Medical resource scarcity and triage in disasters

Medical resource scarcity and triage in disasters concern more than deciding which patient receives a limited bed, machine, or supply. The account connects those choices to decisions made before an emergency, to the uneven ability of patients and communities to reach help, and to the values used to define a fair allocation. The crisis at Memorial Medical Center during Hurricane Katrina shows how unprepared systems can force exhausted staff to make consequential decisions without adequate guidance; later examples ask whether advance protocols, public participation, and practical improvisation can prevent scarcity from becoming a choice between patients’ lives.

Scarcity Before the Bedside

At the time of the mandatory evacuation order in New Orleans, many residents had no cars or other way to leave, while the city’s transport plans had serious gaps, including too few drivers. The New Orleans Superdome was designated a shelter of last resort, and Mayor Ray Nagin urged people with medical conditions to move there without specifying what care they could expect. Acadian dispatchers were overwhelmed by requests to move patients from hospitals, nursing homes, and homes. Many of the roughly two dozen ambulances made available were stuck in traffic, and some began taking patients to the Superdome rather than out of the city. These circumstances made access to safety depend not simply on medical need, but on transportation, congestion, and whether a functioning rescue service could reach a patient.

Hospitals faced a related danger in their own infrastructure. Before Katrina, officials knew that most New Orleans hospitals had generators, electrical switches, or both at ground level. At Memorial, Eric Yancovich had studied storm-surge estimates and sketched the height of hospital entrances and critical equipment against the street elevation. His calculations indicated that less than four feet of flooding on Magnolia Street could carry water over the loading dock. Although the backup generators were on the second floor, parts of the emergency power distribution system were at ground level or only inches to a few feet above it. Yancovich warned that relatively little water could disable much of the medical center and predicted that patients might have to be moved to the Clara Wing. Floodgates and a raised vent for underground diesel tanks were among the improvements made, but proposed work to raise transfer switches and pumps for medical air and suction was not funded. The resulting vulnerability was therefore not an unforeseen technical detail: a known weak link connected flood risk to the loss of equipment patients depended on.

The wider crisis also exposed unequal vulnerability. Jannie Burgess had worked for thirty-five years as a nurse in New Orleans hospitals and nursing homes, but because of racial segregation she had once been able to care for patients at private hospitals where she could not herself receive treatment. Southern Baptist Hospital, later Memorial, resisted federal programs that conditioned reimbursement on nondiscrimination and was among the last Southern hospitals to integrate. During the flood crisis, patients’ age, illness, dependence on equipment, and ability to travel intersected with institutional arrangements and the history of access to care. The disaster’s effects accumulated across these conditions rather than falling evenly on everyone in the city.

Triage at Memorial

As evacuation proceeded, doctors at Memorial made an exception to the usual priority for the sickest patients and those dependent on machines: patients with Do Not Resuscitate orders would be evacuated last. Four patients in the intensive care unit had such orders, including Jannie Burgess and Helen Breckenridge. A DNR order meant that a patient whose heartbeat or breathing had stopped should not be revived. It was not the same as a Louisiana living will authorizing the withholding or withdrawal of life-sustaining treatment in a terminal and irreversible condition. Richard Deichmann later explained that he believed DNR patients had the least to lose if calamity struck, in part because he thought the law required them to have a certified terminal or irreversible condition. The distinction between an instruction about resuscitation and a judgment about evacuation became consequential when a medical order was used to rank patients for rescue.

The effects of that policy became visible in the Clara Wing. Transporters stopped moving a bed-bound patient after seeing her DNR order and maneuvered other patients around her. Staff were told not to tell her daughter why she had been left behind. When the daughter learned that her mother was at risk of being abandoned, she asked for the order to be canceled. A nurse explained that a patient or proxy could request its discontinuation; the daughter said that agreeing not to resuscitate her mother after cardiac arrest had not meant agreeing that she should not be rescued. The order was removed, and the patient was moved downstairs toward the helipad. On LifeCare’s seventh floor, Angela McManus likewise tried to rescind her mother Wilda’s DNR order after hearing that DNR patients would wait, but staff told her no doctor was available to do so. The contrast shows how the formal possibility of changing an order could be of little use when the crisis made a doctor inaccessible.

Evacuation decisions were also shaped by the limits of what people could see and by the strain inside the hospital. Karen Wynn, caring for patients awaiting rescue, encountered an outpatient who had made her way through floodwater for a scheduled chemotherapy appointment. Wynn had to explain that treatment would not be given. She also received a complaint from a Baton Rouge hospital about a psychiatric patient transferred there with neonates, while she had little sense of how many displaced patients had arrived in the capital. Meanwhile, people in the hospital’s evacuation line sometimes hesitated to leave when they did not know their destination or could not take belongings. The local pressure on staff and patients could make the outside emergency difficult to grasp, even as people elsewhere faced their own shortages.

That pressure extended beyond admitted patients. A family with small children rowed up to Memorial and was turned away. Bryant King objected to the decision and urged the hospital’s leadership to help them. The account notes that the people refused were nearly all African American, while also distinguishing Wynn’s view from King’s: Wynn saw no racism in the refusals, whereas King was offended in large part because the people being turned away had dark skin. The hospital was already sheltering patients, staff, families, and animals, but the refusal placed the question of whom a hospital could help in direct conflict with the urgency of people seeking refuge. At Charity Hospital, by contrast, staff continued caring for patients in their rooms until the end despite flooding, failed utilities, a disrupted evacuation, and a larger patient load relative to staff; fewer than ten patients died there. The comparison complicates any account that treats extreme conditions alone as determining how care is delivered.

Protocols, Consent, and Public Values

Later planning sought to keep frontline clinicians from having to invent allocation rules at the height of a crisis. New York State published a ventilator-allocation protocol in 2008 after concern about a severe influenza pandemic and the arrests of health professionals connected to Memorial. The planning arose from exercises in which officials repeatedly ran short of ventilators. Hospitals could conserve equipment by canceling elective surgery, and New York purchased and stockpiled additional ventilators, but neither step could meet demand in a severe, 1918-scale pandemic. Experts therefore developed rationing plans intended to save as many lives as possible within an ethical framework. The plan represented a departure from ordinary care, which emphasizes doing everything possible for each individual patient, and was meant to establish criteria before clinicians were exhausted and under pressure.

The New York group adapted an approach proposed in Ontario during SARS, using the Sequential Organ Failure Assessment score to help guide access to intensive care. But the score had not been designed to predict survival and was not validated for assessing children. Its use reflected a shortage of appropriate alternatives, not a settled measure of who would benefit. More broadly, a protocol can make decisions more consistent without making its criteria self-evidently just. The question is not only how to distribute a scarce resource, but whether the measure used to rank patients is sound and whether the process can be explained to those affected.

The experience of Nathalie LeBrun in Haiti sharpened the problem of consent. At a field hospital in Port-au-Prince, an American team gave her oxygen when her blood oxygen level fell dangerously low. The tank ran out overnight; staff found another, but the supply chain remained unreliable, and an electric oxygen concentrator repeatedly overheated while generator fuel was scarce. A liaison officer reasoned that LeBrun might need oxygen indefinitely, whereas earthquake patients might need it only temporarily for surgery. He decided to withdraw her oxygen, a decision approved by the head doctor. LeBrun was not told that her oxygen was to be withdrawn or consulted about the plan; she was told only that she would be transferred. At the receiving hospital, Paul Auerbach found some oxygen and treated her with diuretics, stabilizing her for a time without it. The sequence demonstrates that a calculation about aggregate benefit does not resolve the separate question of whether a capable patient should be informed and involved when potentially lifesaving care is denied.

Bellevue: Allocation and Improvisation

During Superstorm Sandy, Laura Evans faced a related problem at Bellevue Hospital Center. Flooding threatened the pumps that supplied oxygen and suction and those that carried fuel from underground tanks to backup generators. If city power failed, the limited fuel reserves attached to generators on the thirteenth floor might last only two hours. Nearly all patients in Evans’s fifty-six-bed adult intensive care unit relied on electrically powered equipment, while a better-protected generator in a neighboring building could supply exactly six outlets. Hospital leaders asked Evans to identify who would receive them.

Evans had studied the New York guidelines and adapted them for the emergency. She assembled an ad hoc committee of doctors, nurses, and ethicists who had no patients under consideration, so that primary clinicians could continue advocating for their own patients. The committee communicated its decisions and reasons to staff, some of whom challenged the choices but accepted them. Within about two hours, it produced a list and moved patients between beds and outlets. Evans wanted a transparent process that could be applied consistently and justified later. Yet the committee did not include patients or family representatives, and staff delayed communicating the allocation to families because they feared how to tell those who would not receive outlets. In retrospect, Evans believed families had a right to know; the team had also failed to consider whether a selected patient might volunteer to give up an outlet.

The anticipated scarcity did not become the only available course. When the lights briefly went out, backup power started after about seven seconds. Floodwater was filling the basement, and staff prepared for equipment failure by stationing professionals beside ventilator-dependent patients with manual resuscitation bags, running bright-orange extension cords through corridors, and counting drops on IV pumps in case medication had to be delivered without the usual equipment. Then volunteers formed a human chain to carry fuel up thirteen flights of stairs and feed the generators manually. The improvisation kept backup power functioning, and Evans’s patients remained on it as evacuation proceeded. The contrast is central: the committee’s allocation process addressed the possibility of scarcity, while an improvised supply solution helped prevent the feared choices from having to be implemented.

Limits and Significance

The account treats triage as a question of public values as well as clinical judgment. Decisions may aim to maximize lives saved, years of life, quality of life, fairness, or public trust; those aims can conflict. In a Seattle and King County public-engagement exercise, many participants considered withdrawal of life-sustaining care unacceptable partly because it could undermine trust in medicine. The proposed public review of New York’s ventilator plan had included focus groups with parents, older adults, people with disabilities, and communities of color, but those sessions never took place. Roger Bernier of the Centers for Disease Control and Prevention argued that members of the public are holders of public values and are well placed to weigh competing ones. Their involvement matters especially because an allocation system that appears neutral to experts may distribute burdens differently across patients and communities.

The debate over crisis standards also carries a risk in both directions. Carl Schultz warned that lowering standards could appeal to governments and companies because it relieves pressure to provide better care, and that lowered standards might later be applied when resources are tight outside a disaster. Ann Knebel countered that planning can help professionals reuse and reallocate resources and provide the highest standard possible under the circumstances. The account’s emphasis is not that one formula can eliminate tragedy, but that planning should not become a substitute for investment, communication, or reassessment. Frederick Burkle worried that rigid criteria could continue to ration resources even when the emergency’s conditions changed; he emphasized reviewing supplies frequently, sometimes daily or hourly.

That concern leads to a broader conclusion: the best way to reduce the harm of triage may be to reduce the need to make irreversible trade-offs. Improvised bubble CPAP in India allowed one improving infant to come off a ventilator and the ventilator to be used for another child; colleagues credited Aarti Kinikar’s approach with saving lives. At Bellevue, manual fuel delivery similarly preserved power and helped avert the allocation crisis. These examples do not erase the need for fair procedures when resources truly cannot meet need. They show why disaster planning must include adaptable equipment, reliable infrastructure, situational awareness, and repeated efforts to expand what is available. The catastrophe at Memorial remains a warning that a decision to hasten patients’ deaths proceeded at the moment helicopters arrived in greater numbers and resources were expanding. Triage is therefore inseparable from the continuing obligation to notice when circumstances have changed, seek alternatives, and avoid treating scarcity as fixed when action might still relieve it.