Optimal control for pandemic influenza: The role of limited antiviral treatment and isolation

https://doi.org/10.1016/j.jtbi.2010.04.003Get rights and content

Abstract

The implementation of optimal control strategies involving antiviral treatment and/or isolation measures can reduce significantly the number of clinical cases of influenza. Pandemic-level control measures must be carefully assessed specially in resource-limited situations. A model for the transmission dynamics of influenza is used to evaluate the impact of isolation and/or antiviral drug delivery measures during an influenza pandemic. Five pre-selected control strategies involving antiviral treatment and isolation are tested under the “unlimited” resource assumption followed by an exploration of the impact of these “optimal” policies when resources are limited in the context of a 1918-type influenza pandemic scenario. The implementation of antiviral treatment at the start of a pandemic tends to reduce the magnitude of epidemic peaks, spreading the maximal impact of an outbreak over an extended window in time. Hence, the controls’ timing and intensity can reduce the pressures placed on the health care infrastructure by a pandemic reducing the stress put on the system during epidemic peaks. The role of isolation strategies is highlighted in this study particularly when access to antiviral resources is limited.

Introduction

The innate ability of the influenza virus to generate secondary cases of infection over short windows in time mean that the timely implementation of pandemic containment strategies is critical (Colizza et al., 2007, Ferguson et al., 2005, Ferguson et al., 2006, Germann et al., 2006, Longini et al., 2004, Longini et al., 2005). The identification, evaluation, and implementation of effective regional, national or global pandemic mitigation plans can be enhanced with the assistance of mathematical frameworks and the extensive simulations and/or analysis of appropriate submodels. Preparedness plans dealing with the allocation of antiviral medications must account for a multitude of factors including the pathogens’ virulence (defined in terms of case fatality rates in the population) which plays a central role in the assessment of the size of the antiviral medication stockpile for a community (Lipsitch et al., 2009). These levels of assessment take on a new meaning when the antiviral drug supplies are insufficient (Bar et al., 2009, Merler et al., 2009) and consequently, the use of non-pharmaceutical interventions or treatment measures that include the distribution of face masks and/or the availability of ventilators must also be factored in any preparedness plan (Ferguson et al., 2006, Merler et al., 2009, Tracht et al., 2010).

The world's insufficient capacity to produce antiviral drugs and vaccines (specially the new 2009 H1N1pdm influenza vaccine) during an emergency such as the one posed by the 2009 H1N1pdm influenza virus raises concerns at multiple levels (Fedson, 2003, Gostin and Berkman, 2007, Kotalik, 2005, Ulmer and Liu, 2002). The stockpiles of antiviral drugs (and H1N1pdm influenza vaccine) during a pandemic event are expected to be in the hands of the industrialized nations. Countries with high population densities and limited access to quality health care like Mexico and/or India do not have the infrastructure to produce antiviral drugs to meet their needs during this type of emergencies. Poor nations did not get timely access to what one would consider minimally adequate drug stockpiles, equipment or vaccine supplies. In fact, without the efforts of the World Health Organization (WHO) a large number of nations would have had no access to the most basic pandemic medical supplies at all. The unusual levels of morbidity and mortality among young adults have raised additional concerns (e.g., Chowell et al., 2009a, Nishiura et al., 2009, Reichert et al., 2010). Who should be vaccinated first? The young, the elderly, expecting women, or emergency personnel? The usefulness of the World Health Organization's (WHO) definition of pandemic is now being questioned since it appears that the severity of this pandemic appears to be lower than that associated with past influenza pandemics.

The task of identifying “optimal” control strategies that minimize the impact of influenza pandemics through the judicious use of a limited antiviral drug supplies in combination with measures like isolation, is the focus of this manuscript. The parameters used in our dynamical systems model are initially calibrated using influenza pandemic data of the 1918 influenza pandemic (Chowell et al., 2006), our severe pandemic baseline scenario. The usefulness of intervention alternatives that involve the distribution of antiviral drugs and/or the isolation of hospitalized patients are explored. Optimal control theory (Fleming and Rishel, 1975, Lenhart and Workman, 2007, Pontryagin et al., 1962), with a history of successful applications in biological, medical and industrial problems (Behncke, 2000, Blayneh et al., 2009, Jung et al., 2002, Rowthorn et al., 2009), is the primary tool used in our analysis.

This paper is organized as follows: Section 2 describes the model including several control functions and defines the objective functionals used in the optimal control framework. We present and compare the results of numerical simulations for five scenarios in Section 3, and our thoughts and conclusions are summarized in Section 4.

Section snippets

Influenza pandemic model with controls

Optimal control theory is used to explore the impact of antiviral treatment and isolation strategies in situations that mimic 1918-like influenza pandemic scenarios (Chowell et al., 2006). We calibrate our model using parameter estimates that correspond to the worst influenza pandemic in record (Andreasen et al., 2008, Chowell et al., 2006, Mills et al., 2004). Intervention strategies (policies) are modeled by the functions ui(t) (i=1,2,3) that externally control the number of clinical cases

Numerical results

In this section we present the results of selected simulations generated by the numerical implementation of the intervention strategies described in Section 2 under unlimited and finite resource scenarios. Results of the sensitivity analyses on some of the model parameters in System (1) are given in Section 3.2. In this section we only report the sensitivity analyses associated with the unlimited antiviral resources scenarios.

Discussion

Mitigating the impact and spread of the ongoing influenza A (H1N1) pandemic was on the minds of every newscaster, government official, aspiring politician, public health officials and literally billions of individuals around the world just a couple of months ago. The rate of growth/spread of influenza A(H1N1) news reports even surpassed the rate of growth of this new strain of influenza A around the world (Ginsberg et al., 2009). The rather unusual age-dependent patterns of spread, morbidity

Acknowledgements

This project have been partially supported by grants from the National Science Foundation (NSF-Grant DMS-0502349), the National Security Agency (NSA-Grant H98230-06-1-0097), the Alfred T. Sloan Foundation and the Office of the Provost of Arizona State University. Finally, we would like to thank the anonymous referees for their valuable comments.

References (47)

  • L.M. Bettencourt et al.

    Towards Real Time Epidemiology: Data Assimilation, Modeling and Anomaly Detection of Health Surveillance Data Streams. Lecture Notes in Computer Science

    (2007)
  • K. Blayneh et al.

    Optimal control of vector-borne disease: treatment and prevention

    Dis. Cont. Dyn. Syst. B

    (2009)
  • F. Brauer et al.

    Mathematical Models in Population Biology and Epidemiology

    (2000)
  • G. Chowell et al.

    Severe respiratory disease concurrent with the circulation of H1N1 influenza

    N. Engl. J. Med.

    (2009)
  • G. Chowell et al.

    Adaptive vaccination strategies to mitigate pandemic influenza: Mexico as a case study

    PLoS ONE

    (2009)
  • V. Colizza et al.

    Modeling the worldwide spread of pandemic influenza: baseline case and containment interventions

    PLoS Med.

    (2007)
  • O. Diekmann et al.

    Mathematical Epidmeiology of Infectious Diseases: Model Building, Analysis and Interpretation

    (2000)
  • D.S. Fedson

    Pandemic influenza and the global vaccine supply

    Clin. Infect. Dis.

    (2003)
  • N.M. Ferguson et al.

    Strategies for containing an emerging influenza pandemic in Southeast Asia

    Nature

    (2005)
  • N.M. Ferguson et al.

    Strategies for mitigating an influenza pandemic

    Nature

    (2006)
  • W.H. Fleming et al.

    Deterministic and Stochastic Optimal Control

    (1975)
  • R. Gani et al.

    Potential impact of antiviral use during influenza pandemic

    Emerg. Infect. Dis.

    (2005)
  • T.C. Germann et al.

    Mitigation strategies for pandemic influenza in the United States

    Proc. Natl. Acad. Sci. USA

    (2006)
  • Cited by (105)

    • Nonlinear optimal control strategies for a mathematical model of COVID-19 and influenza co-infection

      2022, Physica A: Statistical Mechanics and its Applications
      Citation Excerpt :

      COVID-19, a deadly disease, became a pandemic as a result of misunderstandings about its symptoms and the time of year (winter) when influenza is most common. The use of a vaccine, case isolation, therapeutic and prophylactic antiviral treatment, air traffic reduction, global influenza pandemic vaccination program or quarantine policy, and ventilation are some of the control strategies used in mitigating the spread of influenza [11–13]. COVID-19 is of severe magnitude due to the seasonality of influenza disease, which is determined by several factors, including the vaccine’s effectiveness and the characteristics of the circulating virus.

    View all citing articles on Scopus
    View full text