Implementation of ICU Palliative Care Guidelines and Procedures"

CRITICAL CARE MEDICINE Implementation of ICU Palliative Care Guidelines and Procedures" A Quality Improvement Initiative Following an Investigation of Alleged Euthanasia Ware G. Kuschner, MD, FCCP; David A. Gruenewald, MD; Nancy Clum, RN, MN; Alice Beal, MD, FCCP; and Stephen C. Ezeji-Okoye, MD Ethical conflicts are commonly encountered in the course of delivering end-of-life care in the ICU. Some ethical concerns have legal dimensions, including concerns about inappropriate hastening of death. Despite these concerns, many ICUs do not have explicit policies and procedures for withdrawal of life-sustaining treatments. We describe a US Office of Inspector General (OIG) investigation of end-of-life care practices in our ICU. The investigation focused on care delivered to four critically ill patients with terminal diseases and an ICU nurse's concern that the patients had been subjected to euthanasia. The OIG investigation also assessed the validity of allegations that patient flow in and out of our ICU was inappropriately influenced by scheduled surgeries and that end-of-iife care policies in our ICU were not clear. Although the investigation did not substantiate the allegations of euthanasia or inappropriate ICU patient flow, it did find that the policies that discuss end-of-iife care issues were not clear and allowed for wide-ranging interpretations. Acting on the OIG recommendations, we developed a quality improvement initiative addressing end-of-iife care in our ICU, intended to enhance communication and understanding about palliative care practices in our ICU, to prevent ethical conflicts surrounding end-of-iife care, and to improve patient care. The initiative included the introduction of newly developed ICU comfort care guidelines, a physician order set, and a physician template note. Additionally, we implemented an educational program for lcu staff. Staff feedback regarding the initiative has been highly favorable, and the nurse whose concerns led to the investigation was satisfied not only with the investigation but also the policies and procedures that were subsequently introduced in our lcu. (CHEST 2009; 135:26-32) Key words: end-of-iife care; guidelines; ICU; life support; palliative care Abbreviations: OHI = Office of Healthcare Inspections; OIG = Office of InspectorGeneral: VAPARCS = VeteransAffairs Palo Alto Health Care System; WOLS =withdrawal of life-sustaining treatment physicians and nurses frequently confront ethical dilemmas in the course of treating dying patients in the ICU, including decisions about withholding or withdrawing life support. In fact, most ICU deaths occur after a decision to limit or withhold lifesustaining interventions, While limitation of therapy is common, euthanasia-the intentional ending of the life of a person suffering from an incurable or painful disease-is widely viewed as ethically problematic and almost universally illegal. l -3 Palliative care is now recognized as an important area of expertise in the ICU that requires knowledge 26 and competence-, however, ICU health-care practitioners may experience unease when the goals of care shift from cure to palliation. Health-care providers and the general public may make ethical distinctions between withdrawing and withholding life-sustaining treatments from terminally ill patients, adding complexity to the delivery of palliative care in the ICU.5-7 In the setting of withdrawal of life-sustaining treatment (WOLS), it may be difficult to draw clear distinctions between treatments administered solely to relieve pain and suffering with interventions that

are intended to hasten death. In a study" of end-oflife care practices in ICUs in 17 European countries, the use of medications intentionally to shorten the dying process was rare, but the doses used in these situations were within the range of doses commonly used for symptom relief. Furthermore, evidence was found that physicians may give much larger doses of medication than needed for relief of pain or suffering so that the patient can die with dignity, but these physicians do not call this practice euthanasia.? The authors speculated that physicians administering these treatments may consciously or unconsciously be practicing a "disguised form of mercy killing." The administration of very high doses of palliative drugs may be construed as good palliative care by some and as euthanasia or homicide by others. Ethical conflicts often occur between critical care physicians and nurses in the course of end-of-life care. These conflicts can result as a consequence of multiple factors including physician withdrawal from physical and emotional contact with the patient and family, a feeling among some nurses that they have a more moral position because they provide minute to IT,Iinute c?ntinuity of care, and the fact that physicians wnte orders regarding end-of-life management, while nurses implement them. 10 We report four cases of critically ill patients with terminal diseases who died in the Veterans Mfairs Palo Alto Health Care System (VAPAHCS) ICU (hereafter referred to as our ICU) after WOLS and the initiation of palliative care. One staffnurse in our ICU raised concerns that the patients' deaths were intentionally hastened through the administration of high dosages of opioids and the withdrawal of oxygen. A second allegation raised by the nurse was that the deaths of three patients were intentionally hastened in order to make beds available in our ICU for patients scheduled for elective surgery. As a conse- "From the Medical Service (Dr. Kuschner), Pulmonary and Critical Care Medicine Section, US Department of Veterans Affairs Palo Alto Health Care System, Palo Alto, CA; Geriatrics and Extended Care Service (Dr. Gruenewald), US Department of Veterans Affairs Puget Sound Health Care System. Seattle, WA; US Department of Veterans Affairs Palo Alto Health Care System (Ms. Clum), Palo Alto, CA; Medical Service (Dr. Beal), US Department of Veterans Affairs New York Harbor Health Care System, Brooklyn, NY; and Medical Service and Chief of Staff Office (Dr. Ezeji-Okoye), US Department of Veterans Affairs Palo Alto Health Care System, Palo Alto, CA. The authors have no conflicts of interest to disclose. Manuscript received July 8, 2008; revision accepted September 22,2008. Reproduction ~f this article is prohibited without written permission from the Amencan College of Chest Physicians (www.chestjoumal. oiw'misc/reprints.shtml). Correspondence to: Ware G. Kuschner, MD, FCCP, US Department of Veterans Affairs Palo Alto Health Care System. 3801 Miranda Ave, Mail Code: ll1p. Palo Alto, CA 94304; e-mail: ware.kuschnennd@va.gov DOl: lo.13781chest.08-1685 quence of these allegations, an investigation was conducted to determine whether patient deaths were a c:onsequence of euthanasia. We describe the investigation, and we detail the quality improvement initiative introduced into the VAPAHCS as a result of the findings and recommendations ofthe investigation. Patient A CASE REVIEWS The patient was a man in his late 50s. Fourteen months prior to hospital admission, lung cancer metastatic to the brain was diagnosed along with a malignant pericardial effusion. He presented with an arterial embolism of the left leg resulting in a below-the-knee amputation. Two weeks after the operation, he suffered an intracerebral hemorrhage complicated by respiratory failure. He was managed with mechanical ventilation, but his overall condition did not improve. He was unresponsive and received no sedating medications. His prognosis was assessed as extremely poor. On hospital day 4, after consultation with the clinical team, the family decided that c~ci~s should provide palliative care only. Clinicians disconnected the ventilator but left the endotrachealtube in place for airway support. The patient was given an IV injection of 10 mg of morphine sulfate; and over the next 4 h, the morphine was increased to 20 mglh by IV infusion to treat respiratory distress manifested by labored breathing and tachypnea. On hospital day 6, the endotracheal tube was removed at the family's request. Oxygen saturation fell from an average of 95% to a range of 54 to 85%. Heart rate increased from an average of 100 beats/min to 154 beats/min. At approximately 4:00 PM that day, the nurse believed that the tachycardia was a manifestation of patient discomfort and that additional sedation was indicated. After discussion with the physician, morphine was increased to 30 mglh. Morphine continued at this rate for approxi- ~ately 5 h, and the heart rate decreased. At approximately 9:30 PM, the assigned nurse on the next shift believed that the morphine infusion rate of30 mg/h might ~e excessive and that supplemental oxygen would Improve oxygenation and might make the patient more comfortable. She obtained a physician's order to administer oxygen and began supplemental oxygen. Oxygen saturation increased, and the nurse decreased the morphine gradually to 9.5 mglh during the morning of the next day. The patient died at 8:50 AM the day after the endotracheal tube was removed. Patient B The patient was a man in his mid 60s. One year prior to hospital admission, stage IV lung cancerwas www.chesljoumal.org CHEST/135/1/ JANUARY, 2009 27

diagnosed with widespread metastases that was treated with chemotherapy. He presented with renal failure, pneumonia, the ARDS, and sepsis. He was managed with antibiotics, mechanical ventilation, and vasopressors. He was adequately sedated with infusions of fentanyl (100 J.L~) and midazolam (2 mg/h). His condition worsened, and 1 week after hospital admission the family decided that clinicians should provide comfort care only. Clinicians continued mechanical ventilation, decreased the fractional inspired oxygen concentration to 21 from 100%, and discontinued vasopressors. The patient died within minutes of these maneuvers. Patient C The patient was a man in his mid 60s. One week prior to hospital admission, advanced lung cancer was diagnosed and the patient was scheduled for radiation treatment beginning approximately 2 weeks later. However, he was admitted to the hospital a week after diagnosis with respiratory distress, pneumonia, and sepsis. He was managed with antibiotics, vasopressors, and mechanical ventilation. His course was complicated by lung collapse, renal failure, and progressively worsening hypoxia. After a week of treatment, the family requested that clinicians provide comfort care only. Clinicians continued mechanical ventilation, decreased the fractional inspired oxygen concentration to 21%, and discontinued vasopressors while maintaining sedation with fentanyl and midazolam infusions. The patient died within 2 h of these maneuvers, which was the same day as patient B died. Patient D The patient was a man in his mid 50s. One year prior to hospital admission, colon cancer was diagnosed that had metastasized to the liver and he was treated with chemotherapy. He presented with respiratory failure, neutropenia, sepsis, hypotension, and extreme obesity. He was treated with antibiotics, mechanical ventilation, and vasopressors. It was difficult to oxygenate and ventilate the patient. He was paralyzed (with cisatracurium) and sedated with fentanyl (200 J..L~) and midazolam (6 mg/h IV infusions). He remained critically ill, and on hospital day 2 the family requested that clinicians provide comfort care only. Clinicians discontinued vasopressors, neuromuscular blockade, and mechanical ventilation. Fentanyl and midazolam infusion continued. He died within minutes of withdrawal of vasopres- SOTS and mechanical ventilation. This was also the same day that patients Band C died. Following the deaths of these patients, an employee voiced her concerns to the ICU nurse manager that the deaths of the four patients were intentionally hastened through the use of high dosages of narcotics and insufficient administration of oxygen and that these actions were tantamount to euthanasia. This concern was presented to our Chief of Quality Management, who alerted senior managers, who contacted the us Department of Veterans Affairs Office ofinspector General (OIG). THE INVESTIGATION US Department ofveterans Affairs OIG and the Office of Healthcare Inspections: Organization and Mission The us Department of Veterans Affairs OIG is an independent organization with the goal of minimizing fraud, waste, and abuse in the Department of Veterans Affairs. The OIG has its headquarters in Washington, DC, and offices in 24 cities throughout the country. The OIG is organized into several line elements that include the Office of Healthcare Inspections (OHI), which monitors health care provided to veterans. The responsibilities ofthe OHI include the review of patient care, quality assurance issues, and hotline allegations involving medical care. The OHI is legally authorized to gain access to all records, reports, audits, reviews, documents, papers, recommendations, or other pertinent materials in order to carry out inspections. More information about the OIG is available at its Web site: http://www.va.gov/ oigldefault.asp. Purpose The purpose of the OIG inspection was to determine the validity of the following allegations: (1) The deaths of four patients in the ICU were hastened through the use of high dosages of narcotics and too little oxygen. Specifically, in the case of patient A, it was alleged that the dosage of 30 mglh of morphine was much higher than usual and intended to hasten the patient's death, and that supplemental oxygen should have been administered as a comfort measure. In the cases of patients B, C, and D, it was alleged that medications were used to hasten death, and comfort care was initiated because of pressure to open ICU beds for other patients. (2) End-of-life care policies were not clear or comprehensive. (3) Patient flow in and out of the leu was inappropriately influenced by upcoming major surgeries. Scope and Methodology The complainant was interviewed by the OIG by telephone. The OIG conducted a site visit at 28

VAPAHCS on January 8 to 9,2008, and interviewed staff nurses, staff physicians, a respiratory therapist, a pharmacist, and several managers. The OIG reviewed documents, including medical records, policies, meeting minutes, and reports. The validity of the allegations was established by comparing the approaches of the involved clinicians with palliative care guidelines and recommendations in the literaturey,1l-13 The scope of the review was limited to the allegations made by the complainant. The inspection was conducted in accordance with the Quality Standards for Inspections published by the President's Council on Integrity and Efficiency. This publication is available at http://www.ignet.gov/pande/standards/ igstds.pdf. DIG Inspection Findings and Recommendations The complete findings and recommendations of the OIG inspection are in the public domain and are available at http://www.va.gov/oig/54/reportsivaoig- 08-00509-103.pdf. Presented below is a summary of the findings and recommendations published in the executive summary of the investigation report. We did not substantiate that any of the four patients' deaths were intentionally hastened. In each case, the family members, after consultation with the clinical team, had requested that clinicians withhold any further interventions and provide comfort care only. The orders were changed, interventions (including oxygen) were withheld or withdrawn, and sedative medications were provided for comfort. We substantiated that the policies that discuss end of life care issues were not clear and allowed for wide-ranging interpretations. Because end of life care is not the norm and is infrequently given in the ICU, written guidelines for the ICU will likely reduce the levels of disagreement between staff members. On receipt of the complaint, the Deputy Chief of Staff (COS) had assigned the task of drafting comfort care guidelines to the VAPAHCSBioethics Committee Chairman. We did not substantiate the allegation that patient flow into and out of the ICU was inappropriately influenced by upcoming major surgeries. However, the Deputy COS acknowledged that bed utilization issues throughout the VAPAHCS, including the ICU, have been identified and that several task forces are actively working to address them. We recommended that the VAPAHCS complete and implement the end of life care guidelines and provide training to all staff on all shifts in the ICU. Also, the Chief Nurse Executive should add end of life care competency to the initial and annual ICU nurse skills checklists. ICU Comfort Care Quality Improvement Initiative In response to the OIG finding that "the policies that discussend of life care issues were not clear and allowed for wide-ranging interpretations," a quality improvement initiative was developed and introduced at VAPAHCS in order to improve ICU comfort care. The focus of this initiative was limited to the period after a decision has been made to limit the use of life-sustaining treatments, although we acknowledge that ideally palliative care for many ICU patients begins at or before ICU admission with an assessment of goals of care and the decision to begin a therapeutic trial of ICU care. The initiative was a collaborative effort that included contributions from ICU physicians and nurses, palliative health-care physicians and nurses, pharmacists, and respiratory therapists. We sought the perspectives of clinicians based at VAPAHCS and other medical centers. We reviewed published materials and palliative care instruments used at other facilities within and outside of the Veterans Affairs Health Care System. Each of the OIG recommendations was adopted. The initiative included development of comfort care guidelines, a comfort care physician order set, and a comfort care template physician note. Comfort Care Guidelines: We reviewed peerreviewed guidelines and other published literature on palliative care with special attention to leu palliative care 4,8-19 (online supplementary material, document 1). We summarized highlights from these documents in order to provide a concise, readerfriendly overview of general concepts and procedures relevant to ICU palliative care. The guideline is now accessible through a computerized patient record system. The guideline is divided into four main sections. The "General Principles" section describes ethical tenets addressing ICU palliative care. The "Special Considerations" section addresses specific areas that have the potential to generate ethical tension: (1) clarifying goals of care, (2) medication administration, (3) psychosocial and spiritual care, and (4) artificial nutrition and hydration. The "Procedures" sections details specific steps that should be taken in the course of shifting a patient's treatment goals from cure to comfort care and WOLS. The "Resources" section lists publications which provided substantial content for the guidelines. Comfort Care Physician Template Note: We developed a template physician note which physicians can now access on our computerized patient record system (online supplementary material, document 2). The ICU physician can use this instrument to create a note for the patient's electronic medical record. The physician may select ("point and click") prepared text and add original text ("free text") into the electronic note. The template note has the www.chestjournal.org CHEST /135/1/ JANUARY, 2009 29

following functions: (1) it shows the physician suggested text that can effectively serve as a procedural checklist to ensure a series of necessary actions has occurred, or will take place, before a patient's treatment goals transition from cure to palliation; (2) it provides clear and consistent documentation of assessments and plans; and (3) it simplifies the documentation process. The template note prompts the physician to address the following domains: (1) diagnosis; (2) prognosis; (3) patient's decision-making capacity; (4) patient/family/surrogate treatment goals, including plans to pursuecomfortcare; and (5) patient's candidacy for organ and tissue donation. Comfort Care Order Set: We developed an electronic ICU comfort care physician order set (online supplementary material, document 3). The order set includes domains for opioid and benzodiazepine infusions and mechanical ventilator and supplemental oxygen management. The order set also prompts the physician to confirm a do not resuscitate/do not intubate order has been entered and to discontinue previous orders that do not advance the treatment goal of comfort care. The comfort care order set provides a procedural checklist to ensure that appropriate sedation and analgesia orders are in place prior to discontinuation of mechanical ventilator support, including continuous infusion and bolus orders for manifestations of physical discomfort such as grimacing, purposeless movements, and moaning. StaffSatisfaction With ICU Comfort Care Quality Improvement Initiative Although we did not measure outcomes such as staff satisfaction with a standardized instrument, anecdotal reports from our ICU nurses and physicians strongly indicate the instruments are helpful and are being well received. The nurse whose concerns led to the investigation subsequently expressed satisfaction with the investigation and the quality improvement initiative that has now been introduced into our ICU. DISCUSSION Death in the ICU is common. Approximately 20% of all deaths in the United States now occur after admission to an ICU.20 Many if not most of the deaths occurring in ICUs are preceded by a decision to withhold or withdraw life-sustaining treatments. 1 As a general principle, when the goals of care cannot be achieved with aggressive life-sustaining treatments such as mechanical ventilation, it is appropriate to withdraw these treatments and to allow death 30 to occur.!? However, in the absence of explicit ground rules governing WOLS in ICUs, interventions such as administration of large doses of sedatives and analgesics or discontinuation ofmechanical ventilation may be interpreted as hastening the patient's death rather than relieving pain and suffering or allowing the patient to die from the underlying illness. Such ethical tensions arising from doubt and misunderstanding about palliative ICU care are a common challenge facing many ICU health-care practitioners.s! In our ICU, one nurse's concern regarding the intentions of ICU physicians treating dying patients resulted in allegations of euthanasia. In order to promote high-quality ICU palliative care and prevent the commission of acts that may be construed as unethical or illegal, we acted on the recommendations of an OIG investigation and developed ICU comfort care guidelines, a physician order set, and a physician template note for use in the management of ICU patients whose goals of care have shifted from cure to comfort. The goals of this quality improvement initiative are to improve communication about and documentation of end-of-life care in the ICU and, in so doing, strengthen the quality of patient care. Quality improvement initiatives in the ICU setting involving the use of standardized orders alone have been shown to be helpful to clinicians, although improving the quality of the dying process may require more comprehensive interventions.p-w By coupling standardized orders with easily accessible guidelines and a note template for documentation, we provided tools for our ICU clinicians that serve as a checklist to help ensure that high-quality care consistent with ethical norms is provided with a minimum of variability between patients and in adherence with current guidelines for WOLS in the ICU setting. Each ofthe newly developed documents function as a prompt and guide for ICU physicians to provide clearly communicated, ethically robust, and consistent care for ICU patients whose management includes WOLS and comfort care. The physician template note and order set, in particular, function both as action checklists and as "clinician-friendly" instruments that support consistency in ICU comfort care practice and clear documentation of assessments, goals, and plans. Anecdotal reports from our ICU clinicians indicate that they are satisfied with these care tools, including the nurse who Originally raised the allegations ofeuthanasia. In concert with our experience, an increasing body of literature4.13.18.19,23 supports a systematic, standardized checklist-oriented approach to improve end-of-life care in ICUs. However, it is unclear how

widespread the incorporation of such checklist procedures into end-of-life care in ICUs has become. We anticipate that practice guidelines and the requirements of regulatory bodies will increasingly drive facilities to develop policies and procedures for WOLS in the ICU setting: Across health care, there is ever-increasing evidence that protocols and checklists promote safety and homogeneity of preferred practices. The landmark study by Pronovost et al 24 demonstrating the use of a central line checklist by ICU practitioners resulted in > 100 million dollars and 1,500 lives saved over an 18-month period is just one example of the potential of checklists to improve multiple health-care outcomes. However, there are limited data to drive decisions about what should be included in end-of-life care protocols and checklists. For example, the role ofoxygen to alleviate suffering has not been well studied, and its role in end-of-life care is uncertain. Hall and colleagues's described an experience similar to ours, in which legal action against a Canadian physician after a WOLS (a first-degree murder charge, later dismissed) led to development of an initiative to change the conduct of WOLS in two teaching hospital tertiary care medical surgical ICUs. The initiative included the following: (1) the development of a "physician orders for the level of medical treatment" checklist and a "withdrawal of life support checklist"; (2) alteration of nursing progress notes to reflect more complete documentation ofpatient discomfort and its management; and (3) prohibition of the use of open-ended medication orders. The investigators-" reported that following implementation of this ICU initiative they observed a significant reduction in the median cumulative dose of diazepam in the 12-h period immediately preceding death. Additionally, there were significant reductions in the time from ICU admission to WOLS and increased involvement ofpastoral care in WOLS discussions. Eighty percent of nurses in the study ICUs believed that the checklists improved end-of-life care.f' Methodologic constraints complicate end-of-life care clinical research and, in tum, efforts to develop best practices in ICU comfort care. Specifically, patient perspectives about quality of care will always be incomplete because they may not be obtained if the intervention (ICU comfort care) is associated with obtundation, as is typical, and they cannot be obtained after comfort care ends because patients are dead. Accordingly, surrogate end points ofphys- iologic suffering (eg, tachycardia and tachypnea) and other quality indicators (eg, access to pastoral care, family involvement, and families' perception of the experience)25 are necessarily central in end-of-care outcomes studies. In the United States, the courts have established in the ease of Karen Ann Quinlan and others the right ofpatients (and their surrogate decision makers) to refuse unwanted treatments even if they are life sustaining, thereby allowing the patient to die.' All ICU health-care practitioners must abide by the ethical imperatives to respect patient autonomy, including requests by patients or their surrogate decision makers to limit care, and to relieve suffering. Medications to relieve physical and psychological discomfort are commonly administered to dying patients for whom the goals of care have shifted from cure to comfort even when they will hasten death. At the same time, euthanasia, physician-assisted suicide, and mercy killing are ethically problematic and widely illegal. Not surprisingly, a "gray area" is often encountered in end-of-life care between treatments administered to relieve pain and suffering and those intended to shorten the dying process." Of note, opioids or benzodiazepines used to treat discomfort after withdrawal of ventilator support do not appear to hasten death in critically ill patients in the ICU.26 The doses of IV morphine that patient A received are commonly used during the withdrawal of ventilator support," but the important principle is that opioids and sedating medications should be titrated to achieve the desired effect of comfort. There is no arbitrary upper limit on dosage if the goal of relieving the patient's distress has not been achieved. 17 In addition to clear policies and procedures regarding the delivery of comfort care in the ICU, open communication, shared decision making, and a respect for the different perspectives of all healthcare practitioners can prevent many of the ethical conflicts that arise in the course ofend-of-life care. It is important that all members of the ICU care team have an opportunity to discuss the care plan before WOLS. In the cases described in this report, proactively including nursing staff in discussions regarding WOLS and exploring areas of disagreement might have allayed the concerns ofthe nurse who alleged that these patients' deaths were hastened. To strengthen communication and promote consensus building, we introduced a palliative care educational program into our ICU. The program features the following: (1) screening of an educational film, "Integrating palliative and critical care: an educational video about improving palliative care in the ICU," produced by faculty from the University of Washington School of Medicine and the National Institute of Nursing Research, available at http://w\\.w. ninr.nih.govlnewsandinformationipodcastmultimedia/, and (2) a 4-h live didactic lecture series on palliative www.chestjournal.org CHEST/135/1 / JANUARY, 2009 31

care. Completion of the program is now mandatory for all ICU nurses at our facility. In sum, as a result of one nurse's concern regarding end-of-life care in our ICU, an OIG investigation was carried out at our medical center. The conclusion of the investigation was that allpatients received standard of care treatments, but palliative care communication and training in our ICU needed improvement. As a consequence of that investigation, we developed a quality improvement initiative addressing ICU palliative care. We believe the ICU comfort care guidelines, physician template order set, and physician template note now available at our medical center advance the goal of reducing ethical tension associated with the management of patients for whom a decision has been made to shift treatment goals from cure to comfort. ACKNOWLEDGMENT: The authors would like to acknowledge Drs. T. Randall Curtis and Gordon Rubenfeld who developed guidelines and order forms for the WOLS in the ICUs that served as a model for the guidelines and physician order set described in this article. The authors also thank Dr. lames Hallenbeck for his thoughtful comments that informed the development of the instruments described in this article. REFERENCES 1 Prendergast TJ, Luce JM. Increasing incidence of withholding and withdrawal of life support from the criticallyill. Am J Respir Crit Care Med 1997; 155:15-20 2 Eidelman LA, Jakobson DJ, Pizov R, et al. Foregoing lifesustaining treatment in an Israeli ICU. Intensive Care Med 1998; 24:162-166 3 Keenan SP, Busche KD, Chen LM, et al. A retrospective reviewof a large cohort of patients undergoing the process of withholding or withdrawal of life support. Crit Care Med 1997; 22:1020-1025 4 Truog RD, Campbell ML, Curtis JR, et al. Recommendations for end-of-life care in the intensive care unit: a consensus statement by the American Academy of Critical Care Medicine. Crit Care Med 2008; 36:953-963 5 Rydvall A, Lyntie N. Withholding and withdrawing lifesustaining treatment: a comparative study of the ethical reasoning of physicians and the general public. Crit Care 2008; 12:R13 6 Pochard F, Kentish-Barnes N, Azoulay E. Withholding and withdrawing life-sustaining treatment: the necessity of discrepancies in ethical reasoning. Crit Care 2008; 12:418 7 Vincent JL. Withdrawing may be preferable to withholding. Crit Care. 2005; 9:226-229 8 Sprung CL, Cohen SL, Sjokvist P, et al. ETHICUS Study Group: end-of-lifepractices in European intensive care units: the Ethicus Study. JAMA 2003; 290:790-797 9 Sprung CL, Ledoux D, Bulow HH, et al. ETHICUS Study Group: relieving suffering or intentionally hastening death: where do you draw the line? Crit Care Med 2008; 36:8-13 10 Gavrin JR. Ethical considerations at the end of life in the intensive care unit. Crit Care Med 2007; 35(suppJ):S85-S94 11 HawryluckLA, Harvey WR, Lemieux-Charles L, et al, "Consensus guidelines on analgesia and sedation in dying intensive care unit patients." BMC Medical Ethics 2002; 3:E3 12 National Center for Ethics in Health Care. The ethics of palliative sedation: a report by the National Ethics Committee of the Veterans Health Administration, 2006 13 Curtis JR. Interventions to improve care during withdrawal of life-sustaining treatments. J Palliat Med 2005; 8:116-131 14 Lanken PN, Terry PB, Delisser HM, et al. ATS End-of-Life Care Task Force: an official American Thoracic Society clinical policy statement; palliative care for patients with respiratory diseases and critical illnesses. Am J Respir Crit Care Med 2008; 177:912-927 15 Lorenz KA, Lynn J, Dy SM, et al. Evidence for improving palliative care at the end of life: a systematic review. Ann Intern Med 2008; 148:147-159 16 Qaseem A, Snow V, Shekelle P, et al. Evidence-based interventions to improve the palliative care of pain, dyspnea, and depression at the end oflife: a clinical practice guideline from the American College of Physicians. Ann Intern Med 2008; 148:141-146 17 Rubenfeld GD. Principles and practice of withdrawing lifesustaining treatments. Crit Care Clin 2004; 20:435-451 18 Treece PD, Engelberg RA, Crowley L, et al. Evaluation of a standardized order fonn for the withdrawal of life support in the intensive care unit. Crit Care Moo 2004; 32:1141-1148 19 von Gunten CF, Weissman DE. Ventilator withdrawal protocol: fast fact and concept #33, 34 and 35. 2nd ed, July 2005. End-of-Life Palliative Education Resource Center. Available at: http://www.eperc.mcw.edu/ff_index.htm. Accessed July 8, 2008 20 Angus DC, Barnato AE, Linde-Zwirble WT, et al. Robert Wood Johnson Foundation ICU End-Of-Life Peer Group: use of intensive care at the end of life in the United States; an epidemiologic study. Crit Care Med 2004; 32:638-643 21 Breen CM, Abernethy AP, Abbott KH, et al. Conflict associated with decisions to limit life-sustaining treatment in intensive care units. J Cen Intern Med 2001; 16:283-289 22 Curtis JR, Treece PD, Nielsen EL, et al, Integrating palliative and critical care: evaluation of a quality improvement intervention. Am J Respir Crit Care Med 2008 May 14 [Epub ahead of print) 23 Hall RI, Rocker GM, Murray D. Simple changes can improve conduct of end-of-life care in the intensive care unit. Can J Anaesth 2004; 51:631-636 24 Pronovost P, Needham D, Berenholtz S, et al. An intervention to decrease catheter-related bloodstream infections in the ICU. N Engl J Med 2006; 355:2725-2732 25 Wenger NS, Rosenfeld K. Quality indicators for end-of-life care in vulnerable elders. Ann Intern Med 2001; 135:677-68.5 26 Chan JD, Treece PD, Engelberg RA, et al. Narcotic and benzodiazepine use after withdrawal of life support: association with time to death? Chest 2004; 126:286-293 32