Cases & Commentaries
Palliative Care: Comfort vs. Harm
- Recognize errors may be difficult to identify in palliative care.
- State that medication errors and errors in communication are common in palliative care.
- Describe key principles regarding the use of opioids in palliative care.
- Recognize the principle of double effect and the implications for potential medication errors.
An 83-year-old man with chronic kidney disease and end-stage congestive heart failure (CHF) with a severely reduced ejection fraction was admitted for an exacerbation of his CHF. It was his third admission in the previous 4 weeks. During this admission, he improved slightly with treatment, but his overall prognosis remained poor. In discussions with the patient and his son, the patient expressed that he would like to focus on comfort and pursue hospice care. His life expectancy was believed to be weeks to months, based on his comorbidities and frailty.
While he was still hospitalized, comfort measures were initiated. His diuretics, statins, beta-blockers, and other treatments for his heart failure were stopped. He seemed more at peace after making the decision and was able to spend time with his family. Wanting to ensure the patient's comfort, the inpatient team wrote for a standing dose of intravenous hydromorphone every 4 hours.
On the night shift, the bedside nurse went to see the patient at 3:00 AM to administer the next scheduled dose of hydromorphone. The patient appeared to be sleeping but woke up when the nurse came in the room. She asked him, "Are you having any trouble breathing?" He responded, "A little bit," but he remained drowsy. The nurse administered the scheduled dose of intravenous hydromorphone.
The following morning, the palliative care attending saw the patient at 7:00 AM and found him obtunded, with shallow respirations and a low respiratory rate. She believed he was close to death and contacted his family, who came to see him. The patient continued to be obtunded and unarousable, but his respiratory rate and effort improved. He survived for another 5 days in the hospital, requiring a few doses of intravenous hydromorphone for comfort. He died peacefully with his family at the bedside.
In reviewing the case, the palliative care team felt that the overnight dose of hydromorphone may not have been indicated and may have been excessively large. Although the patient did not suffer and died peacefully, they believed that his life may have been shortened. At the same time, the review team remained sensitive to the bedside nurse's desire to relieve suffering in a patient who wanted to focus on comfort. They wondered how best to educate providers about the balance between providing adequate pain control at or near the end of life and minimizing harm.
by Ralf Jox, MD, PhD
This case beautifully illustrates the current challenges in palliative care in three ways. First, it highlights that the last phase of most people's lives is an extended period of geriatric multimorbidity.(1) Second, it shows the difficulty of making treatment decisions for patients with chronic illnesses due to the unpredictable disease trajectory and potential prognostic uncertainty.(2) Third, it underscores the potential for medical errors in end-of-life care and the difficulty in evaluating these errors.
Caring for patients at the end of their lives is demanding, necessitating an interdisciplinary and multiprofessional approach. In palliative care, awareness of the potential for medical errors is less developed, critical incident reporting systems are rare, and empirical research on patient safety is scant.(3) There is no reliable study on the prevalence of critical incidents (errors and near-misses) in palliative care.
One unique feature of palliative care is that the harmful consequences from errors may be harder to identify than in other types of care. Insufficient symptom control is often explained by the seemingly refractory nature of the symptoms, adverse medication effects are attributed to the underlying disease, and hastening death (if at all recognized as such) is viewed as an expected outcome and may comport with the patient's wish to die.(4) While communication errors are common, these too may be hard to identify given the challenges in assessing quality of communication in a structured way.(5,6) Palliative medicine providers have highlighted that patient safety has generally not been part of their training.(5,6)
Given the complex nature of end-of-life care, many situations are prone to errors. The very decision to change the goal of care from sustaining life to a comfort-only approach is one such situation. Errors surrounding these decisions can be made in ethicolegal justification, communication, and medical management. Ethically and legally, foregoing life-sustaining measures can be justified by the lack of beneficial treatment effects to the patient, the explicit or presumed autonomous refusal of the patient, or both of these reasons.(7) To base treatment or the withholding of treatment on financial or personal reasons, disregarding or even contradicting the patient's autonomous will, would clearly be a form of medical error. In the present case, the decision to change the goal of care was made appropriately, based on both the questionable benefit of life-sustaining crisis management and the clear expression of the wishes of the well-informed patient.
The decision-making process ideally involves thorough and clear communication about the details of comfort-oriented end-of-life care. Health professionals should inform the patient and, with the agreement of the patient, the family about what they can likely expect.(8) This includes the way of dying, expected symptoms, and the options to ease these symptoms. Examples of errors in communication would be to tell a patient an exact number of days or weeks he or she will live, or to promise a patient that dying will be pain-free and any distressing symptom can be completely prevented.
Individual patient and family wishes (including in the psychosocial and spiritual domains) should be proactively explored. For example, by asking: "What do I need to know about you as a person to take the best possible care of you?" Advance care planning should document the way patients want to be treated in the event that they lose decision-making capacity, which often occurs in the last days of life.(9) This is particularly relevant in patients with chronic organ failure like ischemic heart disease or renal insufficiency.(10,11) In palliative care, it is crucial that patient rounds be done with physicians and nurses together and that regular multiprofessional case discussions with adequate documentation convey the crucial information to everybody on the care team, including the night shift. Failure to explore patient and family wishes, document advance care planning, or communicate key decisions should all be seen as adverse events.
The practical management of comfort-oriented care includes limiting diagnostic and therapeutic measures focused on prolonging life, and stopping all medications that are not necessary to promote comfort.(12,13) Long-acting, preventive drugs (e.g., statins) are usually not indicated in end-of-life care. Drugs with a life-sustaining effect should be stopped unless they are indispensable to ensure comfort. In the present case, it was reasonable to withdraw the statin, beta-blocker, and diuretic.
On the other hand, effective management of symptoms often requires the addition of new medications. In general, starting new comfort-oriented medication should be done to treat symptoms, not as a preventive standard procedure in case symptoms might occur. The exception may be specific situations, such as weaning a patient on mechanical ventilation, in which significant acute distress (in this case, acute dyspnea) is predictable. In fact, palliative care professionals often observe that the first few days after the withdrawal of life-sustaining medication are a time of symptom relief and increasing well-being. Regular screening for symptoms should be done and any existing symptoms should be meticulously assessed, ideally with the help of validated symptom assessment instruments such as the Edmonton Symptom Assessment Score.(14) The case description casts doubts on whether the treatment with hydromorphone was based on an adequate symptom assessment.
Dying of end-stage heart failure and renal failure can be relatively peaceful, but it may also require specialized palliative care. In this case, withdrawing diuretics and beta-blockers might result in rebound pulmonary edema leading to dyspnea, rebound lower extremity edema could lead to discomfort, or new myocardial ischemia could result in chest pain. Therefore, symptom-focused medications may be necessary. For this set of symptoms, an opioid that has antidyspneic as well as analgesic effects would be indicated. In patients with renal failure, hydromorphone is an acceptable choice as it is mainly metabolized by the liver, but its metabolite can accumulate and may cause neurotoxic adverse effects.(15) Other opioids for patients with chronic renal insufficiency are tramadol (for mild to moderate pain), fentanyl, alfentanil, and methadone (for strong pain).(16,17)
Many different errors may arise in prescribing opioids in palliative care. In general, and especially for a geriatric patient with renal insufficiency, opioid treatment should be started at low doses (even more so in an opioid-naïve patient), and any increases should be titrated based on the symptoms ("start low, go slow"). Whenever possible, opioids should be administered orally or subcutaneously, not intravenously, as the need for peripheral or other intravenous access increases the risk of infection and can limit desired patient activity. Considering that hydromorphone is 7 times more potent than morphine, the starting dose for hydromorphone in an older patient with renal insufficiency would be 1.3 or 0.65 mg orally every 4 hours or 2–4 mg subcutaneously over a period of 24 hours (to convert from parenteral to oral, multiply by 2.5).(18) We are not informed of the dose administered in this case, but it may have been excessive for this patient.
In addition, it is usually not necessary, and may even be counterproductive, to wake the patient at night in order to administer an opioid. If the drug is administered orally, a somewhat higher dose or an extended-release formula can be given late in the evening before sleeping. If it is administered subcutaneously, the best way is to use a syringe driver that continuously releases the medication, both day and night. One of the most common errors in end-of-life opioid treatment is giving intravenous opioids in a standing dose without careful titration of the dosage against the symptoms.(19)
The team had concerns whether the hydromorphone had shortened the life of the patient. This is difficult to assess and depends on the dosage as well as on the general health condition of the patient. Generally, opioids are rather safe drugs. If they are accurately dosed to treat pain or dyspnea, they do not hasten death.(20) In the United States, as in many other countries, it is legally allowed to use opioids as symptom treatment even if they could, as an unintended consequence, shorten the dying process.(21) This appraisal is based on the relative importance of relieving suffering in dying and on the so-called doctrine of double effect, according to which it is legitimate to accept an unintended harmful effect (shortening life) as a potential consequence of an action that pursues an intended beneficial effect (symptom relief). Although double effect has been deemed to be ethically appropriate, it may obfuscate the fact that health care professionals also have the obligation to minimize unnecessary harm, including harm they do not intend but merely foresee.(22)
Adverse effects of opioids should always be carefully monitored. In the present case, the possibility that the patient was obtunded due to an overdose was identified late, but it is hard to evaluate drowsiness when waking a patient in the middle of the night. In the end, it is unclear if the patient experienced an adverse event. He died peacefully according to his wishes. It is not known whether he wished to stay alert longer to discuss any important matters with his family. However, one could imagine a scenario in which giving opioids to an obtunded patient did rise to the level of adverse event.
Even if there was no clear harm to the patient or the family in this case, preventing errors and increasing patient safety is crucial in end-of-life care. The fact that the care team chose to review the case means that they've already taken the first step toward patient safety. Critical incident reporting systems can be an effective way of preventing errors, but it may be prudent to develop systems specific for palliative and end-of-life care.(23) Palliative education for health workers should include explicit training in assessing and reacting to critical incidents and learning how to communicate with patients, their family caregivers, and colleagues about errors. Palliative medicine providers may also benefit from web-based interactive training tools focused on errors and patient safety.(24)
- Errors are prevalent in palliative care, but may be harder to identify than in other health care fields.
- Palliative care is particularly prone to medications and errors in communication.
- At the end of life, opioids should only be given based on thorough symptom assessment, titrated based on the symptoms, and administered in the least invasive route, preferably orally or subcutaneously.
- The doctrine of double effect can legitimize using opioids for symptom control while accepting a potential life-shortening side effect, but it should not obviate the general goal of preventing unnecessary harm to patients and families.
Ralf Jox, MD, PhD Assistant Professor Institute of Ethics, History and Theory of Medicine University of Munich Munich, Germany
Faculty Disclosure: Dr. Jox has declared that neither he, nor any immediate members of his family, has a financial arrangement or other relationship with the manufacturers of any commercial products discussed in this continuing medical education activity. In addition, the commentary does not include information regarding investigational or off-label use of pharmaceutical products or medical devices.
1. van den Block L, Albers G, Pereira SM, Onwuteaka-Philipsen B, Pasman R, Deliens L, eds. Palliative Care for Older People: A Public Health Perspective. Oxford: Oxford University Press; 2015. ISBN: 9780198717614.
2. Holley JL. Palliative care in end-stage renal disease: illness trajectories, communication, and hospice use. Adv Chronic Kidney Dis. 2007;14:402-408. [go to PubMed]
3. Dietz I, Borasio GD, Schneider G, Jox RJ. Medical errors and patient safety in palliative care: a review of current literature. J Palliat Med. 2010;13:1469-1474. [go to PubMed]
4. Kiesewetter I, Schulz C, Bausewein C, Fountain R, Schmitz A. Patients' perception of types of errors in palliative care—results from a qualitative interview study. BMC Palliat Care. 2016;15:75. [go to PubMed]
5. Dietz I, Borasio GD, Molnar C, et al. Errors in palliative care: kinds, causes, and consequences: a pilot survey of experiences and attitudes of palliative care professionals. J Palliat Med. 2013;16:74-81. [go to PubMed]
6. Dietz I, Plog A, Jox RJ, Schulz C. "Please describe from your point of view a typical case of an error in palliative care": qualitative data from an exploratory cross-sectional survey study among palliative care professionals. J Palliat Med. 2014;17:331-337. [go to PubMed]
7. Borasio GD, Jox RJ. Choosing wisely at the end of life: the crucial role of medical indication. Swiss Med Wkly. 2016;146:w14369. [go to PubMed]
8. Kissane DW, Bultz BD, Butow PN, Bylund CL, Noble S, Wilkinson S, eds. Oxford Textbook of Communication in Oncology and Palliative Care. 2nd ed. Oxford: Oxford University Press; 2017. ISBN: 9780198736134.
9. Sudore RL, Lum HD, You JJ, et al. Defining advance care planning for adults: a consensus definition from a multidisciplinary Delphi panel. J Pain Symptom Manage. 2017;53:821-832.e1. [go to PubMed]
10. Denvir MA, Murray SA, Boyd KJ. Future care planning: a first step to palliative care for all patients with advanced heart disease. Heart. 2015;101:1002-1007. [go to PubMed]
11. Combs SA, Davison SN. Palliative and end-of-life care issues in chronic kidney disease. Curr Opin Support Palliat Care. 2015;9:14-19. [go to PubMed]
12. Lindsay J, Dooley M, Martin J, Fay M, Kearney A, Barras M. Reducing potentially inappropriate medications in palliative cancer patients: evidence to support deprescribing approaches. Support Care Cancer. 2014;22:1113-1119. [go to PubMed]
13. Bemben NM. Deprescribing: an application to medication management in older adults. Pharmacotherapy. 2016;36:774-780. [go to PubMed]
14. Hui D, Bruera E. The Edmonton Symptom Assessment System 25 years later: past, present, and future developments. J Pain Symptom Manage. 2017;53:630-643. [go to PubMed]
15. Paramanandam G, Prommer E, Schwenke DC. Adverse effects in hospice patients with chronic kidney disease receiving hydromorphone. J Palliat Med. 2011;14:1029-1033. [go to PubMed]
16. King S, Forbes K, Hanks GW, Ferro CJ, Chambers EJ. A systematic review of the use of opioid medication for those with moderate to severe cancer pain and renal impairment: a European Palliative Care Research Collaborative opioid guidelines project. Palliat Med. 2011;25:525-552. [go to PubMed]
17. Murtagh FE, Chai MO, Donohoe P, Edmonds PM, Higginson IJ. The use of opioid analgesia in end-stage renal disease patients managed without dialysis: recommendations for practice. J Pain Palliat Care Pharmacother. 2007;21:5-16. [go to PubMed]
18. Reddy A, Vidal M, Stephen S, et al. The conversion ratio from intravenous hydromorphone to oral opioids in cancer patients. J Pain Symptom Manage. 2017;54:280-288. [go to PubMed]
19. Heneka N, Shaw T, Rowett D, Phillips JL. Quantifying the burden of opioid medication errors in adult oncology and palliative care settings: a systematic review. Palliat Med. 2016;30:520-532. [go to PubMed]
20. Sykes N, Thorns A. The use of opioids and sedatives at the end of life. Lancet Oncol. 2003;4:312-318. [go to PubMed]
21. McIntyre A. The double life of double effect. Theor Med Bioeth. 2004;25:61-74. [go to PubMed]
22. Beauchamp TL, Childress JF. Principles of Biomedical Ethics. 7th ed. Oxford: Oxford University Press; 2012.
23. Dietz I, Borasio GD, Schneider G, Jox RJ, Müller-Busch C. Errors and critical incidents in palliative care—establishment of an anonymous critical incident reporting system. Palliativmedizin. 2012;13:197-207. [Available at]
24. Tamang S, Kopec D, Shagas G, Levy K. Improving end of life care: an information systems approach to reducing medical errors. Stud Health Technol Inform. 2005;114:93-104. [go to PubMed]