Extubation During Pediatric Extracorporeal Membrane Oxygenation: A Single-Center Experience* Pilar Anton-Martin, MD1; Marita T. Thompson, MD2; Paul D. Sheeran, MD2; Anne C. Fischer, MD, PhD3; Donna Taylor, RRT4; James A. Thomas, MD5
Objectives: Describe aspects of one center’s experience extubating infants and children during extracorporeal membrane oxygenation. Design: Retrospective review of medical records. Setting: Seventy-one-bed critical care service (PICU and cardiovascular ICU) in a large urban tertiary children’s hospital. Patients: Pediatric and neonatal patients supported on extracorporeal membrane oxygenation between 1996 and 2013 who were either not intubated or extubated greater than 24 hours during their extracorporeal membrane oxygenation course. Interventions: None. Measurements and Main Results: Sixteen of 511 patients on extracorporeal membrane oxygenation were extubated for at least 24 hours during their extracorporeal membrane oxygenation courses. Fourteen had respiratory failure and two had cardiac disease. Five patients died while on extracorporeal membrane oxygenation, but the cause of death was not related to complications associated *See also p. 907. 1 Department of Pediatrics, Division of Pediatric Critical Care, University of Texas Southwestern Medical Center, Dallas, TX. 2 Department of Pediatrics, Division of Pediatric Critical Care, University of Missouri, Kansas City/Children’s Mercy Hospital, Kansas City, MO. 3 Division of Pediatric Surgery, William Beaumont Oakland University Medical School, Royal Oak, MI. 4 Department of Critical Care Services, Children’s Medical Center, Dallas, TX. 5 Department of Pediatrics, Section of Pediatric Critical Care, Baylor College of Medicine/Texas Children’s Hospital, Houston, TX. Presented, in part, at the 2012 ELSO Conference in Seattle, WA, September 14–16, 2012, where it was awarded the Robert Bartlett award for best original presentation. Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s website (http://journals.lww.com/pccmjournal). Dr. Anton-Martin received support from the Fundación Alfonso Martín Escudero. The remaining authors have disclosed that they do not have any potential conflicts of interest. For information regarding this article, E-mail: [email protected] Copyright © 2014 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies DOI: 10.1097/PCC.0000000000000235
Pediatric Critical Care Medicine
with extubation. Extubated patients were supported a median of 19.7 days on extracorporeal membrane oxygenation, with a median extubation latency (time between cannulation and first extubation) of 6.2 days and a median extubation duration of 5.5 days. Mean time extubated was 43% of the total time on extracorporeal membrane oxygenation. Two patients were reintubated briefly or had a laryngeal mask airway placed for decannulation (n = 1). The remaining patients were extubated within 5 days of decannulation, weeks afterward (n = 2), transferred to outside facilities (n = 2), or died during extracorporeal membrane oxygenation support (n = 5). We also observed no complications directly attributable to extubation and spontaneous reaeration of consolidated lungs in acute respiratory distress syndrome in extubated patients on extracorporeal membrane oxygenation. Conclusion: Extubation and discontinuation of mechanical ventilation appear feasible in patients requiring long-term extracorporeal membrane oxygenation. Emergency procedure planning may need to be modified in extubated patients on extracorporeal membrane oxygenation. (Pediatr Crit Care Med 2014; 15:861–869) Key Words: extracorporeal membrane life support; extracorporeal membrane oxygenation; extubation; mechanical ventilation; ventilator-induced lung injury
E
xtracorporeal membrane oxygenation (ECMO) developed as a rescue therapy for intractable respiratory, cardiac, or combined cardiopulmonary failure (1). In general, the goals of ECMO therapy are to support cellular respiration by improving tissue oxygenation or Co2 removal until resolution of the primary disease process or as a “bridge” to organ transplantation or another mechanical support modality suitable for long-term use. Because of its invasive nature and potential for catastrophic complications, however, ECMO has been considered a last-resort therapy to be used only when conventional treatments have failed and only long enough until the patient can be returned to more “standard” support modalities (mechanical ventilation, inotropic support, or both). Neither strategy—cannulation after failure of conventional support or minimal duration of ECMO—has been compared www.pccmjournal.org
861
Anton-Martin et al
Table 1.
Patient, Disease, and Extracorporeal Membrane Oxygenation Characteristics
Patient
1
2
ECMO type
Pediatric respiratory
Pediatric respiratory
Pediatric respiratory
Pediatric cardiac
Pediatric respiratory
Pediatric respiratory
Neonatal respiratory
Neonatal respiratory
Age
11 yr
5 yr
4 yr
17 yr
12 yr
3 yr
(35 wk + 5 d)/2 d
(38 wk)/2 d
Sex
Male
Male
Female
Female
Male
Male
Male
Male
Primary disease
Mediastinal mass
Mediastinal mass
Interstitial lung disease
Cardiogenic shock
Aspiration pneumonitis
RSV pneumonia
Meconium aspiration
Right CDH
Secondary disease
Respiratory failure
Respiratory failure
ARDS, air leak
Cardiomyopathy
ARDS
ARDS, air leak
ARDS
Respiratory failure
Maximum oxygenation index-minimum Pao2/Fio2
–/–
–/–
36/61
2/80
37/72
27/55
34/49
14/90
ECMO Mode
VA + V
VA + V
VV + V
VA + V
DLVV + V
DLVV
DLVV
VA
Venous site(s)
RFV
RFV
RFV-LFV
LFV
RIJV
RIJV
RIJV
RIJV
Arterial site
LFA
RFA
—
LFA
Not applicable
Not applicable
Not applicable
RCCA
Additional sites?
Yes (RFV)
Yes (LFV)
Yes (RIJV)
Yes (RFV)
Yes (RFV)
No
No
No
ECMO duration
5
4
54 (84 d total)
20
25
5
18
19
Complications on ECMO
Bleeding
No
Bleeding, infection
Bleeding, clotting in circuit
Bleeding, clotting circuit, delirium
No
Infection, bleeding, cardiac arrest
Flow reversed after cannulation, bleeding.
Sedation preextubation
Not applicable
Not applicable
Fentanyl, midazolam, dexmedetomidine
Fentanyl, midazolam
Fentanyl, midazolam
Fentanyl, midazolam
Fentanyl, midazolam
Fentanyl, midazolam
Fentanyl
Fentanyl, midazolam
Sedation postextubation
3
4
5
6
7
8
Pentobarbital, propofol
Fentanyl, midazolam
Fentanyl, midazolam
Fentanyl, midazolam
Fentanyl, midazolam
Fentanyl, midazolam
Fentanyl
Ketamine
Ketamine
Propofol, ketamine
WD
Pentobarbital, propofol, ketamine
Dexmedetomidine, propofol
WD
WD
WD
ARDS = acute respiratory distress syndrome, CDH = congenital diaphragmatic hernia, ECMO = extracorporeal membrane oxygenation, LFA = left femoral artery, DLVV = double lumen venovenous, LFV = left femoral vein, RCCA = right common carotid artery, RFA = right femoral artery, RFV = right femoral vein, RIJV = right internal jugular vein, RSV = respiratory syncytial virus, VA = venoarterial, VV = venovenous, + V = additional venous cannula. Sedation used: first-line agents (fentanyl and midazolam); second-line agents (dexmedetomidine, ketamine, propofol, and pentobarbital); and withdrawal agents (WD = diazepam, lorazepam, methadone, and morphine).
to alternative approaches that might minimize injury from conventional support modalities and maximize the opportunity for recovery from tissue injury. The adoption of newer circuit technologies (biocompatible surfaces, new pump, and oxygenator technology) and advances in ECMO practice (dedicated teams, standardized approaches to anticoagulation, and early initiation of renal replacement) have led to the perception that ECMO as currently practiced is safer than ECMO practiced at the end of the 1990s (2). This perception also helps explain the increasing prevalence of longer ECMO courses, particularly for severe respiratory disease (3, 4). If in fact ECMO is safer today than 15 years ago, then cannulation 862
www.pccmjournal.org
and support strategies inherited from the early years of ECMO may be outdated as well. One practice that may deserve reexamination is the role of intubation and mechanical ventilation while on extracorporeal support. In most ECMO centers, once a patient has been cannulated, ventilator support is decreased but sedation is maintained to prevent extubation or decannulation. Although there may be a role for maintaining an endotracheal tube (ETT) in place to support pulmonary toilet adjuncts, such as suctioning or bronchoscopy, there is no clinical indication for continued mechanical ventilation when ECMO alone can meet the patient’s tissue metabolic demands. As long as patients remain intubated, the November 2014 • Volume 15 • Number 9
Extracorporeal Support
9
10
11
12
13
14
15
16
Pediatric respiratory
Pediatric respiratory
Neonatal respiratory
Pediatric ECPR
Neonatal respiratory
Neonatal respiratory
Pediatric respiratory
Pediatric respiratory
3 yr
6 yr
(41 wk)/5 d
5 mo
(39 wk)/1 d
19 d
2 yr
5 yr
Male
Male
Male
Female
Male
Female
Female
Male
Pertussis pneumonia
Mycoplasma pneumonia
Meconium aspiration, supraglottic edema
Aortic coarctation, ECMO cardiopulmonary resuscitation
Meconium aspiration
Left CDH
Influenza RSV pneumonia, air pneumonitis leak
ARDS
ARDS
ARDS
Airway disruption/ edema
ARDS
Respiratory failure
ARDS
ARDS
48/42
54/52
32/54
–/–
91/26
50/40
–/–
52/58
DLVV → VA
DLVV
DLVV → VA
VA → DLVV
DLVV
DLVV
DLVV
DLVV
RIJV
RIJV
RIJV/double lumen
Right atrium
RIJV
RIJV
RIJV
RIJV
RCCA
Not applicable
RCCA
Aorta
Not applicable
Not applicable
Not applicable
Not applicable
No
No
No
Switched to DLVV (RIJV)
No
No
Yes (RFV)
Yes (Cephalad)
71
18
15
14
9
7
21
10
Bleeding, clotting, pericardial tamponade
Bleeding, sepsis
Bleeding, recurrent pneumothorax, septic shock, renal failure
Cardiac arrest
Bleeding
No
Bleeding, cannula displaced, delirium
Bleeding
Fentanyl
Fentanyl, midazolam
Fentanyl, midazolam
Fentanyl
Fentanyl, midazolam
Fentanyl, midazolam
Fentanyl, midazolam
Fentanyl, midazolam
Dexmedetomidine
Dexmedetomidine, propofol
Fentanyl, midazolam
Fentanyl, midazolam
Dexmedetomidine, ketamine
Propofol, ketamine
WD
WD
Dexmedetomidine, pentobarbital Fentanyl, midazolam
Fentanyl, midazolam
Fentanyl, midazolam
Fentanyl, midazolam
Fentanyl, midazolam
Dexmedetomidine, pentobarbital, propofol WD
ETT remains an irritant, necessitating sedation with its attendant complications, including sedation habituation and dose escalation, delirium, neuromuscular deconditioning, and prolonged absence of social interaction. Although recent strategies to decrease cumulative sedative doses or to withhold continuous sedation may decrease the need for polypharmacy (5), removal of the primary irritant may be a more effective intervention. Centers providing ECMO have begun to report their experience with extubated, awake, and spontaneously breathing adult patients. Most publications examine the experience with adult lung transplantation (2, 6–9) and acute respiratory distress syndrome (ARDS) (10, 11). Two lung transplantation Pediatric Critical Care Medicine
Fentanyl, midazolam
WD
WD
WD
centers, in Germany and Italy, have recently shared larger case series. The aggregate experience suggests that spontaneously breathing ECMO patients may be better transplant candidates and enjoy better outcomes than those requiring mechanical ventilation prior to transplantation (6, 9). Garcia et al (7, 10) have reported a device to allow extubated patients to ambulate and perform physical activity while on extracorporeal support and thereby avoid the consequences of being bedridden (12). Very few studies describe extubation experience in children and are focused primarily on airway issues (13–15). There is a single published report of three extubated pediatric patients on ECMO awaiting lung transplantation www.pccmjournal.org
863
Anton-Martin et al
State behavioral Scale (SBS) is a behavioral assessment tool with a numeric rating scale that describes the sedation-agitation continuum of pediatric patients supported on mechanical ventilation. It was derived from ratings of seven content dimensions that include descriptors that have been previously described in the literature to be associated with agitation (17). In this way, undersedation is a proxy for activity and/or alertness. To gauge the level of oversedation, appropriate sedation, Figure 1. Distribution of extubated patients on extracorporeal membrane oxygenation (ECMO) over time. The and undersedation, the numnumber of extubated patients on ECMO is plotted against the year in which they were extubated. The majority of ber of days each patient scored extubations occurred in 2011, 2012, and 2013. either –1 to –3 (responsive to gentle touch to unresponfrom the Hannover center, reflecting the incorporation of sive), 0 (awake and able to calm), or +1 (restless and difficult strategies from their evolving adult practice into their care to calm) to +2 (agitated) were tallied. To compare sedation in of children (16). extubated and intubated ECMO patients, the scores of the extuThis report summarizes our experience with patients who bated patients were compared with the scores of the intubated were either cannulated without prior intubation or extubated patients. for more than 24 hours while on ECMO. The Institutional Review Board at the University of Texas Southwestern Medical Center approved this study and waived the need for informed consent. MATERIALS AND METHODS Summary data were collected and reported as percentages, Records of patients supported on ECMO at Children’s Medinormally distributed data expressed as means ± sd, and noncal Center Dallas between 1996 and March 2013 were reviewed normal data as medians and interquartile range (IQR). to determine if they had been extubated for more than 1 day during their ECMO course. The year 1996 was chosen as the RESULTS start date for the review because that was the year an electronic medical record was instituted in the PICU. Throughout Of 511 patients who underwent ECMO at our institution this period, a dedicated bedside nurse and an ECMO special- between 1996 and early 2013, 16 were either cannulated withist assigned solely to that patient cared for ECMO patients. out prior intubation (n = 2) or extubated during their ECMO course (n = 14). Selected details about these patients, including Equipment did, however, change during this period, with serial adoption of hollow fiber oxygenators (2009), biocompat- age, sex, primary and secondary disease processes, severity of respiratory disease, ECMO data (type, mode, cannulation sites ible circuits (2010), armored double lumen venovenous (VV) catheters (2010), and centrifugal pump technology (2011). A duration, and complications), and sedation pre- and postexdetailed retrospective chart review of each extubated patient tubation, are presented in Table 1. The median age of all extubated patients is 4.28 years, and the median duration of ECMO was undertaken, and demographic data, reason for ECMO support was 21.6 days. support, type of ECMO support, extubation latency (time The last 14 patients were extubated between the end of 2010 to extubation after cannulation), time extubated on ECMO, and outcomes and complications were all collected. Because and early 2013 (Fig. 1). One of 37 patients (2.7%) was extubated safety was a primary focus, the review sought complications in 2010, two of 34 (5.8%) in 2011, nine of 49 (18.4%) in 2012, and attributable to extubation, lowered sedation, or circuit mal- two of 16 (through June, 12.5%) in 2013. Although no specific function (hypoxemia, increased movement, agitation, seda- algorithm drove or guided the decision to extubate, formation of tion withdrawal, dyspnea, bleeding, catheter displacement, and a close-knit ECMO physician team in 2008 fostered an innovative circuit complications [rupture, clotting, air entrainment, and approach to caring for ECMO patients, including selective extubation. As longer ECMO courses became the norm for severe respimechanical failure]). Because these complications can occur with intubated patients on ECMO, only those incidents that ratory disease, the team faced greater difficulty keeping patients occurred as a result of extubation and its downstream effects adequately sedated and free of delirium and sedation toxicity. So (less sedation, more movement) were deemed complications irritant removal—extubation in this case—was raised as a possible solution to the difficult-to-sedate, long-term ECMO patient. of extubation. 864
www.pccmjournal.org
November 2014 • Volume 15 • Number 9
Extracorporeal Support
Extubation was considered only in patients fully supported on the ECMO circuit, whether for cardiovascular or respiratory disease. In general, “fully supported” meant that ECMO alone provided sufficient oxygenation (VV patients with Sao2 > 75% and Svo2 > 55–60%; venoarterial (VA) patients with Sao2 > 90% and Svo2 > 60–65%). Patients in whom ventilator support was required to achieve above oxygenation goals were not considered for extubation. As comfort with extubation grew, however, any ECMO patient who was fully supported on the extracorporeal circuit was considered a potential candidate during the period spanning the end of 2010 through the first half of 2013. Candidate respiratory patients for extubation also underwent serial bronchoscopies until large airways were clear of mucus plugs prior to ETT removal. The final decision to extubate a patient rested with that patient’s primary PICU attending. If the decision to extubate was reached, the parents were informed and all questions answered and concerns addressed. No parent opposed extubation, and most welcomed the opportunity for increased interaction with their children. Table 2 summarizes characteristics of each patient’s extubation, reintubation, procedures, and outcomes Of the 16 extubated patients, two were cannulated for cardiac reasons (cardiogenic shock and ECMO cardiopulmonary resuscitation) and 14 for respiratory failure. Causes of respiratory failure are summarized in Table 1. The two mediastinal mass patients were never intubated and both were supported on VA ECMO with an additional venous cannula because of greater comfort with this mode in the early 2000s. Although never actually extubated, they provided the first proof of concept that full ECMO support obviated the need for intubation and ventilation, gave us confidence to undertake purposeful extubation several years later, and were therefore included in this series. Eleven of them were male patients and five were female patients. Eleven were originally cannulated for VV ECMO, whereas five patients underwent primary VA cannulation. Two patients initially cannulated for VV were converted to VA ECMO because of myocardial dysfunction, whereas one patient initially cannulated centrally for low cardiac output after aortic arch repair was converted to peripheral VV ECMO with recovery of cardiac function but intercurrent ARDS from a viral respiratory infection. Excluding those patients cannulated for mediastinal masses or cardiac disease, the 12 patients cannulated for lung disease all had severe gas exchange abnormalities, although only 11 had arterial blood gases prior to cannulation (Table 1). The principal disorder leading to cannulation was hypoxemia, although some patients had combined oxygenation and ventilation disturbances. The maximum oxygenation index in these 11 patients ranged from 14 to 91, with a maximum median value of 37 (IQR, 32–52). In contrast, their Pao2/Fio2 ratios varied from 90 to 26, with a median of 54 (IQR, 61–42). Figure 2 shows a representative radiographic progression of a patient with severe ARDS who was extubated 5 days after cannulation (patient 10, see supplemental case reports, Supplemental Digital Content 1, http://links.lww.com/PCC/A114). Only one respiratory patient (patient 15) was cannulated without a prior blood gas. She had influenza pneumonia and severe air leak causing hemodynamic compromise. Pediatric Critical Care Medicine
Of the 16 patients who were extubated during ECMO, 11 (68%) survived their ECMO course (Table 2). The cause of death in nonsurvivors was related to their primary disease process and not to complications associated with extubation. In fact, three had been reintubated well prior to their terminal event, the fourth was reintubated immediately during efforts to resuscitate from a sudden arrhythmia, and the fifth (who had been receiving intrathecal chemotherapy for leukemia) suffered an intracranial hemorrhage and was reintubated prior to decannulation and parent-requested withdrawal of life-sustaining support. Two additional ECMO survivors died in the neonatal ICU of longterm complications related to their primary disease process. ECMO duration for patients who were extubated during extracorporeal support ranged from 4 to 71 days (Table 1). One patient (patient 3) who was cannulated for interstitial lung disease at our center and extubated for 22 days, and who was later transferred to a lung transplantation center on day 54 of ECMO, remained cannulated for a total of 84 days (she remained on ECMO for 30 additional days at the other center). The median ECMO course for patients who underwent extubation was 19.7 days (IQR, 8.5–20.25 d). Extubation latency, the interval between cannulation for ECMO and first extubation, ranged from 0 to 19 days with a median of 6 days (IQR, 1–11) (Table 2). Two patients who were never intubated and one with severe air leak due to positive pressure ventilation who was extubated immediately after cannulation accounted for the shortest latency. Lack of familiarity or discomfort with the concept of extubation during ECMO among some intensivists in our group explained longer delays between cannulation and extubation. Often patients were not immediately extubated because they underwent serial bronchoscopies to ensure adequate large airway clearance and removal of mucus plugging before extubation. To characterize the proportion of an ECMO course during which a patient remained extubated, the number of days extubated was divided by the total days on ECMO (Table 2). Scores approaching 1 indicated that the patient remained extubated for most of the ECMO run, whereas lower scores indicated that extubation time was a smaller fraction of the total time on ECMO. The proportion of extubation time to total ECMO time ranged from 1 to 0.14, with a median of 0.43 (IQR, 0.81–0.29) in our 16 patients. Overall, seven patients were extubated for more than 50% of their ECMO course, whereas nine patients were extubated for less than half of their time on ECMO. When extubated, patients required less sedation and were more active and interactive. Overall, fentanyl use dropped by 67% (average dose before extubation 145 μg/kg/d vs 48.0 μg/ kg/d during extubation) and midazolam requirements decreased by 55% (8.34 mg/kg/d preextubation vs 3.73 mg/kg/d postextubation) after extubation. This decline was not offset by increased doses of second-line agents (dexmedetomidine, ketamine, propofol, and pentobarbital). The state behavioral score is used to monitor the sedation level of ECMO patients in our institution. During the period starting with the extubation of patient 3 in this series and ending in March 2013, 110 patients were supported on ECMO. Extubated ECMO patients (38.5%) spent a larger www.pccmjournal.org
865
Anton-Martin et al
Characterization of Extubation, Reintubation, Procedures, and Outcomes on Extubated Extracorporeal Membrane Oxygenation Patients
Table 2.
Patient
1
2
3
4
5
6
7
Extubation latency (d)
0
0
8
8
14
0
10
Time extubated (d)
5
4
22
11
11
5
6
Extubated days/ECMO days
1
1
0.41
0.55
0.44
1
0.33
Support while extubated
NC
Continuous positive airway pressure
NC → RA
BiPAP, RA
RA → NC
HHFNC
RA → NC
Radiography during EP
↓Mass size
↓Mass size
↓Air leak, ↑aeration
↓Pulmonary edema
↑Aeration
↓Air leak, ↑aeration
↑Aeration
Symptoms during EP
↓Distress
↓Distress
↓Distress
↓Distress
↓Distress
↓Distress
↓Distress
Reintubation on ECMO
No
No
Yes
No
Yes (briefly)
Yes (briefly)
Yes
Reason for reintubation
Not applicable
Not applicable
Recruitment, procedure
Not applicable
Procedure
Procedure
Procedure
Type of procedure during reintubation
Not applicable
Not applicable
Bronchoscopy
Not applicable
Bronchoscopy, CT chest
Bronchoscopy
Circuit change
Bronchoscopy (n)
No
No
Yes (5)
No
Yes (8)
Yes (2)
Yes (3)
Chest CT (n)
Yes (1)
No
Yes (4)
No
Yes (3)
No
No
Airway status at death/ decannulation
Extubated
Extubated
Unknown (transferred)
Unknown (transferred)
Laryngeal mask airway only for decannulation
Reintubated only for decannulation
Remained reintubated
Time to extubation after decannulation
Not applicable
Not applicable
Unknown
Unknown
Minutes
4 hr
Not applicable
Outcome
Survival
Survival
Survival
Survival
Survival
Survival
Death
Time from decannulation to discharge floor/ home (d)
2/4
1/4
Transfer on ECMO for lung transplant eval
Transfer on ECMO for VAD
6/14
2/5
Not applicable
BiPAP = bilevel positive airway pressure, CPR = cardiopulmonary resuscitation, ECMO = extracorporeal membrane oxygenation, EP = extubation period, HFNC = high-flow nasal cannula, NC = nasal cannula, RA = room air, VAD = ventricular assist device, WDC = withdrawal of care. a Survived ECMO but died before discharge home. Extubation latency is the interval (d) between cannulation and extubation.
percentage of their ECMO run appropriately sedated with an SBS of 0 as compared with intubated ECMO patients (28%). The percentage of time oversedated was also decreased in the extubated patients (24.5%) compared with the intubated patients (33.5%). However, the duration of time agitated was the same for both extubated (37%) and intubated patients (38.5%) (Fig. 3). Four nonneonatal patients with ARDS and bilateral pulmonary opacification manifested dyspneic breathing behavior, with tachypnea/hyperpnea, paradoxical abdominal breathing, and head bobbing, despite normal blood gases. The dyspneic behavior in older children was not suppressible pharmacologically, unless sufficient sedation to cause apnea or neuromuscular blockade was used. Eleven of the extubated patients (68.75%) were reintubated during their ECMO course. Criteria for reintubation varied and 866
www.pccmjournal.org
were determined by the primary intensivist in conversations with consultants (surgeons, pulmonologists, and ECMO physicians). Three patients were reintubated briefly for procedures such as bronchoscopy or surgery, whereas others who survived their hospital course remained intubated through decannulation for a median of 4.5 days. Patients who ultimately died remained intubated until death. Of note, five patients had sufficient respiratory function by the time of decannulation that they were not intubated (n = 2), reintubated briefly (n = 2, < 24 hr), or had an laryngeal mask airway placed for decannulation and quickly transitioned to heated high-flow nasal cannula (n = 1). Two additional patients were extubated in the first week following decannulation (3 and 5 days after decannulation), whereas the remaining patients were either extubated weeks after decannulation (n = 2), transferred to an outside November 2014 • Volume 15 • Number 9
Extracorporeal Support
8
9
10
11
12
13
14
15
16
1
19
4
1
15
3
2
13
1
3
10
14
4
4
6
2
5
9
0.32
0.14
0.78
0.27
0.29
0.67
0.29
0.24
0.9
RA
HHFNC
RA → HHFNC
RA
RA
RA → HHFNC
HHFNC
NC
HHFNC, BiPAP
↑Aeration
↑Aeration
↑Aeration
↑Aeration
↓Air leak, ↑aeration
↑Aeration
↑Aeration
↑Aeration
↑Aeration
↓Distress
↓Distress
↓Distress
↓Distress
↓Distress
↓Distress
↓Distress
↓Distress, ↑cough
↓Distress, ↑cough
Yes
Yes
Yes (briefly)
Yes
Yes
No
Yes
Yes
Yes
CPR, procedure, recruitment
Recruitment
Procedure
Recruitment
CPR
Not applicable
Procedure
Apneic episode
Procedure
Bronchoscopy
Not applicable
Chest tube
Not applicable
Not applicable
Not applicable
Bronchoscopy
Not applicable
Remove ECMO support
Yes (2)
Yes (9)
Yes (5)
No
Yes (5) + bronchogram
Yes (1)
Yes (2)
Yes (3)
Yes (2)
No
Yes (3)
Yes (2)
No
No
No
No
No
No
Remained reintubated
Remained reintubated
Reintubated only for decannulation
Remained reintubated
Reintubated for arrest
Reintubated only for decannulation
Remained reintubated
Remained reintubated
Reintubated to remove ECMO support
Not applicable
Not applicable
8 hr
Not applicable
Not applicable
5d
Not applicable
3d
Not applicable
Survivala
Death
Survival
Survivala
Death
Survival
Death
Survival
Death
WDC on hospital day # 100 for gastrointestinal problems
Not applicable
11/27
WDC on hospital day# 53 due to recurrent air leaks
Not applicable
10/42
Not applicable
6/20
Not applicable
facilities where details about postdecannulation extubation are unknown (n = 2), or died during ECMO support (n = 5). All five who died during ECMO were intubated at the time of stopping support. Three had been reintubated several days prior, one was reintubated for cardiac arrest and one was reintubated prior to decannulation and life support withdrawal.
DISCUSSION We present a series of 16 children who successfully remained extubated during ECMO support. They represented a broad range of cardiopulmonary diseases and included patients supported with both VV and VA modes. This experience, while still small, demonstrates the feasibility of extubation while receiving ECMO support under a variety of conditions. Pediatric Critical Care Medicine
The practice of extubating ECMO patients evolved quickly when purposeful extubation began in 2010. The only consistent requirement for extubation candidacy was that the patient be fully supported by the ECMO circuit and not require additional respiratory support (i.e., no additional therapeutic benefit to intubation/ventilation). Some patients would remain intubated for an additional day or two until bronchoscopy succeeded in clearing the large airways. The ECMO physician would discuss possible extubation with the primary intensivist, who made the ultimate decision. If the primary intensivist agreed to extubation, the family was informed and all questions answered. Although many families had questions about the procedure, none objected to or opposed extubation. No management protocols were used in this early experience. www.pccmjournal.org
867
Anton-Martin et al
expose the patients in this series to additional risks. Our experience does suggest, however, that safety procedures may require modification in the extubated patient on ECMO. For example, we began to keep an appropriately sized ETT, laryngoscope, and blade taped at the head of each extubated patient’s bed in E F G H case of pump failure and the need to activate the Pediatric Advanced Life Support algorithm after patient 8’s circuit failed suddenly. We also made a number of interesting observations in our extubated patients that will require further systemI J atic study to determine their validity. Our inability to draw firm conclusions at this stage of experience is due to two considerations. First, as we grew more comfortable with the practice of extubating, we began to perform it earlier in subsequent patient’s courses, Figure 2. Radiographic progression of acute respiratory distress syndrome in a representative patient. A, Chest uncovering potential advanradiograph prior to cannulation for venovenous extracorporeal membrane oxygenation (ECMO). B, Immediately tages not previously enviafter cannulation. C, Chest CT on day 3 of ECMO. D, Day 5 of ECMO, immediately after extubation. E, Postextubation day 3 (ECMO day 8). F, Postextubation day 5 (ECMO day 10). G, Postextubation day 7 sioned (e.g., possibly lower (ECMO day 12). H, Postextubation day 10 (ECMO day 15). I, Postdecannulation day 6. J, Day of ICU discharge sedation requirements and faster resolution of pulmonary inflammation). Our perspective on certain observations Furthermore, ECMO survival was close to 70%, and none of the ECMO deaths was attributable to extubation while on changed with greater experience (e.g., dyspneic breathing ECMO. Death was related to failure of primary disease to reverse, behavior). complication of the primary disease, ECMO complication, or a Extubated patients were generally more interactive with their combination of these. Extubation on ECMO does not seem to family, caregivers, and environment. They spent more time awake, with their eyes open, and with increasing age, responded to family members verbally, watched TV, and made their needs known to caregivers. Furthermore, these patients generally required less sedation, although this observation requires some qualification. Patients with longer extubation latencies (greater time between cannulation and first extubation) tended to have higher sedative requirements than those who were extubated shortly after cannulation, although this is difficult to quantify in such a small series (18). The association between extubation and lower Figure 3. Distribution of State behavioral Scale (SBS) scores of extubated and intubated extracorporeal sedation also suggested that the ETT membrane oxygenation (ECMO) patients, showing percentage of ECMO course patients spent and mechanical ventilation may have appropriately sedated, oversedated, or undersedated.
A
868
B
www.pccmjournal.org
C
D
November 2014 • Volume 15 • Number 9
Extracorporeal Support
been more powerful irritants than ECMO, driving the bulk of the need for sedation with its attendant side effects. The SBS was our institution’s tool to assess sedation level in PICU patients, including those on ECMO. Although useful, it cannot distinguish agitation from hypermotor delirium, a consequence of long-term sedation in the ICU (19, 20). We realized that we might be underestimating the prevalence of delirium in our long-term ECMO patients when we extubated an older agitated patient and found him to be hallucinating. Whether delirium was responsible for SBS scores of +1 or +2 in some agitated patients is impossible to ascertain. It nonetheless bears noting that many long-term ECMO patients were agitated, difficult to sedate, and possibly delirious. With increasing emphasis on long-term cognitive and functional impairment of patients after delirium in the ICU (21, 22), improving our recognition of this complication and minimizing noxious stimulation that requires sedation may be one strategy to prevent it and its lingering effects. Extubation during ECMO permitted increased physical activity in cannulated patients. Initially, we were hesitant to move our first patients because of concerns for cannula dislodgement, but this proved not to be an issue, and each subsequent patient received increasingly vigorous physical and occupational therapy, up to and including limited ambulation. The amount of physical activity relates to both patient age and cannulation site, but extubation provides the opportunity to prevent the physical deconditioning associated with long ICU stays (usually also associated with long periods of intubation and ventilation). Spontaneous, negative pressure ventilation may have been effective in promoting spontaneous reaeration, cough, and secretion mobilization in patients with severe pulmonary consolidation. In many patients, resolution of consolidation and subsequent aeration occurred in the absence of positive pressure ventilation. What the present series cannot answer, however, is whether extubation speeds or delays resolution of lung injury and alveolar reinflation. Although “rest settings” are likely less harmful than aggressive ventilator settings, it is unknown whether positive pressure of any sort is harmful or helpful in an injured lung. Our study has several limitations. The series is small, with 16 patients. The data derive from a single center and the analysis was performed retrospectively. Most of the observations will require follow-up in larger series at several centers to confirm.
CONCLUSIONS We present the largest series of extubated pediatric ECMO patients to date and demonstrate both the feasibility of extubation in a broad range of patients on both VV and VA ECMO. Furthermore, we share preliminary evidence that extubation was associated with reduced patient sedation requirements and increased opportunity to avoid deconditioning.
ACKNOWLEDGMENTS We thank the extracorporeal membrane oxygenation (ECMO) specialists and bedside care staff for the wonderful care provided to all our ECMO patients, including the extubated ones and the children and their families Pediatric Critical Care Medicine
REFERENCES
1. Registry Report. Ann Arbor, MI, Extracorporeal Life Support Organization, 2010 2. MacLaren G, Combes A, Bartlett RH: Contemporary extracorporeal membrane oxygenation for adult respiratory failure: Life support in the new era. Intensive Care Med 2012; 38:210–220 3. Hirshberg E, Miller RR 3rd, Morris AH: Extracorporeal membrane oxygenation in adults with acute respiratory distress syndrome. Curr Opin Crit Care 2013; 19:38–43 4. Rehder KJ, Turner DA, Bonadonna D, et al: State of the art: Strategies for extracorporeal membrane oxygenation in respiratory failure. Expert Rev Respir Med 2012; 6:513–521 5. Prins S, van Dijk M, Tibboel D: Sedation and analgesia in the PICU: Many questions remain. Intensive Care Med 2006; 32:1103–1105 6. Fuehner T, Kuehn C, Hadem J, et al: Extracorporeal membrane oxygenation in awake patients as bridge to lung transplantation. Am J Respir Crit Care Med 2012; 185:763–768 7. Garcia JP, Iacono A, Kon ZN, et al: Ambulatory extracorporeal membrane oxygenation: A new approach for bridge-to-lung transplantation. J Thorac Cardiovasc Surg 2010; 139:e137–e139 8. Olsson KM, Simon A, Strueber M, et al: Extracorporeal membrane oxygenation in nonintubated patients as bridge to lung transplantation. Am J Transplant 2010; 10:2173–2178 9. Nosotti M, Rosso L, Tosi D, et al: Extracorporeal membrane oxygenation with spontaneous breathing as a bridge to lung transplantation. Interact Cardiovasc Thorac Surg 2013; 16:55–59 10. Garcia JP, Kon ZN, Evans C, et al: Ambulatory veno-venous extracorporeal membrane oxygenation: Innovation and pitfalls. J Thorac Cardiovasc Surg 2011; 142:755–761 11. Schmid C, Philipp A, Hilker M, et al: Venovenous extracorporeal membrane oxygenation for acute lung failure in adults. J Heart Lung Transplant 2012; 31:9–15 12. Turner DA, Cheifetz IM, Rehder KJ, et al: Active rehabilitation and physical therapy during extracorporeal membrane oxygenation while awaiting lung transplantation: A practical approach. Crit Care Med 2011; 39:2593–2598 13. Collar RM, Taylor JC, Hogikyan ND, et al: Awake extracorporeal membrane oxygenation for management of critical distal tracheal obstruction. Otolaryngol Head Neck Surg 2010; 142:618–620 14. Raake J, Johnson B, Seger B, et al: Extracorporeal membrane oxygenation, extubation, and lung-recruitment maneuvers as rescue therapy in a patient with tracheal dehiscence following slide tracheoplasty. Respir Care 2011; 56:1198–1202 15. Wickiser JE, Thompson M, Leavey PJ, et al: Extracorporeal membrane oxygenation (ECMO) initiation without intubation in two children with mediastinal malignancy. Pediatr Blood Cancer 2007; 49:751–754 16. Schmidt F, Sasse M, Boehne M, et al: Concept of “awake venovenous extracorporeal membrane oxygenation” in pediatric patients awaiting lung transplantation. Pediatr Transplant 2012; 17:224–230 17. Curley MA, Harris SK, Fraser KA, et al: State Behavioral Scale: A sedation assessment instrument for infants and young children supported on mechanical ventilation. Pediatr Crit Care Med 2006; 7:107–114 18. Anton-Martin P, Darnell-Bowens C, Raman L, et al: Sedation and analgesic requirements in extubated pediatric patients on ECMO (abstract presented). 24th Annual ELSO Conference. Philadelphia, PA, September 19–22, 2013 19. Reade MC, Finfer S: Sedation and delirium in the intensive care unit. N Engl J Med 2014; 370:444–454 20. Smith HA, Berutti T, Brink E, et al: Pediatric critical care perceptions on analgesia, sedation, and delirium. Semin Respir Crit Care Med 2013; 34:244–261 21. Pandharipande PP, Girard TD, Jackson JC, et al; BRAIN-ICU Study Investigators: Long-term cognitive impairment after critical illness. N Engl J Med 2013; 369:1306–1316 22. Brummel NE, Jackson JC, Pandharipande PP, et al: Delirium in the ICU and subsequent long-term disability among survivors of mechanical ventilation. Crit Care Med 2013; 42:369–377 www.pccmjournal.org
869
Objectives: Describe aspects of one center’s experience extubating infants and children during extracorporeal membrane oxygenation. Design: Retrospective review of medical records. Setting: Seventy-one-bed critical care service (PICU and cardiovascular ICU) in a large urban tertiary children’s hospital. Patients: Pediatric and neonatal patients supported on extracorporeal membrane oxygenation between 1996 and 2013 who were either not intubated or extubated greater than 24 hours during their extracorporeal membrane oxygenation course. Interventions: None. Measurements and Main Results: Sixteen of 511 patients on extracorporeal membrane oxygenation were extubated for at least 24 hours during their extracorporeal membrane oxygenation courses. Fourteen had respiratory failure and two had cardiac disease. Five patients died while on extracorporeal membrane oxygenation, but the cause of death was not related to complications associated *See also p. 907. 1 Department of Pediatrics, Division of Pediatric Critical Care, University of Texas Southwestern Medical Center, Dallas, TX. 2 Department of Pediatrics, Division of Pediatric Critical Care, University of Missouri, Kansas City/Children’s Mercy Hospital, Kansas City, MO. 3 Division of Pediatric Surgery, William Beaumont Oakland University Medical School, Royal Oak, MI. 4 Department of Critical Care Services, Children’s Medical Center, Dallas, TX. 5 Department of Pediatrics, Section of Pediatric Critical Care, Baylor College of Medicine/Texas Children’s Hospital, Houston, TX. Presented, in part, at the 2012 ELSO Conference in Seattle, WA, September 14–16, 2012, where it was awarded the Robert Bartlett award for best original presentation. Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s website (http://journals.lww.com/pccmjournal). Dr. Anton-Martin received support from the Fundación Alfonso Martín Escudero. The remaining authors have disclosed that they do not have any potential conflicts of interest. For information regarding this article, E-mail: [email protected] Copyright © 2014 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies DOI: 10.1097/PCC.0000000000000235
Pediatric Critical Care Medicine
with extubation. Extubated patients were supported a median of 19.7 days on extracorporeal membrane oxygenation, with a median extubation latency (time between cannulation and first extubation) of 6.2 days and a median extubation duration of 5.5 days. Mean time extubated was 43% of the total time on extracorporeal membrane oxygenation. Two patients were reintubated briefly or had a laryngeal mask airway placed for decannulation (n = 1). The remaining patients were extubated within 5 days of decannulation, weeks afterward (n = 2), transferred to outside facilities (n = 2), or died during extracorporeal membrane oxygenation support (n = 5). We also observed no complications directly attributable to extubation and spontaneous reaeration of consolidated lungs in acute respiratory distress syndrome in extubated patients on extracorporeal membrane oxygenation. Conclusion: Extubation and discontinuation of mechanical ventilation appear feasible in patients requiring long-term extracorporeal membrane oxygenation. Emergency procedure planning may need to be modified in extubated patients on extracorporeal membrane oxygenation. (Pediatr Crit Care Med 2014; 15:861–869) Key Words: extracorporeal membrane life support; extracorporeal membrane oxygenation; extubation; mechanical ventilation; ventilator-induced lung injury
E
xtracorporeal membrane oxygenation (ECMO) developed as a rescue therapy for intractable respiratory, cardiac, or combined cardiopulmonary failure (1). In general, the goals of ECMO therapy are to support cellular respiration by improving tissue oxygenation or Co2 removal until resolution of the primary disease process or as a “bridge” to organ transplantation or another mechanical support modality suitable for long-term use. Because of its invasive nature and potential for catastrophic complications, however, ECMO has been considered a last-resort therapy to be used only when conventional treatments have failed and only long enough until the patient can be returned to more “standard” support modalities (mechanical ventilation, inotropic support, or both). Neither strategy—cannulation after failure of conventional support or minimal duration of ECMO—has been compared www.pccmjournal.org
861
Anton-Martin et al
Table 1.
Patient, Disease, and Extracorporeal Membrane Oxygenation Characteristics
Patient
1
2
ECMO type
Pediatric respiratory
Pediatric respiratory
Pediatric respiratory
Pediatric cardiac
Pediatric respiratory
Pediatric respiratory
Neonatal respiratory
Neonatal respiratory
Age
11 yr
5 yr
4 yr
17 yr
12 yr
3 yr
(35 wk + 5 d)/2 d
(38 wk)/2 d
Sex
Male
Male
Female
Female
Male
Male
Male
Male
Primary disease
Mediastinal mass
Mediastinal mass
Interstitial lung disease
Cardiogenic shock
Aspiration pneumonitis
RSV pneumonia
Meconium aspiration
Right CDH
Secondary disease
Respiratory failure
Respiratory failure
ARDS, air leak
Cardiomyopathy
ARDS
ARDS, air leak
ARDS
Respiratory failure
Maximum oxygenation index-minimum Pao2/Fio2
–/–
–/–
36/61
2/80
37/72
27/55
34/49
14/90
ECMO Mode
VA + V
VA + V
VV + V
VA + V
DLVV + V
DLVV
DLVV
VA
Venous site(s)
RFV
RFV
RFV-LFV
LFV
RIJV
RIJV
RIJV
RIJV
Arterial site
LFA
RFA
—
LFA
Not applicable
Not applicable
Not applicable
RCCA
Additional sites?
Yes (RFV)
Yes (LFV)
Yes (RIJV)
Yes (RFV)
Yes (RFV)
No
No
No
ECMO duration
5
4
54 (84 d total)
20
25
5
18
19
Complications on ECMO
Bleeding
No
Bleeding, infection
Bleeding, clotting in circuit
Bleeding, clotting circuit, delirium
No
Infection, bleeding, cardiac arrest
Flow reversed after cannulation, bleeding.
Sedation preextubation
Not applicable
Not applicable
Fentanyl, midazolam, dexmedetomidine
Fentanyl, midazolam
Fentanyl, midazolam
Fentanyl, midazolam
Fentanyl, midazolam
Fentanyl, midazolam
Fentanyl
Fentanyl, midazolam
Sedation postextubation
3
4
5
6
7
8
Pentobarbital, propofol
Fentanyl, midazolam
Fentanyl, midazolam
Fentanyl, midazolam
Fentanyl, midazolam
Fentanyl, midazolam
Fentanyl
Ketamine
Ketamine
Propofol, ketamine
WD
Pentobarbital, propofol, ketamine
Dexmedetomidine, propofol
WD
WD
WD
ARDS = acute respiratory distress syndrome, CDH = congenital diaphragmatic hernia, ECMO = extracorporeal membrane oxygenation, LFA = left femoral artery, DLVV = double lumen venovenous, LFV = left femoral vein, RCCA = right common carotid artery, RFA = right femoral artery, RFV = right femoral vein, RIJV = right internal jugular vein, RSV = respiratory syncytial virus, VA = venoarterial, VV = venovenous, + V = additional venous cannula. Sedation used: first-line agents (fentanyl and midazolam); second-line agents (dexmedetomidine, ketamine, propofol, and pentobarbital); and withdrawal agents (WD = diazepam, lorazepam, methadone, and morphine).
to alternative approaches that might minimize injury from conventional support modalities and maximize the opportunity for recovery from tissue injury. The adoption of newer circuit technologies (biocompatible surfaces, new pump, and oxygenator technology) and advances in ECMO practice (dedicated teams, standardized approaches to anticoagulation, and early initiation of renal replacement) have led to the perception that ECMO as currently practiced is safer than ECMO practiced at the end of the 1990s (2). This perception also helps explain the increasing prevalence of longer ECMO courses, particularly for severe respiratory disease (3, 4). If in fact ECMO is safer today than 15 years ago, then cannulation 862
www.pccmjournal.org
and support strategies inherited from the early years of ECMO may be outdated as well. One practice that may deserve reexamination is the role of intubation and mechanical ventilation while on extracorporeal support. In most ECMO centers, once a patient has been cannulated, ventilator support is decreased but sedation is maintained to prevent extubation or decannulation. Although there may be a role for maintaining an endotracheal tube (ETT) in place to support pulmonary toilet adjuncts, such as suctioning or bronchoscopy, there is no clinical indication for continued mechanical ventilation when ECMO alone can meet the patient’s tissue metabolic demands. As long as patients remain intubated, the November 2014 • Volume 15 • Number 9
Extracorporeal Support
9
10
11
12
13
14
15
16
Pediatric respiratory
Pediatric respiratory
Neonatal respiratory
Pediatric ECPR
Neonatal respiratory
Neonatal respiratory
Pediatric respiratory
Pediatric respiratory
3 yr
6 yr
(41 wk)/5 d
5 mo
(39 wk)/1 d
19 d
2 yr
5 yr
Male
Male
Male
Female
Male
Female
Female
Male
Pertussis pneumonia
Mycoplasma pneumonia
Meconium aspiration, supraglottic edema
Aortic coarctation, ECMO cardiopulmonary resuscitation
Meconium aspiration
Left CDH
Influenza RSV pneumonia, air pneumonitis leak
ARDS
ARDS
ARDS
Airway disruption/ edema
ARDS
Respiratory failure
ARDS
ARDS
48/42
54/52
32/54
–/–
91/26
50/40
–/–
52/58
DLVV → VA
DLVV
DLVV → VA
VA → DLVV
DLVV
DLVV
DLVV
DLVV
RIJV
RIJV
RIJV/double lumen
Right atrium
RIJV
RIJV
RIJV
RIJV
RCCA
Not applicable
RCCA
Aorta
Not applicable
Not applicable
Not applicable
Not applicable
No
No
No
Switched to DLVV (RIJV)
No
No
Yes (RFV)
Yes (Cephalad)
71
18
15
14
9
7
21
10
Bleeding, clotting, pericardial tamponade
Bleeding, sepsis
Bleeding, recurrent pneumothorax, septic shock, renal failure
Cardiac arrest
Bleeding
No
Bleeding, cannula displaced, delirium
Bleeding
Fentanyl
Fentanyl, midazolam
Fentanyl, midazolam
Fentanyl
Fentanyl, midazolam
Fentanyl, midazolam
Fentanyl, midazolam
Fentanyl, midazolam
Dexmedetomidine
Dexmedetomidine, propofol
Fentanyl, midazolam
Fentanyl, midazolam
Dexmedetomidine, ketamine
Propofol, ketamine
WD
WD
Dexmedetomidine, pentobarbital Fentanyl, midazolam
Fentanyl, midazolam
Fentanyl, midazolam
Fentanyl, midazolam
Fentanyl, midazolam
Dexmedetomidine, pentobarbital, propofol WD
ETT remains an irritant, necessitating sedation with its attendant complications, including sedation habituation and dose escalation, delirium, neuromuscular deconditioning, and prolonged absence of social interaction. Although recent strategies to decrease cumulative sedative doses or to withhold continuous sedation may decrease the need for polypharmacy (5), removal of the primary irritant may be a more effective intervention. Centers providing ECMO have begun to report their experience with extubated, awake, and spontaneously breathing adult patients. Most publications examine the experience with adult lung transplantation (2, 6–9) and acute respiratory distress syndrome (ARDS) (10, 11). Two lung transplantation Pediatric Critical Care Medicine
Fentanyl, midazolam
WD
WD
WD
centers, in Germany and Italy, have recently shared larger case series. The aggregate experience suggests that spontaneously breathing ECMO patients may be better transplant candidates and enjoy better outcomes than those requiring mechanical ventilation prior to transplantation (6, 9). Garcia et al (7, 10) have reported a device to allow extubated patients to ambulate and perform physical activity while on extracorporeal support and thereby avoid the consequences of being bedridden (12). Very few studies describe extubation experience in children and are focused primarily on airway issues (13–15). There is a single published report of three extubated pediatric patients on ECMO awaiting lung transplantation www.pccmjournal.org
863
Anton-Martin et al
State behavioral Scale (SBS) is a behavioral assessment tool with a numeric rating scale that describes the sedation-agitation continuum of pediatric patients supported on mechanical ventilation. It was derived from ratings of seven content dimensions that include descriptors that have been previously described in the literature to be associated with agitation (17). In this way, undersedation is a proxy for activity and/or alertness. To gauge the level of oversedation, appropriate sedation, Figure 1. Distribution of extubated patients on extracorporeal membrane oxygenation (ECMO) over time. The and undersedation, the numnumber of extubated patients on ECMO is plotted against the year in which they were extubated. The majority of ber of days each patient scored extubations occurred in 2011, 2012, and 2013. either –1 to –3 (responsive to gentle touch to unresponfrom the Hannover center, reflecting the incorporation of sive), 0 (awake and able to calm), or +1 (restless and difficult strategies from their evolving adult practice into their care to calm) to +2 (agitated) were tallied. To compare sedation in of children (16). extubated and intubated ECMO patients, the scores of the extuThis report summarizes our experience with patients who bated patients were compared with the scores of the intubated were either cannulated without prior intubation or extubated patients. for more than 24 hours while on ECMO. The Institutional Review Board at the University of Texas Southwestern Medical Center approved this study and waived the need for informed consent. MATERIALS AND METHODS Summary data were collected and reported as percentages, Records of patients supported on ECMO at Children’s Medinormally distributed data expressed as means ± sd, and noncal Center Dallas between 1996 and March 2013 were reviewed normal data as medians and interquartile range (IQR). to determine if they had been extubated for more than 1 day during their ECMO course. The year 1996 was chosen as the RESULTS start date for the review because that was the year an electronic medical record was instituted in the PICU. Throughout Of 511 patients who underwent ECMO at our institution this period, a dedicated bedside nurse and an ECMO special- between 1996 and early 2013, 16 were either cannulated withist assigned solely to that patient cared for ECMO patients. out prior intubation (n = 2) or extubated during their ECMO course (n = 14). Selected details about these patients, including Equipment did, however, change during this period, with serial adoption of hollow fiber oxygenators (2009), biocompat- age, sex, primary and secondary disease processes, severity of respiratory disease, ECMO data (type, mode, cannulation sites ible circuits (2010), armored double lumen venovenous (VV) catheters (2010), and centrifugal pump technology (2011). A duration, and complications), and sedation pre- and postexdetailed retrospective chart review of each extubated patient tubation, are presented in Table 1. The median age of all extubated patients is 4.28 years, and the median duration of ECMO was undertaken, and demographic data, reason for ECMO support was 21.6 days. support, type of ECMO support, extubation latency (time The last 14 patients were extubated between the end of 2010 to extubation after cannulation), time extubated on ECMO, and outcomes and complications were all collected. Because and early 2013 (Fig. 1). One of 37 patients (2.7%) was extubated safety was a primary focus, the review sought complications in 2010, two of 34 (5.8%) in 2011, nine of 49 (18.4%) in 2012, and attributable to extubation, lowered sedation, or circuit mal- two of 16 (through June, 12.5%) in 2013. Although no specific function (hypoxemia, increased movement, agitation, seda- algorithm drove or guided the decision to extubate, formation of tion withdrawal, dyspnea, bleeding, catheter displacement, and a close-knit ECMO physician team in 2008 fostered an innovative circuit complications [rupture, clotting, air entrainment, and approach to caring for ECMO patients, including selective extubation. As longer ECMO courses became the norm for severe respimechanical failure]). Because these complications can occur with intubated patients on ECMO, only those incidents that ratory disease, the team faced greater difficulty keeping patients occurred as a result of extubation and its downstream effects adequately sedated and free of delirium and sedation toxicity. So (less sedation, more movement) were deemed complications irritant removal—extubation in this case—was raised as a possible solution to the difficult-to-sedate, long-term ECMO patient. of extubation. 864
www.pccmjournal.org
November 2014 • Volume 15 • Number 9
Extracorporeal Support
Extubation was considered only in patients fully supported on the ECMO circuit, whether for cardiovascular or respiratory disease. In general, “fully supported” meant that ECMO alone provided sufficient oxygenation (VV patients with Sao2 > 75% and Svo2 > 55–60%; venoarterial (VA) patients with Sao2 > 90% and Svo2 > 60–65%). Patients in whom ventilator support was required to achieve above oxygenation goals were not considered for extubation. As comfort with extubation grew, however, any ECMO patient who was fully supported on the extracorporeal circuit was considered a potential candidate during the period spanning the end of 2010 through the first half of 2013. Candidate respiratory patients for extubation also underwent serial bronchoscopies until large airways were clear of mucus plugs prior to ETT removal. The final decision to extubate a patient rested with that patient’s primary PICU attending. If the decision to extubate was reached, the parents were informed and all questions answered and concerns addressed. No parent opposed extubation, and most welcomed the opportunity for increased interaction with their children. Table 2 summarizes characteristics of each patient’s extubation, reintubation, procedures, and outcomes Of the 16 extubated patients, two were cannulated for cardiac reasons (cardiogenic shock and ECMO cardiopulmonary resuscitation) and 14 for respiratory failure. Causes of respiratory failure are summarized in Table 1. The two mediastinal mass patients were never intubated and both were supported on VA ECMO with an additional venous cannula because of greater comfort with this mode in the early 2000s. Although never actually extubated, they provided the first proof of concept that full ECMO support obviated the need for intubation and ventilation, gave us confidence to undertake purposeful extubation several years later, and were therefore included in this series. Eleven of them were male patients and five were female patients. Eleven were originally cannulated for VV ECMO, whereas five patients underwent primary VA cannulation. Two patients initially cannulated for VV were converted to VA ECMO because of myocardial dysfunction, whereas one patient initially cannulated centrally for low cardiac output after aortic arch repair was converted to peripheral VV ECMO with recovery of cardiac function but intercurrent ARDS from a viral respiratory infection. Excluding those patients cannulated for mediastinal masses or cardiac disease, the 12 patients cannulated for lung disease all had severe gas exchange abnormalities, although only 11 had arterial blood gases prior to cannulation (Table 1). The principal disorder leading to cannulation was hypoxemia, although some patients had combined oxygenation and ventilation disturbances. The maximum oxygenation index in these 11 patients ranged from 14 to 91, with a maximum median value of 37 (IQR, 32–52). In contrast, their Pao2/Fio2 ratios varied from 90 to 26, with a median of 54 (IQR, 61–42). Figure 2 shows a representative radiographic progression of a patient with severe ARDS who was extubated 5 days after cannulation (patient 10, see supplemental case reports, Supplemental Digital Content 1, http://links.lww.com/PCC/A114). Only one respiratory patient (patient 15) was cannulated without a prior blood gas. She had influenza pneumonia and severe air leak causing hemodynamic compromise. Pediatric Critical Care Medicine
Of the 16 patients who were extubated during ECMO, 11 (68%) survived their ECMO course (Table 2). The cause of death in nonsurvivors was related to their primary disease process and not to complications associated with extubation. In fact, three had been reintubated well prior to their terminal event, the fourth was reintubated immediately during efforts to resuscitate from a sudden arrhythmia, and the fifth (who had been receiving intrathecal chemotherapy for leukemia) suffered an intracranial hemorrhage and was reintubated prior to decannulation and parent-requested withdrawal of life-sustaining support. Two additional ECMO survivors died in the neonatal ICU of longterm complications related to their primary disease process. ECMO duration for patients who were extubated during extracorporeal support ranged from 4 to 71 days (Table 1). One patient (patient 3) who was cannulated for interstitial lung disease at our center and extubated for 22 days, and who was later transferred to a lung transplantation center on day 54 of ECMO, remained cannulated for a total of 84 days (she remained on ECMO for 30 additional days at the other center). The median ECMO course for patients who underwent extubation was 19.7 days (IQR, 8.5–20.25 d). Extubation latency, the interval between cannulation for ECMO and first extubation, ranged from 0 to 19 days with a median of 6 days (IQR, 1–11) (Table 2). Two patients who were never intubated and one with severe air leak due to positive pressure ventilation who was extubated immediately after cannulation accounted for the shortest latency. Lack of familiarity or discomfort with the concept of extubation during ECMO among some intensivists in our group explained longer delays between cannulation and extubation. Often patients were not immediately extubated because they underwent serial bronchoscopies to ensure adequate large airway clearance and removal of mucus plugging before extubation. To characterize the proportion of an ECMO course during which a patient remained extubated, the number of days extubated was divided by the total days on ECMO (Table 2). Scores approaching 1 indicated that the patient remained extubated for most of the ECMO run, whereas lower scores indicated that extubation time was a smaller fraction of the total time on ECMO. The proportion of extubation time to total ECMO time ranged from 1 to 0.14, with a median of 0.43 (IQR, 0.81–0.29) in our 16 patients. Overall, seven patients were extubated for more than 50% of their ECMO course, whereas nine patients were extubated for less than half of their time on ECMO. When extubated, patients required less sedation and were more active and interactive. Overall, fentanyl use dropped by 67% (average dose before extubation 145 μg/kg/d vs 48.0 μg/ kg/d during extubation) and midazolam requirements decreased by 55% (8.34 mg/kg/d preextubation vs 3.73 mg/kg/d postextubation) after extubation. This decline was not offset by increased doses of second-line agents (dexmedetomidine, ketamine, propofol, and pentobarbital). The state behavioral score is used to monitor the sedation level of ECMO patients in our institution. During the period starting with the extubation of patient 3 in this series and ending in March 2013, 110 patients were supported on ECMO. Extubated ECMO patients (38.5%) spent a larger www.pccmjournal.org
865
Anton-Martin et al
Characterization of Extubation, Reintubation, Procedures, and Outcomes on Extubated Extracorporeal Membrane Oxygenation Patients
Table 2.
Patient
1
2
3
4
5
6
7
Extubation latency (d)
0
0
8
8
14
0
10
Time extubated (d)
5
4
22
11
11
5
6
Extubated days/ECMO days
1
1
0.41
0.55
0.44
1
0.33
Support while extubated
NC
Continuous positive airway pressure
NC → RA
BiPAP, RA
RA → NC
HHFNC
RA → NC
Radiography during EP
↓Mass size
↓Mass size
↓Air leak, ↑aeration
↓Pulmonary edema
↑Aeration
↓Air leak, ↑aeration
↑Aeration
Symptoms during EP
↓Distress
↓Distress
↓Distress
↓Distress
↓Distress
↓Distress
↓Distress
Reintubation on ECMO
No
No
Yes
No
Yes (briefly)
Yes (briefly)
Yes
Reason for reintubation
Not applicable
Not applicable
Recruitment, procedure
Not applicable
Procedure
Procedure
Procedure
Type of procedure during reintubation
Not applicable
Not applicable
Bronchoscopy
Not applicable
Bronchoscopy, CT chest
Bronchoscopy
Circuit change
Bronchoscopy (n)
No
No
Yes (5)
No
Yes (8)
Yes (2)
Yes (3)
Chest CT (n)
Yes (1)
No
Yes (4)
No
Yes (3)
No
No
Airway status at death/ decannulation
Extubated
Extubated
Unknown (transferred)
Unknown (transferred)
Laryngeal mask airway only for decannulation
Reintubated only for decannulation
Remained reintubated
Time to extubation after decannulation
Not applicable
Not applicable
Unknown
Unknown
Minutes
4 hr
Not applicable
Outcome
Survival
Survival
Survival
Survival
Survival
Survival
Death
Time from decannulation to discharge floor/ home (d)
2/4
1/4
Transfer on ECMO for lung transplant eval
Transfer on ECMO for VAD
6/14
2/5
Not applicable
BiPAP = bilevel positive airway pressure, CPR = cardiopulmonary resuscitation, ECMO = extracorporeal membrane oxygenation, EP = extubation period, HFNC = high-flow nasal cannula, NC = nasal cannula, RA = room air, VAD = ventricular assist device, WDC = withdrawal of care. a Survived ECMO but died before discharge home. Extubation latency is the interval (d) between cannulation and extubation.
percentage of their ECMO run appropriately sedated with an SBS of 0 as compared with intubated ECMO patients (28%). The percentage of time oversedated was also decreased in the extubated patients (24.5%) compared with the intubated patients (33.5%). However, the duration of time agitated was the same for both extubated (37%) and intubated patients (38.5%) (Fig. 3). Four nonneonatal patients with ARDS and bilateral pulmonary opacification manifested dyspneic breathing behavior, with tachypnea/hyperpnea, paradoxical abdominal breathing, and head bobbing, despite normal blood gases. The dyspneic behavior in older children was not suppressible pharmacologically, unless sufficient sedation to cause apnea or neuromuscular blockade was used. Eleven of the extubated patients (68.75%) were reintubated during their ECMO course. Criteria for reintubation varied and 866
www.pccmjournal.org
were determined by the primary intensivist in conversations with consultants (surgeons, pulmonologists, and ECMO physicians). Three patients were reintubated briefly for procedures such as bronchoscopy or surgery, whereas others who survived their hospital course remained intubated through decannulation for a median of 4.5 days. Patients who ultimately died remained intubated until death. Of note, five patients had sufficient respiratory function by the time of decannulation that they were not intubated (n = 2), reintubated briefly (n = 2, < 24 hr), or had an laryngeal mask airway placed for decannulation and quickly transitioned to heated high-flow nasal cannula (n = 1). Two additional patients were extubated in the first week following decannulation (3 and 5 days after decannulation), whereas the remaining patients were either extubated weeks after decannulation (n = 2), transferred to an outside November 2014 • Volume 15 • Number 9
Extracorporeal Support
8
9
10
11
12
13
14
15
16
1
19
4
1
15
3
2
13
1
3
10
14
4
4
6
2
5
9
0.32
0.14
0.78
0.27
0.29
0.67
0.29
0.24
0.9
RA
HHFNC
RA → HHFNC
RA
RA
RA → HHFNC
HHFNC
NC
HHFNC, BiPAP
↑Aeration
↑Aeration
↑Aeration
↑Aeration
↓Air leak, ↑aeration
↑Aeration
↑Aeration
↑Aeration
↑Aeration
↓Distress
↓Distress
↓Distress
↓Distress
↓Distress
↓Distress
↓Distress
↓Distress, ↑cough
↓Distress, ↑cough
Yes
Yes
Yes (briefly)
Yes
Yes
No
Yes
Yes
Yes
CPR, procedure, recruitment
Recruitment
Procedure
Recruitment
CPR
Not applicable
Procedure
Apneic episode
Procedure
Bronchoscopy
Not applicable
Chest tube
Not applicable
Not applicable
Not applicable
Bronchoscopy
Not applicable
Remove ECMO support
Yes (2)
Yes (9)
Yes (5)
No
Yes (5) + bronchogram
Yes (1)
Yes (2)
Yes (3)
Yes (2)
No
Yes (3)
Yes (2)
No
No
No
No
No
No
Remained reintubated
Remained reintubated
Reintubated only for decannulation
Remained reintubated
Reintubated for arrest
Reintubated only for decannulation
Remained reintubated
Remained reintubated
Reintubated to remove ECMO support
Not applicable
Not applicable
8 hr
Not applicable
Not applicable
5d
Not applicable
3d
Not applicable
Survivala
Death
Survival
Survivala
Death
Survival
Death
Survival
Death
WDC on hospital day # 100 for gastrointestinal problems
Not applicable
11/27
WDC on hospital day# 53 due to recurrent air leaks
Not applicable
10/42
Not applicable
6/20
Not applicable
facilities where details about postdecannulation extubation are unknown (n = 2), or died during ECMO support (n = 5). All five who died during ECMO were intubated at the time of stopping support. Three had been reintubated several days prior, one was reintubated for cardiac arrest and one was reintubated prior to decannulation and life support withdrawal.
DISCUSSION We present a series of 16 children who successfully remained extubated during ECMO support. They represented a broad range of cardiopulmonary diseases and included patients supported with both VV and VA modes. This experience, while still small, demonstrates the feasibility of extubation while receiving ECMO support under a variety of conditions. Pediatric Critical Care Medicine
The practice of extubating ECMO patients evolved quickly when purposeful extubation began in 2010. The only consistent requirement for extubation candidacy was that the patient be fully supported by the ECMO circuit and not require additional respiratory support (i.e., no additional therapeutic benefit to intubation/ventilation). Some patients would remain intubated for an additional day or two until bronchoscopy succeeded in clearing the large airways. The ECMO physician would discuss possible extubation with the primary intensivist, who made the ultimate decision. If the primary intensivist agreed to extubation, the family was informed and all questions answered. Although many families had questions about the procedure, none objected to or opposed extubation. No management protocols were used in this early experience. www.pccmjournal.org
867
Anton-Martin et al
expose the patients in this series to additional risks. Our experience does suggest, however, that safety procedures may require modification in the extubated patient on ECMO. For example, we began to keep an appropriately sized ETT, laryngoscope, and blade taped at the head of each extubated patient’s bed in E F G H case of pump failure and the need to activate the Pediatric Advanced Life Support algorithm after patient 8’s circuit failed suddenly. We also made a number of interesting observations in our extubated patients that will require further systemI J atic study to determine their validity. Our inability to draw firm conclusions at this stage of experience is due to two considerations. First, as we grew more comfortable with the practice of extubating, we began to perform it earlier in subsequent patient’s courses, Figure 2. Radiographic progression of acute respiratory distress syndrome in a representative patient. A, Chest uncovering potential advanradiograph prior to cannulation for venovenous extracorporeal membrane oxygenation (ECMO). B, Immediately tages not previously enviafter cannulation. C, Chest CT on day 3 of ECMO. D, Day 5 of ECMO, immediately after extubation. E, Postextubation day 3 (ECMO day 8). F, Postextubation day 5 (ECMO day 10). G, Postextubation day 7 sioned (e.g., possibly lower (ECMO day 12). H, Postextubation day 10 (ECMO day 15). I, Postdecannulation day 6. J, Day of ICU discharge sedation requirements and faster resolution of pulmonary inflammation). Our perspective on certain observations Furthermore, ECMO survival was close to 70%, and none of the ECMO deaths was attributable to extubation while on changed with greater experience (e.g., dyspneic breathing ECMO. Death was related to failure of primary disease to reverse, behavior). complication of the primary disease, ECMO complication, or a Extubated patients were generally more interactive with their combination of these. Extubation on ECMO does not seem to family, caregivers, and environment. They spent more time awake, with their eyes open, and with increasing age, responded to family members verbally, watched TV, and made their needs known to caregivers. Furthermore, these patients generally required less sedation, although this observation requires some qualification. Patients with longer extubation latencies (greater time between cannulation and first extubation) tended to have higher sedative requirements than those who were extubated shortly after cannulation, although this is difficult to quantify in such a small series (18). The association between extubation and lower Figure 3. Distribution of State behavioral Scale (SBS) scores of extubated and intubated extracorporeal sedation also suggested that the ETT membrane oxygenation (ECMO) patients, showing percentage of ECMO course patients spent and mechanical ventilation may have appropriately sedated, oversedated, or undersedated.
A
868
B
www.pccmjournal.org
C
D
November 2014 • Volume 15 • Number 9
Extracorporeal Support
been more powerful irritants than ECMO, driving the bulk of the need for sedation with its attendant side effects. The SBS was our institution’s tool to assess sedation level in PICU patients, including those on ECMO. Although useful, it cannot distinguish agitation from hypermotor delirium, a consequence of long-term sedation in the ICU (19, 20). We realized that we might be underestimating the prevalence of delirium in our long-term ECMO patients when we extubated an older agitated patient and found him to be hallucinating. Whether delirium was responsible for SBS scores of +1 or +2 in some agitated patients is impossible to ascertain. It nonetheless bears noting that many long-term ECMO patients were agitated, difficult to sedate, and possibly delirious. With increasing emphasis on long-term cognitive and functional impairment of patients after delirium in the ICU (21, 22), improving our recognition of this complication and minimizing noxious stimulation that requires sedation may be one strategy to prevent it and its lingering effects. Extubation during ECMO permitted increased physical activity in cannulated patients. Initially, we were hesitant to move our first patients because of concerns for cannula dislodgement, but this proved not to be an issue, and each subsequent patient received increasingly vigorous physical and occupational therapy, up to and including limited ambulation. The amount of physical activity relates to both patient age and cannulation site, but extubation provides the opportunity to prevent the physical deconditioning associated with long ICU stays (usually also associated with long periods of intubation and ventilation). Spontaneous, negative pressure ventilation may have been effective in promoting spontaneous reaeration, cough, and secretion mobilization in patients with severe pulmonary consolidation. In many patients, resolution of consolidation and subsequent aeration occurred in the absence of positive pressure ventilation. What the present series cannot answer, however, is whether extubation speeds or delays resolution of lung injury and alveolar reinflation. Although “rest settings” are likely less harmful than aggressive ventilator settings, it is unknown whether positive pressure of any sort is harmful or helpful in an injured lung. Our study has several limitations. The series is small, with 16 patients. The data derive from a single center and the analysis was performed retrospectively. Most of the observations will require follow-up in larger series at several centers to confirm.
CONCLUSIONS We present the largest series of extubated pediatric ECMO patients to date and demonstrate both the feasibility of extubation in a broad range of patients on both VV and VA ECMO. Furthermore, we share preliminary evidence that extubation was associated with reduced patient sedation requirements and increased opportunity to avoid deconditioning.
ACKNOWLEDGMENTS We thank the extracorporeal membrane oxygenation (ECMO) specialists and bedside care staff for the wonderful care provided to all our ECMO patients, including the extubated ones and the children and their families Pediatric Critical Care Medicine
REFERENCES
1. Registry Report. Ann Arbor, MI, Extracorporeal Life Support Organization, 2010 2. MacLaren G, Combes A, Bartlett RH: Contemporary extracorporeal membrane oxygenation for adult respiratory failure: Life support in the new era. Intensive Care Med 2012; 38:210–220 3. Hirshberg E, Miller RR 3rd, Morris AH: Extracorporeal membrane oxygenation in adults with acute respiratory distress syndrome. Curr Opin Crit Care 2013; 19:38–43 4. Rehder KJ, Turner DA, Bonadonna D, et al: State of the art: Strategies for extracorporeal membrane oxygenation in respiratory failure. Expert Rev Respir Med 2012; 6:513–521 5. Prins S, van Dijk M, Tibboel D: Sedation and analgesia in the PICU: Many questions remain. Intensive Care Med 2006; 32:1103–1105 6. Fuehner T, Kuehn C, Hadem J, et al: Extracorporeal membrane oxygenation in awake patients as bridge to lung transplantation. Am J Respir Crit Care Med 2012; 185:763–768 7. Garcia JP, Iacono A, Kon ZN, et al: Ambulatory extracorporeal membrane oxygenation: A new approach for bridge-to-lung transplantation. J Thorac Cardiovasc Surg 2010; 139:e137–e139 8. Olsson KM, Simon A, Strueber M, et al: Extracorporeal membrane oxygenation in nonintubated patients as bridge to lung transplantation. Am J Transplant 2010; 10:2173–2178 9. Nosotti M, Rosso L, Tosi D, et al: Extracorporeal membrane oxygenation with spontaneous breathing as a bridge to lung transplantation. Interact Cardiovasc Thorac Surg 2013; 16:55–59 10. Garcia JP, Kon ZN, Evans C, et al: Ambulatory veno-venous extracorporeal membrane oxygenation: Innovation and pitfalls. J Thorac Cardiovasc Surg 2011; 142:755–761 11. Schmid C, Philipp A, Hilker M, et al: Venovenous extracorporeal membrane oxygenation for acute lung failure in adults. J Heart Lung Transplant 2012; 31:9–15 12. Turner DA, Cheifetz IM, Rehder KJ, et al: Active rehabilitation and physical therapy during extracorporeal membrane oxygenation while awaiting lung transplantation: A practical approach. Crit Care Med 2011; 39:2593–2598 13. Collar RM, Taylor JC, Hogikyan ND, et al: Awake extracorporeal membrane oxygenation for management of critical distal tracheal obstruction. Otolaryngol Head Neck Surg 2010; 142:618–620 14. Raake J, Johnson B, Seger B, et al: Extracorporeal membrane oxygenation, extubation, and lung-recruitment maneuvers as rescue therapy in a patient with tracheal dehiscence following slide tracheoplasty. Respir Care 2011; 56:1198–1202 15. Wickiser JE, Thompson M, Leavey PJ, et al: Extracorporeal membrane oxygenation (ECMO) initiation without intubation in two children with mediastinal malignancy. Pediatr Blood Cancer 2007; 49:751–754 16. Schmidt F, Sasse M, Boehne M, et al: Concept of “awake venovenous extracorporeal membrane oxygenation” in pediatric patients awaiting lung transplantation. Pediatr Transplant 2012; 17:224–230 17. Curley MA, Harris SK, Fraser KA, et al: State Behavioral Scale: A sedation assessment instrument for infants and young children supported on mechanical ventilation. Pediatr Crit Care Med 2006; 7:107–114 18. Anton-Martin P, Darnell-Bowens C, Raman L, et al: Sedation and analgesic requirements in extubated pediatric patients on ECMO (abstract presented). 24th Annual ELSO Conference. Philadelphia, PA, September 19–22, 2013 19. Reade MC, Finfer S: Sedation and delirium in the intensive care unit. N Engl J Med 2014; 370:444–454 20. Smith HA, Berutti T, Brink E, et al: Pediatric critical care perceptions on analgesia, sedation, and delirium. Semin Respir Crit Care Med 2013; 34:244–261 21. Pandharipande PP, Girard TD, Jackson JC, et al; BRAIN-ICU Study Investigators: Long-term cognitive impairment after critical illness. N Engl J Med 2013; 369:1306–1316 22. Brummel NE, Jackson JC, Pandharipande PP, et al: Delirium in the ICU and subsequent long-term disability among survivors of mechanical ventilation. Crit Care Med 2013; 42:369–377 www.pccmjournal.org
869
