Vol.

-

No.

- -

2011

Journal of Pain and Symptom Management 1

Original Article

The Infusion of Opioids During Terminal Withdrawal of Mechanical Ventilation in the Medical Intensive Care Unit Mark A. Mazer, MD, Chad M. Alligood, PharmD, and Qiang Wu, PhD Division of Pulmonary, Critical Care and Sleep Medicine (M.A.M.), The Brody School of Medicine, and Department of Biostatistics (Q.W.), East Carolina University; and Palliative Care Service (M.A.M.) and Department of Pharmacy (C.M.A.), Pitt County Memorial Hospital, Greenville, North Carolina, USA

Abstract Context. Most deaths in intensive care units occur after limitation or withdrawal of life-sustaining therapies. Often these patients require opioids to assuage suffering; yet, little has been documented concerning their use in the medical intensive care unit. Objectives. To determine the dose and factors influencing the use of opioids in patients undergoing terminal withdrawal of mechanical ventilation in this setting. Methods. Data were prospectively collected from 74 consecutive patients expected to die soon after extubation. The doses of morphine, effect on time to death, and relation of dose to diagnostic categories were analyzed. Results. The mean (
standard deviation) dose of morphine given to patients during the last hour of mechanical ventilation was 5.3 mg/hour. Patients dying after extubation received 10.6 mg/hour just before death. Immediately before extubation, the dose correlated directly with chronic medical opioid use and sepsis with respiratory failure and inversely with coma after cardiopulmonary resuscitation or a primary neurological event. After terminal extubation, the final morphine dose correlated directly with the presence of sepsis with respiratory failure and chronic pulmonary disease. The mean time to death after terminal extubation was 152.7
229.5 minutes without correlation with premorbid diagnoses. After extubation, each 1 mg/hour increment of morphine infused during the last hour of life was associated with a delay of death by 7.9 minutes (P ¼ 0.011). Conclusion. Premorbid conditions may influence the dose of morphine given to patients undergoing terminal withdrawal of mechanical ventilation. Higher doses of morphine are associated with a longer time to death. J Pain Symptom Manage 2011;-:-e-. Ó 2011 U.S. Cancer Pain Relief Committee. Published by Elsevier Inc. All rights reserved.

No financial support was used for this study. The authors declare no conflicts of interest. Address correspondence to: Mark A. Mazer, MD, Division of Pulmonary, Critical Care and Sleep Medicine, The Ó 2011 U.S. Cancer Pain Relief Committee Published by Elsevier Inc. All rights reserved.

Brody School of Medicine, East Carolina University, 600 Moye Boulevard, Greenville, NC 27834, USA. E-mail: [email protected] Accepted for publication: October 9, 2010. 0885-3924/$ - see front matter doi:10.1016/j.jpainsymman.2010.10.256

2

Mazer et al.

Vol.

-

No.

- -

2011

Key Words Opioids, mechanical ventilation, palliative care, withholding treatment, terminal care, guidelines

Introduction Most patients dying in contemporary intensive care units succumb after limitation or withdrawal of life-sustaining therapy rather than during aggressive therapeutic care.1e5 Withdrawal of mechanical ventilatory support is often an integral component of the orchestration of end-of-life care of critically ill, dying patients. Caring does not cease during these final moments of life,6 and caregivers must have a heightened sense of vigilance and be prepared to intervene if these patients demonstrate any signs of distress. Intensive care end-of-life guidelines have been published7e9 and reiterate that compassionate care during the penultimate moments of life includes offering opioids to control suffering from pain and dyspnea. Significant variation in physicians’ attitudes and practice relating to withholding and withdrawing mechanical ventilatory support has been documented.10 Despite ethical concerns relating to shortening the dying process, proper interpretation of the principle of double effect is clearly consistent with offering opioids to relieve pain and other distressing symptoms, such as dyspnea, even if unintended consequences, such as hypotension or hypoventilation, occur.11 The Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatment underscored that clinicians have significant opportunity to improve current practice in regard to pain control as 50% of conscious dying patients were reported to suffer moderate-to-severe pain.12 Although some clinicians may be reluctant to offer escalating doses of opioids and sedatives in the setting of withdrawal of mechanical ventilation for fear of causing unintended harm or foreshortening life, previous reports suggest that these agents have no such consequence and may, in fact, be associated with longer survival after terminal extubation.13,14 Many of our terminally ill patients succumb in the medical intensive care unit relatively soon after withdrawal of mechanical ventilation, before transfer to the palliative care unit for end-of-life care. As a quality surveillance

initiative, we decided to investigate the administration of opioids to these patients to determine if dosing was commensurate with the published guidelines and reports. Campbell15 recently reviewed the limited published literature relating to morphine use during terminal withdrawal of mechanical ventilation and documented differences not only in how dosing is reported but also in the doses of morphine actually used by clinicians. There are scant data concerning the use of opioids specifically in the medical intensive care unit population. The primary objectives of this study were to determine the dose and factors influencing the use of opioids administered to these patients around the time of death.

Methods Study Design We conducted a prospective observational study focused on the administration of opioids to terminally ill patients undergoing withdrawal of mechanical ventilation in a medical intensive care unit. Our Institutional Review Board approved the use of these quality data in the format of a research study and waived consent.

Study Population and Data Collection Over a six-month period, data were prospectively collected from 74 consecutive patients undergoing terminal weaning from mechanical ventilation and extubation in the 16-bed medical intensive care unit of a single, tertiary, referral academic medical center. Only patients considered to be terminally ill and expected to die soon after withdrawal of mechanical ventilation were included in the study. Pertinent data included age, gender, and time to death after terminal extubation. The length of time to death after extubation was calculated starting from the moment of extubation until the patient expired. The mean and median opioid infusion rates given during the last hour before extubation and during the last hour of life after extubation

Vol.

-

No.

- -

2011 Infusion of Opioids During Terminal Withdrawal of Mechanical Ventilation

were determined by review of the medication administration record. The absolute change in dose, if any, was calculated. Although most patients received a continuous infusion of intravenous morphine, fentanyl or hydromorphone infusions also were used. For simplicity and the convenience of analysis, doses were analyzed as morphine-equivalent milligrams. The main reasons for intensive care, medical histories, and chronic medical use of opioids also were recorded. For a variety of reasons, such as coma, the terminal disease process, and medications, these patients were unable to cognitively communicate or self-report symptoms, such as pain or dyspnea. Therefore, we used a behavioral pain scale to assess the pain and discomfort of sedated or otherwise uncommunicative, mechanically ventilated patients.16 This scale focuses on facial grimaces, upper extremity movements, and patient-ventilator synchrony. To gauge patient’s comfort and adjust opioid dosing, our general practice includes close observation of the patient for signs of discomfort and assessment using the above-referenced behavioral pain scale after discontinuation of vasopressors, supplemental oxygen, and positive end-expiratory pressure before terminal extubation. Patients are weaned to minimal or no ventilator support as rapidly as possible while assuring comfort and then extubated. Morphine is titrated to effect using the behavioral pain scale. Other agents such as benzodiazepines are rarely used, and if used, they are not titrated after extubation. In this series, only four patients were extubated on a lorazepam infusion. In many cases, withdrawing these therapeutic modalities is quickly followed by death before extubation. Thus, we gathered data on morphine use in patients who died before terminal extubation and in those who survived to expire postextubation.

Statistical Analysis Statistical analysis was performed using SPSS version 16.0 (Chicago, IL). Descriptive statistics included means
standard deviations (SDs) and medians for opioid doses, including ranges. Percentages were calculated for categorical measures. Simple linear regression was used to examine the relation between time to death and opioid doses. Comparisons were done with Mann-Whitney U tests for continuous variables

3

and chi-squared tests for categorical variables. In all cases, P < 0.05 was considered significant.

Results Patients and Outcomes Of the 74 patients, 34 (46%) were women and 40 (54%) men. The ages ranged from 27 to 87 years, with a mean of 59.6 years. The patients fit into four broad diagnostic categories: 20 (27%) had a chronic pulmonary condition, such as chronic obstructive pulmonary disease, asthma, or obstructive sleep apnea; 17 (23%) suffered from an acute primary neurological event; 32 (43.2%) had sepsis with concomitant acute respiratory failure, such as acute lung injury (ALI) or acute respiratory distress syndrome (ARDS); and 14 (18.9%) were persistently comatose secondary to hypoxic ischemic encephalopathy in the wake of cardiopulmonary resuscitation. Twenty-four patients (32.4%) fit into two or more of the diagnostic categories, whereas five patients (6.8%) did not fit into any of these categories. Thirteen patients (17.6%) used medically prescribed opioids on a chronic basis. Twenty-seven patients (36.5%) died during terminal weaning before extubation, whereas 47 patients (63.5%) died after terminal extubation (Table 1).

Dose of Morphine The mean (
SD) dose of morphine administered to all patients during the last hour of mechanical ventilation was 5.3
6.8 mg/ hour, with a median dose of 2.5 mg/hour (range 0e25 mg/hour). Those who died before terminal extubation received mean and median doses of 5.5
7.3 and 2.5 mg/hour (range 0e25 mg/hour), respectively. Patients who died after extubation received mean and median doses of 5.2
6.5 and 2.5 mg/hour (range 0e20 mg/hour) before extubation and mean and median rates of 10.6
10.7 and 10 mg/hour (range 0e53.3 mg/hour) after extubation during the last hour before death. This difference is statistically significant, with the change of mean and median doses from before extubation to the terminal rates after extubation being 5.4 and 7.5 mg/hour, respectively (P < 0.001) (Table 2). Thirty-six patients (48.6%) were not receiving any opioids before terminal weaning or extubation. Considering the 38 patients (51.4%)

4

Mazer et al.

Vol.

-

No.

- -

2011

Table 1 Gender, General Diagnostic Categories, and Time to Death After Terminal Extubation Patients and Diagnosis Total patients Male Female Chronic medical opiate use Chronic pulmonary condition Acute neurological event Sepsis and ARDS/ALI Hypoxic ischemic encephalopathy Other

n (%) 74 40 34 13 20 17 32 14 5

(100) (54) (46) (17.6) (27) (23) (43.2) (18.9) (6.8)

Died After Extubation, n (%) 47 30 17 9 15 12 17 7 4

(63.5) (75) (50) (69.2) (75) (70.6) (53.1) (50) (80)

Time to Death (
SD) After Extubation (minutes) 152.7
164.0
131.0
210.4
277.7
65.5
161.2
146.7
181.5


229.5 229.2, 235.4, 258.0, 294.0, 109.8, 233.5, 232.2, 270.0,

ns ns ns ns ns ns ns ns

ns ¼ not significant. These parameters did not influence time to death after extubation. In regard to the diagnoses, the sum of the percentages exceeds 100% because several patients had more than one diagnosis.

receiving morphine, the mean and median doses were 10.7
6.9 and 8.8 mg/hour (range 2.5e25 mg/hour), respectively, for the 14 patients (36.8%) who died before extubation. Concerning the 24 patients (63.2%) who died after extubation, differences between the mean and median doses of 10.2
5.6 and 10 mg/hour (range 2.5e20 mg/hour) before extubation and 15.5
12.3 and 11 mg/hour (range 2e 53.3 mg/hour) before death after extubation were significant (P ¼ 0.022). In this case, the change in mean and median doses was 5.3 and 1 mg/hour, respectively (Table 2).

Association with Diagnostic Categories The dose of morphine given before extubation did not correlate with age, gender, or a chronic pulmonary diagnosis but did correlate directly with home opioid use (mean dose 9.4
8.1 vs. 4.5
6.2 mg/hour, P ¼ 0.019) and sepsis with concomitant respiratory failure

(mean dose 8.9
6.3 vs. 2.6
5.8 mg/hour, P < 0.001). An inverse relation with the presence of persistent coma after cardiopulmonary resuscitation (mean dose 2.0
3.4 vs. 6.1
7.1 mg/ hour, P ¼ 0.036) and a primary neurological diagnosis (mean dose 0.6
1.7 vs. 6.8
7.1 mg/ hour, P < 0.001) was noted (Table 3). The dose of morphine infused during the last hour of life after extubation correlated with sepsis with concomitant respiratory failure (mean dose 15.4
13.8 vs. 7.9
7.4 mg/hour, P ¼ 0.038) and chronic pulmonary disease (mean dose 15.0
12.6 vs. 8.5
9.1 mg/hour, P ¼ 0.034). The absolute change in dose before extubation to the preterminal dose did not correlate any of the variables (Table 3).

Time to Death Overall, the mean (
SD) time to death after terminal extubation was 152.7
229.5 minutes (median time 42 minutes), with the earliest

Table 2 Dose of Morphine During the Last Hour Before Extubation and/or Death Patients, n (%)

Mean Dose (
SD) of Morphine (mg/hour)

Median Dose (Range) of Morphine (mg/hour)

All patients, last hour of mechanical ventilation, 74

5.3
6.8

2.5 (0e25)

Deceased before terminal extubation, 27 (36.5)

5.5
7.3

2.5 (0e25)

5.2
6.5 10.6
10.7 5.4, P < 0.001

2.5 (0e20) 10.0 (0e53.3) 7.5, P < 0.001

10.7
6.9

8.8 (2.5e25)

10.2
5.6 15.5
12.3 5.3, P ¼ 0.022

10.0 (2.5e20) 11.0 (2.0e53.3) 1.0, P ¼ 0.022

Deceased after terminal extubation, 47 (63.5) Dose preextubation Dose postextubation just before death Change of dose Patients received morphine, any dose, 38 (51.4) Deceased before extubation, 14 (36.8) Deceased after extubation, 24 (63.2) Dose preextubation Dose postextubation just before death Change of dose

The increase in dose given during the last hour of life after extubation compared with the dose given just before extubation is significant.

Vol.

-

No.

- -

2011 Infusion of Opioids During Terminal Withdrawal of Mechanical Ventilation

death four minutes and the latest death 934 minutes after extubation. The mean times to death ranged from 65.5
109.8 minutes for those with a neurological diagnosis to 277.7
294.0 minutes for those with a chronic pulmonary diagnosis (Table 1). However, there was lack of significant correlation with any of the specific variables. The dose of opioid before extubation did not correlate with the time to death. However, higher doses of opioid given during the hour before death after extubation did correlate with a longer time to death (P ¼ 0.011), with each 1 mg/hour of morphine associated with a delay in the time to death of 7.9 minutes. The time to death also directly correlated with the change in dose (P ¼ 0.002) from before to after extubation just before death, with each 1 mg/ hour increment of morphine associated with a 12.2 minute increase in the time to death.

Discussion This study demonstrates that antemortem diagnoses potentially affect opioid dosing during terminal withdrawal of mechanical ventilation. Furthermore, higher doses of opioids are associated with longer survival after terminal extubation. The usual infusion rate of morphine recommended for critically ill patients including those undergoing withdrawal of life-sustaining

5

therapy is 0.05e0.5 mg/kg/hour.7e9 This is a wide range, offering much latitude, and ideally, clinicians should have an understanding of what doses are actually used in their personal practice in comparison with clinical guidelines and the published literature. In our study, the overall mean dose of morphine during the last hour of mechanical ventilation in a medical intensive care unit was 5.3
6.8 mg/hour, with a median dose of 2.5 mg/hour (range 0e25 mg/hour). Those who died after extubation received mean and median rates of 10.6
10.7 and 10 mg/hour (range 0e53.3 mg/hour), respectively, during the last hour before death. These doses and ranges are generally consistent with previous reports describing the use of opioids during withdrawing or withholding life-sustaining measures in various hospital settings and intensive care units.13,17e26 Of note, most authors report doses generally well within the lower-to-middle range of current practice guidelines. Previous studies have reported no relation between the duration of survival and the dose of morphine given both before and after withdrawal of mechanical ventilation.19e22 Campbell et al.20 reported no difference in dosing in relation to baseline Glasgow Coma Scale scores, the duration of weaning, or the time to death after extubation. Chan et al.13 noted no association

Table 3 Mean Dose of Morphine Before Terminal Extubation and Before Death Postextubation Mean Dose (
SD) of Morphine (mg/hour) Patient Category

Preextubation

P Value

Postextubation

P Value

Gender Male Female

5.8
6.8 4.8
6.8

0.346

11.3
11.9 9.4
8.4

0.730

Chronic medical opiate use Yes No

9.4
8.1 4.5
6.2

0.019

8.8
7.6 11.0
11.3

0.780

Chronic pulmonary condition Yes No

6.8
7.8 4.8
6.3

0.262

15.0
12.6 8.5
9.1

0.034

Acute neurological event Yes No

0.6
1.7 6.8
7.1