++
Mr N's daughter, Ms N, his primary care physician, Dr W, and his surgeon, Dr V, were interviewed by a Perspectives editor.
++
MS N (DAUGHTER): For about 2 weeks prior to the final event there was a rapid change in his health. He'd been complaining about some abdominal pains and there was a lot of bloating. One weekend he was in the hospital, and the next weekend he was back in the hospital again. He was really distended. He was in a lot of pain with pressure…We just watched his stomach swell. A lot of gas and bloating. They suspected, but were never able to conclude, [he had] pancreatic cancer. But, once he had the surgery and they opened his abdominal cavity, it had spread and every major organ was affected by the cancer. Even the lining of his abdominal cavity was laden with the cancer.
We had discussed with a doctor at the hospital about a surgery to relieve an obstruction in the intestine because he was unable to pass gas, have a bowel movement, or anything. There was always talk about it, but I was discouraged because there was never action taken, even though he was going through the distention and he hadn't had a movement for weeks. So, he just kept swelling, and his stomach was really big, and then the next morning his bowel actually burst. So, they took him up for emergency surgery, and he came through the surgery, but they couldn't sew him back up because of the cancer. So, they put a mesh covering over him and left him open and sent him to ICU to recover. Then they started a respirator for life support and IV [intravenous] fluid because his blood pressure had dropped. A whole lot was going on. [A]fter the surgery, they were hooking him up to everything and they couldn't give him pain meds because…his heart would stop. They just kept him hanging on in agonizing pain for hours, even though I had the do-not-resuscitate document in my hand.
DR W (PRIMARY CARE PHYSICIAN): I first met Mr N in my primary care clinic. He had some stomach pain, and a few weeks later, about 2 months before his final hospitalization, he had an episode of gastrointestinal bleeding and went to another hospital where CT scans showed he had a pancreatic mass extending to his large colon. I asked him what his understanding of what his diagnosis was. I said that…the most likely diagnosis was pancreatic cancer…I asked him about what his thoughts were about his prognosis, and he said that he was uncertain. I said that if it did turn out to be pancreatic cancer that had spread to his liver the therapies that we had to extend his life would be very limited, and his prognosis could be measured in months, not years.
At that first meeting we were alone, and he took the news as disturbing. I asked him if he wanted me to call a family member, and he was quite clear that I should not. We made an appointment within a few weeks. I referred him the palliative care program here…The plan was that he would undergo a biopsy elsewhere. He would tell me the results of that biopsy. While I would serve as his primary care doctor, his specialty care would be outside the hospital. In a hypothetical way, he [had] said that he would not want his life prolonged if it meant that he could not be independent and live at home and speak with his family. He valued comfort greatly, over longevity…In the next visit we talked about the possibility of a bowel obstruction. He was complaining of bloating and abdominal pain. I said I was concerned that the tumor might be obstructing his large colon and that he might need a surgical consultation and we should talk about whether he really wanted surgery or not in such a situation. He said, again, that he wanted to talk about it with his outside doctor. I recommended that he see a surgeon. He said that he would try to get that done, but his main concern at that point was to get a biopsy because we still didn't have a tissue diagnosis. During his stay in the hospital, I began to talk with him in a more urgent way about what he would like done if his situation should deteriorate acutely, in particular, if he developed a perforation or a true obstruction. Would he consider surgery now? He was very reluctant to make up his mind, even after extensive discussion. He said he wasn't sure if he would want emergency surgery. I told him emergency surgery was a lot more risky, and said I hoped it wouldn't come to that. He said he hoped it wouldn't come to that, too. Here, again, I think I could have done a better job. The chances of a patient recovering from emergency surgery after a perforation when they have metastatic pancreatic cancer, I think are pretty slim. I think I should have made that a little bit more clear to him, in retrospect.
DR V (SURGEON): The surgical team was not fully aware of anything that was going on with his condition. Once we opened, we found that he had free air. He had almost a liter of ascites. There was a perforation in his cecum. He had hard cancer, just about everywhere, such that I couldn't even mobilize his intestine. We stapled off the perforation because I couldn't mobilize the cecum and the right colon safely. I dissected the small bowel peripheral to that and made a loop ileostomy. It looked also like he had a near full obstruction of his left colon, so at that point we brought up a looped colostomy and we put in a G-tube. After the operation was the first time I had a chance to really talk and get a feeling of what [the family] knew and what they didn't know. I told them…that the prognosis was not very good. [Ms N] said that…her dad was going to be very disappointed if he ever woke up because this aggressive care was against his wishes. I tried to hide my shock and frustration.
++
Patients and their families struggle with myriad choices concerning the medical care that precedes death, ranging from the aggressiveness of treatment to the site of care. These struggles may generate conflict and dissatisfaction with the medical team.1,2
++
Clear decision making contributes to quality of life at the end of life, and its absence may lead to worse outcomes.3,4 Despite a series of conversations and, ultimately, the presence of a written advance directive, Mr N, his family, and his physicians struggled with decisions about his care, ending with traumatic conflict. Confusion about goals of care and the inability to make certain decisions meant that Mr N experienced unnecessary pain. (See also Chapter 4, Dealing With Conflict in Caring for the Seriously Ill.)
++
Even when physicians and patients try to plan for the future, advance directives are easily misunderstood or misinterpreted.5 This discussion reviews challenges in creating and implementing advance directives; elucidates factors that affect clear decision making, including trust, uncertainty, affect, hope, and the presence of multiple clinicians; and offers practical techniques to ease decision making at the end of life.
++
Advance Directives—Promising But Insufficient
++
DR W: [The cardinal issues are] to discuss with the patient, early and often, the likely consequence of his disease and the decision points that are coming up and get some strong signals as to what his preferences would be in those circumstances. I think physicians are much better at predicting that than patients are. Patients are often reluctant to make the decisions…In particular, if he ended up with a perforated viscus and emergency surgery was being considered, I should have recommended that he not undergo that procedure, because it was not consistent with his goals.
MS N: We had decided about 2 and a half months [earlier], do not resuscitate. We had consulted with an attorney who had drawn up papers. If my father could not have good quality of life, we agreed that it was selfish to have him here in a vegetative state or in pain constantly. We discussed this over and over again.
++
Advance care planning is the process by which patients, together with their families and health care practitioners, consider their values and goals and articulate preferences for future care. Written advance directives formalize these preferences and include living wills or other statements of patient preferences and durable powers of attorney for health care, which name health care proxies. DNR orders are written by physicians to operationalize one specific set of preferences articulated by patients and their proxies. Unfortunately, with few exceptions, the introduction of advance directives and the US federal Patient Self-Determination Act have had little demonstrated impact on resuscitation events.6-13 Some of the barriers to successful implementation have been procedural when, for example, documents are not available when needed. More importantly, problems arise with deciding in advance about specific interventions,14 the adequacy of communication,15 the willingness of clinicians to follow patient preferences,11,16 and patient and family misunderstandings about the process.
++
Both procedural and communication difficulties occurred with Mr N. First, the document his daughter referred to as a DNR order was in fact an advance directive form that stated only global preferences. Mr N clearly valued quality of life over quantity of life, but his physician was unsure whether the specific instance of perforated bowel met the patient's criteria for withholding aggressive treatment.
++
This case is unusual in the degree of conflict between family and physicians, but it helps to illustrate that advance directives are not the same as DNR orders and generally do not provide instruction for specific circumstances. Rather, they should facilitate discussions of goals of care and consideration of all of the elements of quality at the end of life from symptom control to limitation of treatment. Because some of this care may occur after the patient has lost decision-making capacity, such discussions about advance care planning should be held with the patient, family, and physician together to permit discussion and clarification of treatment goals specific to the patient's condition.
++
Decision Making Must Be Responsive to Changing Scenarios
++
DR W: While he was awaiting his CT scan,…his pressure dropped. He became unresponsive and was intubated. [After the surgery, the family was] distraught to see him in the ICU with a tube in his mouth. They were questioning if this was what he would really want.
DR V: The primary team attending came by, and I actually talked to him, as we were just about to walk out of the [emergency department]. He told me that the father didn't really want anything aggressive, but he hadn't really finished his conversation to the fullest extent. So, at that point, that was as much as I knew.
++
This patient's outcome, including intubation in an ICU, resulted from his stated preferences. Mr N had been unable to decide against the surgery, perhaps holding out hope that he did not have metastatic cancer or that the operation would buy him some time. When confronted emergently about whether to operate, his daughter made choices consistent with her father's expressed wishes. On the other hand, despite the patient's presumed diagnosis before surgery, the likelihood of widespread carcinomatosis and slim chance of recovery apparently were not discussed with the daughter, making the postoperative course a disturbing surprise for the family. In the emergent situation, the surgeons likely knew little about his case, but discussing with the family the possibility that Mr N would not recover from surgery and the palliative care options would have better prepared the family for the eventual outcome. (See also Chapter 19, Complexities in Prognostication in Advanced Cancer.)
++
After the surgery, the prognosis was clear. Because the family did not expect the eventual outcome, and the patient could not participate in the decision-making process, a family meeting was in order to allow clinical status, prognosis, and treatment goals to be reassessed and, if necessary, modified. As new information emerges, decision making must be fluid.7-19 Time-limited trials of therapy can be very useful to assess whether particular interventions are effective while defining an end to that treatment if goals are not met.20 Although communication could have been substantially improved and the family better prepared, ultimately Ms N appears to have acted as her father would have in consenting to the surgery, but then wanting withdrawal of support.21 The tragedy occurred when the withdrawal of support was delayed, and the patient underwent an unnecessary and traumatic resuscitation attempt.
++
When Is a Document Needed?
++
DR W: [After the surgery, the family] mentioned that he had an advance directive, that a lawyer was working on it. I said that it would be really important to get that advance directive now. I went to my office, thinking I would get the fax. He went into an abnormal rhythm, and because he was still full code,…the MICU [medical ICU] team began resuscitation procedures.
MS N: He kept saying he was in pain, he was in pain. They called a code blue,…and I just started screaming, "Why won't you listen? I am giving the document that says he does not want this. He is in pain. He's talking through morphine telling you to let him die, and you're ignoring his wishes." It was not until Dr W came in and kind of took over on his own. He was only a consulting physician, and he was not supposed to come in and take over. But it wasn't until he came in and tried to relieve my father's pain that everything happened. He removed the intubation tube and minutes later my father died. I was there. I stayed until the end. Although I'm grieving, and I'm sad he's not here, it was a tremendous relief to see him out pain.
DR W: [After the resuscitation] I had the advance directive. A no-code order was written by the MICU doctor. In one of the most horrendous experiences of my time as a physician, he kept trying to pull the tube out, saying that he was in a great deal of pain. The family decided that he didn't want the tube any more, and they decided to extubate him, which of course, was upsetting to the MICU team, which had just met him. I advised the MICU team that this was not a spur-of-the-moment decision on the part of the family or the patient. The tube was removed, and he died not long after that.
++
Unfortunately, DNR orders are frequently not written even when this is the patient's preference.16,22,23 If one thing could have been done differently to change the outcome of this case (apart from providing adequate pain control), it would have been to write the DNR order much earlier—ideally before the surgery, or certainly immediately after, when the diagnosis was clear.24,25 Resuscitation outcomes are extremely poor in the setting of metastatic malignant disease.26-28 Once the operation was completed and the patient was in the ICU, there was no reason to hold up the DNR order waiting for the advance directive document.
++
Written advance directives are useful when there is disagreement within a family, when there is a conflict between the family and health care team, or when the patient assigns a nontraditional family member (eg, friend or same-sex partner) as the surrogate. Family members may choose more aggressive life-sustaining measures than patients prefer. A systematic review of 16 studies found that surrogates predicted patients' treatment preferences with only 68% accuracy, and 2 recent studies from Japan and Taiwan suggest that such lack of concordance between patients and family members is common in these Asian countries as well.29-31 If the patient's preferences are known and understood by the family and team through an oral advance directive, in most states the written document is superfluous. However, to facilitate communication among members of the health care team, it is important to document the patient's preferences and how they were determined. In the United States, if the patient has completed a durable power of attorney for health care form, this form can help health care professionals know which family member to rely on for inpatient medical decisions when the patient cannot make these decisions.
++
Key Elements of Communication
++
Although Mr N's physicians talked to him and his family on multiple occasions about his illness, the likely prognosis, and options for care, ultimately the patient's decisions contributed to a tragic outcome. Factors in addition to words influence how the messages of a medical encounter are interpreted and can complicate a seemingly straightforward transaction.
++
MS N: I think, in all honesty, that they were more interested in protecting their own jobs, not wanting to be responsible for my father's death.
++
Trust, the confidence that a clinician is acting unfailingly in the patient's interest, is fundamental to effective medical care, particularly at the end of life. Ms N ascribed unprofessional motives to the physicians' actions, a belief that may have provoked her extraordinary response to the resuscitation attempt. Although conflict among health care professionals, patients, and families is not uncommon,32 the establishment of trust fosters successful resolution of conflict through open and honest communication.2,20,32,33
++
Frequently, a medical team and a hospitalized patient have not had a lengthy relationship, so physicians must rapidly create a trusting environment. In one study, women with cancer stated that feeling that a physician cared for, understood, and respected them; gave them enough time; listened; and was open and honest encouraged them to trust physicians and accept their recommendations.34 Such communication can be accomplished quickly. Women who observed an oncologist add 40 seconds of compassionate language to his usual encounter felt less anxious and thought the physician cared more about the patient.35 The manner in which one interacts with a patient appears to be the most important factor to establishing trust.36 An empathic, patient-centered style and a forum that permits an open discussion of grievances may be the best approach (Box 2-1).37
++
++
DR W: One of the things I learned is that presenting them with information is one thing, but assisting them in the decision making is another thing. I should have made much stronger recommendations.
++
Had Mr N known for certain that he had metastatic cancer, he likely would have been inclined to accept a purely palliative approach to care and a DNR order. Had Ms N known about the likely outcome of surgery, she may have declined the intervention. However, the diagnosis and prognosis were only certain after surgery, and then she was able to make the decision to withhold further treatment.
++
Uncertainty characterizes all medical decision making, and physicians must help patients manage it.38-40 Surveys consistently show that most patients wish to receive as much information as possible,41,42 perhaps as a way to cope with uncertainty.39,40 However, patients who are at either extreme of preferences for decision making—those who want to be entirely in control of decisions and those who defer entirely to their physicians—are less satisfied43 and are more anxious than those who take an intermediate role.44
++
This work suggests that most patients prefer to participate in decision making but wish to receive a physician's advice about recommended options. Physicians must find the balance between conveying the ambiguity that at times may cloud medical practice and helping patients find the best options for them.
++
On reflection, Mr N's physician believed that he could have given a clearer recommendation. Physicians tend to discuss uncertainty but frequently do not give recommendations,45 likely because of concerns about appearing overly paternalistic. However, such guidance is often welcome,42,44 particularly when based on the patient's elicited values.46 In this case, the physician, aware of the patient's goals, might have advised him of the very low likelihood that certain courses of treatment would allow him to achieve these goals and could have strongly recommended a DNR order.
++
MS N: It wasn't until I got emotional and started screaming and crying that anyone noticed. I don't think until then that they heard the concerns.
++
Attention to affect, the feelings associated with the content of conversation, is key to resolving communication difficulties at the close of life.47 After receiving bad news, most people are so emotionally overwhelmed that they are unable to comprehend very much about the details of the illness or treatment plan.48,49 Emotion affects processing; people who are in negative moods may pay more attention to the delivery than to the content of a message.50,51 Clinicians should attend to affect while sharing information to increase the likelihood that patients will hear what they are trying to say. Unfortunately, conversations between physicians and patients often transpire only in the cognitive realm; physicians frequently miss opportunities to enter the ambit of emotions.49-56 Rather than using facilitative communication techniques, such as open-ended questions or empathic responses when inquiring about psychosocial issues, studies have shown that physicians and nurses often impede discussion by changing the subject or ignoring these concerns.57,58 One study showed that, even in a hospice setting, only 40% of patient concerns were elicited.59 Patients with cancer tend to disclose fewer than half of their concerns,58,59 leading physicians to inaccurately assess distress.60 Communication techniques that are open-ended and focus on patients' and family members' emotional states are more likely to elicit patients' concerns about symptoms, which can then be treated (Box 2-2).61,62
++
++
In the present case, whether the patient's or his daughter's emotions were ever acknowledged directly is unknown. It is easy to imagine that Mr N's indecision was a function of coping, at a relatively young age, with a sudden and devastating diagnosis. Similarly, his daughter was grieving terribly as she was about to lose her father and yet was distracted with worry about contacting her lawyer to get the advance directive faxed. Statements such as "Are you feeling scared?" "This must be terribly hard," or "I can't imagine what you're going through, but I'm impressed with how you've been able to cope" name and acknowledge the emotion and could have been applied effectively earlier on. They align the physician with the patient before engaging in the difficulties of decision making.
++
DR V: The surgical team and the primary care team talked to the daughter and told her that we were going to explore and if there was anything that we could fix, then we'd fix it…She herself said that it was a selfish thing on her part because she wanted her dad around a little longer.
++
Virtually no one wants to die. We may be able to promise the most peaceful of final days, but 55-year-old patients rarely acquiesce to death without a fight. Hope is the frame within which they construct their future. It may be a desire for a particular outcome, or it may be, more broadly, trust or reliance. Physicians rightfully struggle to promote hope in the patient with advanced disease and to support a positive outlook.41 Yet such an approach is frequently accompanied by an unjustified fear that discussing death may distress patients.63-68 Physicians frequently convey overly optimistic prognoses or do not give this information at all.69 One study of ICU family conferences found that in more than one-third of cases, physicians did not discuss the patient's prognosis for survival at all.70 Fearing the loss of hope, patients frequently cope by expressing denial and may be unwilling to hear what is said.71 Patients with more optimistic assessments of their own prognosis are more likely to choose aggressive therapies at the end of life.18,72
++
Physicians should recognize that it is not their job to "correct" the patient's hope for a miracle.73 The key question is whether this hope is interfering with appropriate planning and behavior. Clinicians, at their best, can provide an empathic, reflective presence that will help patients marshal and draw strength from their existing resources (Box 2-3). A recent systematic review documented that many patients can maintain a sense of hope despite acknowledging the terminal nature of their illness.74 More practically, together, the physician and the patient can "hope for the best but prepare for the worst."75 Helping the patient and family manage their hope and their resources in a realistic way may leave the family in the best possible shape after their loss.
++
++
Communication Among Multiple Clinicians
++
MS N: I think that if the doctors who were working in the ICU and the emergency department had more communication with the palliative care team that would have been better because we had discussed over and over again my father's wish for a good quality of life. It kind of goes null and void when you have a whole new set of doctors who…just come in and start treating the patient.
DR V: Most times in surgery, unless we are specifically, explicitly told not to go ahead and give the most care to keep somebody alive, that's what we do. Since I was so late to this situation, I didn't know enough to step back and say, "What are we doing?" When I talked to another doctor who had seen the patient about 2 weeks earlier, he said that they had the discussions with the family, but they were never completed. This could have been prevented. His suffering could have been prevented. The terrible taste in everyone's mouth could have been prevented.
++
Sharing medical information and coordinating care by multiple physicians can be challenging. Working collaboratively to honor a patient's preferences is even more difficult when clinicians become involved only late in the illness and during emergencies. Collaboration requires clear, consistent, physician-to-physician communication. Medical record notes should include lucid discussions about patient preferences and goals for care and, during critical illness, regular family meetings with the entire care team.76,77 It is useful to begin such meetings with a statement such as "Over the past couple of days, a lot of doctors have been involved with your (or your loved one's) care. Tell me what others have been telling you about the illness."
++
Although the burden of maintaining continuity in our health care system ought not fall on patients and families, physicians can suggest that they keep notes from meetings with their physicians, share these with subsequent clinicians, and request that physicians communicate with each other to increase the likelihood that preferences will be honored.
++
From a systems perspective, electronic medical records provide a potential solution.78 For example, the US Veterans Health Administration electronic medical record system allows placement of clinical warnings that link to progress notes, DNR orders, and scanned advance directive documents. Another successful intervention is the Physician Order for Life Sustaining Treatment (POLST), a preprinted and signed physician's order specifying treatment instructions in the event of serious illness, including cardiopulmonary resuscitation, levels of medical intervention, antibiotics, IV fluids, and feeding tubes.52,79 Used for community-dwelling, frail, elderly persons and those living in nursing homes, care has been consistent with the form in more than 80% of cases studied,52,79 and their specificity yields greater success at achieving palliative care goals than do advance directives.