Mathematical Modeling and Analysis of COVID-19 pandemic in Nigeria

2.1 Model Formulation

The model is developed by splitting the total human population at time t, denoted by N (t), into the mutually-exclusive compartments of susceptible (S(t)), exposed (E(t)), symptomatically-infectious (I_s(t), asymptomatically-infectious (I_a(t)), hospitalized (I_h(t)) and recovered (R(t)) individuals. Thus,

It should be emphasized that the epidemiological compartment I_a also contain individuals who may show mild symptoms of the disease. Furthermore, the compartment I_h for hospitalization also includes those with clinical symptoms of COVID-19 who are self-isolating at home. The model is given by the following deterministic system of nonlinear differential equations (where a dot represents differentiation with respect to time t):

A flow diagram of the model is depicted in Figure 2. The state variables and parameters of the model are described in Tables 1 and 2, respectively.

• Download figure
• Open in new tab

Figure 2:

Flow diagram of the COVID-19 model (2.1).

In the model (2.1), the parameter β_s(β_a) represents the effective contact rate (i.e., contacts capable of leading to COVID-19 transmission) for symptomatically-infectious (asymptomatically-infectious) individuals, 0 < c_M ≤ 1 is the proportion of members of the public who wear face-masks (correctly and consistently) in public and 0 < ϵ_M ≤ 1 is the efficacy of the face-masks. This formulation (i.e., β_s ≠ β_a) allows us to account for the possible heterogeneity in the contact rates of infectious with (I_s) or without (I_a) clinical symptoms of COVID-19. The parameter σ represents the progression rate of exposed individuals (i.e., 1/σ is the intrinsic incubation period of COVID-19). A proportion, 0 <r ≤ 1, of exposed individuals show no clinical symptoms of COVID-19 (and move to the class I_a) at the end of the incubation period. The remaining proportion, 1 − r, show clinical symptoms and move to the I_s class. The parameter γ_s(γ_a)(γ_h) represents the recovery rate for individuals in the I_s(I_a)(I_h) class. Similarly, ϕ_s is the hospitalization (or self-isolation) rate of individuals with clinical symptoms of COVID-19. Finally, the parameter δ_s(δ_h) represents the COVID-induced mortality rate for individuals in the I_s(I_h) class.

It is convenient, for housekeeping purposes, to define the state variable D(t) for the number of COVID-deceased individuals, to keep track of COVID-19 related deaths. It then follows from some of the equations of the model (2.1) that the rate of change of the population of COVID-deceased individuals is given by:

2.2 Data Fitting and Parameter Estimation

Estimates for some of the parameters of the model (2.1) are available from the literature (as indicated in Table 4). Other parameters, such as the community contact rate parameters (β_a and β_s), the hospitalization rate (ϕ_s) and the COVID-induced mortality parameters (δ_h and δ_s), are unknown. Consequently, we use the data published by the NCDC [9] to estimate these unknown parameters for the entire Nigerian nation, the States of Kano and Lagos and the FCT. Furthermore, we estimate the respective basic reproduction number (ℛ₀) for each of the four communities. In particular, we fit the cumulative number of deaths generated from the model (2.1) to the observed cumulative COVID-19 mortality data for the four communities reported by the NCDC [9]. The data-fitting process involves implementing the Markov Chain Monte Carlo (MCMC) sampling method in the statistical software package R. This process produces 50,000 different parameter sets that are now used with a nonlinear least squares method to estimate the values of the aforementioned unknown parameters as well as the associated 95% confidence intervals (CIs). The nonlinear least squares method allows the determination of the set of parameters that minimizes the sum of the squares of the differences between the cumulative deaths predicted by the model (2.1) and the observed cumulative deaths (for the entire Nigerian nation, the two states of Kano and Lagos, and the Federal Capital Territory, Abuja). The estimated values of the unknown parameters and the basic reproduction numbers, and their 95% confidence intervals, are tabulated in Table 3. Figure 3 depicts the fitting of the observed and predicted cumulative deaths for the four communities.

• Download figure
• Open in new tab

Figure 3:

Data fitting of the model (2.1) using COVID-19 cumulative case data for (a) Nigeria (b) Lagos State (c) Kano State (d) FCT Abuja. Cumulative case data obtained from NCDC [9].

3.1 Analytical Results: Asymptotic Stability Analysis of Continuum of Disease-free Equilibria

The model (2.1) has a continuum of disease-free equilibria (DFE), given by where N (0) is the initial total sizes of the population and 0 ≤ R^∗ ≤ N (0). The next generation operator method [25, 26] can be used to analyse the asymptotic stability property of the DFE. In particular, using the notation in [25], it follows that the associated next generation matrices, F and V, for the new infection terms and the transition terms, are given, respectively, by and,

It is convenient to define the quantity ℛ_c by: where ρ is the spectral radius,

The quantity ℛ_c is the control reproduction number of the model (2.1). It measures the average number of new COVID-19 cases generated by a typical infectious individual introduced into a population where a certain fraction is protected. It is the sum of the constituent reproduction numbers associated with the number of new cases generated by symptomatically-infectious humans and asymptomatically-infectious individuals. The reproduction number ℛ_c will now be interpreted epidemiologically below.

3.2 Computation of Final Size of the Pandemic

In this section, the final size relation for the COVID-19 pandemic will be rigorously derived. Using the notation in [28], let , and z IR₊ represent the sets of infected, susceptible and recovered components of the model. Thus,it follows from the model (2.1), that x(t)= (E(t), I_s(t), I_a(t), I_h(t))^T, y(t)= S(t) and z(t)= R(t). Further, following Arino et al. [28], let D be the m× m diagonal matrix whose diagonal entries, denoted by σ_i(i = 1, 2, …, m), are the relative susceptibilities of the corresponding susceptible class. It is convenient to define Π to be an n× n matrix with the property that the (i, j) entry represents the fraction of the j^th susceptible compartment that goes into the i^th infected compartment upon becoming infected. Let b be an n dimensional row vector of relative horizontal transmissions. Using the notation in [28], let the infection rate, λ, of the model (2.1) be represented by β. That is, λ = β(x, y, z). Define the m dimensional vector Γ = [Γ₁, Γ₂, …, Γ_m] = βbV ⁻¹ΠD [28]. It follows, in the context of the model (2.1), that

Using the above change and variables and definitions, the model (2.1) reduces to: where W is a k× n matrix with the property that the (i, j) entry represents the rate at which individuals of the j^th infected compartment transition into the recovered (i^th z) compartment upon recovery and the matrix V is as defined in Section 3.1. It is worth stating that the reproduction number (ℛ_c, of the model (2.1) (or, equivalently, (3.4)), can be recovered using the definition ℛ_c = β(0, y₀, z₀)bV ⁻¹ΠDy₀ given in Theorem 2.1 of [28] (it should be noted that this theorem also allows for recovering the local asymptotic stability result for the family of disease-free equilibria of the model (2.1), given in Section 3.1). Furthermore, the results below, for the final size relations of the model (2.1) (or, equivalently, (3.4)), can be established using Theorem 5.1 of [28].

Theorem 3.3. Consider the epidemic model (2.1) (or, equivalently, (3.4)). The final size relations are given by

It is worth mentioning that, by setting I_a(0) = 0, with S(0) > 0 and I_s(0) > 0, the final size relations, given by the inequalities in (3.5), reduce to:

4.1 Effect of Social-distancing

The effect of social-distancing on the dynamics of COVID-19 is monitored by simulating the model, using the baseline parameter values tabulated in Tables 3 and 4 (with baseline mask compliance and efficacy), for various effectiveness levels of social-distancing. For simulation purposes, the following three effectiveness levels of social-distancing are considered (it should be mentioned that, the case with no social-distancing is equivalent to using the baseline values of the community contact rate parameters, β_s and β_a, given in Table 3):

1. Mild effectiveness level of social-distancing: This entails reducing the community contact rates (β_s and β_a) by 10% from their baseline values in Table 3. That is, β_s = 0.6597 (Nigeria), 0.7648 (Lagos State), 0.1832 (Kano State) and 0.2520 (FCT Abuja) per day. Similarly, β_a = 0.0913 (Nigeria), 0.1599 (Lagos State), 1.3390 (Kano State) and 0.5913 (FCT Abuja) per day.

2. Moderate effectiveness level of social-distancing: This entails reducing the community contact rates (β_s and β_a) by 30% from their baseline values in Table 3. That is, β_s = 0.5131 (Nigeria), 0.5949 (Lagos State), 0.1425 (Kano State), 0.1960 (FCT Abuja) per day. Similarly, β_a = 0.0710 (Nigeria), 0.1244 (Lagos State), 1.0415 (Kano State) and 0.4599 (FCT Abuja) per day.

3. Strict effectiveness level of social-distancing: This entails reducing the community contact rates (β_s and β_a) by 50% from their baseline values in Table 3. That is, β_s = 0.3665 (Nigeria), 0.4249 (Lagos State), 0.1018 (Kano State), 0.1400 (FCT Abuja) per day, and β_a = 0.0507 (Nigeria), 0.0889 (Lagos State), 0.7439 (Kano State), 0.3285 (FCT Abuja) per day.

Figure 4 depicts the daily cases and cumulative COVID-19 mortality for the four communities, for various levels of social-distancing effectiveness. In particular, Figure 4 (a) (blue curve) shows that, when no social-distancing measures are implemented in the country (i.e., when we simulate the worst-case scenario), Nigeria will attain a peak in the daily number of cases (of about 4,800) around July 28, 2020. This corresponds to a peak cumulative mortality of about 225,000, which is attained by February 2021 (Figure 4 (b), blue curve). It should be mentioned that this estimate for the cumulative COVID-19 mortality for Nigeria is a lot higher than what one would expect using the current mortality figures reported by the NCDC. There are many reasons to explain this major discrepancy. First of all, our cumulative mortality projections are for the worst-case scenario where the pandemic is spreading in the country and no public health interventions (such as social-distancing, community lockdown, quarantine and isolation measures etc.) are implemented in the nation. Hence, it is certainly intuitive that, in the absence of any public health intervention measures, the COVID-19 pandemic will be spreading exponentially fast, causing such unprecedented level of cases and mortality [5, 13]. The so-called “mysterious deaths” of hundreds of people in Kano during the month April, 2020 provide some justification to this claim. Secondly, the number of COVID-19 tests carried out in Nigeria is infinitesimally small. Data from the NCDC shows that only 35, 345 COVID-19 tests have been carried out in Nigeria as of May 18, 2020 [9]. Thirdly, no real autopsy or testing of deceased individuals is (universally) carried out in Nigeria. This is particularly more pronounced in the northern part of the country (and in some other faith communities in the southern part of the country), where deceased persons are buried within hours of death (and no autopsy or testing is carried out to determine the cause of death). Taking these factors in totality, it seems reasonably plausible to assume that the COVID-19 data provided by the NCDC represents a gross under-representation of the actual data. In other words, in the absence of sizable percentage of COVID-19 diagnostic testing conducted in Nigeria, the NCDC data most likely represents a gross under-estimation of the real burden of the COVID-19 pandemic in Nigeria. Hence, the predictions of reasonably-realistic mathematical models, such as the one we formulated in this study, may represent far more reliable representation of the true burden of the pandemic than the NCDC data portrays. Another point to emphasize is the possibility that the pandemic may well be spreading in Nigeria a lot earlier than the suggested February 27, 2020 date of initial introduction (i.e., index case), as stated in Section 3.1.1.

• Download figure
• Open in new tab

Figure 4:

Simulations of the model (2.1) for various effectiveness of social-distancing control measures in the absence of mask usage. Mild social-distancing corresponds to 10% decrease in the values of β_a and β_s from their baseline values. Moderate social-distancing corresponds to 30% decrease in the values of β_a and β_s from their baseline values. Strict (high) social-distancing corresponds to 50% decrease in the values of β_a and β_s from their baseline values.(a) Daily cases for Nigeria, (b) cumulative deaths for Nigeria, (c) Daily cases for Lagos State, (d) cumulative deaths for Lagos State (e) Daily cases for Kano State, (f) cumulative deaths for Kano State, (g) Daily cases for FCT Abuja, (h) cumulative deaths for FCT Abuja. Parameter values used are as given in Tables 3 and 4.

We now simulate the model (2.1) to assess the impact of social-distancing on COVID-19 dynamics in Nigeria. Our simulations show that when mild social-distancing measures are implemented in the nation, the pandemic peak shifts to August 23, 2020, and the peak daily case decreased to about 3,900 (representing a 19% reduction from the peak daily cases corresponding to the worst-case scenario described above) (Figure 4 (a), red curve). The cumulative mortality, for this setting, reduced by 5%, to 213,500 (Figure 4 (b), red curve). Thus, while the implementation of mild social-distancing measures in Nigeria significantly decreases peak daily cases (by about 19%), it only marginally reduces COVID-19 mortality.

More dramatic reductions are observed when moderate social-distancing measures are implemented (Figures 4 (a) and (b), gold curve). In particular, the peak daily cases shifts to January 2021, with a peak daily case of 2000 (representing a whopping 58% reduction in the peak daily case, in comparison to the worst-case scenario baseline). Furthermore, the peak cumulative mortality reduced to 171,800 (attained around April 2021). This represents a 24% reduction in cumulative mortality nationwide. Thus, a dramatic reduction in COVID-19 burden will be achieved in Nigeria if moderate levels of social-distancing measures are implemented nationwide (with projected number of COVID-induced mortality not exceeding 171,800). If strict social-distancing measures were implemented (and maintained) during the early stage of the pandemic, our simulations show that the pandemic will fail to take off in the country (Figures 4 (a) and (b), purple curves…. which are superimposed with the x-axis). Thus, this study shows that COVID-19 can be effectively curtailed in Nigeria using moderate or strict social-distancing measures. In particular, COVID-19 would have had no chance to thrive in Nigeria if strict social-distancing measures were implemented early during the pandemic (and maintained for a certain period of time).

Simulations for Lagos State, for the worst-case scenario (with no social-distancing), show a peak daily case of about 2,500, attained by around August 2020 (Figure 4 (a), blue curve). This corresponds to a cumulative mortality of about 136,000, attained by May 2021 (Figure 4 (b), blue curve). Mild social-distancing decreases the peak daily cases and cumulative mortality by 24% and 7%, respectively (Figures 4 (c) and (d), red curve), in relation to the baseline worst-case scenario. Furthermore, moderate social-distancing reduces the peak daily cases and cumulative mortality by 69% and 33% (Figures 4 (c) and (d)), respectively, in comparison to the worst-case scenario. Here, too, the pandemic will fail to take off if strict social-distancing was implemented (and maintained) during the early stages of the pandemic.

For Kano State, the peak daily cases for the worst-case scenario (of about 8,800) is attained around June 1, 2020 (Figure 4 (e)). This corresponds to a peak cumulative mortality of about 276,500, attained around July 2021 (Figure 4 (f)). While mild social-distancing reduces the peak and cumulative mortality by 14% and 3%, in comparison to the worst-case scenario, the implementation of moderate social-distancing in the state reduces these numbers by 47% and 12%. The implementation of strict social-distancing in the state reduces these numbers by 84% and 42%, respectively. Our simulations show that strict social-distancing may not be enough to prevent the pandemic from taking off in Kano State.

Finally, for the Federal Capital Territory, simulations depicted in Figures 4 (g) and (f) show that the peak daily cases for the worst-case scenario (of about 3,000) is achieved around August 2020 (Figure 4 (g), blue curve). This corresponds to a cumulative mortality of 174,100, attained around October 2021 (Figure 4 (h), blue curve). When mild social-distancing measures are implemented, the peak daily cases shifts to October 2020, and 23% and 8% reductions in the peak daily cases and cumulative mortality are, respectively, achieved (Figures 4 (g) and (h), red curve). Although moderate social-distancing may not be adequate to prevent the pandemic from taking off in the FCT, the implementation of strict social-distancing from the onset of the COVID-19 pandemic would have effectively suppressed the pandemic from taking off in the FCT.

4.2 Effect of Wearing Face Masks in Public

The community-wide effect of using face masks in public is monitored by simulating the model (2.1), using the baseline parameters tabulated in Tables 3 and 4, using various levels of face mask compliance and efficacy. For these simulations, social-distancing is implemented at the baseline level (given by the baseline values of the community contact rate parameters, β_s and β_a, given in Table 3). Specifically, contour plots of the control reproduction number (ℛ_c), as a function of face mask efficacy (ε_M) and compliance (c_M), are generated for the four communities (Figure 5). These contours show a decrease in the control reproduction number with increasing face mask efficacy and compliance, as expected. In particular, Figure 5 (a) shows that although the use of face masks, with the assumed baseline face mask efficacy of 50%, in Nigeria, Lagos and Kano States can reduce the COVID-19 burden (by reducing ℛ_c), it fails to reduce ℛ_c to a value less than unity (which is needed for the elimination of the pandemic, in line with Theorem 3.1) even if everyone wears mask in public. However, COVID-19 can be eliminated in the FCT, if at least 96% of the residents of the community wear face masks in public. Thus, this study shows that COVID-19 can be effectively controlled in Nigeria using face masks of moderate efficacy (e.g., face masks with 50% efficacy) if the compliance in their usage is high enough (in the extremely high range of 80% to 95%). These face mask compliance figures (of 80% to 95%) are certainly not realistically-attainable in Nigeria, or anywhere else in the world. Hence, the control strategy based on the use of face masks (with moderate efficacy) in the public should be complemented with other non-pharmaceutical interventions, such as social-distancing and community lockdown, to achieve realistic effective control of the pandemic in Nigeria.

• Download figure
• Open in new tab

Figure 5:

Contour plots of the control reproduction number (ℛ_c) of the model (2.1), as a function of mask coverage (c_M) and mask efficacy (ε_M), for the special case of the model with no social-distancing. (a) the entire Nigerian nation. (b) Lagos State. (c) Kano State. (d) FCT Abuja. Parameter values used are as given in Tables 3 and 4.

Figure 6 depicts the cumulative mortality recorded in the four communities with various efficacy and coverage levels of face masks. For low quality face masks (such as home-made cloth masks, with estimated efficacy of about 25%), our simulations show that, for the worst-case scenario where masks are not used in public, up to 225,000 deaths can be recorded in Nigeria by February 2021 (Figure 6 (a)). This number decreases, of course, when face masks are used in public, and with increasing compliance. For instance, the cumulative mortality associated with the use of face masks of the low efficacy of 25% reduces to 210,000 (representing 7% reduction, in comparison to the worst-case scenario with no mask usage) if 50% of the Nigerian populace wear face masks in public. The reduction increases to 11% if 75% of Nigerians wear face masks in public. Similar reductions are observed in the States of Lagos (Figure 6 (b)), Kano (Figure 6 (c)) and the FCT (Figure 6 (d)). Thus, as noted by Eikenberry et al. [4] and Ngonghala et al. [5], the use of face masks in public, even those with low efficacy (such as the home-made masks with efficacy of about 25%), is useful. However, their impact in reducing the cumulative COVID-19 mortality is somewhat marginal (since it only resulted in 8% to 14% reduction in the cumulative mortality, in relation to the worst-case scenario).

• Download figure
• Open in new tab

Figure 6:

Simulations of the model (2.1), showing cumulative deaths, as a function of time, for 25% face-mask efficacy (ϵ_M) and coverage (c_M). (a) the entire Nigerian nation. (b) Lagos State. (c) Kano State. (d) FCT Abuja. Parameter values used are as given in Tables 3 and 4.

If masks of moderate efficacy are used (such as masks of efficacy of 50%), our simulations show a more dramatic reduction in the cumulative mortality in the four communities with increasing masks compliance (Figure 7). In particular, Figure 7 (a) shows that, while up to 225,000 deaths could be recorded in Nigeria by February 2021 under the worst-case scenario (Figure 7 (a), red curve), the cumulative deaths can decrease to as low as 141,000 by the same time if 75% of Nigerians wear the moderately-effective face masks in public (Figure 7 (a), green curve). This represents a whopping 37% reduction (in cumulative mortality) from the worst-case scenario baseline. For Lagos State, our simulations show that the use of the moderately-effective (i.e., 50% efficacious) face mask can lead to a reduction of up to 24% if 50% of Lagosians wear face masks in public (Figure 7 (b), purple curve). Further, there is a reduction of up to 81% if 75% of Lagosians wear the moderately-effective face mask in public (Figure 7 (b), green curve). Similar results are obtained for the COVID-19 dynamics in Kano State (Figure 7 (c)) and the FCT (Figure 7 (d)). Even more dramatic reductions in COVID-19 burden in the four communities are recorded if face masks of very high efficacy (e.g., masks of 75% efficacy, such as the N95 respirators) are used (Figure 8). Thus, in summary, these simulations show that although the use of face masks in public can significantly decrease COVID-19 burden in Nigeria (if the face mask efficacy and compliance are high enough), it may fail to eliminate the disease nationwide. However, such elimination is feasible in Kano and Lagos States, as well as in the FCT, if residents wear face masks in public even if the efficacy of the face masks, and compliance in their usage, are moderate.

• Download figure
• Open in new tab

Figure 7:

Simulations of the model (2.1), showing cumulative deaths, as a function of time, for 50% face-mask efficacy (ϵ_M) and coverage (c_M). (a) the entire Nigeria. (b) Lagos State. (c) Kano State. (d) FCT Abuja. Parameter values used are as given in Tables 3 and 4.

• Download figure
• Open in new tab

Figure 8:

Simulations of the model (2.1), showing cumulative deaths, as a function of time, for 75% face-mask efficacy (ϵ_M) and coverage (c_M). (a) the entire Nigeria. (b) Lagos State. (c) Kano State. (d) FCT Abuja. Parameter values used are as given in Tables 3 and 4.

4.3 Effect of Early Relaxation/Lifting of Community Lockdown Measures

The community-wide impact of the length (or duration) of the community lockdown measures implemented in Nigeria is monitored by simulating the model (2.1), using the baseline parameter values in Tables 3 and 4 in the absence of face mask usage in public, for various time duration of the community lockdown. It should be recalled that the index case was reported in Nigeria on February 27, 2020, and lockdown measures were initially announced in Nigeria on March 30, 2020. The purpose of these simulations is to determine when it would be safe to consider easing or lifting the lockdown measures in Nigeria. In other words, when can we consider any kind of lifting or easing of the strict lockdown currently in place without triggering the possibility of a second wave of the pandemic that may be as (or even more) devastating as the first wave. For these simulations, we consider the following three levels of lifting of lockdown measures:

1. Mild lifting (easing) of lockdown measures: this entails increasing the community contact rates (β_s and β_a) by 10% from their baseline values in Table 3.

2. Moderate lifting (easing) of lockdown measures: this entails increasing the community contact rates (β_s and β_a) by 30% from their baseline values in Table 3.

3. Full lifting (easing) of lockdown measures: this entails increasing the community contact rates (β_s and β_a) by 70% from their baseline values in Table 3.

We first simulated the case when the lockdown measures are lifted/eased after a month of their implementation. That is, we simulated the case when lockdown measures are eased/lifted on April 30, 2020. The results obtained, depicted in Figure 9, showed that the cumulative mortality increases with increasing level of the lifting/easing allowed. In particular, in the case when no social-distancing and lockdown measures are implemented in the country (i.e., when we simulate the worst-case scenario of the model with the baseline values of the community contact rates in Table 3), Nigeria will attain a peak cumulative mortality of about 225,000 by February 2021. Figure 9 (a) shows that when the lifting is mild, up to 232,400 cumulative deaths will be recorded by December 2020 (Figure 9 (a), blue curve). This exceeds the aforementioned cumulative mortality for the worst-case scenario (of 225,000 deaths). If the lifting is moderate or full, the projected cumulative deaths increased (from the worst-case scenario baseline) by 8% and 11%, respectively. This, this study shows that lifting/easing the lockdown in Nigeria after a month of its implementation will trigger a second wave that will be more devastating (8% to 11%) than the first.

• Download figure
• Open in new tab

Figure 9:

Simulations of the model (2.1), showing cumulative deaths, as a function of time in the absence of mask usage, for the entire Nigerian nation. (a) one month after implementing the lockdown on March 30, 2020 (b) two months after implementing the lockdown on March 30, 2020 (c) three months after implementing the lockdown on March 30, 2020 (d) four months after implementing the lockdown on March 30, 2020. The solid dotted vertical lines indicate the start of the social-distancing lockdown regimen (March 30, 2020), dashed vertical lines indicate when the current strict social-distancing is terminated, and the purple horizontal line segments indicate the duration of the strict social-distancing regimen. Parameter values used are as given in Tables 3 and 4.

When the lockdown measures are lifted after two months of their implementation (on May 30, 2020), our simulations show up to 231,200 cumulative deaths will be recorded by December 2020 if the lifting is mild (Figure 9 (b)). Again, this exceeds the cumulative mortality recorded under the worst-case scenario. Furthermore, if the lifting is moderate or full, the projected cumulative deaths increased by 7% and 11%, respectively. When the lockdown measures are lifted/eased after three months of implementation (on June 30, 2020), our simulations show that mild lifting will generate up to 188,500 cumulative deaths by December 2020 (Figure 9 (c)). This represent a 16% decrease in the worst-case scenario projection (of 225,000 deaths). Thus, running the lockdown measures for at least three months period will significantly lower the burden of the pandemic in Nigeria. Furthermore, if the lifting is moderate or full, the projected cumulative deaths decreased by 12% and 8%, respectively, from the worst-case scenario baseline. Finally, when the lockdown measures are lifted after four months of their implementation (on July 30, 2020), our simulations show that up to 37,500 cumulative deaths will be recorded by December 2020 if the lifting/easing is mild (Figure 9 (d)). This represents a 83% decrease from the worst-case scenario baseline. If the lifting is moderate or full, the projected cumulative deaths decreased by 79% and 75%, respectively, from the worst-case scenario baseline. Thus, this study shows that extending the lockdown to at least four months will greatly help in curtailing the COVID-19 pandemic in Nigeria. In fact, our study shows that if the lockdown measures are extended until the end of 2020, the projected cumulative deaths decreases by 100%, 96% and 92% if the lifting is mild, moderate and full. In other words, we can actually eliminate the pandemic if the lockdown measures are extended until the end of the year. Similar results are obtained for Lagos State, Kano State and the FCT (not reported here to save space). In summary, this study shows that running the lockdown measures to longer periods (at least four months) will greatly enhance the prospects of effectively curtailing or eliminated the pandemic. However, the early lifting or relation of the social-distancing and lockdown measures, in an effort to re-open the economy and the country, could trigger a second wave of the COVID-19 pandemic that may be more devastating than the first wave.