Hospital DNR Orders – WTF (Why They Failed)

A thought provoking article just came out ahead of print in the Journal of General Internal Medicine. It is titled “Hospital Do-Not-Resuscitate Orders: Why They Have Failed and How to Fix Them” and authored by Jackie Yuen, Carrington Reid, and Michael D. Fetters (a quick disclosure – I’m unbelievably excited that Jackie is coming to our UCSF Geriatrics Fellowship next year – we need more bright minds like her in our field). The paper lays out four main failures of DNR orders and four possible solutions.

Let’s start with the least controversial part of the paper – the four failures of DNR orders. To sum it up – discussions rarely occur, and when they do occur it happens much too late with doctors not really giving enough information for patients to make a true informed decision. Worst of all, physicians make unfounded assumptions based on the presence of a DNR order that may limit other treatments (please see my rant on AND orders for issues around extrapolation and assumptions around end-of-life orders).

The solutions proposed by Dr. Yuen and her colleagues are a little bit more controversial than the failures, but overall are pretty spot on.

The authors first recommend promoting a culture change in hospital to improve patient-centered end-of-life care. This recommendation is meant to address our current culture that praises technology and interventions. I completely agree with the authors, although I would like to see more specifics on what hospitals can do, as culture change is no small feat (damn you journal word limits!)

The second recommendation addresses inadequate hospital policies that fail to set standards for DNR discussions. The authors recommend new Joint Commission standards for DNR discussions and a requirement that institutions demonstrate compliance with these standards in order to be accredited. The standards used in the article include making the ‘Attending physician’ responsible for leading DNR discussions with appropriate patients or their surrogates within 72 hours of hospitalization (and with change in clinical condition). I find this standard fascinating and a little concerning. It’s fascinating as I’m guessing attending’s rarely have these conversations in most academic hospitals. It’s concerning as I am not sure if attending physicians would be any better at it than junior physicians who probably have a greater amount of training in communication skills.

The last two recommendations include establishing formal communication skills training programs in goals of care and DNR discussions, and payment reform that no longer rewards volume and intensity of care, but rather uses financial incentives that use patient satisfaction and/or the quality of DNR discussions as performance measures. I again, couldn’t agree more.

There are aspects of hospital DNRs that are not addressed by the authors. The main recommendation that I think is missing is one addressing the complete lack of portability of most DNR orders outside that of the POLST paradigm. For example, nursing home residents may have a DNR in the nursing home, but the second EMS picks them up to go to the hospital, these very same residents become full code again. Same thing happens when that patient who is DNR in the hospital gets admitted to the nursing home. Hospitals and policy makers cannot think of illness as confined to discrete time periods. Hospitals no longer ‘fix’ people. Rather, people continue to live with chronic, progressive, and advanced illnesses even after hospital discharge. I would encourage any future reform to look at this issue from a much larger community approach than just focusing on the hospital setting.

This is a small point though in a much larger and well written article that I would encourage everyone to read.

by: Eric Widera