Patient evacuation from Memorial Medical Center during Hurricane Katrina

The evacuation of Memorial Medical Center during Hurricane Katrina was not simply a movement of patients out of a threatened building. It became a test of a hospital’s known but incompletely addressed flood risks, a strained emergency response, and a policy that put patients with do-not-resuscitate orders last. The hospital completed its evacuation by September 1, but the decisions made during that effort—and allegations about medication given to patients—continued to shape the aftermath.

Known Vulnerabilities and Delayed Preparation

The danger to Memorial’s electrical system had been identified before Katrina. After Tropical Storm Allison flooded Houston in 2001 and disabled power at hospitals in the Texas Medical Center, New Orleans health officials asked local hospitals whether they could withstand fifteen feet of water and whether they would consider raising generators. The effort stalled amid the cost of improvements. The day before Katrina arrived, emergency officials were told that nearly all New Orleans hospitals had generators, electrical switches, or both at ground level; Memorial was among them.

After Hurricane Ivan narrowly missed the city in 2004, Eric Yancovich, Memorial’s plant operations director, reviewed storm-surge estimates and studied how water could reach the hospital. Memorial sat three feet below sea level, and the estimates suggested that a major hurricane could bring twelve to fifteen feet of water to the area. Yancovich’s own calculations showed that less than four feet of street flooding could send water over the loading dock. Although the generators themselves were on the second floor, important components of the emergency power distribution system, along with pumps for medical air and suction, were at or near ground level. He warned that flooding could disable most of the medical center and recommended raising vulnerable equipment. The electrical work alone was estimated to cost more than a quarter of a million dollars; lacking capital, the hospital deferred it.

From Flood Threat to Power Crisis

As floodwaters approached, Yancovich and Memorial’s emergency leadership recognized that evacuation was likely. Susan Mulderick raised the prospect with CEO L. René Goux, who contacted his superiors at Tenet Healthcare. The hospital’s backup system, however, depended on more than its generators: the power-distribution equipment that delivered electricity through the building was its weak point.

Three large generators had already been running for two days, though Memorial’s disaster plans anticipated using them for three. They were designed to support essential functions for minutes or hours, not to carry a hospital through prolonged operation. On-site engineers and electricians were not generator mechanics. When one engine shut down, staff working by flashlight attributed it to overheating, added water to its radiator, and restarted it. Keeping the machines running could not remove the underlying risk to the rest of the system. The evacuation therefore unfolded while staff were trying to preserve power in a building whose emergency electrical infrastructure had long been recognized as vulnerable to flooding.

Evacuation Priorities and DNR Patients

The hospital’s doctors made a consequential exception to the usual priority of moving the sickest patients and those dependent on machines: patients with DNR orders would be evacuated last. Four such patients were in the intensive care unit, including Helen Breckenridge and Jannie Burgess. A DNR order directs that a patient whose breathing or heartbeat stops not be resuscitated; it does not, by itself, mean that the patient has a terminal or irreversible condition. Richard Deichmann, Memorial’s medical chairman, later said he believed the law required that condition for a DNR order and considered those patients to have the least to lose if disaster struck. The evacuation rule thus rested, in part, on a mistaken understanding of what the order meant.

The distinction became urgent on a fifth-floor hallway. Staff moving an oxygen-dependent, bed-bound patient stopped when they noticed her DNR order, and nurses were told not to explain that this was why she had been left behind. When an order came for staff to leave the floor, nurse Michelle Pitre-Ryals told the patient’s daughter that she could request cancellation of the DNR. The daughter had originally agreed that her ninety-three-year-old mother should not be resuscitated after cardiac arrest; facing the possibility that she would not be rescued at all, she asked for the order to be withdrawn. The patient was moved downstairs toward the helipad and survived her immediate ordeal. On LifeCare’s seventh floor, Angela McManus likewise tried to rescind the DNR order for her mother, Wilda, but staff told her that no doctor was available to do so. These episodes showed that a decision about resuscitation was being made to govern access to evacuation, although the two decisions were not equivalent.

Completion and Contested Aftermath

Memorial’s patients were transferred by Thursday, September 1. That timing matters to the later dispute: as the retrospective account notes, the plan to inject patients went ahead just as helicopters arrived in force, expanding the rescue resources available. The evidence does not erase the severe conditions or establish that every patient could have been saved. It does, however, make the sequence of decisions central to evaluating what alternatives remained as the evacuation progressed.

After the disaster, interim CEO Curtis Dosch told investigators that Mulderick had described the patients as being made comfortable. He did not volunteer that he had seen a doctor inject a patient, had asked Goux whether he knew what was happening, or had seen Memorial’s chief nursing officer crying. Dosch later said he had wondered whether the patients might have lived if they had been evacuated, and whether someone else had the right to decide they would not make it. Nurses described an immediate refusal to discuss the events—a Code of Silence that left questions about what had happened unresolved even among hospital staff.

An arrest affidavit later assembled witness accounts and toxicology results concerning four LifeCare patients. Witnesses reported that Anna Pou had said the patients would receive a lethal dose and that LifeCare staff should leave; one nurse described seeing another nurse inject Rose Savoie. Tests found morphine and midazolam in the four patients named in the affidavit, and pathologist Cyril Wecht concluded that each had died from a lethal amount of the drugs. A judge signed warrants alleging that Pou, Cheri Landry, and Lori Budo had intentionally killed the patients. These were allegations in a criminal case, not a finding established by the evacuation itself, but they made the handling of patients during the hospital’s final transfer a subject of investigation and public conflict.

Meaning of the Evacuation

The book places Memorial’s crisis within failures that extended beyond the hospital: the city’s delayed evacuation, inadequate transport for people without cars, uncoordinated rescue efforts, poor communications, and insufficient backup power and water at medical facilities. Memorial becomes a concentrated instance of that wider breakdown, where compromised infrastructure and divided decision-making converged on patients who depended on others to move them. The comparison with Charity Hospital sharpens the point: Charity flooded and endured severe operational failures, yet staff continued providing care until its patients were transferred on September 2; fewer than ten patients died there, despite roughly twice as many patients across its campuses and a lower staff-to-patient ratio than at Memorial.

The evacuation’s significance lies not only in its completion, but in the choices made under pressure and the assumptions embedded in them. DNR status was transformed from a limit on resuscitation into a reason for lower evacuation priority; known engineering risks had not been fully corrected; and decisions about medication coincided with an expansion of rescue capacity. The account does not reduce the outcome to one cause. It shows instead how preparedness, available resources, interpretation of medical orders, and individual decisions interacted—and how the questions raised during evacuation persisted after the patients had left the building.