Assessing the dynamic impacts of non-pharmaceutical and pharmaceutical intervention measures on the containment results against COVID-19 in Ethiopia.

The rapid spread of COVID-19 in Ethiopia was attributed to joint effects of multiple factors such as low adherence to face mask-wearing, failure to comply with social distancing measures, many people attending religious worship activities and holiday events, extensive protests, country election rallies during the pandemic, and the war between the federal government and Tigray Region. This study built a system dynamics model to capture COVID-19 characteristics, major social events, stringencies of containment measures, and vaccination dynamics. This system dynamics model served as a framework for understanding the issues and gaps in the containment measures against COVID-19 in the past period (16 scenarios) and the spread dynamics of the infectious disease over the next year under a combination of different interventions (264 scenarios). In the counterfactual analysis, we found that keeping high mask-wearing adherence since the outbreak of COVID-19 in Ethiopia could have significantly reduced the infection under the condition of low vaccination level or unavailability of the vaccine supply. Reducing or canceling major social events could achieve a better outcome than imposing constraints on people's routine life activities. The trend analysis found that increasing mask-wearing adherence and enforcing more stringent social distancing were two major measures that can significantly reduce possible infections. Higher mask-wearing adherence had more significant impacts than enforcing social distancing measures in our settings. As the vaccination rate increases, reduced efficacy could cause more infections than shortened immunological periods. Offsetting effects of multiple interventions (strengthening one or more interventions while loosening others) could be applied when the levels or stringencies of one or more interventions need to be adjusted for catering to particular needs (e.g., less stringent social distancing measures to reboot the economy or cushion insufficient resources in some areas).

1 Introduction and background

Since the emergence of the first case of COVID-19 on 13 March 2020 in Ethiopia, as of 6 January 2022, this country had reported 436,586 confirmed cases and 6,988 deaths [1]. Furthermore, during the same period, due to the global COVID-19 vaccine inequity [2, 3], there were only 1.4% and 6.6% of the total population being fully vaccinated and partially vaccinated, respectively [4, 5].

As one of the ancient countries with strong religious people, in Ethiopia, 43.5% of the population are Orthodox Christians, 33.9% of the population are Muslims, and the rest are protestants and traditional religious followers [6]. Ethiopian Orthodox Churches encourage the communities to pray more in groups together by reinforcing a feeling of unity in its people. In Orthodox Church, people made greetings to each other by shaking hands and having cheek-to-cheek kisses, and face masks were not used in the church because the church is deemed a sacred place [6]. Muslim communities also have similar greeting etiquettes. Such experiences were critical contributing factors to the increased risk of COVID-19 transmission. Moreover, Ethiopia is the second-most populous country in Africa next to Nigeria, and it is prevalent for people to live together with their extended families under one roof, eat together from one plate daily, and move in groups on the very narrow paved road since this country has solid social solidarity [7]. Therefore, the practices mentioned above of religion, cultural, and social interactions in Ethiopia posed significant challenges in effectively containing and controlling COVID-19.

In the meantime, Ethiopia is one of the poorest countries in the world, with a per capita GDP of $936.34 in 2020 [8]. As of 2021, nearly 23% of the population lives in extreme poverty [9]. A study confirmed that hospital preparedness in the selected state was tremendously insufficient as per World Health Organization measurement, with one out of eight hospitals admitting COVID-19 patients [10]. What is more serious is that, according to statistics from the World Bank, the standard hospital beds per thousand people ratio of 0.33 in 2016 in Ethiopia was far from the WHO minimum standard (3 beds per 1000) [11], which rendered the unadmitted COVID-19 patients helpless and forced them to become the source of infection.

With the advent of COVID-19, scientists in public health worldwide have been trying to investigate the gaps and challenges in the containment measures and, consequently, evaluate a particular country’s preparedness for the new emerging disease [12-15]. As non-pharmaceutical interventions (NPIs) such as social distancing, hand washing, and face mask-wearing have been proved effective in providing necessary help, scholars have examined the knowledge, attitude, and practices of face mask utilization affecting them in Ethiopia [16-21]. Ayele Tadesse Awoke et al.’s survey in the Amhara region of Ethiopia revealed that, at the early stage of the COVID-19 pandemic, levels of adherence regarding hand hygiene, physical distancing, and mask utilization were 12.0%, 13.00%, and 26%, respectively, which demonstrated regional variation. They suggested that community-based education would increase the practices mentioned above [22]. The surveys conducted by Endriyas et al. and Haftom & Petrucka showed that the levels of mask utilization were more than 50% [23, 24]. Ayele Wondimu et al. evaluated the spread dynamics of COVID-19 in Ethiopia under the assumptions of face mask utilization of 20%, 40%, and 60%. They also assessed the combined effects of enforcing social distancing and increasing the adherence to mask-wearing and found significant reductions in infections [25]. Studies from Tucho and Kumsa, and Zewude et al. indicated the challenges of keeping certain compliance levels of mask-wearing [26, 27]. Ejigu et al.’s research predicted the infections under the assumptions of implementing different NPIs including social distancing, mask-wearing, and sanitary measures [28].

Bushira used geospatial techniques and the CHIME model to evaluate the impacts of 25%, 75%, and 95% social distancing interventions on flattening and delaying the curve [29]. Deressa examined the practices of social distancing of government employees in Addis Ababa and the results exhibited a 96% adherence level to mask-wearing [30]. Fikrie et al.’s study identified that knowledge and attitude were contributing factors causing poor practices of social distancing in the West Guji Zone of Ethiopia [31]. Hailu et al. investigated the barriers and driving factors for influencing the compliance level of social distancing measures [17]. In Tolu et al.’s study, the calculated ReadyScore was 52% indicating that more measures needed to be implemented, where they recommended social distancing measures, increasing case tracing, sanitary measures, etc [15].

Apart from assessing the impacts of NPIs on the spread dynamics in Ethiopia, Suthar et al. evaluated the vaccination (i.e., pharmaceutical interventions-PIs) on the transmission dynamics of COVID-19 [32]. Deressa and Duress used the mathematical epidemiological model to evaluate and identify the optimal combination of multiple measures-i.e., public health education, personal protective measures, and hospitalization of the infected [33].

Researchers have used different models to predict the spread trend of COVID-19 over time. Abebe used an exponential smoothing model for the new COVID-19 infections [34]. Eticha employed a case-based autoregressive integrated moving average model to predict the new cases of COVID-19 [35]. In a study done by Gebretensae and Asmelash, the researchers used Box–Jenkins modeling framework, i.e., ARIMA (p, d, q), to forecast the trend of COVID-19 spread in Ethiopia [36]. Gebremeskel et al. applied a compartmental epidemic model to predict the transmission dynamics of COVID-19 in Ethiopia [37]. Habenom et al. adopted a model with fractional differential equations to analyze the transmission dynamics of COVID-19 in Ethiopia [38].

Given the capability of system dynamics (SD) models to capture nonlinearity between cause and effect and integrate time delays and feedback loops prevalent in disease progression and population health, SD models have been widely used in multiple areas of health-related research since the 1970s [39]. Its broad applications have been seen in studying the dynamics of both infectious diseases [40, 41] and non-communicable chronic diseases [42-52]. SD’s applications in health-related fields were also often witnessed in evaluating the impacts of policies and interventions [53-61]. SD models were also widely used in studies related to health service improvement (e.g., hospital management) [62-64]. SD models were well recognized and practiced in research in developing and evaluating national health policy [54, 65–69] and investigating complexity and uncertainties in healthcare and health-related socioeconomic systems [55, 62, 70–75]. In the past two years, scholars have extensively used SD models to understand the transmission dynamics, impacts of containment measures, and prediction of COVID-19 spread [76-87].

Majorities of previous research have focused on the evaluations of limited interventions and their combinations without simultaneously capturing the impacts of NPIs and PIs. For example, a typical SEIR model was used by Ayele Wondimu et al. to assess the impacts of social distancing measures and different compliance levels of face mask-wearing [25]. The SEIR models employed by Ejigu et al. and Taye et al. evaluated the impacts of interventions such as social-distancing measures, mask-wearing, and handwashing [28, 88]. Our extended SEIR further considered the impacts of a lot more factors including hospitalized patients, un-hospitalized patients, patients with mild and severe symptoms, asymptomatic patients, vaccine administration levels, hospital/quarantine hospital capacity, and different levels of compliance in face mask-wearing.

To gain a better understanding of issues that existed in the containment measures against COVID-19 disease at the previous stage and gain insights into the pandemic trend under different intervention scenarios, this study mainly focuses on the application of the SD model to conduct counterfactual analysis (16 scenarios) to identify issues and gaps in the past containment. It also aims to control measures and capture the possible future transmission dynamics of COVID-19 under various combinations of interventions, including NPIs and PIs (264 scenarios). In the NPI, we evaluated the stringency of social distancing measures (three levels of stringency for social events and routine activities represented by average contact rates-see Table 2 and Appendix C in S1 Appendix [25]), the adherence level of face mask-wearing (three levels with 48% [18, 19] in baseline, 60%, and 70% in the proposed assumptions), and hospital beds (20000 beds in baseline [89, 90]-including temporary beds in isolation centers, 12000–60%, 16000–80%, 40000–200% in the proposed assumptions). Regarding PIs, we considered the capacity of vaccine supply and administration rate (30%- and 20% of the total population being administered the 1st & 2nd dose, 20% and 10% of the total population being administered the 1st & 2nd dose in the proposed assumptions), vaccine efficacy (68.4% for the first dose and 80% for the second dose [91, 92], 40% for the first dose and 60% for the second dose in the proposed assumptions), and immunological period (240 months in the baseline [93-96], proposed assumption of 180 months).

Table 2

Scenario settings in the counterfactual analysis.

Scenarios	Medical resources	Vaccine supply and administration				Interventions
	Hospital beds	Vaccine administration	Vaccine supply	✓Vaccine efficacy	Immuno-logical period (days)	† Social event-Average contact rate	†† Imposing social distancing- Average contact rate	*Symptomatic infectivity- weighted by mask- wearing adherence	**Asymptomatic infectivity- weighted by mask-wearing adherence
S 1	20000	No change	No change	68.4%,80%	240	Table A1 (S1 Appendix)	14	Table A4 (S1 Appendix)	Table A5 (S1 Appendix)
S 2	20000	No change	No change	68.4%,80%	240	22	Table A1	Table A4 (S1 Appendix)	Table A5 (S1 Appendix)
S 3	20000	No change	No change	68.4%,80%	240	22	14	Table A4 (S1 Appendix)	Table A5 (S1 Appendix)
S 4	20000	No change	No change	40%, 60%	240	Table A1 (S1 Appendix)	14	Table A4 (S1 Appendix)	Table A5 (S1 Appendix)
S 5	20000	No change	No change	40%, 60%	240	22	Table A1	Table A4 (S1 Appendix)	Table A5 (S1 Appendix)
S 6	20000	No change	No change	40%, 60%	240	22	14	Table A4 (S1 Appendix)	Table A5 (S1 Appendix)
S 7	20000	No change	No change	68.4%,80%	240	Table A1 (S1 Appendix)	14	55%	55%
S 8	20000	No change	No change	40%, 60%	240	Table A1 (S1 Appendix)	14	55%	55%
S 9	40000	No change	No change	68.4%,80%	240	Table A1 (S1 Appendix)	14	Table A4 (S1 Appendix)	Table A5 (S1 Appendix)
S 10	40000	No change	No change	68.4%,80%	240	22	Table A1	Table A4 (S1 Appendix)	Table A5 (S1 Appendix)
S 11	40000	No change	No change	68.4%,80%	240	22	14	Table A4 (S1 Appendix)	Table A5 (S1 Appendix)
S 12	40000	No change	No change	40%, 60%	240	Table A1 (S1 Appendix)	14	Table A4 (S1 Appendix)	Table A5 (S1 Appendix)
S 13	40000	No change	No change	40%, 60%	240	22	Table A1	Table A4 (S1 Appendix)	Table A5 (S1 Appendix)
S 14	40000	No change	No change	40%, 60%	240	22	14	Table A4 (S1 Appendix)	Table A5 (S1 Appendix)
S 15	40000	No change	No change	68.4%,80%	240	Table A1 (S1 Appendix)	14	55%	55%
S 16	40000	No change	No change	40%, 60%	240	Table A1 (S1 Appendix)	14	55%	55%

Notes

✓ Based on the literature, the assumed efficacies of the first dose and second dose COVID-19 vaccine are 68.4% and 80%, respectively, in the baseline setting. In the counterfactual analysis, we want to see the impact of low vaccine efficacy on the spread dynamics. Of course, we also combine this factor with other proposed changes in the analysis.

Here, we consider the situations by incorporating social events and imposing social distancing containment measures. In the counterfactual analysis, we use the average contact rates of 22 and 14 for the simulation periods with social events and the average period with enhanced social distancing containment.

* and ** We referred to the mask-wearing data for Ethiopia in the historical data coming from IHME (Institute for Health Metrics and Evaluation) https://covid19.healthdata.org/ethiopia?view=mask-use&tab=trend.

It was assumed that 48% mask-wearing adherence in the baseline settings and 55% mask-wearing adherence in some scenarios of the counterfactual analysis were used.

This paper is structured as follows. Section 1 presents the background and literature review for this study. In Section 2, methodology and formations are provided. Section 3 shows the three types of simulation results, including calibration and parameter estimation, counterfactual analysis, and trend analysis under different containment strategies. Section 4 concludes this study with a discussion of the results. A statement on the limitations and opportunities for future research is also discussed.

2 Materials and methodology

2.1 Context and design for the study

This study intends to investigate the COVID-19 spread dynamics from the emergence of the first case till Nov 5, 2021 (by doing counterfactual analysis) and to predict the transmission trend of the disease till Nov 5, 2022, under different possible containment and control measures in Ethiopia. We built a model in this research attempting to capture factors including COVID-19 characteristics (infectivity, incubation period, fractions of symptomatic and asymptomatic infections, rate of severe case), major social events (religious events, war in Tigray, election campaign), the stringency of containment measures (social distancing measures, adherence level of face mask-wearing), and vaccine (efficacy, supply, and administration capacity, immunological period). The model considers the 114.9 million total population of Ethiopia.

2.2 Model structure and formulations

In this study, we built a system dynamics model with the first graph (Fig 1) demonstrating a causal loop diagram that illustrates the feedback loops and causal process tracing of variables and the second graph showing main stock flow variables (Fig 2). Fig B1 in Appendix B (S1 Appendix) also provides a detailed view of the model structure. Built on previous literature, this SD model is an extended SEIR compartmental model which includes state variables as S(t)-susceptible population, E(t)-exposed population, SYC(t)-symptomatic patients, AYC(t)-asymptomatic patients, CM(t)-confirmed mild cases, SC(t)-severe cases, D(t)-deaths, and R(t)-recovered cases. The model also captures structures related to untreated cases and vaccination, which has UCM(t)-untreated mild cases, USC(t)-severe untreated cases, UR(t)-untreated recovered cases, UD(t)-untreated deaths, SV1(t)-population administered 1st dose vaccine, SV2(t)-population administered 2nd dose vaccine, TA1(t)-total available 1st dose vaccine reserve, and TA2(t) -total available 2nd dose vaccine reserve.

Fig 1

Causal loop diagram for the system dynamics model.

Fig 2

Simplified stock flow diagram for the Ethiopia SEIR model.

2.3 Data collection and quality

The COVID-19-related infections, recovery, deaths, and vaccination data were obtained from online published reports of the World Health Organization and Ministry of Health-Ethiopia [1, 4] (https://covid19.who.int/region/afro/country/et; https://www.trade.gov/country-commercial-guides/ethiopia-healthcare; https://covid19.healthdata.org/ethiopia?view=resource-use&tab=trend&resource=all_resources). Data and parameters relate to average contact rate [25], the infectivity of both symptomatic and asymptomatic cases [97, 98], the ratio of symptomatic cases [99, 100], incubation period [101, 102], and time from severe symptomatic to recovery (treatment) [99, 103] were all retrieved from literature and confirmed in calibration. As for the setting and calibrations for other parameters, please refer to Table 1 for details.

Table 1

Model calibration results.

Variable	Variable name	Initial value	Data sources	Units
S(t)	Susceptible population	1.14964e+008	Empirical data	Person
E(t)	Exposed population	0	Empirical data	Person
SYC(t)	symptomatic case	1	Empirical data	Person
AYC(t)	asymptomatic case	0	Empirical data	Person
CM(t)	confirmed mild case	0	Empirical data	Person
SC(t)	severe case	0	Empirical data	Person
D(t)	Death	0	Empirical data	Person
R(t)	recovered population	0	Empirical data	Person
UCM(t)	Untreated confirmed mild case	0	Empirical data	Person
USC(t)	Untreated severe case	0	Empirical data	Person
UD(t)	Untreated death population	0	Empirical data	Person
UR(t)	Untreated recovered population	0	Empirical data	Person
SV1(t)	Successfully Vaccinated population with 1st Dose	0	Empirical data	Person
SV2(t)	Additional successfully vaccinated population with 2nd dose	0	Empirical data	Person
TA1(t)	Total Available 1st Dose Vaccine Reserve	0	Empirical data	dose
TA2(t)	Total Available 2nd Dose Vaccine Reserve	0	Empirical data	dose
Q	national hospital capacity (CoVID-19)	20000	Empirical data	bed
Φ	Total population	1.14964e+008	Empirical data	Person
α	contact rate per day	Appendix 1	[25] and calibrated	Person/day
β 1	Infectivity of symptomatic cases	Appendix 1	[97, 98] and calibrated	Dmnl
β 2	infectivity of asymptomatic cases	Appendix 1	[97, 98] and calibrated	Dmnl
γ	ratio from Exposed to symptomatic	0.95	[99, 100] and calibrated	Dmnl
δ	incubation period	5.2	[101, 102]	day
δ1	time from asymptomatic to symptomatic	14–5.2 = 8.8	[99]	day
υ1	diagnosis time	0.321721	calibrated	day
υ2	time from symptomatic cases to untreated mild cases	3	calibrated	day
ρ1	time from confirmed mild symptomatic to severe symptomatic (treatment)	3	calibrated	day
σ1	recovery time from mild to recovery (treatment)	15	calibrated	day
τ1	ratio from mild to severe	0.0117848	calibrated	Dmnl
φ1	time from severe symptomatic to recovery (treatment)	Appendix 1	[99, 103] and calibrated	day
ω1	time from severe symptomatic to death (treatment)	Appendix 1	calibrated	day
θ1	death ratio (treatment)	Appendix 1	calibrated	Dmnl
ρ2	time from mild symptomatic to severe case (untreated)	7.28445	calibrated	day
σ2	the recovery period from untreated mild cases	17	calibrated	day
τ2	ratio from untreated mild symptomatic to severe symptomatic	0.0919931	calibrated	Dmnl
φ2	The recovery period from untreated severe cases	Appendix 1	calibrated	day
ω2	The period from untreated severe Symptomatic to death (untreated)	Appendix 1	calibrated	day
θ2	death ratio without treatment	Appendix 1	calibrated	Dmnl
λ1	supply of 1st dose vaccine from various sources	Table 1 Sheet 1	Empirical data	dose
λ2	supply of 2nd dose vaccine from various sources	Table 1 Sheet 1	Empirical data	dose
ε	average immunological memory period	240	Assumption	day
κ1	actual administering capacity of 1st dose vaccine per day	Table 1 Sheet 2	Empirical data	dose
κ2	actual administering capacity of 2nd dose vaccine per day	Table 1 Sheet 2	Empirical data	dose
η1	effectiveness of 1st dose	0.684	Assumption	Dmnl
η2	effectiveness of 2nd dose	0.8	Assumption	Dmnl
q	Delay time in administering 2nd dose vaccine	14	Assumption	day
ζ	Time 1	1	calibrated	day

3 Results

3.1 Model calibration

The model was calibrated against official data obtained from WHO and the Ministry of Health-Ethiopia by covering the period from March 13th, 2020, to November 5th, 2021, which is a total of 603 days (refer to Fig 3) [1]. The estimated parameters listed in Table 1 were aligned with references and expert inputs.

Fig 3

Baseline results by calibrating the model against historical data.

Based on the calibrated model, simulation scenarios in the following section were designed by assuming different values in relevant parameters-i.e., medical resources (mainly considering beds), level of mask-wearing adherence, vaccine supply, doses of vaccine administered, efficacies of 1st and 2nd dose of vaccine, immunological period, average contact rates under the circumstance of having social events and imposition of social distancing measures (SDMs). The level of mask-wearing adherence was also used to calculate weighted symptomatic and asymptomatic infectivity (refer to Table 2). Counterfactual analysis and future transmission trend analysis were then conducted to gain a better understanding of the impacts of different interventions to inform policy development and decision-making.

3.2 Counterfactual analysis

The counterfactual analysis covered the period from March 13th, 2020, to November 5th, 2021. In Fig 4, for all four scenarios 1, 4, 9, and 12, the most significant impacts were from the measure of imposing social distancing on routine activities, which could have reduced the cumulative infections from more than 360,000 to less than 240,000. Under the same circumstance, the reduction in cumulative death (40.7%) was more significant than cumulative infections (34.75%). Doubling hospital capacity while keeping all other factors constant could slightly reduce the number of cumulative infections (1.2%, S1 vs. S9). It was because the average number of infections per day did not exceed the hospital capacity, and not all COVID-19 patients could be admitted to the hospital given the accessibility issues of hospitals in some regions, where the impact was a lot smaller than that of the measures of implementing social distancing (34.7% reduction, S1 vs. historical data). Due to the meager vaccination rate (as of November 6, 2021, with 6.07% partially vaccinated and 1.23% fully vaccinated) [4], the reduced efficacy of COVID-19 (S1 vs. S4) could only slightly increase the incidence of infections by 0.088%. Doubling hospital capacity could reduce cumulative deaths by 1.8%, slightly larger than the reduction in accumulative infections.

Fig 4

Impacts of hospital capacity, vaccine efficacy, and social distancing in counterfactual analysis.

In Fig 5, lowering the contact rate in different social events (e.g., reducing the scale of those events by canceling the events, or emphasizing social distancing during social events) while holding other factors unchanged could achieve a significant reduction in the cumulative infections (89.01%, S2 vs. Historical data). Due to the significant reduction in cumulative infections, increasing hospital capacity did not reduce cumulative infections and deaths (S2 vs. S10). Similarly, due to the low vaccination rate, the reduced efficacy of the vaccine increased the number of infections by 25 (S2 vs. S5).

Fig 5

Impacts of enforced social distancing for social events in counterfactual analysis.

In Fig 6, this group of scenarios demonstrates the impacts of enforcing social distancing for main social events and routine life activities. The results showed not much different from those of values in Fig 5, which obtained an 89.35% of reduction in cumulative infections (S3 vs. Historical data). Other impacts followed the pattern in Fig 5.

Fig 6

Impacts of enforced social distancing for both social events and routine life activities in counterfactual analysis.

In Fig 7, this group of scenarios assumed that the adherence level to wearing face masks was 55% since the first COVID-19 infections in Ethiopia. The results revealed a considerable reduction in the cumulative infections, less than 1500. Furthermore, the cumulative deaths dropped to 20 in this setting.

Fig 7

Impacts of increasing face mask-wearing in counterfactual analysis.

In Fig 8, this group of scenarios compared the impacts of enforcing social distancing on either main social events or routine life activities and their combination (S1, S2, S3). Fig 8 shows that enforcing social distancing on major social events can achieve much more impact in reducing infections than enforcing the same containment measures on routine life activities (S1: S2: S3 = 34.75%: 89.01%:89.35%).

Fig 8

Impacts of imposing social distancing on major social events and routine life activities in counterfactual analysis.

3.3 COVID-19 transmission dynamics in Ethiopia under different scenarios

In evaluating the impacts of different possible future scenarios on COVID-19 spread in Ethiopia in the coming period, we proposed and simulated 264 scenarios (Refer to Tables C1 through C7 in Appendix C (S1 Appendix)). Factors considered in the scenario analysis include hospital capacity (doubling, 80%, and 60% of current capacity), vaccine administration and supply (as is-same as what had been administering; 1st dose 20% population, 2nd dose 10% population; 1st dose 30% population, 2nd dose 20% population), vaccine efficacy (1st dose 68.4%, 2nd dose 80%; 1st dose 40%, 2nd dose 60%), immunological period of vaccine, average contact rates for social events and everyday activities, and adherence level of mask-wearing (48%, 60%, and 70%). The prediction simulation periods covered from November 6th, 2022, to November 5th, 2022. For the sake of increasing visualization effects for comparison among different scenarios, we set the bound minimums of cumulative infections and cumulative deaths at 360,000 and 6000, respectively, because the calibrated data for corresponding variables (i.e., cumulative infections and cumulative deaths) were the same for all scenarios below the chosen minimums.

3.3.1 Impacts of increasing mask-wearing adherence level

Fig 9 illustrates the impacts of vaccine efficacy and adherence level of mask-wearing on the COVID-19 spread over the next year. With different adherence levels for mask-wearing, while keeping other factors fixed, the cumulative infections at the end of the simulation period (i.e., Nov 5, 2022) could range from 393,304 (adherence level 70%)(S41) to 661,696 (adherence level 48%). The cumulative deaths were 15,697 in S17 (mask-wearing 48%) and 7867 in S41 (mask-wearing70%), which led to a reduction of 49.88%. As for the impact of efficacy, the higher the mask-wearing adherence level, the lower the impact of reduced vaccine efficacy on the increase of cumulative infections was, which ranged from 0.013% to 0.96%.

Fig 9

Impacts of increasing adherence level of mask-wearing in scenario analysis.

3.3.2 Impacts of the immunological period under low vaccination conditions

Fig 10 demonstrates the impacts of the immunological period on cumulative infections and cumulative deaths. Similar to the pattern in Fig 9, the higher the mask-wearing adherence level, the lower the impact of the reduced immunological period on the increase of accumulative infections was, which ranged from 0.01% to 0.84%.

Fig 10

Impacts of the immunological period on cumulative infection and cumulative deaths in scenario analysis (48% mask-wearing).

3.3.3 Impacts of increasing stringencies of social distancing measures

Under the assumption of a 48% mask-wearing adherence level, this group of scenarios compares the impacts of different stringency levels of SDMs on the spread dynamics of COVID-19 (refer to Fig 11). With the most stringent measure in SDMs, the cumulative infections could reach 424,681 at the end of the simulation (S25), which was 31,377 more than what could have been achieved in the case of 70% mask-wearing adherence level while having laxly implemented SDMs (S41). The cumulative deaths were 8823 in S25 (the most stringent measure in SDMs), which had a 43.79% reduction compared with S17.

Fig 11

Impacts of increasing stringency of social distancing measures in scenario analysis.

3.3.4 Impacts of hospital capacity

This scenario group shows the impacts of different hospital capacities on the cumulative infections and cumulative deaths over time. Since the population size is very large and the results were very sensitive to the change of some conditions in our assumptions, the cumulative infections (S173 & S245) would exceed the number of total populations (because of repetitive infections) if the hospital capacity were smaller than 60% of the current level. These very large results just provide a reference here, which will not be used to compare with other variables. Changes in hospital capacity did not cause different accumulative infections (S17, S53, S137) as long as the hospital capacity was larger than 80% of the current one since the daily infections do not exceed hospital capacities. Compared to S137, the reductions in cumulative infections and cumulative deaths in S209 (11.89% and 13.88%) caused by increasing vaccine supply and vaccination capacity were noticeable.

3.3.5 Impacts of combined NPIs and PIs

Figs 12–14 show the impacts of the immunological period and vaccine efficacy under different mask-wearing adherence levels and levels of vaccination across the nation on the spread of COVID-19 in Ethiopia, which were calculated by taking the difference for the cumulative infections and cumulative deaths between two corresponding scenarios (e.g., S68-67). By comparing Figs 12 and 13, we can see that reduced vaccine efficacy produced larger effects than the lowered immunological period in causing increased cumulative infections and cumulative deaths. The effect was especially obvious for the situation of higher vaccination rate (i.e., 30% population vaccinated 1st dose, 20% population vaccinated 2nd dose) and low mask-wearing adherence level (48%), which had the ratio of (S102-S101)/S101:(S103-S101)/S101 = 3.99%:1.61% (21181 vs. 8556 in number). Furthermore, in the situation mentioned above, the combined effects of both reduced immunological period and reduced vaccine efficacy could cause 5.6% (29737) more infections (Fig 14). However, the adverse effects of reduced immunological period and vaccine efficacy were to a large extent offset by increased mask-wearing adherence level, which is changed to the ratio of (S126-S125)/S101:(S127-S125)/S101 = 0.09%:0.025% (352 vs. 96 in number) with mask-wearing adherence level of 70%.

Fig 12

Impacts of the immunological period on COVID-19 considering mask-wearing adherence and increasing vaccination capacity.

Fig 14

Impacts of immunological period and vaccine efficacy on COVID-19 considering mask-wearing adherence and increasing vaccination capacity.

Fig 13

Impacts of vaccine efficacy on COVID-19 spread considering mask-wearing adherence and increasing vaccination capacity.

In Fig 15, S133 is the scenario having minimum cumulative infections and cumulative deaths (384094 and 7586) among all scenarios with different assumed interventions, where the interventions include: 30% and 20% of the total population being administered 1st and 2nd doses of COVID-19 vaccine, respectively, keeping 70% mask-wearing adherence level, adopting the most stringent SDMs, and holding vaccine efficacy and immunological period and hospital capacity unchanged. The maximum cumulative infections and cumulative deaths (680231 and 16235) happened in S139 where the interventions contain: reducing hospital capacity to 80% of the original level, laxly implementing SDMs, assuming vaccine supplying and administering pace calculated in historical data, keeping 48% mask-wearing adherence level, and adopting reduced vaccine efficacy (from 68.4% and 80% to 40% and 60% for the 1st and 2nd doses vaccine, respectively), and shortened immunological period (from 240 to 180 days). The maximum and minimum values differences were 77.10% and 114% of minimum cumulative infections and cumulative deaths, respectively. Scenarios S21, S141, and S177 represented the medium level of cumulative infections and cumulative deaths (482358 and 10549, please refer to C1, C4, and C5 for their corresponding settings.).

Fig 15

The best-medium-worst outcomes among the defined scenarios.

3.3.6 Offsetting effects among multiple interventions

In Fig 16, the results show that moderately increasing mask-wearing adherence level in the case of lax implementation of SDMs can achieve approximately the same effect (S25 vs. S29 and S37 vs. S41) as the situation of implementing stringent SDMs while having a low adherence level of mask-wearing (or due to insufficient supply of mask).

Fig 16

Offsetting less stringent social distancing measures with increasing mask-wearing adherence level.

Fig 17, Fig 18, Fig 19 show the impacts of increasing vaccination and implementing SDMs on the cumulative infections and cumulative deaths under three mask-wearing adherence levels-i.e., 48%, 60%, and 70%. The impact of increasing vaccination to the extent of administering 1st dose for 30% population and 2nd dose for 20% of the population cannot achieve the effect that is achieved by increasing mask-wearing adherence level to 70% (showed in Fig 16). Furthermore, the effect was smaller than that of implementing more stringent SDMs. The impacts of increasing vaccination were lessened with more stringent SDMs and increasing the adherence level of mask-wearing because the cumulative infections and cumulative deaths were significantly reduced.

Fig 17

Impacts of changing vaccination levels and implementing social distancing measures with mask-wearing adherence level of 48%.

Fig 18

Impacts of changing vaccination levels and implementing social distancing measures with mask-wearing adherence level of 60%.

Fig 19

Impacts of changing vaccination levels and implementing social distancing measures with mask-wearing adherence level of 70%.

Fig 20, Fig 21, Fig 22 show the effect comparisons between implementing stringent SDMs and the scenarios of combining less stringent SDMs while increasing vaccination levels in the whole population of Ethiopia under different mask-wearing adherence levels-i.e., 48%, 60%, and 70%. As in the case of mask-wearing adherence level of 48%, implementing the most stringent SDMs (S25) had cumulative infections of 424,681 and cumulative deaths of 8,823. And the scenario of vaccinating 30% and 20% total population with 1st and 2nd doses, respectively, and implementing the least stringent SDMs (for both major events and routine activities) (S101) had cumulative infections of 531083 and cumulative deaths of 12027, which were 25.05% and 36.31% more than of the S25. When considering the factor of reduced vaccine efficacy and immunological period, the differences were enlarged to be 32.13 and 46.07%, respectively. When the vaccination coverage was reduced to administering 1st and 2nd doses for 20% and 10% of the total population, respectively (S65), the cumulative infections and cumulative deaths of S65 were 37.29% and 53.21% more than that of S25. When considering the reductions in vaccine efficacy and immunological period, the differences were enlarged to 44.27% and 62.54% (S68 vs. S25). Under the assumption of a mask-wearing adherence level of 70%, implementing the most stringent SDMs (S49) had cumulative infections of 384,644 and cumulative deaths of 7,062. In comparison, the scenario of vaccinating 30% and 20% total population with 1st and 2nd doses, respectively, and implementing the least stringent SDMs (S125) had the cumulative infections of 391,735 and cumulative deaths of 7,819, whereas the latter (S125) were 1.84% and 2.86% more than of the former (S49). As for the vaccination cover of 20% and 10% of the total population having 1st and 2nd doses, respectively (S89), the cumulative infections and cumulative deaths of S89 were 2.04% and 3.17% more than that of S49. Our study did not propose a higher vaccination level given that the issues in vaccine acquisition in Ethiopia cannot be overcome shortly.

Fig 20

Implementing stringent SDMs versus less stringent SDMs and increased vaccination levels under the mask-wearing adherence level of 48%.

Fig 21

Implementing stringent SDMs versus less stringent SDMs and increased vaccination levels under the mask-wearing adherence level of 60%.

Fig 22

Implementing stringent SDMs versus less stringent SDMs and increased vaccination levels under the mask-wearing adherence level of 70%.

In Fig 23, by comparing scenarios S37, S89, and S125, we found that increasing the vaccination rate and adherence level of mask-wearing under the circumstance of less stringent SDMs can achieve similar control results as that of the measure of lower-level vaccination rate and mask-wearing adherence with more stringent SDMs. Comparing scenarios S25 and S77 can further approve the observation mentioned above.

Fig 23

Increasing vaccination levels and mask-wearing adherence as compensation for the less stringent SDMs.

4 Discussion and conclusion

This SD model served as a framework for understanding the issues and gaps in the containment measures against COVID-19 in the past period and the spread dynamics of the infectious disease over the next year under different interventions and their combinations. This study simulated the results of 280 scenarios considering the vaccination level, efficacies of both 1st and 2nd vaccines, the immunological period for both vaccinated people and infected patients, stringency levels in enforcing social distancing, adherence level of face mask-wearing, and hospital capacity. With evaluations and comparisons for alternative interventions, the results can inform policy and implementation science regarding the path, scale, duration, and stringency level of different interventions and their optimal combinations.

The fast spread of COVID-19 in Ethiopia can be attributed to joint effects of multiple factors including but are not limited to: (1) low adherence to face mask-wearing and hand hygiene practices; (2) failure to comply with social distancing; (3) most religious people attending more than three days a week in churches to pray together; (4) special religion holidays celebration events were held for more than 7 days across the country by Orthodox Christian; (5) more than three holidays events by Muslim followers; (6) extensive protests in a different state in 2020 during a pandemic; (7) massive rally during the country election on 21 June 2021 [6, 8–10]; and (8) the war between the federal government and Tigray Region since November 4, 2020, which not only increased COVID-19 infection risk of fighters and almost two million displaced migrant but also disrupted the vaccine supply and administration and everyday operations of many hospitals [11].

In the counterfactual analysis, we found that keeping high mask-wearing adherence since the outbreak of COVID-19 in Ethiopia could have significantly reduced the infection under the condition of low vaccination level or unavailability of vaccine (Fig 7). In the trend analysis, higher mask-wearing adherence still played the dominant role in significantly reducing infection, and the best outcome was attained with one condition being a 70% adherence level. In terms of the social distancing measures, reducing or canceling major social events (e.g., religious gatherings and protests) can achieve a better outcome than imposing the same constraints on people’s routine life activities. Moreover, since the daily infections did not exceed hospital treatment capacity in the early stage of the COVID-19 pandemic, it seemed that increasing hospital capacity did not play a pivotal role in significantly reducing infections. However, the significant contributions of hospitalizing the COVID-19 infected patient were to reduce the death toll caused by severe cases and carry out necessary quarantine functions, which significantly reduced the sources of further transmission [104, 105]. This can be testified in Fig 24, where the insufficient capacity for hospitals (including quarantine hospitals) could lead to a considerable number of accumulative infections.

Fig 24

Impacts of hospital capacity in scenario analysis.

In the trend analysis, under the conditions of low vaccination rate, reduced vaccine efficacy, and immunological period had no apparent effects on accumulative infections. Furthermore, increasing mask-wearing adherence and enforcing more stringent social distancing were two main measures that could significantly reduce possible infections. Higher mask-wearing adherence had more significant impacts than enforcing social distancing measures in our settings. As the vaccination rate increases, factors such as vaccine efficacy and immunological period that affect vaccination effectiveness started to take effect, where the reduced efficacy could cause more infections than that of shortened immunological periods. Offsetting effects of multiple interventions (strengthening one or more interventions while loosening others) could be applied when the levels or stringencies of one or more interventions needed to be adjusted for catering to particular needs (e.g., less stringent SDMs to reboot the economy or cushion insufficient resources in some areas).

Like all other research, the current study has certain limitations. First of all, since the number of possible scenarios by combining multiple interventions and their different levels can be very large, this paper only picked and analyzed minimal representative scenarios. Secondly, data used in the simulation were at the aggregate level without incorporating regional heterogeneity. Moreover, the ever-mutating COVID-19 virus means that current analysis might not be able to reflect future pandemic dynamics.

Several possible studies can be conducted in the future research: (1) the model can be easily extended to evaluate impacts of containment measures for emerging infectious diseases; (2) by adding a demand-supply interaction substructure for personal protection equipment (PPE) such as face mask, the revised model can be used to evaluate impact of PPE logistics on the spread dynamics of COVID-19 (considering adherence level and protection ability); (3) this model can help find the optimal combination of containment measures; (4) the model can be expanded to inform decision makers of determining optimal lock-down window and period by jointly considering the supply & demand dynamics of PPE and vaccine, vaccine efficacy and immunological period, effect of social distancing measures, and characteristics of the mutants of the disease; (5) a revised model can also help undertake hospital capacity planning for dealing with public health emergency like COVID-19 pandemic; and (6) the model can also be used to define the necessary vaccination level and consequently the immune barrier in a given period by considering vaccine efficacy and immunological period for both vaccinated and infected people.

In conclusion, SD can be a handy tool for expedited learning for designing and implementing public health emergency policies (or interventions), especially those involving multiple interventions that could have thousands of possible implementation paths. It can help pinpoint issues and gaps in the historical path and investigate and choose the appropriate path for achieving better containment and control outcomes given limited resources, complicated socioeconomic systems, and characteristics of the emerging infectious disease. This evaluation of the NPIs and PIs will help provide constructive inputs to inform policy and decision-making regarding COVID-19 and other emerging infectious diseases.

(DOCX)

Click here for additional data file.

(ZIP)

Click here for additional data file.

18 Apr 2022

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Vinícius Silva Belo

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf  and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. In your Data Availability statement, you have not specified where the minimal data set underlying the results described in your manuscript can be found. PLOS defines a study's minimal data set as the underlying data used to reach the conclusions drawn in the manuscript and any additional data required to replicate the reported study findings in their entirety. All PLOS journals require that the minimal data set be made fully available. For more information about our data policy, please see http://journals.plos.org/plosone/s/data-availability.

Upon re-submitting your revised manuscript, please upload your study’s minimal underlying data set as either Supporting Information files or to a stable, public repository and include the relevant URLs, DOIs, or accession numbers within your revised cover letter. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories. Any potentially identifying patient information must be fully anonymized.

Important: If there are ethical or legal restrictions to sharing your data publicly, please explain these restrictions in detail. Please see our guidelines for more information on what we consider unacceptable restrictions to publicly sharing data: http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions. Note that it is not acceptable for the authors to be the sole named individuals responsible for ensuring data access.

We will update your Data Availability statement to reflect the information you provide in your cover letter.

3. We note that the grant information you provided in the ‘Funding Information’ and ‘Financial Disclosure’ sections do not match.

When you resubmit, please ensure that you provide the correct grant numbers for the awards you received for your study in the ‘Funding Information’ section.

4. Thank you for stating the following in your Competing Interests section:

“The authors declare no conflict of interests”

Please complete your Competing Interests on the online submission form to state any Competing Interests. If you have no competing interests, please state "The authors have declared that no competing interests exist.", as detailed online in our guide for authors at http://journals.plos.org/plosone/s/submit-now

This information should be included in your cover letter; we will change the online submission form on your behalf.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: No

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: No

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: Comments to Authors

Currently, COVID-19 is a major public health issue and assessing the dynamic impacts of non-pharmaceutical and pharmaceutical intervention measures against COVID-19 is an important area of study. The authors try to investigate system dynamics model to capture COVID-19 characteristics, major social events, stringencies of containment measures, and vaccine dynamics. To improve the quality of the paper, I suggest to address the following issues:

Abstract

- Avoid abbreviation from the abstract. i.e. SD, SDM

Background

- The background section didn’t intensively explore literatures related with this study.

- Many statements written without citing references that support the statement. Such us line 64-66, line 72-78, Line

- This section didn’t motivate why doing this study is important and what are the limitation of other similar studies, i.e. Ejigu et al (2021)

- Need proof reading. There are many grammatical issues across the paper.

i.e., line 61 “only were 1.4%....”.

line 161 “This study is to better understand…”

Method

- Line 161 is not a meaningful sentence. “This study is to better understand…”

- Motivation to the considered “Model structure and formulations” were not clearly stated. Under what assumption this model is valid? How you get the model parameter input parameters?

- Add links/hyperlinks (cite websites) from where you got the data under the “Data collection and quality” subsection.

- Further in this section, you need to explain

a) How the Immunological period (days) 240 and 180 determined.

b) How the vaccine efficacy determined

c) Briefly explain how other parameters also fixed

-

Result and discussion

- Many of the assumptions set for the scenarios are far away the reality. For example, line 272-273, it was stated that “.. adherence level in wearing face was assumed 55%” which is really far away from the reality in the context of Ethiopia. As far as I know, 75% of the population in Ethiopia are living in rural areas and majority of them didn’t her about facemask. Further, in line 295 “…. adherence level of mask-wearing (48%, 60%, and 70%).”

- Revise the assumed adherence level related with face masking, social distancing,… in the context of Ethiopia.

- Line 297 “…total of 968 days”, but in line 205 “… total of 603 days” which one is correct.

- Finding of other studies related with this work were not well discussed.

- Limitation of the study not clearly stated. i.e., the quality of the data and others not clearly addressed.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

Submitted filename: Comments to authors.pdf

Click here for additional data file.

13 Jun 2022

First of all, we would like to sincerely thank the editor and anonymous reviewers for their comments and constructive critiques. We appreciate the opportunity to address the comments by submitting a revision. We have considered all comments with care, and we believe the revisions address these comments and have greatly improved the paper. In this revision report, we explain the responses to each comment in detail. We have reproduced the original comments in black and typed our responses in blue for clarity.

Responses to Reviewer #1

1. - Avoid abbreviation from the abstract. i.e. SD, SDM

Response:

Thank you for mentioning this point. We have changed those abbreviations in the abstract to their full names.

2. -The background section didn’t intensively explore literatures related with this study.

Response:

We appreciate that you pointed this out. We further searched relevant studies and summarized them in the revised manuscript. Relevant references are listed below.

References:

1. Ayele W, Tesfaye L, Abagero A, Taye G, Abate B, Habtamu T, et al. COVID 19 Epidemic trajectory modeling results for Ethiopia. Ethiop J Heal Dev. 2021;35: 25–32.

2. Ayele TA, Fentahun N, Tiruneh SA, Haile M, Zeru T, Meseret G, et al. Spatial variation and level of compliance on COVID-19 Prevention strategies in Amhara region, Ethiopia: Observational survey. Ethiop J Heal Dev. 2021;35.

3. Endriyas M, Kawza A, Alano A, Hussen M, Shibru E. COVID-19 prevention practices in urban setting during early introduction of the disease: results from community survey in SNNP Region, Ethiopia. BMJ Open. 2021;11. doi:10.1136/bmjopen-2020-047373

4. Haftom M, Petrucka PM. Determinants of Face Mask Utilization to Prevent Covid-19 Pandemic among Quarantined Adults in Tigrai Region, Northern Ethiopia, 2020. Clin Nurs Res. 2021;30: 1107–1112. doi:10.1177/10547738211013219

5. Tucho GT, Kumsa DM. Universal Use of Face Masks and Related Challenges During COVID-19 in developing countries. Risk Manag Healthc Policy. 2021;14: 511–517. doi:10.2147/RMHP.S298687

6. Zewude B, Melese B, Addis E, Solomon W. Changing patterns of compliance with protective behavioral recommendations in the post first-round COVID-19 vaccine period among healthcare workers in Southern Ethiopia. Risk Manag Healthc Policy. 2021;14: 3575–3587. doi:10.2147/RMHP.S325699

7. Bushira KM. Modeling the effectiveness of social distancing interventions on the epidemic curve of coronavirus disease in Ethiopia. Model earth Syst Environ. 2021; 1–11. doi:10.1007/s40808-021-01190-9

8. Deressa W, Worku A, Abebe W, et al. Social distancing and preventive practices of government employees in response to COVID-19 in Ethiopia. PLOS ONE. 2021; 16(9): e0257112:1-22. https://doi.org/10.1371/journal.pone.0257112

9. Fikrie A, Amaje E, Golicha W. Social distancing practice and associated factors in response to COVID-19 pandemic at West Guji Zone, Southern Ethiopia, 2021: A community based cross-sectional study. PLoS One. 2021;16. doi:10.1371/journal.pone.0261065

10. Hailu W, Derseh L, Hunegnaw MT, Tesfaye T, Angaw DA. Compliance, barriers, and facilitators to social distancing measures for prevention of coronavirus disease 2019 in Northwest Ethiopia, 2020. Curr Ther Res Exp. 2021;94. doi:10.1016/j.curtheres.2021.100632

11. Tolu LB, Ezeh A, Feyissa GT. How prepared is Africa for the COVID-19 pandemic response? The case of Ethiopia. Risk Manag Healthc Policy. 2020;13: 771–776. doi:10.2147/RMHP.S258273

3. - Many statements written without citing references that support the statement. Such us line 64-66, line 72-78, Line

Response:

Thank you for raising this point. We have double-checked the whole paper and added citations at relevant places.

4. - This section didn’t motivate why doing this study is important and what are the limitation of other similar studies, i.e. Ejigu et al (2021)

Response:

Thank you for the suggestions. We have added a paragraph deliberating this point.

5. - Need proof reading. There are many grammatical issues across the paper. i.e., line 61 “only were 1.4%....”. line 161 “This study is to better understand…”

Responses:

Many thanks to the reviewer for indicating this situation. We agree that improving the writing is very important for guaranteeing the quality of the manuscript. We have invited professional English-editing service to edit and polish the manuscripts.

6. - Line 161 is not a meaningful sentence. “This study is to better understand…”

Responses:

Thank you for helping improve the writing of this manuscript, we have revised this sentence and checked other places as well.

7. - Motivation to the considered “Model structure and formulations” were not clearly stated. Under what assumption this model is valid? How you get the model parameter input parameters?

Responses:

Thank you for mentioning this. We did have some elaboration on the motivation for using the extended SEIR model in “Introduction” and subsection of “Model structure and formulations” Built on previous research, our extended SEIR model tends to include not only the state variables of susceptible population, exposed population, symptomatic patients, asymptomatic patients, mild cases, severe cases, deaths, recovered cases, but also the sub-structure of hospitalized patients, un-hospitalized patients, vaccine administration.

Piecewise calibration was conducted to estimate necessary model parameters, given that the cumulative infections and cumulative deaths did not follow an expectedly exponential growth curve. Historical data were split into two parts with one part being used for building the model. And the model was then verified and validated by the other part of the data. As for the model parameters, some of the parameters were obtained through the fitting with historical and they were verified through repetitive calibrations. Besides, the value settings for other parameters were done by referring to literature, which were clearly presented in Table 1.

8. - Add links/hyperlinks (cite websites) from where you got the data under the “Data collection and quality” subsection.

Responses:

Thank you for advising on this point, we have put the link in the relevant place of the “Data collection and quality” subsection.

9. - - Further in this section, you need to explain

a) How the Immunological period (days) 240 and 180 determined.

b) How the vaccine efficacy determined

c) Briefly explain how other parameters also fixed

Responses:

Thank you for mentioning this. For a) and b), we have listed literature that indicated the source of the value ([1,2] for a) and [3-6] for b)). In this revision, we added several more references. Please also refer to the literature provided below. Regarding c), some of the parameters were obtained through the fitting with historical and they were verified through repetitive calibrations. Besides, the value settings for other parameters were done by referring to literature, which was clearly presented in Table 1.

References:

1. Khoury DS, Cromer D, Reynaldi A, et al. Neutralizing antibody levels are highly predictive of immune protection from symptomatic SARS-CoV-2 infection. Nat Med 2021;27: 1205-11

2. Tenforde MW, Self WH, Naioti EA, et al. Sustained effectiveness of Pfizer BioNTech and Moderna vaccines against COVID-19 associated hospitalizations among adults-United States, March, July 2021. Morb Mortal Wkly Rep. 2021; 70(34), 1156.

3. Bernal JL, Andrews N, Gower C, et al. Effectiveness of the Pfizer-BioNTech and Oxford-AstraZeneca vaccines on COVID-19 related symptoms, hospital admissions, and mortality in older adults in England: test negative case-control study. BMJ-British Med J. 2021; 373:n1088. https://doi.org/10.1136/bmj.n1088

4. Dagan N, Barda N, Kepten E, et al. BNT162b2 mRNA Covid-19 vaccine in a nationwide mass vaccination setting. N Engl J Med. 2021; 384(15):1412-1423. https://doi.org/10.1056/NEJMoa2101765

5. Jara A, Undurraga EA, Gonzalez C, et al. Effectiveness of an inactivated SARS-CoV-2 Vaccine in Chile. N Engl J Med. 2021; 385(10):875-884. https://doi.org/10.1056/NEJMoa2107715

6. Shekhar R, Sheikh A, Upadhyay S, et al. COVID-19 vaccine acceptance among health care workers in the United States. Vaccines. 2021;9,119. https://doi.org/10.3390/vaccines9020119

10. - - - Many of the assumptions set for the scenarios are far away the reality. For example, line 272-273, it was stated that “.. adherence level in wearing face was assumed 55%” which is really far away from the reality in the context of Ethiopia. As far as I know, 75% of the population in Ethiopia are living in rural areas and majority of them didn’t her about facemask. Further, in line 295 “…. adherence level of mask-wearing (48%, 60%, and 70%).”

- Revise the assumed adherence level related with face masking, social distancing… in the context of Ethiopia.

Responses:

Thanks to the reviewer for pointing this out. We agree that the compliance level to face mask-wearing in Ethiopia was very low, especially in those poor rural areas at the outset of COVID-19. However, with the supply of face masks, the level of compliance witnessed an obvious increase, which increased from 30% to as high as 62% (IHME). We also searched the literature and found results from some surveys done by Ethiopian scholars showing that the levels of adherence (Endriyas et al., 2021; Haftom and Petrucka, 2020) were more than 50%. Ayele Wondimu et al. evaluated the spread dynamics of COVID-19 in Ethiopia under the assumptions of face mask utilization of 20%, 40%, and 60%.Therefore, the percentages used in our scenarios represented relatively reasonable assumptions.

References:

1. IHME, Mask use in Ethiopia. Accessed on June 5th, 2022 at

https://covid19.healthdata.org/ethiopia?view=mask-use&tab=trend

2. Endriyas M, Kawza A, Alano A, Hussen M, Shibru E. COVID-19 prevention practices in urban setting during early introduction of the disease: Results from community survey in SNNP Region, Ethiopia. BMJ Open. 2021;11. doi:10.1136/bmjopen-2020-047373

3. Haftom M, Petrucka PM. Determinants of Face Mask Utilization to Prevent Covid-19 Pandemic among Quarantined Adults in Tigrai Region, Northern Ethiopia, 2020. Clin Nurs Res. 2021;30: 1107–1112. doi:10.1177/10547738211013219

4. Ayele W, Tesfaye L, Abagero A, Taye G, Abate B, Habtamu T, et al. COVID 19 Epidemic trajectory modeling results for Ethiopia. Ethiop J Heal Dev. 2021;35: 25–32.

11. - Line 297 “…total of 968 days”, but in line 205 “… total of 603 days” which one is correct.

Responses:

Thank you for mentioning this point. Our study has two parts with one focusing on counterfactual analysis that has a total of 603 days and the other focusing on the prediction that covers one year (365 days). Therefore, the total simulation days considering prediction are 968 days. To reduce confusion, we have made changes at the corresponding place.

12. - Finding of other studies related with this work were not well discussed.

Responses:

Thank you so much for pointing this out. We have added relevant studies in the Introduction, Results, and Discussion parts.

13. - - Limitation of the study not clearly stated. i.e., the quality of the data and others not clearly addressed.

Responses:

Thanks for your suggestions. We have added limitations at the end of the discussion part.

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

27 Jun 2022

Assessing the dynamic impacts of non-pharmaceutical and pharmaceutical intervention measures on the containment results against COVID-19 in Ethiopia

PONE-D-22-02713R1

Dear Dr. Liu,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Vinícius Silva Belo

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

8 Jul 2022

PONE-D-22-02713R1

Assessing the dynamic impacts of non-pharmaceutical and pharmaceutical intervention measures on the containment results against COVID-19 in Ethiopia

Dear Dr. Liu:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Vinícius Silva Belo

Academic Editor

PLOS ONE