Withdrawal Of Life-sustaining Interventions Research Articles

Patterns of Pediatric Palliative and End-of-Life Care in Neonatal Intensive Care Patients in the Southern U.S.

Despite high rates of mortality among infants in the Southern U.S., little is known about the timing of pediatric palliative care (PPC), the intensity of end-of-life care, and whether there are differences among sociodemographic characteristics. To describe PPC patterns and treatment intensity during the last 48 hours of life among neonatal intensive care unit (NICU) patients in the Southern U.S. who received specialized PPC. Medical record abstraction of infant decedents who received PPC consultation in two NICUs (in Alabama and Mississippi) from 2009 to 2017 (n = 195) including clinical characteristics, palliative and end-of-life care characteristics, patterns of PPC, and intensive medical treatments in the last 48 hours of life. The sample was racially (48.2% Black) and geographically (35.4% rural) diverse. Most infants died after withdrawal of life-sustaining interventions (58%) and had do not attempt resuscitation orders documented (75.9%); very few infants enrolled in hospice (6.2%). Initial PPC consult occurred a median of 13 days after admission and a median of 17 days before death. Infants with a primary diagnosis of genetic or congenital anomaly received earlier PPC consultation (P = 0.02) compared to other diagnoses. In the last 48 hours of life, NICU patients received intensive interventions including mechanical ventilation (81.5%), CPR (27.7%) and surgeries or invasive procedures (25.1%). Black infants were more likely to receive CPR compared to White infants (P = 0.04). Overall, PPC consultation occurred late in NICU hospitalizations, infants received high-intensity medical interventions in the last 48 hours of life, and there are disparities in intensity of treatment interventions at end of life. Further research is needed to explore if these patterns of care reflect parent preferences and goal concordance.

A different kind of death? Barts NHS Trust v Dance and Battersbee

The case of Archie Battersbee, a 12-year-old boy who suffered a catastrophic hypoxic brain injury, was the subject of several Family Division and Appeal Court hearings between April and August 2022. During the protracted legal process, appeals were made by the family to the Supreme Court, the European Court of Human Rights, and the United Nations’ Committee on the Rights of Persons with Disabilities (CRPD Committee). These were unsuccessful in achieving a stay on the withdrawal of life-sustaining interventions, whose continuance the Family Division of the High Court had found not to be in Archie’s best interests. This commentary focuses on two novel aspects of the proceedings: the Court of Appeal’s overturning of Arbuthnot J’s conclusion that Archie was brainstem dead, and the CRPD Committee’s intervention in response to the family’s appeal.

Non-cardiac Cause of Death And Organ Dysfunction Common Among Patients Admitted To The Cardiac Intensive Care Unit

Introduction Over the past decade, there has been an increase in the proportion of patients admitted to the cardiac intensive care unit (CICU) with non-cardiac comorbidities and admission diagnoses. However, organ system dysfunction and causes of death in CICU patients remains poorly characterized. Given the increasing proportion of admissions to the CICU for non-cardiac diagnoses, we hypothesized that there would be a substantial proportion of non-cardiac organ dysfunction and death due to non-cardiac causes. Methods We identified adult patients admitted to a CICU at a large tertiary care center (2015-2019) who died during the hospitalization. Principal hospitalization diagnosis, demographics, organ system dysfunctions, cause of death, and withdrawal of life support were abstracted via a standardized tool used in prior studies. The primary cause of death was defined as the organ system that most directly contributed to death or withdrawal of life support. Results We identified 258 decedents admitted to the CICU. Most patients (76%) had cardiac dysfunction, defined as cardiogenic shock or sustained ventricular arrhythmia. However, only 11% of patients had cardiac dysfunction alone, 89% experienced non-cardiac organ dysfunction, and 24% died without cardiac dysfunction. Multi-organ failure and sepsis were present in 70% and 54% of patients, respectively. The most common primary causes of death were cardiac (58%), pulmonary (15%), and neurologic dysfunction (11%). The majority of patients (76%) died after withdrawal of life-sustaining interventions. Conclusions In this contemporary cohort of CICU decedents, 3 of 4 patients had cardiogenic shock or sustained ventricular arrhythmias. However, non-cardiac organ dysfunction occurred in the vast majority, and nearly half had a non-cardiac primary cause of death. The high frequency of multi-organ failure and non-cardiac causes of death underscores the importance of broad cardiac and non-cardiac critical care training among physicians practicing in the CICU.

Influence of time to admission to a comprehensive stroke centre on the outcome of patients with intracerebral haemorrhage.

In patients with spontaneous intracerebral haemorrhage, it is uncertain if diagnostic and therapeutic measures are time-sensitive on their impact on the outcome. We sought to determine the influence of the time to admission to a comprehensive stroke centre on the outcome of patients with acute intracerebral haemorrhage. We studied a prospective database of consecutive patients with intracerebral haemorrhage attended at two comprehensive stroke centres (2005-2017). We excluded patients with an unwitnessed time of onset of the intracerebral haemorrhage, or previous modified Rankin Scale >3 or in those in whom withdrawal of life-sustaining interventions were decided <24 h from admission. We recorded the time from the intracerebral haemorrhage onset to admission, demographic, clinical, radiological data, the functional outcome (favourable when modified Rankin Scale ≤3) and mortality at 90 days. We conducted a propensity score-matching analysis to evaluate functional outcome and mortality. We included 487 patients (mean age 72.3 ± 13.9 years), and 53.2% were men. Compared to patients with an admission >110 min, patients who were admitted ≤110 min were significantly younger, and had higher National Institutes of Health Stroke Scale scores. Moreover, patients admitted ≤110 min were more likely to have basal ganglia intracerebral haemorrhage, and to show neurological deterioration. The propensity score groups were well matched. We did not find an association between time to admission and the favourable outcome (OR: 1.42 (95% CI: 0.93-2.16)) or mortality (OR: 0.64 (0.41-0.99)) at 90 days. Our results suggest that in patients with intracerebral haemorrhage and known symptom onset who are admitted to a comprehensive stroke centre, an early admission (≤110 min) does not influence the outcome at 90 days.

Withholding and withdrawal of life-sustaining treatments for neonate in Japan: Are hospital practices associated with physicians' beliefs, practices, or perceived barriers?

ObjectiveTo describe the current status of withholding or withdrawal of life-sustaining interventions (LSI) for neonates in Japan and to identify physician- and institutional-related factors that may affect advance care planning (ACP) practices with parents. Study designA self-reported questionnaire was administered to assess frequency of withholding and withdrawing intensive care at the respondent's facility, the physician's degree of affirming various beliefs about end-of-life care that was compared to 7 European countries, their self-reported ACP practices and perceived barriers to ACP. Three neonatologists at all 298 facilities accredited by the Japan Society for Neonatal Health and Development were surveyed, with 572 neonatologists at 217 facilities responding. ResultsAt 76% of facilities, withdrawing intensive care treatments was “never” done, while withholding intensive care had been done “sometimes” or more frequently at 82% of facilities. Japanese neonatologists differed from European neonatologists regarding their degree of affirmation of 3 out of 7 queried beliefs about end-of-life care. In hospitals that were more likely to “sometimes” (or more often) withdraw treatments, respondents were less likely to affirm beliefs about doing “everything possible” or providing the “maximum of intensive care”. Self-reported ACP practices did not vary between neonatologists based on their hospital's overall pattern of withholding or withdrawing treatments. ConclusionAmong NICU facilities in Japan, 21% had been sometimes withdrawing LSI and 82% had been “sometimes” withholding LSI. Institutional treatment practices may have a strong association with physicians' beliefs that then affect end-of-life discussions, but not with self-reported ACP practices.

Determining Brain Death: Basic Approach and Controversial Issues.

Between 15% and 30% of patients admitted to intensive care units are comatose at some point during their admission as a result of primary neurologic (trauma, stroke, meningitis, neoplasm, seizures) or systemic (cardiopulmonary arrest, toxins, sepsis, metabolic derangements) problems.1 Comatose patients have grossly impaired brain function that renders them unconscious. Although some comatose patients recover, others reacquire minimal environmental awareness (a minimally conscious state), transition to being in a vegetative state, or progress to death.2 Patients who become vegetative have a complete lack of awareness of themselves and their environment, but have sleep-wake cycles and variable preservation of cranial nerve function. After a month of being vegetative, a patient is described as being in a persistent vegetative state, and if there is a high degree of clinical certainty that the condition is irreversible, the patient is deemed to be in a permanent vegetative state.3 Patients who are comatose, have absent brainstem reflexes, and are unable to breathe spontaneously are brain dead and are legally dead throughout the United States.4–6Similar to the general public, nursing students do not have a good understanding of the difference between coma, a vegetative state, and brain death.7–9 Although nurses in intensive care units are exposed to brain death, they, too, often do not understand the concept of brain death and the process of brain death determination.10 Because management of patients with catastrophic brain injuries who are being evaluated for brain death in the intensive care unit can be both medically and ethically challenging, it is imperative for critical care nurses to have a firm understanding of brain death and the controversies that can arise during and after determination of brain death.The American Academy of Neurology (AAN) created guidelines for determination of brain death in adults in 1995 and updated them in 2010.6,11 The AAN requires that a single examination be performed to determine that a patient is comatose, has no brainstem reflexes, and is unable to breathe spontaneously.Before conducting a brain death evaluation, it is necessary to confirm that (1) there is an irreversible and proximate cause of coma according to the medical history and the results of physical examination, neuroimaging, and laboratory testing; (2) drug levels or calculations of drug clearance based on 5 times a drug’s half-life indicate that no sedatives or narcotics are present in the patient’s system (of note, hypothermia or abnormal hepatic or renal function may extend a drug’s half-life); (3) the blood alcohol level is less than 0.08 g/dL; (3) no neuromuscular blocking agents have been administered recently or reversal of paralytic agents has been confirmed by using a train-of-4 test; (4) no severe electrolyte, acid-base, or endocrine derangements are present; (5) the patient’s body temperature is greater than 36°C (this may require the use of a warming blanket); and (6) the patient’s systolic blood pressure is at least 100 mm Hg (this may require the use of vasopressors).Patients who are brain dead have no evidence of responsiveness to painful stimulation of their extremities, supraorbital ridges, or temporomandibular joints but may have spinal reflexes, which can be disconcerting at times (eg, the Lazarus sign, in which a patient appears to grasp for the endotracheal tube spontaneously during apnea testing or with neck flexion).12 Additionally, all brainstem reflexes are absent in brain-dead patients as follows:If a patient meets the prerequisites for a brain death evaluation and is found to be comatose with no brainstem reflexes, apnea testing must be performed to finalize the brain death determination. Before apnea testing, it is necessary to ensure that the patient is normotensive (systolic blood pressure ≥ 100 mm Hg), normothermic (body temperature &gt; 36°C), eucapnic (Paco2 35–45 mm Hg), and adequately preoxygenated (Pao2 &gt; 200 mm Hg; oxygenation on a fraction of inspired oxygen of 1.0 for at least 10 minutes). The AAN guidelines require that, after meeting prerequisites, patients be disconnected from the ventilator for 8 to 10 minutes. During this time, oxygenation is preserved through delivery of 100% oxygen at 6 L/min through an insufflation catheter placed in the endotracheal tube at the level of the carina. If the patient breathes during testing, the procedure is aborted, as the patient is clearly not brain dead. If the patient is unstable during apnea testing (systolic blood pressure &lt; 90 mm Hg, oxygen saturation &lt; 85%, or an arrhythmia develops), the patient should be reconnected to the ventilator and the procedure may be retried when the patient is stable using a T-piece, 10 cm H2O continuous positive airway pressure, and 100% oxygen at 12 L/min. If no evidence of respiratory drive is observed during testing, the patient should be reconnected to the ventilator and a blood sample should be sent for blood gas analysis. If respiratory movements are absent during testing and Paco2 is at least 60 mm Hg or is 20 mm Hg greater than the patient’s prior Paco2 on the repeat blood gas analysis, the apnea test is considered positive (consistent with brain death). If the Paco2 does not increase adequately, the test is inconclusive and may be repeated for 10 to 15 minutes if the patient is hemodynamically stable.If results of the full neurologic examination and apnea testing are consistent with brain death, the patient is declared dead. However, if part of the neurologic examination cannot be completed (eg, in the setting of facial trauma) or if a patient cannot tolerate apnea testing, an ancillary test should be performed to confirm the brain death determination (of note, brain death is a clinical determination and ancillary testing is not a substitute for the clinical evaluation; apnea testing and all portions of the clinical examination that can be conducted must be done and must be consistent with brain death before ancillary testing is performed to confirm the clinical findings). The ancillary tests that the AAN recommends include an electroencephalogram to demonstrate cerebral inactivity or an angiogram or nuclear study to demonstrate lack of intracranial blood flow. Because brain death is a clinical determination, it is important to be aware that no ancillary test is perfect and it is possible to have false-positive or false-negative results.13Guidelines for brain death determination in pediatric patients were written by the American Academy of Pediatrics in 1987 and then updated in conjunction with the Pediatric Section of the Society of Critical Care Medicine and the Child Neurology Society in 2011.5,14 These guidelines are grossly similar to the AAN guidelines for brain death determination in adults except for the following: (1) after an observation period of 24 to 48 hours, 2 examinations and 2 apnea tests should be performed by 2 different physicians (in adults, only 1 examination and apnea test is required and there is no mandatory observation period); (2) systolic blood pressure should be normalized for age (in adults, systolic blood pressure must be ≥ 100 mm Hg); (3) core body temperature should be &gt; 35°C (in adults, body temperature must be &gt; 36°C); (4) sedative and narcotic elimination should be considered, but there is no finite recommendation for the number of half-lives that must pass before a brain death examination can be performed (in adults, at least 5 half-lives must pass after sedative or narcotic administration before a brain death examination can be performed); (5) for an apnea test to be positive, the Paco2 must be at least 60 mm Hg and must be 20 mm Hg greater than baseline (in adults, the Paco2 must be at least 60 mm Hg or must be 20 mm Hg greater than baseline); and (6) ancillary testing may be done if there is uncertainty about the results of the examination or if medication effect interferes with evaluation of the patient (in adults, ancillary testing is done only if apnea testing or a portion of the clinical examination cannot be completed).5,6Because brain dead patients are legally dead, organ support should be discontinued in a reasonably respectful time period after brain death determination. However, it is important to be aware that brain dead patients are candidates for organ donation. Organ procurement organizations (OPOs) should be contacted before the brain death determination in comatose patients if brain death is being discussed or withdrawal of life-sustaining interventions is anticipated. Discussions of organ donation should be initiated with families only by trained OPO personnel, not by members of the treatment team, once the OPO has determined that a patient is a donation candidate. If a family consents to organ donation, hemodynamic support is maintained until organ retrieval can be arranged. Because systemic complications, including neurogenic pulmonary edema, hypotension, arrhythmias, myocardial dysfunction, disseminated intravascular coagulation, and diabetes insipidus, may develop in brain dead patients, it is often necessary to support patients awaiting organ retrieval with dopamine, methylprednisolone, levothyroxine, and vasopressin.15Some controversial aspects of brain death determination include (1) variability in the performance of brain death determinations, (2) management of religious objections to determination of death by neurologic criteria, (3) determination of brain death after therapeutic hypothermia, and (4) determination of brain death for patients on extracorporeal membrane oxygenation (ECMO).Despite clear guidelines on determination of brain death,5,6 institutional protocols for brain death determination vary both within the United States and around the world.16,17 A 2016 review of 508 US hospitals’ protocols for adult brain death revealed variability in prerequisites for a brain death examination, examination requirements, apnea testing requirements, examiner credentials, number of examinations, waiting periods between examinations, indications for ancillary testing, and ancillary tests to perform.16 In a 2015 survey on brain death determination worldwide, only 77% of countries endorsed having institutional brain death protocols, and responses showed variability in examination requirements, apnea testing, and ancillary testing.17Although brain death is widely accepted in the United States and around the world as death on medical, legal, and ethical levels,4,18–21 families sometimes voice religious objections to determination of death by neurologic criteria.22 As a result, 4 states (California, Illinois, New Jersey, and New York) have legal accommodations for families who voice objections to determination of death by neurologic criteria on religious grounds.23 However, other states do not offer legal guidance about management of this situation, and most hospitals do not have protocols that address this scenario. This situation leaves treatment teams bewildered and results in variable management strategies such as continuation of organ support until cardiopulmonary arrest, transfer to other facilities, and discontinuation of organ support without family consent. When these situations arise, management is often affected by fear of lawsuits or media attention.18,24,25Unfortunately, comatose patients with anoxic brain injury after cardiac arrest sometimes deteriorate neurologically despite use of therapeutic hypothermia to optimize neurologic recovery. Determination of brain death can be challenging in this scenario. Webb and Samuels26 described a 55-year-old man who was treated with therapeutic hypothermia and had findings consistent with brain death when examined 24 hours after rewarming. Organ procurement was planned, but when he was taken to the operating room, it was noted that he had regained some brainstem reflexes and thus was not brain dead. It is important to be aware that treatment with therapeutic hypothermia can confound the process of brain death determination, so it is necessary to wait an extended period of time after rewarming to perform a brain death determination in such patients.26When patients require ECMO, the chance of brain death is 10% to 20%.27,28 Performance of apnea testing in this population is complex, controversial, and challenging.29,30Determination of brain death can be both medically and ethically challenging. Critical care nurses must be aware of the intricacies of the determination process to provide adequate care to these patients and support their families. Additionally, critical care nurses should strive to ensure that institutional protocols for brain death adhere to societal guidelines in order to minimize complications and controversies.5,6

Patterns and Outcomes of Care in Children With Advanced Heart Disease Receiving Palliative Care Consultation

ContextAlthough access to subspecialty pediatric palliative care (PPC) is increasing, little is known about the role of PPC for children with advanced heart disease (AHD). ObjectivesThe objective of this study was to examine features of subspecialty PPC involvement for children with AHD. MethodsThis is a retrospective single-institution medical record review of patients with a primary diagnosis of AHD for whom the PPC team was initially consulted between 2011 and 2016. ResultsAmong 201 patients, 87% had congenital/structural heart disease, the remainder having acquired/nonstructural heart disease. Median age at initial PPC consultation was 7.7 months (range 1 day–28.8 years). Of the 92 patients who were alive at data collection, 73% had received initial consultation over one year before. Most common indications for consultation were goals of care (80%) and psychosocial support (54%). At initial consultation, most families (67%) expressed that their primary goal was for their child to live as long and as comfortably as possible. Among deceased patients (n = 109), median time from initial consultation to death was 33 days (range 1 day–3.6 years), and children whose families expressed that their primary goal was for their child to live as comfortably as possible were less likely to die in the intensive care unit (P = 0.03) and more likely to die in the setting of comfort care or withdrawal of life-sustaining interventions (P = 0.008). ConclusionPPC involvement for children with AHD focuses on goals of care and psychosocial support. Findings suggest that PPC involvement at end of life supports goal-concordant care. Further research is needed to clarify the impact of PPC on patient outcomes.

Early Withdrawal Decision-Making in Patients with Coma After Cardiac Arrest: A Qualitative Study of Intensive Care Clinicians.

Neurologists are often asked to define prognosis in comatose patients. However, comatose patients following cardiac arrest are usually cared for by cardiologists or intensivists, and it is their approach that will influence decisions regarding withdrawal of life-sustaining interventions (WLSI). We observed that factors leading to these decisions vary across specialties and considered whether they could result in self-fulfilling prophecies and early WLSI. We conducted a hypothesis-generating qualitative study to identify factors used by non-neurologists to define prognosis in these patients and construct an explanatory model for how early WLSI might occur. This was a single-center qualitative study of intensivists caring for cardiac arrest patients with hypoxic-ischemic coma. Thirty attending physicians (n=16) and fellows (n=14) from cardiac (n=8), medical (n=6), surgical (n=10), and neuro (n=6) intensive care units underwent semi-structured interviews. Interview transcripts were analyzed using grounded theory techniques. We found three components of early WLSI among non-neurointensivists: (1) development of fixed negative opinions; (2) early framing of poor clinical pictures to families; and (3) shortened windows for judging recovery potential. In contrast to neurointensivists, non-neurointensivists' negative opinions were frequently driven by patients' lack of consciousness and cardiopulmonary resuscitation circumstances. Both groups were influenced by age and comorbidities. The results demonstrate that factors influencing prognostication differ across specialties. Some differ from those recommended by published guidelines and may lead to self-fulfilling prophecies and early WLSI. Better understanding of this framework would facilitate educational interventions to mitigate this phenomenon and its implications on patient care.

Ethics and the Law: Practical Applications in the Neonatal Intensive Care Unit

Neonatologists may be unfamiliar with the many legal precedents and federal, state, and local laws that may affect medical practice in the United States. Knowledge of specific laws and court cases pertinent to neonatology is essential for medical practice, particularly in today’s increasingly litigious environment. This article reviews several federal and state laws applicable to neonatology and relevant legal cases that involve neonates that have led to the creation of case law in certain states. Two ethically challenging cases are presented and discussed as theoretical examples of how relevant law and court cases may be applied to neonatal practice. Topics include the attempted resuscitation of an infant or withdrawal of life-sustaining interventions despite parental objection, the right of infants to receive appropriate medical therapy in the emergency department and elsewhere in the hospital (regardless of congenital anomalies or prematurity), and difficulties providing end-of-life care to infants.

Parental Perspectives on Suffering and Quality of Life at End-of-Life in Children With Advanced Heart Disease

To describe parent perspectives regarding the end-of-life experience of children with advanced heart disease. Cross-sectional multicenter survey study of bereaved parents. Two tertiary care pediatric hospitals. Parents of children younger than 21 years with primary cardiac diagnoses who died in the hospital 9 months to 4 years before the survey date. Parents were excluded if they were non-English speakers or had previously denied permission to contact. The Survey for Caring for Children with Advanced Heart Disease was developed, piloted, and then sent to parents of all children who died at two sites. Fifty bereaved parents responded (39% response rate) a mean of 2.7 years after their child's death. Median age at death was 6 months (3.6 d to 20.4 yr). At end-of-life, 86% of children were intubated and 46% were receiving mechanical circulatory support. Seventy-eight percent died during withdrawal of life-sustaining interventions and 16% during resuscitative efforts. Parents realized that their child had no realistic chance of survival a median of 2 days prior to death (0-30 d). According to parents, 47% of children suffered "a great deal," "a lot," or "somewhat" during the end-of-life period. The symptoms parents perceived to be causing the most suffering were breathing and feeding difficulties in children under 2 years and fatigue and sleeping difficulties in older children. Seventy-one percent of parents described the quality of life of their child during the last month of life as "poor" or "fair." Most parents (84%) described the quality of care delivered as "very good" or "excellent." According to their parents, many children with advanced heart disease experience suffering in the end-of-life care period. For most, realization that their child has no realistic chance of survival does not occur until late, some not until death is imminent. Once this realization occurs, however, parents perceive peacefulness, a "good death," and excellent quality of care. Strategies for improved communication around symptom management, quality of life, prognosis, and advance care planning are needed for families of children with advanced heart disease.

Neurocritical care

The care of critically ill brain-injured patients is complex and requires careful balancing of cerebral and systemic treatment priorities. A growing number of studies have reported improved outcomes when patients are admitted to dedicated neurocritical care units (NCCUs). The reasons for this observation have not been definitively clarified. When recently published articles are combined with older literature, there have been more than 40 000 patients assessed in observational studies that compare neurological and general ICUs. Although results are heterogeneous, admission to NCCUs is associated with lower mortality and a greater chance of favorable recovery. These findings are remarkable considering that there are few interventions in neurocritical care that have been demonstrated to be efficacious in randomized trials. Whether the relationship is causal is still being elucidated but potential explanations include higher patient volume and, in turn, greater clinician experience; more emphasis on and adherence to protocols to avoid secondary brain injury; practice differences related to prognostication and withdrawal of life-sustaining interventions; and differences in the use and interpretation of neuroimaging and neuromonitoring data. Neurocritical care is an evolving field that is associated with improvements in outcomes over the past decade. Further research is required to determine how monitoring and treatment protocols can be optimized.

Modes of Death in Pediatrics: Differences in the Ethical Approach in Neonatal and Pediatric Patients

To compare end-of-life decisions for neonatal and pediatric patients. This study involved a chart review of all pediatric deaths occurring over a 2-year period at a large maternal-child university hospital. Modes of death were compared. Of the 220 deaths analyzed, 145 occurred in intensive care units (ICUs), including 77 in the neonatal ICU (NICU) and 68 in the pediatric ICU (PICU). Only 6% of deaths were preceded by cardiopulmonary resuscitation. Dying while on the respirator was the most common mode of death in the PICU (51%) and the least common in the NICU (5%; P<.05). Unstable physiology at time of death was much more common in the PICU (82% vs 47%; P<.05). Withdrawal of life-sustaining interventions (LSI) in stable patients for quality of life reasons was the most common cause of death in the NICU (53% vs 16%; P<.05). Seventy-five children died outside of an ICU because LSI were withheld; neonates died mainly of extreme prematurity, and older children died mainly from terminal illness. The majority of pediatric deaths occur in ICUs. Modes of death in the NICU and the PICU are strikingly different. A greater proportion of deaths in the NICU occur in infants with stable physiology who might not have died had LSI not been withdrawn. Most deaths outside of ICUs are attributable to withholding of LSI. A significant proportion of neonates in whom LSI are withheld have a possibility of intact survival, unlike older patients.

Early mortality following spontaneous intracerebral hemorrhage

The authors reviewed the charts of 1,421 patients with cerebral hemorrhage to determine the cause of death. Limitation or withdrawal of life-sustaining interventions was the most common cause of death (68%) followed by brain death (28%). Neurologic reasons were the most common cause of delayed decisions to withdraw or limit therapy. Brain death was more common in African Americans, whereas life-sustaining interventions were withdrawn or limited early more often in whites.

Private Attending Physician Status and the Withdrawal of Life-Sustaining Interventions

Private Attending Physician Status and the Withdrawal of Life-Sustaining Interventions

Withdrawing Care

Objective. —To describe the process and outcomes of withdrawing life-sustaining interventions in a medical intensive care unit (MICU). Design. —Retrospective case series. Setting. —Medical intensive care unit in a community teaching hospital. Patients. —Consecutive series of 28 patients in whom mechanical ventilation, dialysis, and/or vasopressors were withdrawn. We distinguished physiological, neurological, and functional rationales for care withdrawal. Main Outcome Measures. —Duration of discussions, MICU length of stay, and hospital survival. Results. —Mean ±SD Acute Physiology and Chronic Health Evaluation (APACHE II) score was 27.1 ±7.3 on MICU admission, and average ±SD predicted hospital mortality was 61%±22%. Discussions leading to withdrawal of care occurred over an average ±SD of 5.2±5.5 days, with decisions achieved soonest in cases with poor neurological prognosis. Average ±SD MICU length of stay was 1.4±1.8 days following a decision to withdraw MICU care, and only four patients received more than 48 hours of additional MICU care. Four patients were discharged alive from the hospital. Conclusions. —Patients and their surrogates willingly considered outcomes in addition to mortality when considering withdrawal of life-sustaining interventions. Finding an accommodation between physician judgments and patient preferences took time and effort but was an effective means of limiting ineffective life-sustaining efforts. Withdrawing futile or unwanted care was not always fatal. ( JAMA . 1994;271:1358-1361)

Withdrawing care. Experience in a medical intensive care unit.

To describe the process and outcomes of withdrawing life-sustaining interventions in a medical intensive care unit (MICU). Retrospective case series. Medical intensive care unit in a community teaching hospital. Consecutive series of 28 patients in whom mechanical ventilation, dialysis, and/or vasopressors were withdrawn. We distinguished physiological, neurological, and functional rationales for care withdrawal. Duration of discussions, MICU length of stay, and hospital survival. Mean +/- SD Acute Physiology and Chronic Health Evaluation (APACHE II) score was 27.1 +/- 7.3 on MICU admission, and average +/- SD predicted hospital mortality was 61% +/- 22%. Discussions leading to withdrawal of care occurred over an average +/- SD of 5.2 +/- 5.5 days, with decisions achieved soonest in cases with poor neurological prognosis. Average +/- SD MICU length of stay was 1.4 +/- 1.8 days following a decision to withdraw MICU care, and only four patients received more than 48 hours of additional MICU care. Four patients were discharged alive from the hospital. Patients and their surrogates willingly considered outcomes in addition to mortality when considering withdrawal of life-sustaining interventions. Finding an accommodation between physician judgments and patient preferences took time and effort but was an effective means of limiting ineffective life-sustaining efforts. Withdrawing futile or unwanted care was not always fatal.

Withdrawing Care

<h3>Objective.</h3> —To describe the process and outcomes of withdrawing life-sustaining interventions in a medical intensive care unit (MICU). <h3>Design.</h3> —Retrospective case series. <h3>Setting.</h3> —Medical intensive care unit in a community teaching hospital. <h3>Patients.</h3> —Consecutive series of 28 patients in whom mechanical ventilation, dialysis, and/or vasopressors were withdrawn. We distinguished physiological, neurological, and functional rationales for care withdrawal. <h3>Main Outcome Measures.</h3> —Duration of discussions, MICU length of stay, and hospital survival. <h3>Results.</h3> —Mean ±SD Acute Physiology and Chronic Health Evaluation (APACHE II) score was 27.1 ±7.3 on MICU admission, and average ±SD predicted hospital mortality was 61%±22%. Discussions leading to withdrawal of care occurred over an average ±SD of 5.2±5.5 days, with decisions achieved soonest in cases with poor neurological prognosis. Average ±SD MICU length of stay was 1.4±1.8 days following a decision to withdraw MICU care, and only four patients received more than 48 hours of additional MICU care. Four patients were discharged alive from the hospital. <h3>Conclusions.</h3> —Patients and their surrogates willingly considered outcomes in addition to mortality when considering withdrawal of life-sustaining interventions. Finding an accommodation between physician judgments and patient preferences took time and effort but was an effective means of limiting ineffective life-sustaining efforts. Withdrawing futile or unwanted care was not always fatal. (<i>JAMA</i>. 1994;271:1358-1361)