Systems Dynamics Approach for Modelling South Africa's Response to Covid-19: A “What If” Scenario

Abstract

Background: Many countries in the world are still struggling to control COVID-19 pandemic. As of April 28, 2020, South Africa reported the highest number of COVID-19 cases in Sub-Sahara Africa. The country took aggressive steps to control the spread of the virus including setting a national command team for COVID-19 and putting the country on a complete lockdown for more than 100 days. Evidence across most countries has shown that, it is vital to monitor the progression of pandemics and assess the effects of various public health measures, such as lockdowns. Countries need to have scientific tools to assist in monitoring and assessing the effectiveness of mitigation interventions. The objective of this study was thus to assess the extent to which a systems dynamics model can forecast COVID-19 infections in South Africa and be a useful tool in evaluating government interventions to manage the epidemic through ‘what if’ simulations.
Design and Methods: This study presents a systems dynamics model (SD) of the COVID-19 infection in South Africa, as one of such tools. The development of the SD model in this study is grounded in design science research which fundamentally builds on prior research of modelling complex systems.
Results: The SD model satisfactorily replicates the general trend of COVID-19 infections and recovery for South Africa within the first 100 days of the pandemic. The model further confirms that the decision to lockdown the country was a right one, otherwise the country’s health capacity would have been overwhelmed. Going forward, the model predicts that the level of infection in the country will peak towards the last quarter of 2020, and thereafter start to decline.
Conclusions: Ultimately, the model structure and simulations suggest that a systems dynamics model can be a useful tool in monitoring, predicting and testing interventions to manage COVID-19 with an acceptable margin of error. Moreover, the model can be developed further to include more variables as more facts on the COVID-19 emerge.

Full text

Journal of Public Health Research 2021; volume 10:1897
Systems dynamics approach for modelling South Africa’s response
to COVID-19: A “what if” scenario
Shingirirai Savious Mutanga,1,2 Mercy Ngungu,3Fhulufhelo Phillis Tshililo,2,3 Martin Kaggwa4
1Council for Scientific and Industrial Research (CSIR), Smart Place Cluster, Holistic Climate Change-Climate
Services Group, Pretoria; 2Department of Quality and Operations Management, Faculty of Engineering and
Built Environment, University of Johannesburg; 3Human Sciences Research Council Developmental, Capable
and Ethical States, Pretoria; 4Sam Tambani Research Institute, Johannesburg, South Africa
Abstract
Background: Many countries in the world are still struggling
to control COVID-19 pandemic. As of April 28, 2020, South
Africa reported the highest number of COVID-19 cases in Sub-
Sahara Africa. The country took aggressive steps to control the
spread of the virus including setting a national command team for
COVID-19 and putting the country on a complete lockdown for
more than 100 days. Evidence across most countries has shown
that, it is vital to monitor the progression of pandemics and assess
the effects of various public health measures, such as lockdowns.
Countries need to have scientific tools to assist in monitoring and
assessing the effectiveness of mitigation interventions. The objec-
tive of this study was thus to assess the extent to which a systems
dynamics model can forecast COVID-19 infections in South
Africa and be a useful tool in evaluating government interventions
to manage the epidemic through ‘what if’ simulations.
Design and Methods: This study presents a systems dynamics
model (SD) of the COVID-19 infection in South Africa, as one of
such tools. The development of the SD model in this study is
grounded in design science research which fundamentally builds
on prior research of modelling complex systems.
Results: The SD model satisfactorily replicates the general
trend of COVID-19 infections and recovery for South Africa with-
in the first 100 days of the pandemic. The model further confirms
that the decision to lockdown the country was a right one, other-
wise the country’s health capacity would have been overwhelmed.
Going forward, the model predicts that the level of infection in the
country will peak towards the last quarter of 2020, and thereafter
start to decline.
Conclusions: Ultimately, the model structure and simulations
suggest that a systems dynamics model can be a useful tool in
monitoring, predicting and testing interventions to manage
COVID-19 with an acceptable margin of error. Moreover, the
model can be developed further to include more variables as more
facts on the COVID-19 emerge.
Introduction
The coronavirus pandemic is a world-catastrophic event
which has unpredictably put the entire world to a halt.1,2 Tracking
its genesis, many patients with pneumonia of unknown etiology
emerged in Wuhan City, China, in December 2019.3Severe Acute
Respiratory Syndrome Coronavirus 2’ (SARS-CoV-2) was con-
firmed as the causative agent of ‘Coronavirus Disease 2019’ or
COVID-19. As of April 28, 2020, the virus had spread to more
than 213 countries, with a record of 2,883,603 cases and 198,842
deaths.4It became a looming threat to the African continent given
the incidence trends and the underlying vulnerable healthcare sys-
tems.5Evidence across most countries have shown that, it is vital
to monitor the progression of such outbreaks and assess the effects
of various public health measures, such as lockdown on mass
gatherings, extra-ordinary personal hygiene, social distancing
measures in real-time and protective clothing.
The systems dynamics model presented in this paper attempt
to broaden the understanding of South Africa’s epidemic interven-
tions using the “what if” scenario. The main objective of this
study was to assess the extent to which a systems dynamics model
can forecast the COVID-19 infections in South Africa and be a
useful tool in evaluating government interventions to manage the
epidemics through ‘what if’ simulations. The model simulates
infections, deaths and recovery in light of the various strategic
intervention measures based on various assumptions.
Essentially the paper attempts to broadly answer three ques-
tions, namely: i) Can a system dynamics model predict the South
African trend of COVID-19 spread, in particular the levels of
infection, death, and recovery? ii) When is the expected infection
peak period? iii) Based on model simulations, can South Africa’s
COVID-19 interventions be justified and what does this mean
going forward? The model presented provides the feedback
processes based on the various interventions employed by the
South African government in response to the pandemic.
Article
Significance for public health
The paper offers a nuance in the realm of uncertainty through prediction of infectious diseases, which could assist national authorities in decision making
through a multidisciplinary approach of systems dynamics.
[Journal of Public Health Research 2021; 10:1897] [page 51]

[page 52] [Journal of Public Health Research 2021; 10:1897]
COVID-19 intervention strategies
In recognising the fact that COVID-19 is a new virus and there
is currently no vaccine and it might take at least 18 months to
develop, countries around the world have implemented public
health measures to control transmissions.6-10 According to Fong et
al.6and Anderson et al.11 these measures assist in slowing down
the spread of infectious disease, delaying the time of peak infec-
tion, buying time for healthcare systems preparation and develop-
ment of vaccines.
Research evidence has shown that public health measures were
successful in treating the 1918 Spanish flu, SARS epidemic in
2003 and Ebola in 2014.10 Improved control of COVID-19 out-
break in Wuhan, China was credited to public health measures
interventions.12 Public health interventions include social distanc-
ing; quarantine, lockdowns, isolation10 implemented in combina-
tion with personal, hand and respiratory hygiene.13
The main common COVID-19 mitigation strategies used by
countries all over the world have been encouraging personal
hygiene, social distancing in public space, locking down the coun-
try and active testing of the citizenry. These strategies are briefly
expounded in the following section.
Personal hygiene
COVID-19 is transmitted through respiratory droplets, close
contact with an infected person and contact with contaminated sur-
faces.14,15 COVID-19 virus has been detected in patient rooms
(bed rail, table, chair, light switch and floor), toilet (door handle,
bowl, surface, hand rain and sink), soles of medical personnel and
trash cans in hospital wards with infected people.16,17 The above
information supports the need for, hand and respiratory hygiene
and use of personal protective equipment (PPE). The following
measures are recommended: regular washing of hands with soap
and water or with an alcohol-based sanitizer, avoiding touching
eyes, nose and mouth, and covering mouth and nose with bent
elbow or tissue, disposing of used tissue immediately and wearing
of a face mask.18 To prevent the shortage of medical masks for
medical personnel in South Africa, the use of cloth face mask was
recommended for consideration by the general public.
Social distancing and effective quarantine
Social distancing attempts to slow the spread of virus transmis-
sion by maintaining a physical distance between people and reduc-
ing social interactions.19-22 Since COVID-19 is transmitted by res-
piratory droplets, people are required to be in certain proximity for
it to spread.9Furthermore, research has shown that even asymptot-
ic people can transmit COVID-19.23-25 Thus, social distancing
reduces transmission of COVID-19.9,24 Social distancing measures
comprise of keeping physical distance between people (keeping 1
meter apart between people), closure of schools, universities and
workplaces, limiting public transportation, cancellation of mass
gatherings, closure of non-essential workplaces, limiting the num-
ber of shoppers in shops, and limiting close contact with people
outside the households.9,20,26 Literature has shown that social dis-
tancing played a substantial role in containing the spread of
COVID-19 outbreak in China.27,28 Although social distancing has
shown to be effective in slowing the spread of COVID-19,
Anderson et al.11 argues that compliance is of utmost importance
for social distancing to be effective. A study by Ferguson et al.21
shows that lifting social distancing measures which are currently in
place most countries without a vaccine might result in the second
wave of peak infection. It is therefore recommended that social
distancing measures be in place until a vaccine is found and is
accessible to all people around the world.
Quarantine is regarded as one of the oldest and most effective
methods for controlling communicable disease outbreaks.29
Quarantine is defined as confinement and restriction of movement
of people who are alleged to have been exposed to a contagious
disease but are not yet showing symptoms either because they did
not get infected or are still in incubation period.9,10,30-33 In contrast,
isolation is the movement restriction of confirmed infected people
either at home or designated facilities.9It may be an individual or
group either at home or designated facility, voluntary or mandato-
ry.9,30,32 Quarantine and isolation are more effective in reducing the
spread of COVID-19 when done at a designated facility than at
home.34 It is recommended that contacts of COVID-19 patients be
quarantined for at least 14 days from the last day they were
exposed to an ill patient.35 Quarantine is most effective when
detection is swift, and contacts can easily be traced within a short
period.9,10 However, like social distancing, it does not work if peo-
ple are not adhering. Well-timed information on the dos and don’ts
while in quarantine plays an important role in educating people
and may therefore increase adherence.33 In addition, sufficient sup-
plies of food, medication and other essentials should be provided
to all quarantines to improve adherence.
Stopping mass gatherings and effects of lockdown
Mass gatherings can be defined as any occasion either organ-
ised or spontaneous that attracts a sufficient number of people to
strain the planning and response resources of the community, city
or nations hosting the event.36 Examples of mass gatherings
include sporting events such as Olympics, concerts, political ral-
lies, conferences, religious gatherings and cruise ships. Over the
years mass gatherings have shown to be the sources of infectious
diseases which have spread globally37 and respiratory disease are
the most common infections transmitted during mass gatherings.38
Mass gathering poses several health risks such as importation of
infectious diseases, amplification of transmission during events
and international spread of disease.39 Cancellation or suspension of
mass gathering is critical to pandemic mitigations.36 Main aug-
ments for cancellation or postponement of mass gatherings during
the pandemic is that main public health measures for infectious
disease without vaccine usually focus on personal hygiene and
social distancing, and are challenging to carry out at mass gather-
ings.38 Examples of mass gatherings where outbreaks occurred
include cruise ships, church gatherings and funerals. An outbreak
of COVID-19 was reported from Princess Diamond cruise ship off
the coast of Japan and 17 % (617) of the 3700 passengers and crew
were confirmed positive.40 A church gathering in Bloemfontein,
South Africa where there were five international visitors who later
tested positive led to 61 cases of COVID-19 in Bloemfontein.41
Several cases in provinces such as Free State, Kwa-Zulu Natal, and
Northern Cape were traced to church gathering. In Eastern Cape
Province, South Africa, nearly 200 cases were traced to 3 funer-
als.42 Of the three funerals, one was attended by East London
Department of Correctional Service employee and led to nearly 80
cases in prison.42 Many countries around the world such as China,
India, Germany, Italy, France, Poland, United Kingdom, New
Zealand including South Africa have imposed restrictive mass
quarantine or what is known as lockdown as the most important
controlling measure to fight the spread of COVID-19.43-45
Lockdowns have been the most effective strategies to fight
COVID-19.46 Duration of lockdown varies and ranges from 2-4
weeks or more. South Africa entered into its first 21 days of lock-
down from the 26th of March to 16th April, 2020.44 The government
then adopted a risk-adjusted strategy as a way of managing infec-
tions in the country. On the 9th of April, the lockdown was extend-
ed for an additional two weeks until the 30th of April 2020 and was
Article

to be lifted out in phases.47 As part of lockdown, all businesses
were expected to be closed except for those offering essential serv-
ices. The police and soldiers were deployed to carry out patrols.
Violation of lockdown carried a penalty. Since the 1st of May, the
country has moved to level 4 and eased some of the lockdown
restrictions.47 While lockdown may help in curbing the spread of
COVID-19, it has significant social, economic,48 environmen-
tal49,50 and physiological impacts.51 For example, a lockdown for
1.5 months in the UK cost 3.5 % of its GDP.48 Lockdown in
Wuhan, China led to reduced production of automobile parts and
reduced production in countries relying on them for supply such as
Japan.52 Research shows than 5 million people lost their jobs in
China between January and February 2020.48 And it is expected to
occur in other countries including South Africa as more business
are closing because of lockdown. Preliminary results from a study
by the Human Sciences Research Council (2020) on the impacts of
lockdown revealed that during the lockdown, 26 % of the respon-
dents had no money to purchase food. Furthermore, 43% -63 % of
people were no longer employed, thus, they were unable to pay
debts and other necessities.
Testing
As of 11th May 2020, South Africa had conducted 356,067
tests, with 10652 positive confirmed cases and 206 mortalities.47
About 4357 patients had recovered from the virus resulting in
6,089 active cases. At the time of this writing Western Cape
Province had the highest number of cases followed by Kwa-Zulu
Natal (KZN) and Gauteng with 3,908, 722 and 702 active cases,
respectively. Although testing is vital in understanding the pan-
demic and developing intervention strategies, alone it will not help
in stopping the spread of COVID-19, it should be part of the strat-
egy.53 The testing strategy should include rapid diagnosis followed
by a quick and efficient testing with immediate isolation and rigor-
ous contact tracing and self-isolation of all presumed close con-
tacts. A model by Hellewell et al.54 showed that for a reproduction
number of 2.5, 80% of the contacts would need to be traced to con-
trol 90% of the outbreaks. Contact tracing can be quick when the
number of cases is still low, but when they increase dramatically,
it becomes labour intensive and can overload public health sys-
tems.55 A number of digital apps are available for contact tracing,
although they are not widely implemented due to ethical consider-
ations.56 Such apps can collect real-time data such as individual
location, health status and movements. For the testing strategy to
be successful testing centres should be widely available and easily
accessible.57 However, when demand surpasses the capacity, test-
ing strategy may be inffective.18
The South African context of modelling COVID-19
spread
South Africa has not been spared from this pandemic as it
recorded its first case on the 5th of March 20 and has witnessed an
upsurge of 4,793 confirmed cases and a death toll of 90 by the 27th
of April 2020.47 Figure 1 illustrates the spatial distribution of cases
and recoveries across the country with the Western Cape being the
epicenter of the epidemic, followed by Gauteng and KZN respec-
tively. Although the recovery statistics provide an encouraging sce-
nario of the government’s efforts in reducing the spread, the future
is still uncertain given the increasing infections. COVID-19 is a
classic example of how modern-day society grapples with uncer-
tainty.
Modelling is an important way of simplifying a complex
reality but to the extent that one can still get plausible insights on
a phenomenon of interest and on how to influence it. Specific to
South Africa, there are a number of models that have been referred
to, in the country’s projection of COVID-19 spread, explanation of
ramifications thereof and justifying interventions. These models,
however, have not been out in the public domain, hence there is a
limit to which they can be engaged for knowledge purposes,
among other aspects. This study adds to the existing models for
COVID-19 spread for South Africa. The unique aspect of this
model is that it is in the public domain for review, and improve-
ment if need be.
Design and Methods
The study adopted Peffers et al. 59 framework on design sci-
ence research. Fundamentally this builds on prior research of men-
tal modelling of uncertainty and complex systems. The research
design also involved targeted collection of secondary data as
published by the NICD South Africa. In addition, the study also
adopted the Standard SEIR (Susceptible, Exposure, Infective, and
Recovery) Model. The basic structure of the model structure is
based on the modified system dynamics model of epidemic spread
first developed by Kermack and McKendrick in 192760 customised
to the South African situation. A reinforcing feedback loop is
responsible for causing exponential growth in the number of
infected people. Essentially the SEIR model was built on the basic
premise that, when no vaccine is available, the isolation of diag-
nosed infectives, social distancing and hygiene are the only control
measures available.
Model assumptions
• Pandemic is in a short period of time; hence the population
remains constant.
• We also assume that the N mixes homogeneously, regardless
of demographics (age, gender where they live or work) or any
other behaviour traits that the individuals might have.
Proportionality of rates is applied due to homogeneity.
• Rate of increase of infected is proportional to the contacts
between infective and susceptible (manner in which the dis-
ease is transmitted). In this case we assume that when suscep-
tible come in contact with an infected person there is a proba-
bility r=0.8 of transmission per contact.
[Journal of Public Health Research 2021; 10:1897] [page 53]
Article
Figure 1. Map showing the spatial distribution of COVID cases
in South Africa.58

• Lack of infectiousness during the incubation period (14 days)
• There is a constant rate either death or recovery
• There is no re-infection for those who have recovered.
The set of differential equations adopted have thus been sum-
marized follows:
Description of system variables
The analytical solution using system dynamics is thus denoted
as follows:
The model structure as depicted on Figure 2 comprise of 5
major stocks namely Susceptible, Exposure, Infections Deaths and
Recoveries.
Results
The results presented essentially revolve around the three key
questions which this paper attempts to respond to. The first part is
a response to the extent to which COVID-19 spreads in South
Africa. Essentially the focus has been to determine the number of
infections and recovery before and after the government interven-
tions, i.e. the lockdowns.
Can a system dynamics model predict the South
African trend of COVID-19 spread, in particular, the
levels of infection, death, and recovery?
The model simulation of infections and recovery before govern-
ment interventions referred in this study as the Business As Usual
Scenario indicated that by the 23rd day since the first recorded case
of COVID-19, South Africa had approximately 1,187 infected peo-
ple denoted as infectives on Figure 3. The simulated recoveries’ by
then were very low far much less than the cumulative 30 people.
The model depicted a similar trend for recoveries and infec-
tions over the first simulated phase as shown on Figure 3. In spite
of the similar trajectory the estimated figures during the first
month are slightly lower than the actual recorded figures as indi-
cated on Figure 2 with last date infections record of 1,187 and
1,210, respectively. The recovery by then were relatively few as
indicated on Figure 2 for the actual in comparison with the simu-
lated results. The model satisfactorily depicted the trend, and this
was validated using regression curves for the simulated data vs the
actual recordings shown on Figure 4.
When is the expected peak period?
As shown on Figure 5 the model suggests that infections peak
will be experienced around 130 days post the interventions which
is around August 2020. The level of infections will skyrocket to
over 2 million people while the exposed population will be much
higher within the first 100 days. While recoveries will be growing
cumulatively over time the levels of mortality will also rise with an
anticipated zenith of around 16,000 deaths.
Article
Figure 2. Standard SEIR Model used in the study.
Figure 3. Simulated vs recorded infections and recovery before
lockdown (Business as Usual Scenario).
[page 54] [Journal of Public Health Research 2021; 10:1897]

Alternative scenario: model simulation with interven-
tions (lockdown, hygiene, social distancing)
As shown on Figure 5, the model suggests that the peak of
infections would have been experienced much earlier around day
80, which is around June 2020. Infections level could have risen to
nearly 2.5 million. Deaths on the other hand would surge to over
50000 deaths. Lastly an exponential increase in recoveries is also
expected under this scenario. To validate the model regression
curves were developed comparing the simulated results with the
actual recorded cases. The next section provides the regression
curves for infections, recoveries and deaths.
Discussions
Can a SEIR model adapted using systems dynamics be
able to predict infection spread of COVID-19?
Despite the needed room for improvement, the systems
dynamics model presented has been able to depict the general
trend of infections. The model adds value to the body of knowl-
edge on transmission dynamic models, which are an important ini-
tial step towards understanding emerging infectious disease such
as COVID-19.61 The feedback loops between key variables point
towards the heterogeneity of infectiousness. This view point is
mindful of the model’s major underlying assumptions, such as lack
of infectiousness during the incubation period and the net effect of
government interventions to flatten the curve COVID-19. Similar
to SARS, which was characterised by a number of super-spreaders
in which most people infect two to three others, the risk of estab-
lished local transmission with a single imported case is consider-
ably higher.61 This could assist in explaining the burgeoning rate of
infections especially when the lockdown phase in South Africa
was lifted. Similarly, Kucharski62 ascertained that chains of trans-
mission might not take off initially and might require up to four
imported cases to establish transmission. Social distancing thus
remains a great plausible action which the South African govern-
ment undertook to delay and minimise the spread.
The differences in figures between simulated cases and
the actual
While the first few weeks of the simulation reflects generally
low figures compared to the actual recorded cases and later on the
simulated infections become relatively higher compared to the
actual cases. The two contrasting scenarios demonstrate that in one
crucial respect, though, these simulations are not perfect predictors
of reality. Specific to system dynamics modelling, more weight is
put in replicating trends rather than point accuracy.63
This study argues that while the recorded cases may not nec-
essarily translate to the exact total number of cases experienced in
the country, the r squares indicated on Figure 6 points to the abil-
ity of the designed model to satisfactorily predict the infection
trend for South Africa. The relatively lower figures at the begin-
ning few weeks could possibly be explained by the patient zero
effect. The model started simulating from the week of the first
recorded infection, yet a number of people could have already
been exposed and some infected but not recorded or reported by
then. The patient zero effect was envisaged weeks later with
traced 39 patients and 80 staff linked to the hospital had been
infected, and 15 patients died in Kwa-Zulu Natal South Africa.64
Thus the model may still have had the lag for incubation of possi-
ble cases in those few weeks.
Unaccounted for figures may help explain the relatively higher
estimates compared to the actual recordings. Like many other
countries in the world South Africa may not have the full capacity
for testing let alone reporting of all the cases related to COVID-19.
As such a large number of people may have remained unaccounted
for. This corroborates the WHO18 acknowledgement that testing
strategy can be threatened when demand surpasses the capacity. In
addition, the disparity in figures could also be explained by the
sheer inability of the model to provide the exact figures given the
Article
Figure 5. Simulated infections, exposure, recoveries and deaths
(Infection peak period with limited interventions).
Figure 6. Simulated vs actual recorded regression curves.
Figure 4. Simulated infections, exposure, recoveries and deaths.
[Journal of Public Health Research 2021; 10:1897] [page 55]

[page 56] [Journal of Public Health Research 2021; 10:1897]
inherent assumptions and constant rates that have been imputed to
run the simulations. Despite these variations the model still pro-
vides a good idea of the trend in infections, recovery and mortality
thereof. However, simulations are not COVID-19, and these simu-
lations vastly oversimplify the complexity of real life.65
When was the peak expected and what could have been
the alternative scenario had the government not inter-
vened?
According to this model at the time of this writing, the peak is
yet to be experienced. Had it not been the government’s interven-
tions, in particular social distancing and the risk adjusted lockdown
strategy the peak period would have been experienced earlier dur-
ing the peak of winter. The net effect of government intervention
has thus been the peak delay and is thus anticipated to occur during
the second half of the year 2020 giving an allowance for better pre-
paredness. One of the greatest threats posed by COVID-19 has
been the tremendous pressure on the health care systems.
Countries such as Spain, Italy and the US have witnessed enor-
mous pressure on hospitals to support the influx of people with
severe disease, and such countries have experienced high levels of
mortality.
Model limitations
In this study, we acknowledge some limitations in the model.
Among these is the: i) The homogeneous mixing which can lead to
an overestimation of the final pandemic size and the magnitude of
the interventions needed to stop the pandemic. ii) The model
attempts to simulate and predict South Africa's covid situation in
view of the lack of suitable data and the uncertainty of the different
parameters, namely, the variation of the degree of isolation and
social distancing as a function of time, the initial number of
exposed individuals and infected people, the incubation and infec-
tious periods, and the fatality rate. This paper underscores the
school of thought which affirms that “All models are wrong, but
some are useful”.
Conclusions
Tools such as the adapted SEIR model using systems dynamics
can predict infection spread of COVID-19. Such tools provide
valuable insights to decision makers through understanding the
course of the epidemic in especially the levels of infections, recov-
eries, and deaths. Had it not been the government’s interventions,
in particular social distancing and the risk adjusted lockdown strat-
egy the peak period would have been experienced earlier during
the peak of winter. Developing models such as the SD presented in
this paper can be crucial for anticipating resource requirements
particularly during the peak of the epidemic. Planning and level of
preparedness remain crucial cornerstones for a sound disaster risk
of strategy for pandemics such as COVID-19 and thus systems
dynamics model can assist with informed decision making.
Although the South African government’s response strategy can be
commendable and justified given the fairly low mortality rates
recorded, thus far, in comparison to many developed countries, the
country still needs better tools of prediction and analysis to support
it in the next phase of similar pandemics. A systems dynamics
model is one of such needed tools.
References
1. Shim E, Tariq A, Choi W, Lee Y, Chowell G. Transmission
potential and severity of COVID-19 in South Korea. Int J
Infect Dis 2020;93:339-44.
2. Mizumoto K, Kagaya K, Zarebski A, Chowell G. Estimating
the asymptomatic proportion of coronavirus disease 2019
(COVID-19) cases on board the Diamond Princess cruise ship,
Yokohama, Japan, 2020. Eurosurveillance 2020;25:2000180.
3. Li H, Liu S-M, Yu X-H, et al. Coronavirus disease 2019
(COVID-19): current status and future perspective. Int J
Antimicrob Agents 2020;55:105951.
4. WHO. Coronavirus disease 2019 (COVID-19): situation
report, 72. 2020. Available from: https://apps.who.int/iris/han-
dle/10665/331685
5. Loembé M, Tshangela A, Salyer S, et al. COVID-19 in Africa:
the spread and response. Nat Med 2020;26:999-1003.
6. Fong MW, Gao H, Wong JY, et al. Nonpharmaceutical meas-
ures for pandemic influenza in nonhealthcare settings - social
distancing measures. Emerg Infect Dis 2020;26:976-84.
7. Hartley DM, Perencevich EN. Public health interventions for
COVID-19: emerging evidence and implications for an evolv-
ing public health crisis. JAMA 2020;323:1908-9.
Article
Correspondence: Dr. Shingirirai Savious Mutanga, 246 Rivea Street,
Doornport, PO BOX 13502 Sinnoville, 0182 Pretoria Gauteng, South
Africa.
Tel.+27.731631827.
E-mail: smutanga@csir.co.za
Key words: COVID-19, systems dynamics; SEIR.
Contributions: All the authors contributed to a review of COVID-19
response strategies, SEIR model modification, systems dynamics
model design and development (building assumptions, model parame-
terisation, model simulations- scenario building, model validation)
discussion, comparing with South African (COVID-19 statistics),
revision of manuscript, critical review, finalisation of manuscript. All
the authors have read and approved the final version of the manuscript
and agreed to be accountable for all aspects of the work.
Conflicts of interests: The authors declare that they have no compet-
ing interests, and all authors confirm accuracy.
Ethics approval: Not applicable.
Disclaimer: The model has been built based on assumptions and the
limited available accessible data which could be used by the authors.
Significance for public health: The paper offers a nuance in the
realm of uncertainty through prediction of infectious diseases, which
could assist national authorities in decision making through a multi-
disciplinary approach of systems dynamics.
Received for publication: 11 August 2020.
Accepted for publication: 26 December 2020.
©Copyright: the Author(s), 2021
Licensee PAGEPress, Italy
Journal of Public Health Research 2021;10:1897
doi:10.4081/jphr.2021.11897
This work is licensed under a Creative Commons Attribution
NonCommercial 4.0 License (CC BY-NC 4.0).

8. Heymann DL, Shindo N. COVID-19: what is next for public
health? Lancet 2020;395:542-5.
9. Wilder-Smith A, Freedman DO. Isolation, quarantine, social
distancing and community containment: pivotal role for old-
style public health measures in the novel coronavirus (2019-
nCoV) outbreak. J Travel Med 2020;27:taaa020.
10. Wilder-Smith A, Chiew CJ, Lee VJ. Can we contain the
COVID-19 outbreak with the same measures as for SARS?
Lancet Infect Dis 2020. doi: 10.1016/S1473-3099(20)30129-8
11. Anderson RM, Heesterbeek H, Klinkenberg D, Hollingsworth
TD. How will country-based mitigation measures influence the
course of the COVID-19 epidemic? Lancet 2020;395:931-4.
12. Pan F, Ye T, Sun P, et al. Time course of lung changes on chest
CT during recovery from 2019 novel coronavirus (COVID-19)
pneumonia. Radiology 2020;295:200370.
13. WHO. Rational use of personal protective equipment (PPE)
for coronavirus disease (COVID-19): interim guidance, 19
March 2020. Available from: https://apps.who.int/iris/ han-
dle/10665/331498
14. Liu Y, Ning Z, Chen Y, et al. Aerodynamic characteristics and
RNA concentration of SARS-CoV-2 aerosol in Wuhan hospi-
tals during COVID-19 outbreak. BioRxiv 2020. doi:
10.1101/2020.03.08.982637
15. WHO. Modes of transmission of virus causing COVID-19:
implications for IPC precaution recommendations: scientific
brief, 29 March 2020. Available from: https://www.who.int/
news-room/commentaries/detail/modes-of-transmission-of-
virus-causing-covid-19-implications-for-ipc-precaution-rec-
ommendations
16. Guo Z-D, Wang Z-Y, Zhang S-F, et al. Aerosol and surface dis-
tribution of severe acute respiratory syndrome coronavirus 2 in
hospital wards, Wuhan, China, 2020. Emerg Infect Dis
2020;26:1583-91.
17. Ong SWX, Tan YK, Chia PY, et al. Air, surface environmental,
and personal protective equipment contamination by severe
acute respiratory syndrome coronavirus 2 (SARS-CoV-2) from
a symptomatic patient. JAMA 2020;323:1610-2.
18. WHO. Laboratory testing strategy recommendations for
COVID-19: interim guidance, 22 March 2020. Available from:
https://apps.who.int/iris/handle/10665/331509
19. Glass RJ, Glass LM, Beyeler WE, Min HJ. Targeted social dis-
tancing designs for pandemic influenza. Emerg Infect Dis
2006;12:1671.
20. Block P, Hoffman M, Raabe IJ, et al. Social network-based dis-
tancing strategies to flatten the COVID 19 curve in a post-
lockdown world. Nat Hum Behav 2020;4:588-96.
21. Ferguson N, Laydon D, Nedjati Gilani G, et al. Report 9:
Impact of non-pharmaceutical interventions (NPIs) to reduce
COVID19 mortality and healthcare demand. Imperial College
London [Internet]; 2020. Available from: https://www.imperi-
al.ac.uk/mrc-global-infectious-disease-analysis/covid-
19/report-9-impact-of-npis-on-covid-19/
22. Greenstone M, Nigam V. Does social distancing matter?
Becker Friedman Institute Econ [Internet]; 2020. Available
from: https://bfi.uchicago.edu/working-paper/2020-26/
23. Tian L, Li X, Qi F, et al. Pre-symptomatic transmission in the
evolution of the COVID-19 pandemic. arXiv 2020.
arXiv:2003.07353v6 [q-bio.PE].
24. Wei WE, Li Z, Chiew CJ, et al. Presymptomatic transmission
of SARS-CoV-2 - Singapore, January 23-March 16, 2020.
Morb Mortal Wkly Rep 2020;69:411.
25. Qian G, Yang N, Ma AHY, et al. COVID-19 transmission with-
in a family cluster by presymptomatic carriers in China. Clin
Infect Dis 2020;71:861-2.
26.Cowling BJ, Aiello AE. Public health measures to slow com-
munity spread of coronavirus disease 2019. J Infect Dis
2020;221:1749-51.
27.Chen S, Yang J, Yang W, et al. COVID-19 control in China
during mass population movements at New Year. Lancet
2020;395:764–6.
28. Kraemer MUG, Yang C-H, Gutierrez B, et al. The effect of
human mobility and control measures on the COVID-19 epi-
demic in China. Science 2020;368:493-7.
29. Taghrir MH, Akbarialiabad H, Marzaleh MA. Efficacy of mass
quarantine as leverage of health system governance during
COVID-19 outbreak: A mini policy review. Arch Iran Med
2020;23:265-7.
30. Rothstein MA. From SARS to Ebola: legal and ethical consid-
erations for modern quarantine. Ind Heal L Rev 2015;12:227.
31. Brooks SK, Webster RK, Smith LE, et al. The psychological
impact of quarantine and how to reduce it: rapid review of the
evidence. Lancet 2020;395:912-20.
32. Parmet WE, Sinha MS. Covid-19 - the law and limits of quar-
antine. N Engl J Med 2020;382):e28.
33. Webster RK, Brooks SK, Smith LE, et al How to improve
adherence with quarantine: Rapid review of the evidence.
Public Health 2020;182:163-9.
34. Dickens BL, Koo JR, Wilder-Smith A, Cook AR. Institutional,
not home-based, isolation could contain the COVID-19 out-
break. Lancet 2020;395:1541-2.
35. WHO. Considerations for quarantine of individuals in the con-
text of containment for coronavirus disease (COVID-19):
interim guidance, 19 March 2020. Available from:
https://apps.who.int/iris/handle/10665/331497
36. Ahmed QA, Memish ZA. The cancellation of mass gatherings
(MGs)? Decision making in the time of COVID-19. Travel
Med Infect Dis 2020;34:101631.
37. Memish ZA, Ahmed QA, Schlagenhauf P, et al. No time for
dilemma: mass gatherings must be suspended. Lancet
2020;395:1191.
38. Ebrahim SH, Memish ZA. COVID-19–the role of mass gath-
erings. Travel Med Infect Dis 2020;34:101617.
39. Wilder-Smith A, Steffen R. Mass gatherings. In: Travel medi-
cine. Elsevier; 2019. p. 383–6.
40. Rocklöv J, Sjödin H, Wilder-Smith A. COVID-19 outbreak on
the Diamond Princess cruise ship: estimating the epidemic
potential and effectiveness of public health countermeasures. J
Travel Med 2020;27:taaa030.
41. Boonzaaier D, Eybers J. Bloemfontein church gathering: 67
more people test positive for Covid-19. City Press [Internet].
[cited 2020 May 11]. Available from: https://city-
press.news24.com/News/bloemfontein-church-gathering-67-
more-people-test-positive-for-covid-19-20200328
42. McCain N. 200 Covid-19 cases linked to three Eastern Cape
funerals. News24 [Internet]. [cited 2020 May 11]. Available
from: https://www.news24.com/SouthAfrica/News/200-covid-
19-cases-linked-to-three-eastern-cape-funerals-20200425
43. Chatterjee K, Chauhan VS. Epidemics, quarantine and mental
health. Med J Armed Forces India 2020;76:125-7.
44. Hamzelou J. World in lockdown. New Sci 2020;245:7.
45. Muggeo V, Sottile G, Porcu M. Modelling covid-19 outbreak:
segmented regression to assess lockdown effectiveness.
ResearchGate 2020. doi: 10.13140/RG.2.2.32798.28485
46. Das S, Ghosh P, Sen B, Mukhopadhyay I. Critical community
size for COVID-19—a model based approach to provide a
rationale behind the lockdown. arXiv 2020. arXiv200403126.
47. South Africa Department of Health. COVID-19 South African
coronavirus news and information. [Internet]. [cited 2020 May
[Journal of Public Health Research 2021; 10:1897] [page 57]
Article

[page 58] [Journal of Public Health Research 2021; 10:1897]
12]. Available from: https://sacoronavirus.
co.za/?__cf_chl_jschl_tk__=8c3117c2416ea9bdc47a69838e7
56d147d21525b-1589256955-0-AaKZrMvoF-
KnY3JGXBlxyGB1s7w6Jyfo6VkSERB9MMJWHJN-
Nake5dc6IZGQr061I6ujUy8NMKTl3M4WwqB_qdQn_QGf
QalAmTcfN5KfeXX4fy5U2hWCzoyXWEVNQq8yooDdJ1Z-
V6etZoBXsOt3bcUV
48. Fernandes N. Economic effects of coronavirus outbreak
(COVID-19) on the world economy. 2020. Available from:
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3557504
49. Kerimray A, Baimatova N, Ibragimova OP, et al. Assessing air
quality changes in large cities during COVID-19 lockdowns:
The impacts of traffic-free urban conditions in Almaty,
Kazakhstan. Sci Total Environ 2020;730:139179.
50. Muhammad S, Long X, Salman M. COVID-19 pandemic and
environmental pollution: A blessing in disguise? Sci Total
Environ 2020;730:138820.
51. Ahmed MZ, Ahmed O, Aibao Z, et al. Epidemic of COVID-19
in China and associated psychological problems. Asian J
Psychiatr 2020;51:102092.
52. Inoue H, Todo Y. The propagation of the economic impact
through supply chains: The case of a mega-city lockdown
against the spread of COVID-19. 2 PLoS One
2020;15:e0239251.
53. Salathé M, Althaus CL, Neher R, et al. COVID-19 epidemic in
Switzerland: on the importance of testing, contact tracing and
isolation. Swiss Med Wkly 2020;150:w20225.
54. Hellewell J, Abbott S, Gimma A, et al. Feasibility of control-
ling COVID-19 outbreaks by isolation of cases and contacts.
Lancet Global Health 2020;8:488-96.
55. Maghdid HS, Ghafoor KZ. A smartphone enabled approach to
manage COVID-19 lockdown and economic crisis. SN
Comput Sci 2020;1:271.
56. Morley J, Cowls J, Taddeo M, Floridi L. Ethical guidelines for
SARS-CoV-2 digital tracking and tracing systems. 2020.
Available from: https://papers.ssrn.com/sol3/papers.cfm?
abstract_id=3582550
57. Onder G, Rezza G, Brusaferro S. Case-fatality rate and charac-
teristics of patients dying in relation to COVID-19 in Italy.
JAMA 2020;323:1775-6.
58. National Institute for Communicable Disease [Internet].
National COVID-19 Daily Report. [cited 2020 Dec 22].
Available from: https://www.nicd.ac.za/diseases-a-z-
index/covid-19/surveillance-reports/national-covid-19-daily-
report/
59. Peffers K, Tuunanen T, Rothenberger MA, Chatterjee S. A
design science research methodology for information systems
research. J Manag Inf Syst 2007;24:45-77.
60. Schrӧck S. COVID-19 and the SIR-Model. [cited 2020 Aug 8].
Available from: https://www.transentis.com/covid-19-and-the-
sir-model/en/
61. McBryde E. The value of early transmission dynamic studies
in emerging infectious diseases. Lancet Infect Dis
2020;20:512-3.
62. Kucharski AJ, Russell TW, Diamond C, et al. Early dynamics
of transmission and control of COVID-19: a mathematical
modelling study. Lancet Infect Dis 2020;20:553-8.
63. Motzev M. Prediction accuracy-A measure of simulation real-
ity. ResearchGate doi: 10.13140/RG.2.2.18743.75687
64. Nordling L. Study tells ‘remarkable story’about COVID-19’s
deadly rampage through a South African hospital. ScienceMag
[Internet]. 2020. Available from:
https://www.sciencemag.org/news/2020/05/study-tells-
remarkable-story-about-covid-19-s-deadly-rampage-through-
south-african
65. Stevens H. Why outbreaks like coronavirus spread exponen-
tially, and how to “flatten the curve.” Washington Post
[Internet]. 2020. Available from: https://www.washington-
post.com/graphics/2020/world/corona-simulator/.
Article