Skip Navigation
Skip to contents

PHRP : Osong Public Health and Research Perspectives

OPEN ACCESS
SEARCH
Search

Articles

Page Path
HOME > Osong Public Health Res Perspect > Volume 8(6); 2017 > Article
Original Article Effects of Timely Control Intervention on the Spread of Middle East Respiratory Syndrome Coronavirus Infection
Ilsu Choi, Dong Ho Lee, Yongkuk Kim
Osong Public Health and Research Perspectives 2017;8(6):373-376.
DOI: https://doi.org/10.24171/j.phrp.2017.8.6.03
Published online: December 31, 2017
  • 2,684 Views
  • 26 Download
  • 1 Crossref
  • 3 Scopus

aDepartment of Statistics, Chonnam National University, Gwangju, Korea

bDepartment of Mathematics, Kyungpook National University, Daegu, Korea

• Received: September 5, 2017   • Revised: November 2, 2017   • Accepted: November 9, 2017

Copyright ©2017, Korea Centers for Disease Control and Prevention

This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

  • Objectives
    The 2015 Middle East Respiratory Syndrome Coronavirus (MERS-CoV) outbreak in Korea caused major economic and social problems. The control intervention was conducted during the MERS-CoV outbreak in Korea immediately after the confirmation of the index case. This study investigates whether the early risk communication with the general public and mass media is an effective preventive strategy.
  • Methods
    The SEIR (Susceptible, Exposed, Infectious, Recovered) model with estimated parameters for the time series data of the daily MERS-CoV incidence in Korea was considered from May to December 2015. For 10,000 stochastic simulations, the SEIR model was computed using the Gillespie algorithm. Depending on the time of control intervention on the 20th, 40th, and 60th days after the identification of the index case, the box plots of MERS-CoV incidences in Korea were computed, and the results were analyzed via ANOVA.
  • Results
    The box plots showed that there was a significant difference between the non-intervention and intervention groups (the 20th day, 40th day, and 60th day groups) and seemed to show no significant difference based on the time of intervention. However, the ANOVA revealed that early intervention was a good strategy to control the disease.
  • Conclusion
    Appropriate risk communication can secure the confidence of the general public in the public health authorities.
The emergence of Middle East respiratory syndrome coronavirus (MERS-CoV) in South Korea in 2015 exerted huge social and economic tolls. Mathematical models are effective for understanding and controlling the spread of MERS-CoV, and so far, many attempts at applying mathematical models have been made to understand the MERS-CoV outbreak in Korea [19]. The control intervention was conducted during the MERS-CoV outbreak in Korea immediately after the confirmation of the index case. Using a mathematical model, we investigated whether the early risk communication with the general public and mass media is an effective preventive strategy.
The SEIR (Susceptible, Exposed, Infectious, Recovered) model with estimated parameters from the time series data on the daily incidence of MERS-CoV in Korea was considered from May to December 2015. For the 10,000 stochastic simulations, the SEIR model was computed using the Gillespie algorithm. Depending on the time of control interventions on the 20th, 40th, and 60th days since the index case was identified, the box plots of MERS-CoV incidences in Korea were computed, and then analysis of variance (ANOVA) was used to analyze the results.
1. The basic model for MERS-CoV dynamics
The following SEIR model by Lee et al. [10] that categorizes each individual into one of the six epidemiological classes was considered: susceptible (S), exposed (or high-risk latent) (E), symptomatic and infectious (I), infectious but asymptomatic (A), hospitalized (H), and recovered (R).
dSdt=-β(I+l1A+l2H)NSdEdt=β(I+l1A+l2H)NS-κEdIdt=κρE-(γa+γI)IdAdt=κ(1-ρ)E-γIAdHdt=γaI-γrHdRdt=γII+γrH+γIA
It was assumed that not only infectious and hospitalized individuals, but also asymptomatic individuals could infect others. The parameters β, l1, l2, κ, ρ, γa, γI and γr represent human-to-human transmission rate per unit time, the relative transmissibility of asymptomatic and hospitalized classes, the rate of progression from exposed class E to symptomatic I or asymptomatic infectious class A, the proportion of symptomatic infections, the hospitalization rate of symptomatic individuals, the recovery rate without being hospitalized, and the recovery rate of hospitalized patients, respectively.
2. Stochastic simulation methods
We used the Gillespie algorithm to study random interactions occurring in the given system of equations. The stochastic simulation algorithm, suggested by Gillespie [11], is as follows:
For a set of coupled ordinary differential equations
dX(t)dt=j=1Mcjaj(X(t)),
we can construct an exact numerical realization of the process X(t):
  • Step 0: Initialize the time t = t0 and the system’s state X(t0) = X0.

  • Step 1: With the system in state X at time t, evaluate all the aj(X) and their sum a0(X):=i=1Mai(X).

  • Step 2: Draw two random numbers r1 and r2 from the uniform distribution in the unit interval, and take

    τ=1a0(X)ln(1r1)
    j=the smallest integer satisfying i=1jai(X)>a0(X):=i=1Mai(X)
  • Step 3: Replace tt + τ and XX + cj.

  • Step 4: Record (X, t) as desired. Return to Step 1, or else end the simulation.

For the 10,000 stochastic simulations, the SEIR model was computed by using the Gillespie algorithm with initial values S = 100,000; E = 10; I = A = H = R = 0 and the parameter values [10] β = 0.085, l1 = 0.2, l2 = 10, κ = 1/6.6, ρ = 0.585, γa = 0.6403, γI = 1/5, and γr = 1/7. The control measure was used by changing the value l2 from 10 to 8.5. Figure 1 depicts the box plots of incidences I (t) + A (t) + H (t) of the MERS-CoV depending on the time of the control intervention on the 20th, 40th, and 60th days after the identification of the index case.
The box plots showed that there was a significant difference between the non-intervention and intervention groups (the 20th day, 40th day, and 60th day groups) and seemed to show no significant difference based on the time of intervention. However, the ANOVA in Table 1 revealed a significant difference between the averages in the intervention groups and showed that early intervention promotes a good strategy to control the disease. In particular, these results were evident from the average and standard deviation, which were smaller in the early intervention period. The difference was markedly larger 100 days after the identification of the index case, and the difference in the effect of the intervention over time showed a decreasing trend.
The control intervention was conducted during the MERS-CoV outbreak in Korea immediately after the confirmation of the index case and the control measures were carried out on the 20th day after the confirmation of the index case. Using the stochastic simulations of the SEIR model depending on the time of control interventions on the 20th, 40th, and 60th days after the confirmation of the index case, this study investigated whether early risk communication with the general public and mass media is an effective preventive strategy. As a result, the intervention on the 20th day after the identification of the index case was much better than the intervention on the 60th day. Therefore, we conclude that appropriate risk communication can secure the confidence of the general public in the public health authorities.
This study was carried out with the support of the Research Program of Rural Development Administration, Republic of Korea (Project No. PJ011563).

CONFLICTS OF INTEREST

No potential conflict of interest relevant to this article was reported.

Figure 1
Box plot for the control interventions according to the number of days (A, 100 days; B, 200 days; C, 300 days; D, 400 days) after the identification of the index case.
phrp-08-373f1
Table 1
Results of the ANOVA according to the day of intervention after the identification of the index case
Variable Data ANOVA (F-value)a
The 100th day 231.72
 No control 23.0017 ± 30.1585

 The 20th day control 9.4533 ± 14.3353

 The 40th day control 11.4083 ± 16.0851

 The 60th day control 14.5683 ± 19.9566

The 200th day 108.49

 No control 52.6752 ± 82.3164

 The 20th day control 7.3237 ± 16.8129

 The 40th day control 8.6575 ± 18.1072

 The 60th day control 11.2189 ± 21.7501

The 300th day 66.99

 No control 106.4571 ± 164.5254

 The 20th day control 5.5167 ± 16.7547

 The 40th day control 6.3517 ± 17.2051

 The 60th day control 8.4170 ± 20.5291

The 400th day 38.18

 No control 172.5906 ± 241.5982

 The 20th day control 3.9101 ± 14.7193

 The 40th day control 4.4760 ± 15.3066

 The 60th day control 5.8240 ± 17.5711

Value are presented as mean ± standard deviation.

a Degree of freedom (d.f.) (factor) = 2; d.f. (error) = 29,997; p = 0.0000.

Figure & Data

References

    Citations

    Citations to this article as recorded by  
    • Healthcare-associated infections: the hallmark of Middle East respiratory syndrome coronavirus with review of the literature
      J.A. Al-Tawfiq, P.G. Auwaerter
      Journal of Hospital Infection.2019; 101(1): 20.     CrossRef

    • PubReader PubReader
    • Cited
      CITED
      export Copy
      Close
    • XML DownloadXML Download
    Figure

    PHRP : Osong Public Health and Research Perspectives