Hospitals typically employ triage systems to classify and prioritize patients based on their medical need and potential benefit from timely life-support treatment, especially when hospital resources are overwhelmed. Without such a system, patients' risk being admitted in order of arrival, potentially disadvantaging those who arrive later but have a greater chance of survival, ultimately increasing overall mortality.
Before establishing triage criteria, fundamental principles-such as maximizing lives saved and ensuring fairness-must guide decision-making. In this talk, we propose a new analytical triage framework, using an epidemic model as a working example to evaluate efficiency (measured in total lives saved) and fairness across different triage strategies designed to prioritize specific patient groups.
We introduce the concept of return, defined as the benefit (mortality rate reduction between treated and untreated patients) per cost (treatment duration). This represents the instantaneous patient survival rate: the higher the return, the greater the number of lives saved at any given moment.
Using return as a guiding principle, we propose a simple greedy triage strategy that maximizes instantaneous patient survival. We mathematically prove that this strategy is the most effective life-saving approach under realistic and critical conditions. Numerical results confirm its superiority in most cases compared to other strategies studied in our work. Furthermore, results indicate that as hospital overload increases, the difference in lives saved between the greedy triage strategy and the traditional first-come-first-served approach becomes more significant, highlighting the critical importance of our work during a pandemic.
Finally, we examine the balance between efficiency and fairness across various triage strategies, demonstrating that a balanced approach is achievable. This analysis can assist doctors in selecting a strategy that maximizes lives saved while ensuring equitable hospital bed allocation.
