawi
5.624644
1.591987
0.044747
5.647434
2.271744
8.647288
Mali
7.498353
1.576386
0.041329
7.526019
4.210472
10.66046
Mauritania
6.956213
1.478732
0.046762
6.9735
4.086428
9.825301
Mauritius
6.062692
1.969679
0.059298
6.075007
2.36554
9.902099
Morocco
6.959709
1.752861
0.053196
6.933538
3.482169
10.14171
Mozambique
6.459878
1.53219
0.048452
6.500065
3.466867
9.376899
Niger
6.469959
1.517907
0.048
6.490142
3.457315
9.413131
Nigeria
6.383653
1.649187
0.047999
6.360061
2.993698
9.744013
Sao Tome and Principe
6.227326
1.533601
0.048497
6.160694
3.209579
9.239112
Senegal
6.517526
1.512331
0.046891
6.479507
3.551688
9.490925
Sierra Leone
6.134481
1.543039
0.050963
6.135388
3.097228
9.074404
Somalia
6.774352
1.533205
0.045498
6.791852
3.607021
9.871252
South Africa
7.863409
1.569188
0.049855
7.898317
4.739687
10.77602
South Sudan
7.100958
1.581893
0.048637
7.055144
4.049621
10.17683
Sudan
7.760693
1.705331
0.052591
7.787608
4.410964
11.17894
Swaziland
6.750008
1.569939
0.048598
6.720614
3.83395
10.00434
Tanzania
7.359101
1.584761
0.046307
7.323524
4.293065
10.63432
Togo
6.152347
1.515413
0.047922
6.195639
3.161758
9.181919
Tunisia
8.96566
1.753995
0.055829
8.957646
5.632428
12.62017
Zambia
6.575517
1.575097
0.051517
6.607708
3.565797
9.594013
Zimbabwe
6.94222
1.516954
0.053294
6.932369
3.994743
9.909969
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=3657607
Annex 4: Posterior Predictive Summary for Test Statistics
Note: P(T>=T_obs) close to 0 or 1 indicates lack of fit.
Posterior predictive summary MCMC sample size = 1,000
T
Mean
Std. Dev.
E(T_Obs)
P(T>=T_Obs)
mean
6.749488
0.3085233
6.775585
0.462
min
2.818528
1.005077
3.218876
0.389
max
10.89979
1.062004
10.28329
0.694
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=3657607
References
1. World Health Organization, World Health Organization, covid19.who.int/.
2. Verity, Robert, et al. “Estimates of the Severity of Coronavirus Disease 2019: a Model-Based
Analysis.” The Lancet Infectious Diseases, Elsevier, 30 Mar. 2020,
www.sciencedirect.com/science/article/pii/S1473309920302437.
3. Severe Acute Respiratory Syndrome (SARS). World Health Organization, World Health
Organization, 23 July 2015, www.who.int/csr/sars/en/.
4. Garske, Tini, et al. “Assessing the Severity of the Novel Influenza A/H1N1 Pandemic.” The BMJ,
British Medical Journal Publishing Group, 14 July 2009, www.bmj.com/content/339/bmj.b2840.
5. Lipsitch, Marc, et al. “Potential Biases in Estimating Absolute and Relative Case-Fatality Risks
during Outbreaks.” PLOS Neglected Tropical Diseases, Public Library of Science,
journals.plos.org/plosntds/article?id=10.1371%2Fjournal.pntd.0003846.
6. Impact of Non-Pharmaceutical Interventions (NPIs) to Reduce COVID19 Mortality and
Healthcare Demand, by Neil Ferguson Et Al, March 16th, 2020: Comparison of Effects of
Strategies for Mitigation and for Suppression.,
www.reddit.com/r/COVID19/comments/fjssf6/impact_of_nonpharmaceutical_interventions_npis
_to/.
7. Forecasting COVID-19 Impact on Hospital Bed-Days, ICU-Days, Ventilator Days and Deaths by
US State in the next 4 Months. Institute for Health Metrics and Evaluation, 21 Apr. 2020,
www.healthdata.org/research-article/forecasting-covid-19-impact-hospital-bed-days-icu-days-
ventilator-days-and-deaths.
8. Li, Ruiyun, et al. “Substantial Undocumented Infection Facilitates the Rapid Dissemination of
Novel Coronavirus (SARS-CoV-2).” Science (New York, N.Y.), American Association for the
Advancement of Science, 1 May 2020, www.ncbi.nlm.nih.gov/pubmed/32179701.
9. Chan, Jasper Fuk-Woo, et al. “Improved Molecular Diagnosis of COVID-19 by the Novel,
Highly Sensitive and Specific COVID-19-RdRp/Hel Real-Time Reverse Transcription-PCR
Assay Validated In Vitro and with Clinical Specimens.” Journal of Clinical Microbiology,
American Society for Microbiology, 23 Apr. 2020, www.ncbi.nlm.nih.gov/pubmed/32132196.
10. Ng, Yixiang, et al. “Evaluation of the Effectiveness of Surveillance and Containment Measures
for the First 100 Patients with COVID-19 in Singapore - January 2-February 29, 2020.” MMWR.
Morbidity and Mortality Weekly Report, U.S. National Library of Medicine, 20 Mar. 2020,
www.ncbi.nlm.nih.gov/pubmed/32191691.
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=3657607
11. Lipsitch, Marc, et al. “Potential Biases in Estimating Absolute and Relative Case-Fatality Risks
during Outbreaks.” PLOS Neglected Tropical Diseases, Public Library of Science,
journals.plos.org/plosntds/article?id=10.1371%2Fjournal.pntd.0003846.
12. Ghani, A C, et al. “Methods for Estimating the Case Fatality Ratio for a Novel, Emerging
Infectious Disease.” American Journal of Epidemiology, Oxford University Press, 1 Sept. 2005,
www.ncbi.nlm.nih.gov/pubmed/16076827.
13. Serology-Informed Estimates of SARS-CoV-2 Infection Fatality Risk in Geneva, Switzerland.
www.reddit.com/r/COVID19/comments/hrnpoi/serologyinformed_estimates_of_sarscov2_infecti
on/.
14. COVID-19 Infection Fatality Rate Associated with Incidence-A Population-Level Analysis of 19
Spanish Autonomous Communities.
www.reddit.com/r/COVID19/comments/hdhwi3/covid19_infection_fatality_rate_associated_with
/.
15. Srinivasan, Balaji S. “Peer Review of ‘COVID-19 Antibody Seroprevalence in Santa Clara
County, California.’” Medium, Medium, 19 Apr. 2020, medium.com/@balajis/peer-review-of-
covid-19-antibody-seroprevalence-in-santa-clara-county-california-1f6382258c25.
16. World Population Prospects - Population Division. United Nations, United Nations,
population.un.org/wpp/.
17. Li, Rubyuna, et al. “Substantial Undocumented Infection Facilitates the Rapid Dissemination of
Novel Coronavirus (SARS-CoV-2).” Science (New York, N.Y.), American Association for the
Advancement of Science, 1 May 2020, www.ncbi.nlm.nih.gov/pubmed/32179701.
18. Brown, Emma, et al. “US Deaths Soared in Early Weeks of Pandemic, Far Exceeding Number
Attributed to COVID-19.” Anchorage Daily News, Anchorage Daily News, 27 Apr. 2020,
www.adn.com/nation-world/2020/04/27/us-deaths-soared-in-early-weeks-of-pandemic-far-
exceeding-number-attributed-to-covid-19/.
19. Bayesian Data Analysis. Routledge & CRC Press, www.routledge.com/Bayesian-Data-
Analysis/Gelman-Carlin-Stern-Dunson-Vehtari-Rubin/p/book/9781439840955.
20. Estimated Influenza Illnesses, Medical Visits, Hospitalizations, and Deaths in the United States -
2018–2019 Influenza Season. Centers for Disease Control and Prevention, Centers for Disease
Control and Prevention, 8 Jan. 2020, www.cdc.gov/flu/about/burden/2018-2019.html.
21. Verity, Robert, et al. “Estimates of the Severity of Coronavirus Disease 2019: a Model-Based
Analysis.” The Lancet Infectious Diseases, Elsevier, 30 Mar. 2020,
www.sciencedirect.com/science/article/pii/S1473309920302437.
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=3657607
22. Kelland, Kate. “World Must Prepare for Inevitable next Flu Pandemic, WHO Says.” Reuters,
Thomson Reuters, 11 Mar. 2019, www.reuters.com/article/us-health-flu-who/world-must-
prepare-for-inevitable-next-flu-pandemic-who-says-idUSKBN1QS1EP.
23. Baldo, V, et al. “The New Pandemic Influenza A/(H1N1)pdm09 Virus: Is It Really
‘New’?” Journal of Preventive Medicine and Hygiene, Pacini Editore SRL, 2016,
www.ncbi.nlm.nih.gov/pmc/articles/PMC4910438/.
24. Ioannidis, John. “The Infection Fatality Rate of COVID-19 Inferred from Seroprevalence Data.”
MedRxiv, Cold Spring Harbor Laboratory Press, 1 Jan. 2020,
www.medrxiv.org/content/10.1101/2020.05.13.20101253v2.
25. Global Covid-19 Case Fatality Rates. CEBM, 19 June 2020, www.cebm.net/covid-19/global-
covid-19-case-fatality-rates/.
26. The Infection Fatality Rate of COVID-19 in Stockholm – Technical Report.
www.reddit.com/r/COVID19/comments/hakso9/the_infection_fatality_rate_of_covid19_in/.
27. Egypt: First Case of Coronavirus Confirmed in Egypt February 14 /Update 1. GardaWorld,
www.garda.com/crisis24/news-alerts/314286/egypt-first-case-of-coronavirus-confirmed-in-egypt-
february-14-update-1.
28. Gardner, William, et al. “The Coronavirus and the Risks to the Elderly in Long-Term Care.”
Journal of Aging & Social Policy, U.S. National Library of Medicine, 2020,
www.ncbi.nlm.nih.gov/pubmed/32245346.
29. Lima, Carlos Kennedy Tavares, et al. “The Emotional Impact of Coronavirus 2019-NCoV (New
Coronavirus Disease).” Psychiatry Research, Elsevier, 12 Mar. 2020,
www.sciencedirect.com/science/article/pii/S0165178120305163.
30. Tanne, Janice Hopkins, et al. “Covid-19: How Doctors and Healthcare Systems Are Tackling
Coronavirus Worldwide.” The BMJ, British Medical Journal Publishing Group, 18 Mar. 2020,
www.bmj.com/content/368/bmj.m1090.
31. Mikhael, Ehab Mudher, and Ali Azeez Al-Jumaili. “Can Developing Countries Face Novel
Coronavirus Outbreak Alone? The Iraqi Situation.” Public Health in Practice, Elsevier, 26 Mar.
2020, www.sciencedirect.com/science/article/pii/S2666535220300033.
32. Safiya Richardson, MD. “Clinical Characteristics, Comorbidities, and Outcomes Among Patients
With COVID-19 Hospitalized in the NYC Area.” JAMA, American Medical Association, 26 May
2020, jamanetwork.com/journals/jama/fullarticle/2765184.
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=3657607
33. What Are the Health Risks Related to Overcrowding? World Health Organization, World Health
Organization, 29 Aug. 2016,
www.who.int/water_sanitation_health/emergencies/qa/emergencies_qa9/en/.
34. Yang, Xiaobo, et al. “Clinical Course and Outcomes of Critically Ill Patients with SARS-CoV-2
Pneumonia in Wuhan, China: a Single-Centered, Retrospective, Observational Study.” The
Lancet. Respiratory Medicine, Elsevier Ltd., May 2020,
www.ncbi.nlm.nih.gov/pubmed/32105632.
35. Dawei Wang, MD. “Clinical Characteristics of Patients With 2019 Novel Coronavirus (2019-
NCoV)–Infected Pneumonia in Wuhan, China.” JAMA, American Medical Association, 17 Mar.
2020, jamanetwork.com/journals/jama/fullarticle/2761044.
36. Handwashing Data to Inform the COVID-19 Response. UNICEF DATA, 4 June 2020,
data.unicef.org/resources/handwashing-data-covid-19-response/#12047-4.
37. World Health Statistics, 2012. Available from http://apps.who.int/iris/
bitstream/10665/44844/1/9789241564441_eng.pdf
38. Coronavirus Government Response Tracker. Blavatnik School of Government,
www.bsg.ox.ac.uk/research/research-projects/coronavirus-government-response-tracker.
39. Roser, Max, et al. “Coronavirus Pandemic (COVID-19) - Statistics and Research.” Our World in
Data, 4 Mar. 2020, ourworldindata.org/coronavirus.
40. Harding, Andrew. “Coronavirus in South Africa: The Lull before the Surge?” BBC News, BBC,
10 Apr. 2020, www.bbc.com/news/world-africa-52228932.
41. Ritchie, Research and data: Hannah. “Coronavirus (COVID-19) Testing - Statistics and
Research.” Our World in Data, ourworldindata.org/coronavirus-testing#testing-for-covid-19-
background-the-our-world-in-data-covid-19-testing-dataset.
42. Ritchie, Research and data: Hannah. “Coronavirus (COVID-19) Testing - Statistics and
Research.” Our World in Data, ourworldindata.org/coronavirus-testing#testing-for-covid-19-
background-the-our-world-in-data-covid-19-testing-dataset.
43. World Health Organization COVID-19 Briefing Transcript March 30. Rev, 4 Apr. 2020,
www.rev.com/blog/transcripts/world-health-organization-covid-19-briefing-transcript-march-30.
This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=3657607
