Hurricane Katrina and the DNR fallacy

It happened nearly a decade ago, and although the details were there to be read in articles and commentaries, the publication of Sheri Fink’s book ‘Five Days At Memorial‘ has provided an opportunity to explore the tragedy. However, for all the comprehensive detail, the author has been accused by the protagonist with the highest profile, Dr Anna Pou, of creating a fiction. She writes on her website,

                ‘Now several years later Ms Fink has turned her article into a full length novel entitled “Five Days At Memorial [Hospital]” which concludes that scores of patients were euthanized by their doctors. The book is not only an insult to the self sacrificing doctors, nurses and other medical personnel who stayed in harm’s way tending to patients in the most difficult of circumstances, but a disruption to the closure of this tragedy by suggestions to family members of patients who lost their lives that their loved ones were murdered.’

So what did happen? The narrative provided in Fink’s earlier ProPublica article provides a clearer account than that presented in her perhaps over-detailed book, and I recommend it. However, this article is about the decisions that were taken well before the morphine and midazolam injections were given  – namely the policy, hastily agreed in difficult circumstances, of equating DNR status with unsalvageable medical status.

First, a quick overview. Memorial Medical Centre (MMC) attracted attention soon after the waters receded because more people died there (45) than in any other New Orleans hospital. Suspicions grew that something unnatural had occurred. Indeed, one doctor had left during the flood unhappy about a perceived policy of euthanasia; ‘I can’t be part of anything like that.’ (p202) he said to a colleague.

After an investigation by the Medicaid Fraud Control Unit, evidence against Dr Pou was heard before a Grand Jury. It did not indict her on the one count of second-degree murder or the nine counts of ‘conspiracy to commit second degree murder’, and Dr Pou was released. The political atmosphere had become febrile by this time, and the accused doctor (two nurses who had also been accused were offered immunity for their testimony) received a great deal of popular support. Louisiana state’s first term Attorney General, Charles Foti, failed to get re-elected because of his decision to pursue the charges.


The Triage

The hospital weathered the storm itself, but the floodwaters that rose following the failure of the levees isolated it entirely. Doctors and management team members met, and soon decided that the ‘sickest, the ones most dependent on life support or mechanical aids, should go out first.’ (p75) Normal clinical service could not be maintained and a senior physician decided that ‘all but the most essential treatments and care should be discontinued.’ (p81) Rationing of available resources (human and material) was therefore underway. Then (p92) Fink describes how, at the earlier meeting, ‘doctors had established an exception to the protocol of prioritizing the sickest patients and those whose lives relied on machines. They had decided that all patients with Do Not Resuscitate orders would be prioritised last for evacuation.’ Fink describes what DNR means, and emphasises that a ‘DNR order meant one thing: a patient whose heartbeat or breathing stopped should not be revived’, and a few lines later explains ‘but the doctor who suggested at the meeting that DNR patient go last had a different understanding… [He] said he thought the law required patients with DNR orders to have a certified terminal or irreversible condition, and at memorial he believed they should go last because they had “least to lose” compared with other patients if calamity struck.’


As the evacuation process began, patients were triaged explicitly into 3 categories, ‘3’s’ being those judged to be ‘very ill’, or those with DNR orders. Pieces of paper with 1’s, 2’s and 3’s were taped to patients’ clothing, or written directly onto their gowns with thick black pens (p137). This link takes you to a disturbing picture in Fink’s ProPublica article; DNR is scrawled on an obtunded patient’s gown, along with a ‘3’.


The DNR trap

This fallacy, that DNR = end of life, traps and confuses inexperienced nurses and clinicians in everyday clinical practise. Unless a DNR decision is made in the context of imminent death, I usually make a point of saying to the patient, and to ward staff, that ‘this does not affect your treatment, we’re still going to do our best to get you through this illness and home.’ Care must be taken to prevent the impression that DNR means that we will relax, and not bother so much about the details. This danger is encapsulated in a slide taken from a recent presentation I gave on the subject.

dropping a gear

Despite this concern, in the absence of a disaster or a sudden need to ration personnel or equipment, it seems reasonable and achievable to promise that DNR will not affect other aspects of care. The question of limited resources will not come up. Or will it?

There is one scenario where rationing may occur, and that is at the cusp of admission to critical care units. Here, any signal that patients may be near the end of life, or not ‘for everything’, can influence the decision to transfer. It is not uncommon for ward doctors to delay DNR discussions for fear that the signal it gives to those who may be asked to review their patients and decide if organ support has a role. For not everyone can go to ICU. There are not enough beds. This is a form of rationing. Rationing of ICU beds exists, and has been well studied. Although rationing may not be the explicit reason for denying a patient admission, the perennial pressure on beds is very likely (in my view) to raise the bar of anticipated utility in the minds of those who must choose who passes through their unit’s doors.

So perhaps Katrina does offer a lesson to those of us working in functioning health systems. DNR can be used as a label, and if care is not taken its meaning can be amplified into an awful shorthand (at handover for instance) such that it influences the general approach to care . For those of us advocating for more DNR decisions taken earlier after admission, in order to prevent resuscitation of highly frail an comorbid patients by default, the challenge remains – to balance pro-active decision making with protection against lazy thinking. This requires constant vigilance and energetic education.



Note: Double effect

The quintessential point in terms of the guilt, and one that can only be known by the people involved, is whether the primary intention of those involved in injecting patients was to kill or relieve distress. Some would contend that in certain, desperate circumstances, they are one and the same. A patient who is very near the end of life, and who is in pain or severely short of breath, will of course die sooner if morphine is administered. This is the well known double effect.

Note: Reverse triage

There are many reasons why some those triaged to leave last ultimately died. Moving bedbound, morbidly obese or ventilated patients to the helipad proved impossible, for instance. The concept of reverse triage, whereby the sickest go last, is commonly applied in the military setting, or in civilian catastrophes. It is not necessarily controversial. During the years since Katrina, new laws were passed in Louisiana indemnifying doctors against accusations of clinical neglect in the event of patients dying due to this policy.





Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.