Hour 68 of her shift, and the triage officer reviews case 1,847. Respiratory function declining. Cardiovascular instability. The Sequential Organ Failure Assessment score updates: hepatic performance, coagulation, renal output, neurological response. The algorithm calculates priority level P2. Not high enough for immediate ICU access. High enough to keep monitoring.
Day seventeen of the surge. The high-pressure system stalled over the Cascades three weeks ago. Wildfires running through forests dried by consecutive drought years. The western metro area holds at 44°C after sunset, air quality index 287. The regional network has 1,159 ICU beds serving 1,847 patients who need them.
Case 1,847 is a 67-year-old woman with COPD, arrived at County General at 0300 hours with acute respiratory distress. Under the old system—before systematic allocation frameworks—her survival would have depended on which physician happened to be on duty, which hospital she reached first, whether that facility had spare capacity. Under the protocol, her SOFA score determines her priority level. The algorithm doesn't know her socioeconomic status, her race, her perceived social worth. It knows her survival probability given available interventions.
The protocol emerged from documented failure. During the 2029 Phoenix crisis, individual physicians made gut-level decisions about ventilator access. Some prioritized youngest patients. Others focused on best prognosis regardless of age. A few made choices based on which families they knew. Mortality in overwhelmed facilities ran 34% higher than in hospitals with spare capacity—not because physicians were less skilled, but because overwhelmed individual decision-making produces arbitrary variation that kills people.
The triage officer coordinates with University Medical Center. They have two P1 transfers incoming, but capacity for the P2 case if respiratory function continues declining. The system processes 340 cases daily like this, moving patients to where they can receive appropriate care, preventing any single facility from becoming overwhelmed while others have empty beds. Since implementation, outcomes have equalized across the network. A 67-year-old with COPD presenting with acute respiratory distress gets the same priority level whether she arrives at County General or University Medical Center.
The protocol's designers studied heterogeneous ventilator triage policies that half of U.S. hospitals lacked entirely during COVID-19. They examined how crisis standards of care were activated in Idaho in 2021 when resources were exhausted. They built something that could be tested, refined, improved based on measurable outcomes.
Community health coordinators argue this approach erodes trust. They point to neighborhoods where ambulance calls have dropped, where people are dying at home rather than engaging with a system they perceive as a sorting machine. They're not wrong about the trust erosion. The 2029 Phoenix crisis created networks of people who watched the system abandon their neighbors when things got bad enough. Those networks remember.
Trust isn't something you can measure with the precision needed to guide resource allocation during a surge event. You can't calculate a trust score that updates every four hours. You can't coordinate transfers across a regional network based on which patients have the most trusting relationship with their providers. Survival, though—survival you can measure. Whether systematic allocation reduces mortality compared to overwhelmed individual decision-making, that produces data.
Without systematic triage, you get chaos. You get the kind of arbitrary variation that makes healthcare feel like a lottery, where your survival depends on which physician happens to be on duty when you arrive gasping for air. During surge events, when demand exceeds capacity by 60%, triage happens whether you formalize it or not. The choice is between systematic triage based on measurable criteria that can be defended and refined, or ad hoc triage based on individual judgment that introduces bias and produces inconsistent outcomes.
Hour 72. Case 1,847's respiratory function has declined further. SOFA score recalculated. Priority level upgraded to P1. The triage officer coordinates the transfer to University Medical Center, where an ICU bed has just become available. The woman will receive the intensive intervention her condition requires. The algorithm determined she has high survival probability given that intervention.
The heat map still glows red. More cases incoming. The algorithm continues its work of allocating scarce resources according to principles that can be written down, tested against outcomes, and improved based on data. When the next surge arrives—and the projections show more coming, longer duration, higher intensity—the regional network will have a framework that scales, that coordinates across facilities, that treats like cases alike.
The triage officer pulls up case 1,848. Updates every four hours. Respiratory function, cardiovascular stability, hepatic performance. What determines who survives when not everyone can.