Application Story

There have been devastating pandemics in history. In 1350, The Black Death killed over 75 million people. The 1918 Flue pandemic killed 50 million. More recently Influenza claimed more than a million lives. In this article, you will see how ProcessModel optimization increases survivability by 25% during a severe pandemic. See how your chances of survival will be improved.

The Threat

It appears we are getting better at minimizing a pandemic (and we are), but a large part of the formula is out of our control. A severe pandemic happens when the following occurs:

  • A virus mutates into a strain to which general populations lack sufficient immunity
  • Transmissibility is high
  • The incubation period is short
  • The mortality rate is high
  • A vehicle exists for transmission

If the conditions above align for a severe pandemic, it’s possible to have worse outcomes in the future than we’ve had in the past, even if we are getting better at prevention. Health and Preparedness leaders have spoken about the threat of an influenza pandemic.

[a pandemic] is an absolute certainty. When it comes to a pandemic, we are overdue, and we’re under-prepared.
— Mike Leavitt, Secretary, US Health and Human Services

This is a very ominous situation for the globe. It is the most important threat we are facing right now.
— Julie Gerberding, US Centers for Disease Control and Prevention

Emergency department slashes door to discharge time with process simulation.

Even though many of the factors are out of our control, once a pandemic occurs strategic action will significantly reduce the number of deaths. The Center for Disease Control (CDC) has endorsed the idea of triage allocation. But, the CDC won’t recommend any algorithm. The lack of an algorithm means every hospital decides what to do on their own.

Triage is the assignment of degrees of urgency and is used to decide the order of treatment. It means some gain access to scarce resources, and some are turned away — to assure a higher number live. It means saving the most lives.

During a severe pandemic, the system quickly becomes overloaded, and every patient admission limits other admissions. If these decisions remain unplanned, the hospitals will treat patients that will become a casualty and not address those they could have saved. Without using a triage strategy, more people will die. There are proposed strategies, but no one has been able to quantify the value of implementing a strategy…until now.

Alexander Kolker’s Life Saving Optimization

life saving optimization

Alexander Kolker recently provided a formula to reduce the mortality rate experienced during a pandemic. In a published peer review article, in Critical Care Medicine magazine, he demonstrated how to increase survival by almost 25% during a severe pandemic. Alex proposes using a specific triage algorithm. Alex’s research is the first time that a particular triage scheme has been designed using simulation and optimization to find the optimal survival rate.

Flowchart showing the decision structure used to save lives during a severe pandemic.

Optimization Finds the Best Ways of Saving Children’s Lives

This study allowed manipulation of critical thresholds to optimize outcomes. An objective function (a mathematical target) is used to compare the result of all experiments.  The model thresholds are changed automatically using, built-in, evolutionary algorithms to generate various solutions. The solutions must improve the value of the objective function to stay active. The best solution occurs when experiments can’t enhance the value of the objective function. Because of the evolutionary algorithm, only a small subset of many thousand possible solutions need to be run to find the optimal thresholds. Because of the evolutionary algorithm, less time is required to find an optimal solution.

Some of the factors include:

  • Size of the pandemic
  • Pattern of patient arrival
  • Probability of death if treated
  • Likelihood of death if untreated
  • Days of ventilation needed
  • Beds available
  • Etc.

Details of the Objective Function

Optimization converged in simrunnerObjective Function (OF)=F(TPOD, TDV) approaching maximum

OF = w1 × time-averaged occupancy (%) – w2 × mortality untreated (%)–w3 × mortality treated (%)

Where w1, w2, and w3 are relative weights of the corresponding components of the OF. To place a higher priority on minimizing mortality over maximizing bed occupancy, the weights w2 and w3 were assigned twice the weight of w1. Optimizing the combined OF (POD and DV) using simulation was achieved through an evolutionary strategies algorithm built in the ProcessModel.


Large Data Source

This study analyzed close to 1 million historical patient records dealing with similar respiratory conditions to derive the algorithms for the model. Numerous people from healthcare, statistics, and public health participated in the study. The results of the study are striking and difficult to ignore, offing a 25% improvement saving lives during a severe pandemic.

Alex Kolker has published many other groundbreaking articles, including Predictive Analytics for Patient Length of Stay; Optimal Staffing Modeling with Variable Patent Demand; and Interdependency of Hospital Departments and Hospital-Wide Patient Flow, to name a few. ProcessModel, Inc. is indebted to Alex Kolker for his tireless work in creating many improvements in Healthcare and for the research on Maximizing Survival Rate to Save Children’s Lives during a severe pandemic.

Would you like to discuss your experience with ProcessModel? Ping me on the form below, and I will get back to you ASAP.

    Name: *

    Email: *

    Phone: *

    Questions: *