Stockpiling Ventilators for Influenza Pandemics

Hsin-Chan Huang; Ozgur M. Araz; David P. Morton; Gregory P. Johnson; Paul Damien; Bruce Clements; Lauren Ancel Meyers

Disclosures

Emerging Infectious Diseases. 2017;23(6):914-921. 

In This Article

Abstract and Introduction

Abstract

In preparing for influenza pandemics, public health agencies stockpile critical medical resources. Determining appropriate quantities and locations for such resources can be challenging, given the considerable uncertainty in the timing and severity of future pandemics. We introduce a method for optimizing stockpiles of mechanical ventilators, which are critical for treating hospitalized influenza patients in respiratory failure. As a case study, we consider the US state of Texas during mild, moderate, and severe pandemics. Optimal allocations prioritize local over central storage, even though the latter can be deployed adaptively, on the basis of real-time needs. This prioritization stems from high geographic correlations and the slightly lower treatment success assumed for centrally stockpiled ventilators. We developed our model and analysis in collaboration with academic researchers and a state public health agency and incorporated it into a Web-based decision-support tool for pandemic preparedness and response.

Introduction

Diligent preparation and effective countermeasures are critical to mitigating future influenza pandemics. The 1918 influenza pandemic, the most severe in recent history, resulted in ≈50 million deaths globally, of which nearly 675,000 occurred in the United States.[1] The 1957 and 2009 pandemics were less severe, causing ≈70,000 and 9,000–18,000 US deaths, respectively.[1] The US Department of Health and Human Services (HHS) estimated[2] that 865,000 US residents would be hospitalized during a moderate pandemic (as in 1957 and 1968) and 9.9 million during a severe pandemic (as in 1918).

When severe influenza outbreaks cause high rates of hospitalization, a surge of medical resources is required, including critical care supplies, antiviral medications, and personal protection equipment. Given uncertainty in the timing and severity of the next pandemic, as well as the time required to manufacture medical countermeasures, stockpiling is central to influenza preparedness.[3] However, difficulty in forecasting and limited public health budgets often constrain decisions about sizes, locations, and deployment of such stockpiles.

Mechanical ventilators are essential for treating influenza patients in severe acute respiratory failure. Substantial concern exists that intensive care units (ICUs) might have insufficient resources to treat all persons requiring ventilator support. Prior studies argue that current capacities are insufficient to handle even moderately severe pandemics and that sentinel reporting and model-based decision-making are critical for managing limited resources.[4–6] For this reason, the United States has stockpiled mechanical ventilators in strategically located warehouses for use in public health emergencies, such as an influenza pandemic. The Centers for Disease Control and Prevention (CDC) manages this Strategic National Stockpile (SNS) and has plans for rapid deployment to states during critical events.[7]

However, SNS ventilators might not suffice to meet demand during a severe public health emergency. In 2002, the SNS included ≈4,400 ventilators,[8,9] and 4,500 SNS ventilators were added during 2009 and 2010. The American Association for Respiratory Care suggested the SNS inventory should increase to at least 11,000–16,000 ventilators in preparation for a severe influenza pandemic.[10] The American Association for Respiratory Care and CDC[11] provide training on 3 types of SNS ventilators—LP10 (Covidien, Boulder, CO, USA); LTV1200 (CareFusion, Yorba Linda, CA, USA); and Uni-vent Eagle 754 (Impact Instrumentation, Inc., West Caldwell, NJ, USA)—to ensure proper use nationwide. In addition to the nationally held SNS, some US states maintain their own stockpiles.

Successful deployment of central ventilator stockpiles, whether federal or state, requires rapid distribution to healthcare facilities with patients in need, along with adequate bed space, requisite supplies, and trained personnel.[12–14] Robust methods for sizing and locating ventilator stockpiles have not yet been developed.[15] Wilgis[16] discussed the relative merits of central stockpiling of ventilators to be distributed during an emergency versus distributing ventilators to hospitals a priori. Centralized stockpiles benefit from better inventory tracking, more timely repairs, and superior allocation of a limited resource, but hospital-based supplies facilitate staff training, enable immediate use, and avoid the cost and logistical challenges of central storage and deployment.

We developed an optimization framework for allocating mechanical ventilators to central and local stockpiles to ensure adequate surge capacity during a future pandemic. This data-driven method considers the trade-off between risk and stockpiling cost, where risk is measured 2 ways: expected value of unmet demand (EUD; number of influenza patients not receiving required ventilation) and probability of unmet demand (PUD; probability at least 1 patient does not receive required ventilation). For a given set of healthcare providers in a region, we determined the optimal number of mechanical ventilators to stockpile centrally and at each provider site.

As a case study, we considered the US state of Texas under mild, moderate, and severe influenza pandemic scenarios. Based on the Texas Department of State Health Services (DSHS) response to the 2009 influenza A(H1N1) pandemic and planning efforts for future pandemics, we considered stockpiling across 9 sites: a centrally held state stockpile and local stockpiles in each of Texas' 8 health service regions (HSRs; online Technical Appendix Figure 1, https://wwwnc.cdc.gov/EID/article/23/6/16-1417-Techapp1.pdf). We implemented this model in a Web-based decision-support tool for DSHS.[17]

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....