Effective patient prioritization in mass casualty incidents using hyperheuristics and the pilot method

Whenever a mass casualty disaster takes place, the medical infrastructure available has to deal with a surge in the number or patients severely ill or injured. Using triage methods casualties have to be prioritized to receive health care in a limited-resource scenario. Aiming to do the greatest good to the greatest number of people, it has to be determined how to make the best use of these resources. This constitutes a very complex task that has to consider issues such as the current number of casualties, their lifetime expectancy, their resource consumption, etc. We approach this task within the framework of the pilot method and hyperheuristics. We show how these metaheuristics can effectively manage a number of simpler heuristics, providing improved results on an ample set of simulated problem scenarios. An exhaustive empirical evaluation analyzes the influence on performance of factors such as the total number of casualties, the severity of their medical condition, the treatment time, the number of resources available, or the number of triage classes.