Teaching learners about DNR discussions is one of the more difficult tasks in teaching communication skills at the end of life. As with other communication areas, habits are likely to have developed around DNR discussions that are difficult to break. Furthermore, a model of communication has developed which may not achieve the true goals for these discussions and may only serve to create greater misunderstanding. These barriers to learning must be addressed before new skills and approaches can be adopted.

RATIONALE:

Conceptual and skill barriers exist for DNR discussions. Addressing the conceptual barriers will help create space to develop necessary skills.

The DNR discussion has come to have an important legal role in patient care. The legal features of the document are partly to blame for the focus of DNR discussions being on reading lists of treatments to be withheld or provided and on obtaining the patient' signature. Having a signed DNR Order in a patient's chart helps reassure the team that they need not go through the motions of a resuscitation effort that they believe is futile.

However, there are also important ethical considerations for a DNR discussion. With ready access to life-prolonging technology, it can be difficult for the family, the patient, or the health care team to acknowledge when important transitions in the goals of care may have changed. A DNR discussion can help the patient clarify his or her own goals, and allows the family and patient to talk openly about hopes and fears at the time of death.

PITFALLS

• Old habits are hard to break. It can take multiple discussions and attempts before the learner is able to let go of current beliefs and practices. Experiencing success will be the best reinforcement in this case.
• Old models are pervasive. A learner is likely to be surrounded by others who also hold beliefs regarding DNR discussions as legally necessary. Developing the learners' confidence with the new approach will be the best way to have them become effective role models within their programs.

SUGGESTED PROCEDURE:

Assess Pre-Existing Beliefs. All learners will bring assumptions about DNR orders and discussions. Before you begin with a new paradigm, it is important to elicit and address the assumptions that learners currently hold about DNR orders and discussions. A variety of strategies can be used:

• Ask. "How do you normally approach your DNR discussions?"
• Use a pre-test quiz. List common beliefs about DNR orders and discuss whether or not they true.
• Role-play a DNR discussion and debrief observations.
• List common phrases that appear in DNR discussions (e.g. "do you want us to do everything possible to bring you back if your heart stops?"). Discuss the risks and benefits of the phrases.

Often if any of these strategies are used in a small group setting, at least some participants will take the position that the existing approaches to DNR discussions are not effective or comfortable for either the patient or the physician. There are several important teaching points to make during this opening segment:

• Do we ask hospital patients, "if you get an infection, would you like me to treat it with antibiotics?" Why are DNR discussions treated differently than other clinical decisions?
• Do patients really want CPR? Discuss what it is that patients want and what choices it is reasonable for them to make (see discussion below re: the new approach). Are there misconceptions (e.g. CPR success rates on TV) that lead patients to request CPR?
• When is it appropriate for physicians to give recommendations to patients?

Introduce a New Approach. Because of the legal and ethical traditions that emphasize patient's rights to choose and direct their health care, DNR discussions have become more about presenting choices and obtaining patient signatures. Having meaningful choices is an important aspect of respecting patient autonomy; however, in many cases, resuscitative efforts do not represent a reasonable care option for patients. Particularly when a DNR discussion is approached as a menu of treatment options, decisions about individual treatments become even more meaningless (e.g. "I would like chest compressions but no shocks please.") More often than not, what patients want is not the treatment itself, but the outcome of the treatment (e.g. "I would like to be on a breathing tube, but only if it gets me over a temporary bump and I can be restored to my former level of function.") No one really wants CPR, but they might be willing to tolerate it if it achieves their goal of extending life.

With these distinctions in mind, it becomes more productive to focus on the outcomes wished for by the patient. The clinician can then determine if there are any available medical interventions that can help the patient achieve those outcomes. In some cases, the answer might be: "I wish I could promise you a full recovery, but I am afraid I cannot." If the clinician can elicit the patient's goals, hopes, and fears for the end-of-life, the clinician can make an informed recommendation to the patient based on what will be beneficial, from the patient's perspective. The steps of our recommended protocol is presented below in Box 1 and also in the Oncotalk Module: "Talking about Advance Care Plans and Do Not Resuscitate Orders."

Practice or Observe. Address Barriers Again. Even after a discussion, once the learner gets back in the hot seat with the patient, old habits are likely to re-appear. For example:

LEARNER: I wanted to talk with you about decisions that may need to be made around the time of death. It sounds like, and please correct me if I am wrong, but that overall your goals are to stay as comfortable as possible and continue to visit with friends and family as long as you can, is that fair to say? Ok. A lot of times we like to talk to people about what happens if something should come up, if your heart would stop beating and that you would stop breathing, you know, pretty much at the time of death. And there's a couple ways you could go. One of those is to do a full resuscitative effort, which would involve things like chest compressions, potentially even shocks to the heart, being put on a breathing tube, that sort of thing. And that's awfully, that's a lot to go through at the end. Especially when the chance of having a major success is quite low, probably a zero percent chance of bringing you back and that you would be able to live through all that basically. And so, I think in keeping with your goals, at that time if your heart were to stop beating, or you were to stop breathing, maybe letting nature take its course. Would that be a more reasonable approach to take?

In a role-play setting, this is a good place for the faculty person to time-out and check in with the learner to see how it is going. It could be that the learner realizes they are right back into old habits again with the discussion. This learner went back to listing the procedures that could be done in a resuscitative effort and posed the question to the patient as a choice, even as she characterized the choice of full resuscitation as being highly unlikely to succeed. It could be that the learner needs help identifying the sticking point. By diagnosing where the learner is first and eliciting self-reflection, your feedback will be more effective. An example of a faculty intervention could be:

Experience Success. If the learner can experience having a DNR discussion that is led by the patient, rather than by the physician, the teaching points will become much more clear to the learner and to others observing. For example, the second role play might go something like this:

LEARNER: I was just wondering what you have thought about that, or what you've imagined for the time of your death.

STANDARDIZED PATIENT: I'd certainly like to be at home. I hope that it's not painful. And it'd be great to have my boyfriend here. Can you arrange that?

LEARNER: I think that sounds peaceful and very appropriate, and I'd certainly like to help you make that happen. Certainly I would like to keep you comfortable, as you said, and it would be nice if your boyfriend was here. There are things we can do to facilitate that. One thing that comes up a lot of times, and it's really not compatible with what you are describing to me, is that sometimes when someone stops breathing we rush in and do aggressive resuscitation, put people on breathing tubes, do chest compressions, shocks to the heart, things like that. Really I don't think that is a good idea, with what you are telling me here, and just having more of a loving, peaceful, nurturing environment at the time of death sounds more like what you are telling me, rather than doing tubes and machines, and things like that.

STANDARDIZED PATIENT: I'd agree.

And the faculty can help reinforce the positive experience:

Give Permission to Make Mistakes. When learning any new skill, it is easy for learners to get frustrated with early awkward attempts. Faculty can help encourage learners to work through this stage by affirming that everyone takes missteps during patient encounters. The important tool for learners is to know they have the ability to "repair" the conversation. For example:

LEARNER: Now, as you know, your death is pretty close. Have you had any thoughts about what you would want to do, or what your goals would be?

STANDARDIZED PATIENT: It feels strange to talk about goals about dying.

LEARNER: Let me put it this way, what are the things that are important for you right now?

Address the Emotion. DNR discussions are difficult for physicians because they can bring up feels of loss or even failure. Physicians can also worry that by raising the topic of death, they will upset the patient. For the physician and the patient, it is natural to be sad during these discussions. Most patients are very aware that they are dying and welcome the chance to talk about their hopes and fears with their physicians. For example, during one discussion following a DNR discussion with a patient the participants identified the following emotions:

VARIATIONS FOR DIFFERENT TEACHING SETTINGS

Each of the above steps can occur with real-time patient care. Take a few minutes in a quiet meeting room before you see the patient. Ask the learner to describe her usual approach to DNR discussions. Are they comfortable for her? Invite the learner to approach it differently this next time, by focusing on the patient's goals and affirming the treatment being provided to meet those goals. Observe the encounter and give feedback at the conclusion.

PEARLS

• Behavior change is going to take time.
• Embracing underlying assumptions about the purpose of the DNR discussion will help facilitate the transition to new communication skills
• Focusing on patient goals and how the care team can and cannot meet them can help the discussions go much more smoothly for both patient and physician. Respect for patient autonomy can be met more easily through a goal-centered discussion than by obtaining a signature.

REFERENCES

• Diem SJ, Lantos JD, Tulsky JA. Cardiopulmonary resuscitation on television. Miracles and misinformation. N Engl J Med. 1996 Jun 13;334(24):1578-82.
• The AM, Hak T, Koeter G, van der Wal G. "Collusion in doctor-patient communication about imminent death: An ethnographic study." BMJ, 2000;321:1376-81.
• Tulsky JA, Chesney MA and Lo B. "How do medical residents discuss resuscitation with patients?" J Gen Intern Med 1995;10(8):436-42.
• Wenger NS, Phillips RS, Teno JM, Oye RK, Dawson NV, Liu H, Califf R, Layde P, Hakim R, Lynn J. Physician understanding of patient resuscitation preferences: insights and clinical implications. J Am Geriatr Soc. 2000 May;48(5 Suppl):S44-51.