Incidence and Risk Model Development for Severe Tachypnea Following Terminal Extubation

      Background

      Palliative ventilator withdrawal (PVW) in the ICU is a common occurrence.

      Research Question

      The goal of this study was to measure the rate of severe tachypnea as a proxy for dyspnea and to identify characteristics associated with episodes of tachypnea.

      Study Design and Methods

      This study assessed a retrospective cohort of ICU patients from 2008 to 2012 mechanically ventilated at a single academic medical center who underwent PVW. The primary outcome of at least one episode of severe tachypnea (respiratory rate > 30 breaths/min) within 6 h after PVW was measured by using detailed physiologic and medical record data. Multivariable logistic regression was used to examine the association between patient and treatment characteristics with the occurrence of a severe episode of tachypnea post extubation.

      Results

      Among 822 patients undergoing PVW, 19% and 30% had an episode of severe tachypnea during the 1-h and 6-h postextubation period, respectively. Within 1 h postextubation, patients with the following characteristics were more likely to experience tachypnea: no pre-extubation opiates (adjusted OR [aOR], 2.08; 95% CI, 1.03-4.19), lung injury (aOR, 3.33; 95% CI, 2.19-5.04), Glasgow Coma Scale score > 8 (aOR, 2.21; 95% CI, 1.30-3.77), and no postextubation opiates (aOR, 1.90; 95% CI, 1.19-3.00).

      Interpretation

      Up to one-third of ICU patients undergoing PVW experience severe tachypnea. Administration of pre-extubation opiates (anticipatory dosing) represents a key modifiable factor that may reduce poor symptom control.

      Key Words

      Abbreviations:

      aOR (adjusted OR), CMO (comfort measures only), GCS (Glasgow Coma Scale), IQR (interquartile range), MIMIC (Medical Information Mart for Intensive Care), PVW (palliative ventilator withdrawal), SOFA (Sequential Organ Failure Assessment)
      Nearly 25% of deaths in the United States occur in ICUs.
      • Angus D.C.
      • Barnato A.E.
      • Linde-Zwirble W.T.
      • et al.
      Use of intensive care at the end of life in the United States: an epidemiologic study.
      ,
      • Khandelwal N.
      • Benkeser D.
      • Coe N.B.
      • Engelberg R.A.
      • Teno J.M.
      • Curtis J.R.
      Patterns of cost for patients dying in the intensive care unit and implications for cost savings of palliative care interventions.
      The majority of deaths in the ICU reflect a transition from curative to comfort-focused care through a process of palliative ventilator withdrawal (PVW).
      • Cook D.
      • Rocker G.
      Dying with dignity in the intensive care unit.
      Numerous professional societies and patient groups have advocated for improved management of ventilated ICU patients transitioning to palliative care.
      • Carlet J.
      • Thijs L.G.
      • Antonelli M.
      • et al.
      Challenges in end-of-life care in the ICU. Statement of the 5th International Consensus Conference in Critical Care: Brussels, Belgium, April 2003.
      • Holloway R.G.
      • Arnold R.M.
      • Creutzfeldt C.J.
      • et al.
      Palliative and end-of-life care in stroke: a statement for healthcare professionals from the American Heart Association/American Stroke Association.
      • Lanken P.N.
      • Terry P.B.
      • Delisser H.M.
      • et al.
      An official American Thoracic Society clinical policy statement: palliative care for patients with respiratory diseases and critical illnesses.
      • Salinas J.
      • Sprinkhuizen S.M.
      • Ackerson T.
      • et al.
      An international standard set of patient-centered outcome measures after stroke.
      • Steinhauser K.E.
      • Christakis N.A.
      • Clipp E.C.
      • McNeilly M.
      • McIntyre L.
      • Tulsky J.A.
      Factors considered important at the end of life by patients, family, physicians, and other care providers.
      • Teno J.M.
      • Clarridge B.R.
      • Casey V.
      • et al.
      Family perspectives on end-of-life care at the last place of care.
      • Truog R.D.
      • Campbell M.L.
      • Curtis J.R.
      • et al.
      Recommendations for end-of-life care in the intensive care unit: a consensus statement by the American College [corrected] of Critical Care Medicine.
      Research conducted over the past 15 years has informed several aspects of palliative care in the ICU setting, particularly how to improve communication and family-centered care.
      • Truog R.D.
      • Campbell M.L.
      • Curtis J.R.
      • et al.
      Recommendations for end-of-life care in the intensive care unit: a consensus statement by the American College [corrected] of Critical Care Medicine.
      • Carson S.S.
      • Cox C.E.
      • Wallenstein S.
      • et al.
      Effect of palliative care-led meetings for families of patients with chronic critical illness: a randomized clinical trial.
      • Creutzfeldt C.J.
      • Hanna M.G.
      • Cheever C.S.
      • et al.
      Palliative care needs assessment in the neuro-ICU: effect on family.
      A controlled trial to improve care for seriously ill hospitalized patients. The study to understand prognoses and preferences for outcomes and risks of treatment (SUPPORT). The SUPPORT Principal Investigators.
      • Lautrette A.
      • Darmon M.
      • Megarbane B.
      • et al.
      A communication strategy and brochure for relatives of patients dying in the ICU.
      • Nelson J.E.
      • Curtis J.R.
      • Mulkerin C.
      • et al.
      Choosing and using screening criteria for palliative care consultation in the ICU: a report from the Improving Palliative Care in the ICU (IPAL-ICU) Advisory Board.
      • Sullivan A.M.
      • Rock L.K.
      • Gadmer N.M.
      • Norwich D.E.
      • Schwartzstein R.M.
      The impact of resident training on communication with families in the intensive care unit. Resident and family outcomes.
      • Aslakson R.A.
      • Reinke L.F.
      • Cox C.
      • Kross E.K.
      • Benzo R.P.
      • Curtis J.R.
      Developing a research agenda for integrating palliative care into critical care and pulmonary practice to improve patient and family outcomes.
      • Davidson J.E.
      • Aslakson R.A.
      • Long A.C.
      • et al.
      Guidelines for family-centered care in the neonatal, pediatric, and adult ICU.
      • Curtis J.R.
      • Downey L.
      • Engelberg R.A.
      The importance and challenge of measuring family experience with end-of-life care in the ICU.
      • Curtis J.R.
      • Treece P.D.
      • Nielsen E.L.
      • et al.
      Randomized trial of communication facilitators to reduce family distress and intensity of end-of-life care.
      • Seaman J.B.
      • Arnold R.M.
      • Scheunemann L.P.
      • White D.B.
      An integrated framework for effective and efficient communication with families in the adult intensive care unit.
      • White D.B.
      • Ernecoff N.
      • Buddadhumaruk P.
      • et al.
      Prevalence of and factors related to discordance about prognosis between physicians and surrogate decision makers of critically ill patients.
      • White D.B.
      • Cua S.M.
      • Walk R.
      • et al.
      Nurse-led intervention to improve surrogate decision making for patients with advanced critical illness.
      • Zier L.S.
      • Burack J.H.
      • Micco G.
      • Chipman A.K.
      • Frank J.A.
      • White D.B.
      Surrogate decision makers' responses to physicians' predictions of medical futility.
      • White D.B.
      • Engelberg R.A.
      • Wenrich M.D.
      • Lo B.
      • Curtis J.R.
      The language of prognostication in intensive care units.
      • Engelberg R.A.
      • Downey L.
      • Wenrich M.D.
      • et al.
      Measuring the quality of end-of-life care.
      • Curtis J.R.
      • Engelberg R.A.
      • Wenrich M.D.
      • et al.
      Studying communication about end-of-life care during the ICU family conference: development of a framework.
      In contrast, relatively little research has focused on the process of PVW. The practice therefore varies widely across ICUs,
      • Abarshi E.A.
      • Papavasiliou E.S.
      • Preston N.
      • Brown J.
      • Payne S.
      • Euro I.
      The complexity of nurses' attitudes and practice of sedation at the end of life: a systematic literature review.
      • Brinkkemper T.
      • van Norel A.M.
      • Szadek K.M.
      • Loer S.A.
      • Zuurmond W.W.
      • Perez R.S.
      The use of observational scales to monitor symptom control and depth of sedation in patients requiring palliative sedation: a systematic review.
      • Mayer S.A.
      • Kossoff S.B.
      Withdrawal of life support in the neurological intensive care unit.
      • Papavasiliou E.S.
      • Brearley S.G.
      • Seymour J.E.
      • Brown J.
      • Payne S.A.
      • Euro I.
      From sedation to continuous sedation until death: how has the conceptual basis of sedation in end-of-life care changed over time?.
      • van Beinum A.
      • Hornby L.
      • Ward R.
      • Ramsay T.
      • Dhanani S.
      Variations in the operational process of withdrawal of life-sustaining therapy.
      • Robert R.
      • Le Gouge A.
      • Kentish-Barnes N.
      • et al.
      Terminal weaning or immediate extubation for withdrawing mechanical ventilation in critically ill patients (the ARREVE observational study).
      • Campbell M.L.
      How to withdraw mechanical ventilation: a systematic review.
      and little is known about factors associated with better symptom control among terminally ill patients undergoing this procedure.
      Patients undergoing PVW who are able to report symptoms describe the sensation of breathlessness or dyspnea as the most distressing symptom they experienced.
      • Truog R.D.
      • Campbell M.L.
      • Curtis J.R.
      • et al.
      Recommendations for end-of-life care in the intensive care unit: a consensus statement by the American College [corrected] of Critical Care Medicine.
      ,
      • Campbell M.L.
      Dyspnea prevalence, trajectories, and measurement in critical care and at life’s end.
      • Campbell M.L.
      • Bizek K.S.
      • Thill M.
      Patient responses during rapid terminal weaning from mechanical ventilation: a prospective study.
      • Puntillo K.
      • Nelson J.E.
      • Weissman D.
      • et al.
      Palliative care in the ICU: relief of pain, dyspnea, and thirst—a report from the IPAL-ICU advisory board.
      Unfortunately, there are few data to guide the clinical approach to minimizing respiratory distress during PVW. Although evidence suggests lower rates of respiratory distress with gradual reduction of ventilator support (terminal weaning) vs immediate extubation,
      • Robert R.
      • Le Gouge A.
      • Kentish-Barnes N.
      • et al.
      Terminal weaning or immediate extubation for withdrawing mechanical ventilation in critically ill patients (the ARREVE observational study).
      it is unknown whether administering analgesia/sedation pre-extubation (anticipatory dosing)
      • Billings J.A.
      Humane terminal extubation reconsidered: the role for preemptive analgesia and sedation.
      ,
      • Delaney J.W.
      • Downar J.
      How is life support withdrawn in intensive care units: a narrative review.
      relieves distress more effectively than giving these drugs only in response to observed symptoms.
      • Truog R.D.
      • Cist A.F.
      • Brackett S.E.
      • et al.
      Recommendations for end-of-life care in the intensive care unit: the Ethics Committee of the Society of Critical Care Medicine.
      ,
      • Truog R.D.
      • Brock D.W.
      • White D.B.
      Should patients receive general anesthesia prior to extubation at the end of life?.
      Assessing physical discomfort among patients undergoing PVW is challenging. A study of terminally ill ICU patients found that 50% were unable to communicate and 20% to 30% were comatose.
      • Smedira N.G.
      • Evans B.H.
      • Grais L.S.
      • et al.
      Witholding and withdrawal of life support from the critically ill.
      Empirical evidence suggests that many patients deemed comatose retain the ability to experience the sensation of respiratory distress.
      • Campbell M.L.
      Respiratory distress: a model of responses and behaviors to an asphyxial threat for patients who are unable to self-report.
      Moreover, a study of 104 ICU patients, most with neurologic injuries, found that 40% of patients deemed comatose by highly trained physicians actually displayed clinical signs of consciousness.
      • Schnakers C.
      • Vanhaudenhuyse A.
      • Giacino J.
      • et al.
      Diagnostic accuracy of the vegetative and minimally conscious state: clinical consensus versus standardized neurobehavioral assessment.
      Taken together, the current report leverages the Medical Information Mart for Intensive Care III (MIMIC-III) database
      • Johnson A.E.
      • Pollard T.J.
      • Shen L.
      • et al.
      MIMIC-III, a freely accessible critical care database.
      to better examine severe tachypnea in a cohort of ICU patients undergoing PVW. MIMIC is a rich dataset that collects detailed physiologic and clinical data from patients cared for in the seven ICUs at Beth Israel Deaconess Medical Center. The main objectives of the study were to describe the prevalence and identify modifiable factors associated with the occurrence of severe tachypnea among patients who underwent PVW through development of multivariable prediction models.

      Patients and Methods

       Data Source

      Data were ascertained from the MIMIC-III,
      • Johnson A.E.
      • Pollard T.J.
      • Shen L.
      • et al.
      MIMIC-III, a freely accessible critical care database.
      a publicly available dataset managed by the Laboratory for Computational Physiology at the Massachusetts Institute of Technology,
      MIMIC-III Critical Care Database.
      which includes data from seven medical, surgical, and subspecialty ICUs at the Beth Israel Deaconess Medical Center. Data from January 1, 2008, to December 31, 2012, were used for this study. MIMIC-III contains de-identified, high-resolution physiologic and treatment data, including chart notes, continuous telemetry data and vital signs, nursing assessments, and the complete electronic health record.
      • Johnson A.E.
      • Pollard T.J.
      • Shen L.
      • et al.
      MIMIC-III, a freely accessible critical care database.
      The study received exempt status by the Institutional Review Board at Beth Israel Deaconess Medical Center.

       Study Population

      Subjects who were mechanically ventilated and underwent PVW were identified from MIMIC-III based on the following eligibility criteria: (1) age ≥ 18 years; (2) mechanical ventilation followed by extubation; (3) comfort measures only (CMO) documented in an order or physician note prior to ventilator withdrawal (CMO order indicates all therapies are directed toward comfort but does not specify particular treatment approaches); and (4) death within 24 h of extubation. Sample size was determined by convenience sample of all data available for analysis.

       Data Elements

      Patient-level variables were extracted from the dataset as potentially associated with tachypnea or treatment of dyspnea based on prior literature
      • Engelberg R.A.
      • Downey L.
      • Wenrich M.D.
      • et al.
      Measuring the quality of end-of-life care.
      ,
      • Abarshi E.A.
      • Papavasiliou E.S.
      • Preston N.
      • Brown J.
      • Payne S.
      • Euro I.
      The complexity of nurses' attitudes and practice of sedation at the end of life: a systematic literature review.
      ,
      • van Beinum A.
      • Hornby L.
      • Ward R.
      • Ramsay T.
      • Dhanani S.
      Variations in the operational process of withdrawal of life-sustaining therapy.
      • Robert R.
      • Le Gouge A.
      • Kentish-Barnes N.
      • et al.
      Terminal weaning or immediate extubation for withdrawing mechanical ventilation in critically ill patients (the ARREVE observational study).
      • Campbell M.L.
      How to withdraw mechanical ventilation: a systematic review.
      • Campbell M.L.
      Dyspnea prevalence, trajectories, and measurement in critical care and at life’s end.
      ,
      • Truog R.D.
      • Cist A.F.
      • Brackett S.E.
      • et al.
      Recommendations for end-of-life care in the intensive care unit: the Ethics Committee of the Society of Critical Care Medicine.
      ,
      • Soffler M.I.
      • Rose A.
      • Hayes M.M.
      • Banzett R.
      • Schwartzstein R.M.
      Treatment of acute dyspnea with morphine to avert respiratory failure.
      • Binks A.P.
      • Desjardin S.
      • Riker R.
      ICU clinicians underestimate breathing discomfort in ventilated subjects.
      • Cottereau A.
      • Robert R.
      • le Gouge A.
      • et al.
      ICU physicians' and nurses' perceptions of terminal extubation and terminal weaning: a self-questionnaire study.
      • Campbell M.L.
      • Yarandi H.N.
      • Mendez M.A.
      Two-group trial of a terminal ventilator withdrawal algorithm: pilot testing.
      • Deschepper R.
      • Laureys S.
      • Hachimi-Idrissi S.
      • Poelaert J.
      • Distelmans W.
      • Bilsen J.
      Palliative sedation: why we should be more concerned about the risks that patients experience an uncomfortable death.
      • Sanders R.D.
      • Tononi G.
      • Laureys S.
      • Sleigh J.W.
      Unresponsiveness does not equal unconsciousness.
      • Parshall M.B.
      • Schwartzstein R.M.
      • Adams L.
      • et al.
      An official American Thoracic Society statement: update on the mechanisms, assessment, and management of dyspnea.
      • Willms D.C.
      • Brewer J.A.
      Survey of respiratory therapists’ attitudes and concerns regarding terminal extubation.
      • Curtis J.R.
      Interventions to improve care during withdrawal of life-sustaining treatments.
      • O'Mahony S.
      • McHugh M.
      • Zallman L.
      • Selwyn P.
      Ventilator withdrawal: procedures and outcomes. Report of a collaboration between a critical care division and a palliative care service.
      • Cook D.
      • Rocker G.
      • Marshall J.
      • et al.
      Withdrawal of mechanical ventilation in anticipation of death in the intensive care unit.
      • Hawryluck L.A.
      • Harvey W.R.
      • Lemieux-Charles L.
      • Singer P.A.
      Consensus guidelines on analgesia and sedation in dying intensive care unit patients.
      and clinical experience. These variables included: demographic characteristics, insurance status, Glasgow Coma Scale (GCS) score (a score > 8 indicates lack of severe coma),
      • Teasdale G.
      • Jennett B.
      Assessment of coma and impaired consciousness: a practical scale.
      the number of ventilator days (dichotomized at the median of 48 h), and Sequential Organ Failure Assessment (SOFA)
      • Vincent J.L.
      • de Mendonca A.
      • Cantraine F.
      • et al.
      Use of the SOFA score to assess the incidence of organ dysfunction/failure in intensive care units: results of a multicenter, prospective study.
      score (dichotomized at the median > 8) and its components (eg, Pao2/Fio2 ratio most proximal to extubation).
      • Knaus W.A.
      • Draper E.A.
      • Wagner D.P.
      • Zimmerman J.E.
      APACHE II: a severity of disease classification system.
      A binary variable was created indicating the presence of reduced diffusion capacity of lung Pao2/Fio2 < 200, a marker of ARDS. Anticipatory dosing with opioids was defined as the administration (by any route, bolus, or increase in continuous infusion rate) of morphine, fentanyl, or hydromorphone within 1 h pre-extubation.
      • Billings J.A.
      Humane terminal extubation reconsidered: the role for preemptive analgesia and sedation.
      ,
      • Truog R.D.
      • Brock D.W.
      • White D.B.
      Should patients receive general anesthesia prior to extubation at the end of life?.
      ,
      • Mazer M.A.
      • Alligood C.M.
      • Wu Q.
      The infusion of opioids during terminal withdrawal of mechanical ventilation in the medical intensive care unit.
      Fentanyl and hydromorphone were converted to morphine-equivalent doses, and a binary variable was created indicating whether opiates were administered within 60 min prior to extubation. Postextubation opiate administration was indicated by variables created in a similar fashion (morphine and morphine-equivalent doses of fentanyl and hydromorphone) at 1 h and 6 h postextubation, respectively. Variables indicating benzodiazepine (midazolam, lorazepam, and diazepam) and propofol administration were created at the time points of within 60 min prior to and 60 min following extubation. Primary and secondary ICU diagnoses were extracted from the electronic health record. Primary neurologic diagnosis was defined by the presence of any International Classification of Diseases, Tenth Revision, code for ischemic stroke, intracerebral hemorrhage, subarachnoid hemorrhage, status epilepticus, anoxic brain injury, cerebral edema, brain herniation/compression, hydrocephalus, or brain mass.
      • van Drimmelen-Krabbe J.J.
      • Bradley W.G.
      • Orgogozo J.M.
      • Sartorius N.
      The application of the International Statistical Classification of Diseases to neurology: ICD-10.
      Physiologic measures such as heart rate, respiratory rate, and other vital signs were obtained from telemetry data for up to 6 h post-extubation or death. All telemetry data are validated by a patient’s ICU nurse, with inaccurate values corrected or removed as part of routine clinical documentation. ICU type (eg, medical, surgical, subspecialty) was also extracted.

       Outcomes

      The primary outcome of interest was severe tachypnea following PVW. The American Thoracic Society defines dyspnea as the “subjective experience of breathing discomfort that consists of qualitatively distinct sensations that vary in intensity.”
      • Parshall M.B.
      • Schwartzstein R.M.
      • Adams L.
      • et al.
      An official American Thoracic Society statement: update on the mechanisms, assessment, and management of dyspnea.
      Dyspnea is therefore not uniformly assessed, and although tachypnea is not equivalent to dyspnea, it is a measurable objective sign that correlates with breathlessness in controlled settings.
      • Simon P.M.
      • Schwartzstein R.M.
      • Weiss J.W.
      • et al.
      Distinguishable sensations of breathlessness induced in normal volunteers.
      ,
      • Banzett R.B.
      • Pedersen S.H.
      • Schwartzstein R.M.
      • Lansing R.W.
      The affective dimension of laboratory dyspnea: air hunger is more unpleasant than work/effort.
      We used a respiratory rate of > 30 breaths/min (validated from telemetry) to define severe tachypnea,
      • McGee S.
      Evidence-Based Physical Diagnosis.
      which was measured up to 1 h and until 6 h from the time of extubation or death.

       Statistical Analysis

      Frequencies were used to describe all variables. Bivariate analyses using logistic regression measured the association of individual independent variables and the occurrence of a severe tachypnea episode measured at two time points: within 1 h and 6 h postextubation (outcomes). The following patient characteristics (described in the Data Elements section) were included as independent variables: demographic information, GCS score > 8, mechanical ventilation > 48 h prior to PVW, SOFA score > 8, Pao2/Fio2 ratio < 200, anticipatory dosing with opiates, pre-extubation and postextubation benzodiazepine and propofol use, opiate dosing within 1 h and 6 h postextubation, non-neurologic primary diagnosis, and ICU type.
      The process for selection of variables was as follows: independent variables associated with the outcomes at P < .20 in these unadjusted analyses were included in a multivariable logistic regression model for the outcome at 1 h and 6 h, controlling for time at risk (patient survival time). Repeated episodes of severe tachypnea within the same patient, and patients who died, were censored from the analysis of population at risk. Complete case analysis was used for missing data. ORs with 95% CIs were generated from these analyses. A Cox proportional hazards model described time to tachypnea event, and Kaplan-Meier time-to-event curves were generated comparing patients receiving and not receiving anticipatory dosing of opiates. Analyses were performed by using R version 3.6.2 (R Foundation for Statistical Computing) and Stata version 13 (StataCorp) software packages.

      Results

      Among the 17,692 patients in the MIMIC-III database during the years 2008 to 2012, there were 1,251 patients undergoing PVW. The analytical sample consisted of 822 patients with complete CMO and postextubation respiratory rate data available for analysis; 248 of these patients (30%) experienced at least one episode of severe tachypnea following PVW. Median age was 73 years (interquartile range [IQR], 60-82 years) with 68% of patients aged > 65 years (Table 1). Of note, 50% of patients had a GCS score ≤ 8, indicating severe coma, and only 67% had a non-neurologic diagnosis. As shown in Figure 1, the highest proportion of patients experienced an episode of severe tachypnea within the first hour of extubation (19%). Only 20% received anticipatory dosing with opiates. The median morphine-equivalent bolus dose administered within 60 min prior to extubation was 4 mg (IQR, 2-5 mg), and the median change in continuous infusion rate was 5 mg/h (IQR, 3-10 mg/h). Postextubation, the majority of patients received opiates (58%), followed by benzodiazepines (30%) and propofol (8%).
      Table 1Patient Characteristics and Their Unadjusted Association With Episodes of Severe Tachypnea Within 1 h and 6 h of Extubation (N = 822)
      CharacteristicNo. (%) of Patients With CharacteristicNo. (%)
      a Reported proportions are by row and will not sum to 100.
      with Severe Tachypnea at 1 h (n = 153)
      ORNo. (%)
      a Reported proportions are by row and will not sum to 100.
      with Severe Tachypnea at 6 h (n = 248)
      OR
      Age, y
       18-4031 (4)8 (26)1.54 (0.68-3.51)10 (32)1.00
       41-60156 (19)32 (21)1.15 (0.75-1.78)49 (31)0.93 (0.40-2.15)
       61-80381 (47)66 (17)0.84 (0.59-1.20)114 (30)0.86 (0.39-1.93)
       81-100250 (31)47 (18)1.00 (0.69-1.48)73 (29)0.83 (0.37-1.87)
      Male sex449 (55)92 (60)1.30 (0.92-1.84)142 (57)1.20 (0.89-1.62)
      Ethnicity
       White593 (85)106 (17)0.67 (0.41-1.10)168 (28)1.00
       Black41 (6)16 (39)3.00 (1.55-5.80)18 (44)1.83 (0.96-3.48)
       Asian23 (3)3 (13)0.90 (0.30-2.70)9 (39)1.50 (0.64-3.54)
       Hispanic21 (3)4 (19)0.67 (0.41-1.10)5 (23)0.78 (0.28-2.18)
       Other22 (3)3 (14)0.64 (0.20-2.30)7 (32)1.09 (0.44-2.73)
      Insurance
       Medicare571 (70)108 (19)1.00178 (31)1.00
       Medicaid51 (6)3 (18)0.93 (0.44-1.95)15 (29)0.85 (0.45-1.59)
       Private180 (22)34 (19)1.01 (0.66-1.55)49 (27)0.81 (0.55-1.17)
       Uninsured16 (2)2 (13)0.62 (0.14-2.74)4 (25)0.68 (0.22-2.14)
       GCS score > 8406 (50)131 (85)1.65 (1.02-2.66)104 (26)1.47 (0.99-2.17)
       Ventilation > 48 h402 (51)75 (52)1.05 (0.73-1.50)126 (31)1.16 (0.85-1.58)
       SOFA score > 8402 (49)89 (58)1.56 (1.09-2.23)127 (52)1.20 (0.88-1.62)
       Pao2/Fio2 < 200268 (34)76 (49)2.13 (1.50-3.05)100 (40)1.48 (1.09-2.03)
       Pre-extubation benzodiazepine250 (30)43 (28)0.86 (0.59-1.27)72 (29)0.91 (0.66-1.27)
      Pre-extubation propofol94 (11)15 (10)0.80 (0.45-1.44)28 (11)0.99 (0.62-1.60)
      No anticipatory dosing652 (79)128 (83)1.35 (0.85-2.14)128 (20)1.34 (0.86-2.08)
      Postextubation benzodiazepine240 (29)44 (29)0.96 (0.65-1.41)75 (30)1.05 (0.76-1.47)
      Postextubation propofol64 (8)8 (5)0.57 (0.26-1.21)19 (8)0.91 (0.52-1.59)
      No postextubation opiates at 1 h374 (46)57 (37)1.53 (1.07-2.20)105 (42)1.23 (0.90-1.67)
      No postextubation opiates at 6 h304 (37)71 (46)3.87 (2.58-5.82)93 (37)2.39 (1.65-3.46)
      Nonneurologic primary diagnosis548 (67)124 (81)2.39 (1.55-3.65)52 (21)1.76 (1.23-2.51)
      ICU type
       Surgical199 (24)24 (16)1.0050 (26)1.00
       Medical397 (48)95 (62)2.29 (1.41-3.71)134 (36)1.61 (1.09-2.36)
       Trauma104 (13)14 (9)1.13 (0.56-2.30)25 (25)0.97 (0.55-1.68)
       Cardiac81 (10)14 (9)1.52 (0.74-3.12)27 (34)1.51 (0.86-2.65
       Cardiac surgical36 (4)6 (4)1.46 (0.55-3.87)10 (32)1.38 (0.61-3.13)
      Bolded items indicate statistical significance at P < .05. GCS score > 8 indicates lack of severe coma. Anticipatory dosing indicates administration of opiates within 60 min prior to extubation. Pao2/Fio2 < 200 indicates substantial impairment in lung oxygen diffusion capacity and is a marker of ARDS. GCS = Glasgow Coma Scale; SOFA = Sequential Organ Failure Assessment.
      a Reported proportions are by row and will not sum to 100.
      Figure thumbnail gr1
      Figure 1Number of patients with postextubation tachypnea according to time (n = 822).
      In the unadjusted analysis, factors associated with having an episode of tachypnea within 1 h postextubation at the P < .20 level were: male sex, black race, GCS score > 8, SOFA score > 8, Pao2/Fio2 ratio < 200, no anticipatory dosing, no postextubation opiates, non-neurologic primary diagnosis, and ICU type. When these variables were entered into a multivariable model (Table 2), factors that remained independently associated with having a severe episode of tachypnea were: no anticipatory dosing (adjusted OR [aOR], 2.08; 95% CI, 1.03-4.19), no postextubation opiates (aOR, 1.90; 95% CI, 1.19-3.00), GCS score > 8 (aOR, 2.21; 95% CI, 1.30-3.77), and Pao2/Fio2 < 200 (aOR, 3.33; 95% CI, 2.19-5.04). The SOFA score was excluded from the final model to avoid collinearity with the Pao2/Fio2 ratio. To estimate the potential for synergistic effects for anticipatory dosing and postextubation opiate use, an interaction term (eg, the product of these effects) was included. We also explored the potential for interaction between sex and race but observed little evidence in favor of this; thus, the model assumes a sex effect consistent across categories of racial identification, and vice versa. The concordance statistic for the final model was 0.65 (95% CI, 0.60-0.70).
      Table 2Independent Predictors of Severe Tachypnea 1 h and 6 h Following PVW (N = 822)
      Characteristic1 h Post-PVW6 h Post-PVW
      Male sex1.47 (0.98-2.19)1.46 (1.03-2.09)
      Race0.33 (.065-6.29)0.44 (0.03-7.64)
      No anticipatory dosing2.08 (1.03-4.19)1.36 (0.91-1.99)
      No postextubation opiates1.90 (1.19-3.00)1.35 (0.91-1.99)
      ICU type1.73 (0.51-5.91)0.85 (.033-2.17)
      GCS score > 82.21 (1.30-3.77)2.25 (1.45-3.53)
      Pao2/Fio2 < 2003.33 (2.19-5.04)2.34 (1.60-3.41)
      Nonneurologic primary diagnosis3.81 (0.24-7.51)2.76 (0.13-8.32)
      No anticipatory dosing ∗ no postextubation opiates1.20 (0.16-4.34)1.34 (0.21-5.56)
      Data are presented as adjusted OR (95% CI). Bolded items indicate statistical significance at P < .05. Pao2/Fio2 < 200 indicates substantial impairment in lung oxygen diffusion capacity and is a marker of ARDS. Anticipatory dosing indicates administration of opiates within 60 min prior to extubation. No anticipatory dosing ∗ no postextubation opiates indicates the interaction term for these covariates. Covariates listed in the table represent all variables included in the final models for the outcome at both time points. PVW = palliative ventilator withdrawal. See Table 1 legend for expansion of other abbreviation.
      In the unadjusted analysis, factors associated with having an episode of tachypnea within 6 h following PVW at the P < .20 level were: male sex, black race, GCS score > 8, Pao2/Fio2 ratio < 200, no postextubation opiates up to 6 h, non-neurologic primary diagnosis, and ICU type. When these variables were entered into a multivariable model, factors that remained independently associated with having a severe episode of tachypnea were: male sex (aOR, 1.46; 95% CI, 1.03-2.09), GCS score > 8 (aOR, 2.25; 95% CI, 1.45-3.53), and Pao2/Fio2 ratio < 200 (aOR, 2.34; 95% CI, 1.60-3.41).
      Cox proportional hazards modeling including the covariates of male sex, black race, GCS score > 8, Pao2/Fio2 ratio < 200, no anticipatory dosing, no postextubation opiates up to 6 h, non-neurologic primary diagnosis, and ICU type revealed only that patients with a Pao2/Fio2 ratio < 200 were significantly associated with time to tachypnea episode (hazard ratio, 1.48; 1.11-1.97). Kaplan-Meier time-to-event curves comparing patients receiving and not receiving anticipatory dosing revealed no difference in time to death over the full 24 h postextubation period between groups (Fig 2). Log-rank test for equality of survivor functions (P = .28).
      Figure thumbnail gr2
      Figure 2Kaplan-Meier survival curves according to anticipatory dosing of opiates pre-extubation.

      Discussion

      We identified a 30% rate of severe tachypnea during PVW among a mixed population of critically ill patients. The highest incidence of severe tachypnea occurred in the first hour following extubation. A key finding of this analysis is that the odds of severe tachypnea following PVW are more than doubled without anticipatory dosing of opiates. Although a similar strength and direction of association were seen with postextubation opiate dosing, these findings lend empirical support to the theoretical rationale for anticipatory dosing of opiates pre-extubation to prevent uncontrolled symptoms during PVW. There was notably no difference in time to death between patients receiving anticipatory dosing and those who did not. Higher level of consciousness, no opiates given within 1 h of extubation, and lung injury (Pao2/Fio2 ratio < 200) were also associated with greater risk for severe tachypnea within 1 h of PVW.
      Despite the immense resources devoted to the monitoring and care of ICU patients, the relatively high prevalence of this marker of distress during the transition to comfort measures is surprising but consistent with prior estimates.
      • Mayer S.A.
      • Kossoff S.B.
      Withdrawal of life support in the neurological intensive care unit.
      ,
      • Morris D.
      • Galicia-Castillo M.
      Dying with dyspnea in the hospital.
      A small study of 32 terminally extubated patients in a neurologic ICU reported a 59% rate of labored breathing.
      • Mayer S.A.
      • Kossoff S.B.
      Withdrawal of life support in the neurological intensive care unit.
      Apart from the small sample size of the study, the assessment of labored breathing is not fully standardized and may have led to biased estimates. Others have estimated dyspnea rates of 34% among terminally extubated patients.
      • Mayer S.A.
      • Kossoff S.B.
      Withdrawal of life support in the neurological intensive care unit.
      ,
      • Morris D.
      • Galicia-Castillo M.
      Dying with dyspnea in the hospital.
      Previous evaluations of the use of anticipatory dosing of opiates in the setting of PVW have yielded unclear results. Two studies found that anticipatory dosing was not associated with earlier time to death, but the investigators did not examine symptom control or compare it with reactive dosing.
      • Chan J.D.
      • Treece P.D.
      • Engelberg R.A.
      • et al.
      Narcotic and benzodiazepine use after withdrawal of life support—association with time to death?.
      ,
      • Ankrom M.
      • Zelesnick L.
      • Barofsky I.
      • Georas S.
      • Finucane T.E.
      • Greenough III, W.B.
      Elective discontinuation of life-sustaining mechanical ventilation on a chronic ventilator unit.
      A small study of 29 patients described anticipatory dosing as a treatment option for patients who were alert and at risk for poor postextubation symptom control, although characteristics of patients at risk were not delineated.
      • O'Mahony S.
      • McHugh M.
      • Zallman L.
      • Selwyn P.
      Ventilator withdrawal: procedures and outcomes. Report of a collaboration between a critical care division and a palliative care service.
      Finally, a study of 32 patients undergoing different PVW practices advised caution when using anticipatory dosing, particularly among comatose patients.
      • Campbell M.L.
      • Bizek K.S.
      • Thill M.
      Patient responses during rapid terminal weaning from mechanical ventilation: a prospective study.
      Although reviews and guidelines treat the practice of dosing medications in anticipation of end-of-life symptoms as ethically and clinically acceptable, they also cite a lack of detailed empirical evidence.
      • Downar J.
      • Delaney J.W.
      • Hawryluck L.
      • Kenny L.
      Guidelines for the withdrawal of life-sustaining measures.
      ,
      • Rubenfeld G.D.
      Principles and practice of withdrawing life-sustaining treatments.
      There were differences in risk factors identified by the logistic regression and Cox proportional hazard models. These differences are in part attributable to logistic regression models estimating risk at the time points of 1 and 6 h postextubation, whereas the Cox proportional hazards model estimated risk at time of the tachypnea event. Taken together, interpretation of both models suggests that patients with a Pao2/Fio2 ratio < 200 are at highest risk for an episode of tachypnea. This characteristic is readily available in ICU patients prior to extubation and may be used to tailor approaches for reducing the risk of poorly controlled symptoms in the process of PVW. The mechanisms by which patients with higher levels of consciousness and male patients (in the 6-h model) seem to be at risk for higher rates of severe tachypnea are unclear, and they warrant further exploration in a carefully constructed prospective analysis. In the interim, particular attention should be paid to these risk groups during the process of PVW, with carefully titrated symptom-directed treatments.
      A broader context at odds with our findings are studies reporting that the majority of family members of patients dying in the ICU perceive deaths to have occurred with comfort.
      • Rocker G.M.
      • Heyland D.K.
      • Cook D.J.
      • Dodek P.M.
      • Kutsogiannis D.J.
      • O'Callaghan C.J.
      Most critically ill patients are perceived to die in comfort during withdrawal of life support: a Canadian multicentre study.
      However, the family members surveyed in this prospective study of 155 ICU patients undergoing terminal extubation in Canada identified 30% of families reporting that patients were “mostly comfortable,” and 9% of families reporting “slightly uncomfortable” or “very uncomfortable.”
      • Rocker G.M.
      • Heyland D.K.
      • Cook D.J.
      • Dodek P.M.
      • Kutsogiannis D.J.
      • O'Callaghan C.J.
      Most critically ill patients are perceived to die in comfort during withdrawal of life support: a Canadian multicentre study.
      Although family perception of comfort and support during the dying process in the ICU is a core component of high-quality ICU care, the individual patient’s experience remains paramount. Studies using repeated measurement of objective markers of symptom control during PVW are rare in the medical literature.
      Our study offers the advantage of precise calculation of the population at risk to provide robust estimates of episodes of severe tachypnea during PVW. To our knowledge, this study is the largest analysis of symptom control among PVW patients to date. This study is limited by the fact that tachypnea is not equivalent to the subjective experience of dyspnea among patients able to self-report. The respiratory distress observation scale
      • Campbell M.L.
      • Templin T.
      • Walch J.
      A respiratory distress observation scale for patients unable to self-report dyspnea.
      is the only validated measure for dyspnea in this population, but it has yet to be universally adopted. The reporting rate of the respiratory distress observation scale was low in our population, which limited its use for this analysis. Other markers of respiratory distress, such as labored or obstructed patterns of breathing, are used as markers of comfort during the PVW process, but the integration of a validated, systematic approach to the identification and management of patients during PVW remains to be standardized in critical care practice.
      • Campbell M.L.
      How to withdraw mechanical ventilation: a systematic review.
      ,
      • Puntillo K.
      • Nelson J.E.
      • Weissman D.
      • et al.
      Palliative care in the ICU: relief of pain, dyspnea, and thirst—a report from the IPAL-ICU advisory board.
      The use of observational data introduces the potential for uncontrolled confounding. However, analysis of the complete electronic health record, including free text documentation of each patient’s entire hospital course, renders the possibility of overlooking clinically relevant patient or ICU characteristics unlikely. Although ICU nurses validate all vital signs, some validation could occur erroneously. There were no major institutional changes in the process of PVW during the study period. The process of de-identification within the dataset, however, limits formal statistical tests for time period effects, which could alter the reported associations.

      Conclusions

      The prevalence of poorly controlled symptoms as measured by using severe tachypnea during PVW is common. The risk of severe tachypnea may be reduced with anticipatory dosing of opiates. Patients who are male, with higher levels of consciousness, or have more severe lung injury seem to be at greater risk for severe tachypnea. Further prospective study using standardized measurements of dyspnea, and interventions to improve symptom control during PVW, should be directed at these risk groups.

      Acknowledgments

      Author contributions: C. R. F. takes responsibility for the content of the manuscript, including the data and analysis. C. R. F. and M. A. D. L. H. contributed substantially to the original concept, study design, data analysis and interpretation, and writing of the manuscript. L. A. C. provided substantial contributions to programming and interpretation of data. M. L. C., K. F., and A. N. provided substantial contributions to interpretation of study data and critical review. D. B. W. and S. L. M. provided substantial oversight of the study design, data analysis, and provided critical revision of the manuscript. All authors have reviewed and approved the final version of the manuscript.
      Financial/nonfinancial disclosures: None declared.
      Role of the sponsors: The funding sources did not provide any input or contributions in the development of the research or manuscript.

      References

        • Angus D.C.
        • Barnato A.E.
        • Linde-Zwirble W.T.
        • et al.
        Use of intensive care at the end of life in the United States: an epidemiologic study.
        Crit Care Med. 2004; 32: 638-643
        • Khandelwal N.
        • Benkeser D.
        • Coe N.B.
        • Engelberg R.A.
        • Teno J.M.
        • Curtis J.R.
        Patterns of cost for patients dying in the intensive care unit and implications for cost savings of palliative care interventions.
        J Palliat Med. 2016; 19: 1171-1178
        • Cook D.
        • Rocker G.
        Dying with dignity in the intensive care unit.
        N Engl J Med. 2014; 370: 2506-2514
        • Carlet J.
        • Thijs L.G.
        • Antonelli M.
        • et al.
        Challenges in end-of-life care in the ICU. Statement of the 5th International Consensus Conference in Critical Care: Brussels, Belgium, April 2003.
        Intensive Care Med. 2004; 30: 770-784
        • Holloway R.G.
        • Arnold R.M.
        • Creutzfeldt C.J.
        • et al.
        Palliative and end-of-life care in stroke: a statement for healthcare professionals from the American Heart Association/American Stroke Association.
        Stroke. 2014; 45: 1887-1916
        • Lanken P.N.
        • Terry P.B.
        • Delisser H.M.
        • et al.
        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
        • Salinas J.
        • Sprinkhuizen S.M.
        • Ackerson T.
        • et al.
        An international standard set of patient-centered outcome measures after stroke.
        Stroke. 2016; 47: 180-186
        • Steinhauser K.E.
        • Christakis N.A.
        • Clipp E.C.
        • McNeilly M.
        • McIntyre L.
        • Tulsky J.A.
        Factors considered important at the end of life by patients, family, physicians, and other care providers.
        JAMA. 2000; 284: 2476-2482
        • Teno J.M.
        • Clarridge B.R.
        • Casey V.
        • et al.
        Family perspectives on end-of-life care at the last place of care.
        JAMA. 2004; 291: 88-93
        • Truog R.D.
        • Campbell M.L.
        • Curtis J.R.
        • et al.
        Recommendations for end-of-life care in the intensive care unit: a consensus statement by the American College [corrected] of Critical Care Medicine.
        Crit Care Med. 2008; 36: 953-963
        • Carson S.S.
        • Cox C.E.
        • Wallenstein S.
        • et al.
        Effect of palliative care-led meetings for families of patients with chronic critical illness: a randomized clinical trial.
        JAMA. 2016; 316: 51-62
        • Creutzfeldt C.J.
        • Hanna M.G.
        • Cheever C.S.
        • et al.
        Palliative care needs assessment in the neuro-ICU: effect on family.
        Neurocritical Care. 2017; 27: 163-172
      1. A controlled trial to improve care for seriously ill hospitalized patients. The study to understand prognoses and preferences for outcomes and risks of treatment (SUPPORT). The SUPPORT Principal Investigators.
        JAMA. 1995; 274 ([published correction appears in JAMA. 1996;275(16):1232]): 1591-1598
        • Lautrette A.
        • Darmon M.
        • Megarbane B.
        • et al.
        A communication strategy and brochure for relatives of patients dying in the ICU.
        N Engl J Med. 2007; 356: 469-478
        • Nelson J.E.
        • Curtis J.R.
        • Mulkerin C.
        • et al.
        Choosing and using screening criteria for palliative care consultation in the ICU: a report from the Improving Palliative Care in the ICU (IPAL-ICU) Advisory Board.
        Crit Care Med. 2013; 41: 2318-2327
        • Sullivan A.M.
        • Rock L.K.
        • Gadmer N.M.
        • Norwich D.E.
        • Schwartzstein R.M.
        The impact of resident training on communication with families in the intensive care unit. Resident and family outcomes.
        Ann Am Thorac Soc. 2016; 13: 512-521
        • Aslakson R.A.
        • Reinke L.F.
        • Cox C.
        • Kross E.K.
        • Benzo R.P.
        • Curtis J.R.
        Developing a research agenda for integrating palliative care into critical care and pulmonary practice to improve patient and family outcomes.
        J Palliat Med. 2017; 20: 329-343
        • Davidson J.E.
        • Aslakson R.A.
        • Long A.C.
        • et al.
        Guidelines for family-centered care in the neonatal, pediatric, and adult ICU.
        Crit Care Med. 2017; 45: 103-128
        • Curtis J.R.
        • Downey L.
        • Engelberg R.A.
        The importance and challenge of measuring family experience with end-of-life care in the ICU.
        Intensive Care Med. 2016; 42: 1179-1181
        • Curtis J.R.
        • Treece P.D.
        • Nielsen E.L.
        • et al.
        Randomized trial of communication facilitators to reduce family distress and intensity of end-of-life care.
        Am J Respir Crit Care Med. 2016; 193: 154-162
        • Seaman J.B.
        • Arnold R.M.
        • Scheunemann L.P.
        • White D.B.
        An integrated framework for effective and efficient communication with families in the adult intensive care unit.
        Ann Am Thorac Soc. 2017; 14: 1015-1020
        • White D.B.
        • Ernecoff N.
        • Buddadhumaruk P.
        • et al.
        Prevalence of and factors related to discordance about prognosis between physicians and surrogate decision makers of critically ill patients.
        JAMA. 2016; 315: 2086-2094
        • White D.B.
        • Cua S.M.
        • Walk R.
        • et al.
        Nurse-led intervention to improve surrogate decision making for patients with advanced critical illness.
        Am J Crit Care. 2012; 21: 396-409
        • Zier L.S.
        • Burack J.H.
        • Micco G.
        • Chipman A.K.
        • Frank J.A.
        • White D.B.
        Surrogate decision makers' responses to physicians' predictions of medical futility.
        Chest. 2009; 136: 110-117
        • White D.B.
        • Engelberg R.A.
        • Wenrich M.D.
        • Lo B.
        • Curtis J.R.
        The language of prognostication in intensive care units.
        Med Decis Making. 2010; 30: 76-83
        • Engelberg R.A.
        • Downey L.
        • Wenrich M.D.
        • et al.
        Measuring the quality of end-of-life care.
        J Pain Symptom Manage. 2010; 39: 951-971
        • Curtis J.R.
        • Engelberg R.A.
        • Wenrich M.D.
        • et al.
        Studying communication about end-of-life care during the ICU family conference: development of a framework.
        J Crit Care. 2002; 17: 147-160
        • Abarshi E.A.
        • Papavasiliou E.S.
        • Preston N.
        • Brown J.
        • Payne S.
        • Euro I.
        The complexity of nurses' attitudes and practice of sedation at the end of life: a systematic literature review.
        J Pain Symptom Manage. 2014; 47: 915-925.e911
        • Brinkkemper T.
        • van Norel A.M.
        • Szadek K.M.
        • Loer S.A.
        • Zuurmond W.W.
        • Perez R.S.
        The use of observational scales to monitor symptom control and depth of sedation in patients requiring palliative sedation: a systematic review.
        Palliat Med. 2013; 27: 54-67
        • Mayer S.A.
        • Kossoff S.B.
        Withdrawal of life support in the neurological intensive care unit.
        Neurology. 1999; 52: 1602-1609
        • Papavasiliou E.S.
        • Brearley S.G.
        • Seymour J.E.
        • Brown J.
        • Payne S.A.
        • Euro I.
        From sedation to continuous sedation until death: how has the conceptual basis of sedation in end-of-life care changed over time?.
        J Pain Symptom Manage. 2013; 46: 691-706
        • van Beinum A.
        • Hornby L.
        • Ward R.
        • Ramsay T.
        • Dhanani S.
        Variations in the operational process of withdrawal of life-sustaining therapy.
        Crit Care Med. 2015; 43: e450-e457
        • Robert R.
        • Le Gouge A.
        • Kentish-Barnes N.
        • et al.
        Terminal weaning or immediate extubation for withdrawing mechanical ventilation in critically ill patients (the ARREVE observational study).
        Intensive Care Med. 2017; 43 ([published correction appears in Intensive Care Med. 2017;43(12):1942-1943]): 1793-1807
        • Campbell M.L.
        How to withdraw mechanical ventilation: a systematic review.
        AACN Adv Crit Care. 2007; 18: 397-403
        • Campbell M.L.
        Dyspnea prevalence, trajectories, and measurement in critical care and at life’s end.
        Curr Opin Support Palliat Care. 2012; 6: 168-171
        • Campbell M.L.
        • Bizek K.S.
        • Thill M.
        Patient responses during rapid terminal weaning from mechanical ventilation: a prospective study.
        Crit Care Med. 1999; 27: 73-77
        • Puntillo K.
        • Nelson J.E.
        • Weissman D.
        • et al.
        Palliative care in the ICU: relief of pain, dyspnea, and thirst—a report from the IPAL-ICU advisory board.
        Intensive Care Med. 2014; 40: 235-248
        • Billings J.A.
        Humane terminal extubation reconsidered: the role for preemptive analgesia and sedation.
        Crit Care Med. 2012; 40: 625-630
        • Delaney J.W.
        • Downar J.
        How is life support withdrawn in intensive care units: a narrative review.
        J Crit Care. 2016; : 12-18
        • Truog R.D.
        • Cist A.F.
        • Brackett S.E.
        • et al.
        Recommendations for end-of-life care in the intensive care unit: the Ethics Committee of the Society of Critical Care Medicine.
        Crit Care Med. 2001; 29: 2332-2348
        • Truog R.D.
        • Brock D.W.
        • White D.B.
        Should patients receive general anesthesia prior to extubation at the end of life?.
        Crit Care Med. 2012; 40: 631-633
        • Smedira N.G.
        • Evans B.H.
        • Grais L.S.
        • et al.
        Witholding and withdrawal of life support from the critically ill.
        N Engl J Med. 1990; 322: 309-315
        • Campbell M.L.
        Respiratory distress: a model of responses and behaviors to an asphyxial threat for patients who are unable to self-report.
        Heart Lung. 2008; 37: 54-60
        • Schnakers C.
        • Vanhaudenhuyse A.
        • Giacino J.
        • et al.
        Diagnostic accuracy of the vegetative and minimally conscious state: clinical consensus versus standardized neurobehavioral assessment.
        BMC Neurol. 2009; 9: 35
        • Johnson A.E.
        • Pollard T.J.
        • Shen L.
        • et al.
        MIMIC-III, a freely accessible critical care database.
        Sci Data. 2016; 3: 160035
      2. MIMIC-III Critical Care Database.
        https://mimic.physionet.org/about/mimic/
        Date accessed: September 1, 2018
        • Soffler M.I.
        • Rose A.
        • Hayes M.M.
        • Banzett R.
        • Schwartzstein R.M.
        Treatment of acute dyspnea with morphine to avert respiratory failure.
        Ann Am Thorac Soc. 2017; 14: 584-588
        • Binks A.P.
        • Desjardin S.
        • Riker R.
        ICU clinicians underestimate breathing discomfort in ventilated subjects.
        Respir Care. 2017; 62: 150-155
        • Cottereau A.
        • Robert R.
        • le Gouge A.
        • et al.
        ICU physicians' and nurses' perceptions of terminal extubation and terminal weaning: a self-questionnaire study.
        Intensive Care Med. 2016; 42: 1248-1257
        • Campbell M.L.
        • Yarandi H.N.
        • Mendez M.A.
        Two-group trial of a terminal ventilator withdrawal algorithm: pilot testing.
        J Palliat Med. 2015; 18: 781-785
        • Deschepper R.
        • Laureys S.
        • Hachimi-Idrissi S.
        • Poelaert J.
        • Distelmans W.
        • Bilsen J.
        Palliative sedation: why we should be more concerned about the risks that patients experience an uncomfortable death.
        Pain. 2013; 154: 1505-1508
        • Sanders R.D.
        • Tononi G.
        • Laureys S.
        • Sleigh J.W.
        Unresponsiveness does not equal unconsciousness.
        Anesthesiology. 2012; 116: 946-959
        • Parshall M.B.
        • Schwartzstein R.M.
        • Adams L.
        • et al.
        An official American Thoracic Society statement: update on the mechanisms, assessment, and management of dyspnea.
        Am J Respir Crit Care Med. 2012; 185: 435-452
        • Willms D.C.
        • Brewer J.A.
        Survey of respiratory therapists’ attitudes and concerns regarding terminal extubation.
        Respir Care. 2005; 50: 1046-1049
        • Curtis J.R.
        Interventions to improve care during withdrawal of life-sustaining treatments.
        J Palliative Med. 2005; 8: 116-131
        • O'Mahony S.
        • McHugh M.
        • Zallman L.
        • Selwyn P.
        Ventilator withdrawal: procedures and outcomes. Report of a collaboration between a critical care division and a palliative care service.
        J Pain Symptom Manage. 2003; 26: 954-961
        • Cook D.
        • Rocker G.
        • Marshall J.
        • et al.
        Withdrawal of mechanical ventilation in anticipation of death in the intensive care unit.
        N Engl J Med. 2003; 349: 1123-1132
        • Hawryluck L.A.
        • Harvey W.R.
        • Lemieux-Charles L.
        • Singer P.A.
        Consensus guidelines on analgesia and sedation in dying intensive care unit patients.
        BMC Med Ethics. 2002; 3: E3
        • Teasdale G.
        • Jennett B.
        Assessment of coma and impaired consciousness: a practical scale.
        Lancet. 1974; 2: 81-84
        • Vincent J.L.
        • de Mendonca A.
        • Cantraine F.
        • et al.
        Use of the SOFA score to assess the incidence of organ dysfunction/failure in intensive care units: results of a multicenter, prospective study.
        Crit Care Med. 1998; 26: 1793-1800
        • Knaus W.A.
        • Draper E.A.
        • Wagner D.P.
        • Zimmerman J.E.
        APACHE II: a severity of disease classification system.
        Crit Care Med. 1985; 13: 818-829
        • Mazer M.A.
        • Alligood C.M.
        • Wu Q.
        The infusion of opioids during terminal withdrawal of mechanical ventilation in the medical intensive care unit.
        J Pain Symptom Manage. 2011; 42: 44-51
        • van Drimmelen-Krabbe J.J.
        • Bradley W.G.
        • Orgogozo J.M.
        • Sartorius N.
        The application of the International Statistical Classification of Diseases to neurology: ICD-10.
        J Neurol Sci. 1998; 161: 2-9
        • Simon P.M.
        • Schwartzstein R.M.
        • Weiss J.W.
        • et al.
        Distinguishable sensations of breathlessness induced in normal volunteers.
        Am Rev Respir Dis. 1989; 140: 1021-1027
        • Banzett R.B.
        • Pedersen S.H.
        • Schwartzstein R.M.
        • Lansing R.W.
        The affective dimension of laboratory dyspnea: air hunger is more unpleasant than work/effort.
        Am J Respir Crit Care Med. 2008; 177: 1384-1390
        • McGee S.
        Evidence-Based Physical Diagnosis.
        3rd ed. Elsevier, New York, NY2012: 145-155
        • Morris D.
        • Galicia-Castillo M.
        Dying with dyspnea in the hospital.
        Am J Hosp Palliat Care. 2017; 34: 132-134
        • Chan J.D.
        • Treece P.D.
        • Engelberg R.A.
        • et al.
        Narcotic and benzodiazepine use after withdrawal of life support—association with time to death?.
        Chest. 2004; 126: 286-293
        • Ankrom M.
        • Zelesnick L.
        • Barofsky I.
        • Georas S.
        • Finucane T.E.
        • Greenough III, W.B.
        Elective discontinuation of life-sustaining mechanical ventilation on a chronic ventilator unit.
        J Am Geriatrics Soc. 2001; 49: 1549-1554
        • Downar J.
        • Delaney J.W.
        • Hawryluck L.
        • Kenny L.
        Guidelines for the withdrawal of life-sustaining measures.
        Intensive Care Med. 2016; 42: 1003-1017
        • Rubenfeld G.D.
        Principles and practice of withdrawing life-sustaining treatments.
        Crit Care Clin. 2004; 20: 435-451
        • Rocker G.M.
        • Heyland D.K.
        • Cook D.J.
        • Dodek P.M.
        • Kutsogiannis D.J.
        • O'Callaghan C.J.
        Most critically ill patients are perceived to die in comfort during withdrawal of life support: a Canadian multicentre study.
        Can J Anaesth. 2004; 51: 623-630
        • Campbell M.L.
        • Templin T.
        • Walch J.
        A respiratory distress observation scale for patients unable to self-report dyspnea.
        J Palliat Med. 2010; 13: 285-290