CARING FOR THE JAMA-EXPRESS

CRITICALLY ILL PATIENT

Extracorporeal Membrane Oxygenation

for 2009 Influenza A(H1N1)
Acute Respiratory Distress Syndrome
The Australia and New Zealand
Context The novel influenza A(H1N1) pandemic affected Australia and New Zea-
Extracorporeal Membrane land during the 2009 southern hemisphere winter. It caused an epidemic of critical
Oxygenation (ANZ ECMO) Influenza illness and some patients developed severe acute respiratory distress syndrome (ARDS)
Investigators* and were treated with extracorporeal membrane oxygenation (ECMO).

I
N APRIL 2009, THE MEXICAN MIN- Objectives To describe the characteristics of all patients with 2009 influenza A(H1N1)–
istry of Health reported an in- associated ARDS treated with ECMO and to report incidence, resource utilization, and
patient outcomes.
crease in severe pneumonia cases
in young adults.1 The 2009 novel Design, Setting, and Patients An observational study of all patients (n=68) with
swine-origin influenza A(H1N1) vi- 2009 influenza A(H1N1)–associated ARDS treated with ECMO in 15 intensive care
units (ICUs) in Australia and New Zealand between June 1 and August 31, 2009.
rus was identified as its cause and rap-
idly led to a worldwide pandemic.2 This Main Outcome Measures Incidence, clinical features, degree of pulmonary dys-
pandemic began in the northern hemi- function, technical characteristics, duration of ECMO, complications, and survival.
sphere during late spring and early sum- Results Sixty-eight patients with severe influenza-associated ARDS were treated with
mer and appeared to decrease in inten- ECMO, of whom 61 had either confirmed 2009 influenza A(H1N1) (n=53) or influ-
sity within a few weeks.3 Shortly after, enza A not subtyped (n=8), representing an incidence rate of 2.6 ECMO cases per
at the start of the southern hemi- million population. An additional 133 patients with influenza A received mechanical
ventilation but no ECMO in the same ICUs. The 68 patients who received ECMO had
sphere winter, it spread to Australia and
a median (interquartile range [IQR]) age of 34.4 (26.6-43.1) years and 34 patients
New Zealand causing an approxi- (50%) were men. Before ECMO, patients had severe respiratory failure despite ad-
mately 8-fold greater number of con- vanced mechanical ventilatory support with a median (IQR) PaO2/fraction of inspired
firmed cases per head of population oxygen (FIO2) ratio of 56 (48-63), positive end-expiratory pressure of 18 (15-20) cm
than in the United States.4,5 H2O, and an acute lung injury score of 3.8 (3.5-4.0). The median (IQR) duration of
The spread of the virus to Australia ECMO support was 10 (7-15) days. At the time of reporting, 48 of the 68 patients
and New Zealand was also associated (71%; 95% confidence interval [CI], 60%-82%) had survived to ICU discharge, of
with a large number of patients admit- whom 32 had survived to hospital discharge and 16 remained as hospital inpatients.
Fourteen patients (21%; 95% CI, 11%-30%) had died and 6 remained in the ICU, 2
ted to intensive care units (ICUs) across
of whom were still receiving ECMO.
both countries.6 A proportion of these
patients presented with, or developed, Conclusions During June to August 2009 in Australia and New Zealand, the ICUs
severe acute respiratory distress syn- at regional referral centers provided mechanical ventilation for many patients with 2009
influenza A(H1N1)–associated respiratory failure, one-third of whom received ECMO.
drome (ARDS). In some severe cases, These ECMO-treated patients were often young adults with severe hypoxemia and
extracorporeal membrane oxygen- had a 21% mortality rate at the end of the study period.
ation (ECMO) was commenced for the JAMA. 2009;302(17):1888-1895 www.jama.com
treatment of refractory hypoxemia, hy-
percapnia, or both, which occurred de- tory failure, technical characteristics, *Authors/Management and Writing Committee and
spite mechanical ventilation and res- duration of extracorporeal support,
Investigators of the ANZ ECMO Influenza Investiga-
tors are listed at the end of this article.
cue ARDS therapies. complications, and survival in pa- Corresponding Author: Andrew R. Davies, MBBS,
We report herein on the incidence, tients with severe influenza-related
FRACP, FJFICM, Intensive Care Unit, Alfred Hospi-
tal, Commercial Road, Melbourne, Victoria 3004, Aus-
clinical features, severity of respira- ARDS who were treated with ECMO. tralia ([email protected]).
Caring for the Critically Ill Patient Section Editor: Derek
In addition, we discuss the relevance of C. Angus, MD, MPH, Contributing Editor, JAMA
For editorial comment see p 1905.
our findings to the potential ECMO case ([email protected]).

1888 JAMA, November 4, 2009—Vol 302, No. 17 (Reprinted) ©2009 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ by Vitor Monteiro on 05/17/2021

 EXTRACORPOREAL MEMBRANE OXYGENATION FOR INFLUENZA A(H1N1)

load in northern hemisphere coun- leukocytosis [⬎10 000/µL or ⬎15% ECMO for ARDS in the ECMO cen-
tries during their 2009-2010 winter. bands], C-reactive protein ⬎3-fold the ters during the winter of 2008 was ob-
normal upper limit, and a positive cul- tained from each ICU’s registry of cases.
METHODS ture from blood or pleural fluid). We
Study Design and Patient Eligibility identified the presumed infectious Statistical Analysis
We studied adult and pediatric pa- organism from upper and lower respi- Data analysis was descriptive using me-
tients who were treated with ECMO be- ratory tract specimens (polymerase dian and interquartile range (IQR). We
tween June 1 and August 31, 2009. We chain reaction, viral culture, or both), made no assumptions about missing
contacted all 187 ICUs in Australia and blood cultures, or urinary antigens data and adjusted proportions to the
New Zealand and identified the 15 ICUs obtained within the first 72 hours of number of patients with available data.
that provided ECMO support during admission, or from convalescent or When acute lung injury score vari-
this period. We excluded neonates or paired serology testing. ables were missing, the modified score
patients treated with ECMO for pri- We obtained information on the tim- was calculated (dividing the sum of sub-
mary cardiac failure, following heart ing of endotracheal intubation in rela- scores by the number of known vari-
and/or lung transplantation or cardiac tion to presumed onset of symptoms ables). To report our findings rapidly,
surgery. We applied these eligibility cri- and hospital admission, the total du- we censored all outcomes at midnight
teria to capture all confirmed or strongly ration of mechanical ventilation, and on September 7, 2009. Using current
suspected cases of 2009 influenza administration of antiviral and antibi- Australia and New Zealand popula-
A(H1N1)–related respiratory disease. otic medications. We documented tion data,10,11 we calculated the inci-
We also identified and excluded pa- whether the ECMO treatment was com- dence of ECMO use per million people
tients with an alternative diagnosis and menced in the participating hospital or for Australia and New Zealand in total,
who had no virus isolated. whether the patient was retrieved and and for each jurisdiction (Australian
All members of the binational transferred while receiving ECMO from states and New Zealand) that pro-
management committee approved a referral center. vided ECMO support. We also calcu-
the study protocol. Trained research We assessed severity of illness be- lated the incidence of ECMO use for
coordinators or treating clinicians fore endotracheal intubation by docu- confirmed as well as the combination
used a case report form to obtain rel- menting respiratory rate and mea- of confirmed and suspected 2009 in-
evant data. Approval was obtained sures of oxygenation. We assessed fluenza A(H1N1). We also estimated
from the hospital research ethics severity of illness before commence- the ECMO burden by calculating the
committees at all participating cen- ment of ECMO by documenting total number of days on which ECMO
ters. All committees waived the need nonpulmonary vital organ support, se- was provided to all patients and by cal-
for informed consent. verity of hypoxemia, hypercapnia, ven- culating the total number of patients
tilator settings, and use of rescue ARDS treated concurrently in all hospitals in
Data Collection therapies in the 6 hours before ECMO Australia and New Zealand for each day
We collected data retrospectively on commencement. We also obtained data of the winter period.
patient demographics including age, to calculate a modified acute lung in- Comparisons of proportions were
sex, height, weight, and ethnicity, as jury score (range, 0-4) during this made using ␹2 tests for equal propor-
well as the presence of a number of period.9 tion or Fisher exact tests when num-
predefined comorbidities. We assessed We recorded duration of mechani- bers were small. Continuous vari-
whether patients fulfilled criteria dur- cal ventilation, ECMO, ICU and hos- ables were compared using Wilcoxon
ing the period before or at the time of pital stay, mortality, and destination at rank sum tests. All reported P values
presentation to hospital for an influen- hospital discharge. Information on are 2-sided and were not adjusted for
zalike illness based on typical symp- functional status at hospital discharge multiple comparisons. P⬍.05 was
toms7 (defined as ⱖ3 symptoms of in survivors included whether the pa- considered statistically significant.
sore throat, cough, myalgia or arthral- tient was ambulant and the pulse ox- Analysis was performed using SAS
gia, respiratory distress, vomiting or imetry reading on room air. In pa- version 9.1 (SAS Institute Inc, Cary,
diarrhea, and core temperature tients who died during hospital North Carolina).
⬎38°C). We also assessed whether admission, we characterized the mode
they fulfilled criteria for community- of death from a list of predefined RESULTS
acquired pneumonia8 (defined as pres- options. Patient Characteristics
ence of a new or progressive infiltrate Data on demographics, comorbidi- and Use of ECMO
on chest radiograph plus ⱖ2 symp- ties, treatment, and outcome were col- Between June 1 and August 31, 2009,
toms of cough, sputum production, lected on patients with confirmed in- 72 patients were treated with ECMO
core temperature ⬎38°C, auscultatory fluenza A who were not treated with and fulfilled eligibility criteria for the
findings consistent with pneumonia, ECMO in the same ICUs. The use of study in the 15 participating ICUs. Four
©2009 American Medical Association. All rights reserved. (Reprinted) JAMA, November 4, 2009—Vol 302, No. 17 1889

Downloaded From: https://jamanetwork.com/ by Vitor Monteiro on 05/17/2021

 EXTRACORPOREAL MEMBRANE OXYGENATION FOR INFLUENZA A(H1N1)

patients were excluded from analysis For the remaining 68 patients who in meters squared), asthma, and dia-
because 3 patients had alternative di- received ECMO, the median (IQR) age betes mellitus in 34 patients (50%), 19
agnoses (Wegener granulomatosis, con- was 34.4 (26.6-43.1) years and 34 pa- patients (28%), and 10 patients (15%),
nective tissue disease, and cystic fibro- tients (50%) were men. The most com- respectively. Six patients (9%) were
sis) and 1 patient had 2009 influenza mon associated comorbidities were obe- pregnant and 4 patients (6%) were post-
A(H1N1)–associated fulminant myo- sity (body mass index ⬎30, calculated partum (⬍28 days of delivery). Three
carditis without ARDS. as weight in kilograms divided by height children (aged ⬍15 years) and no el-
derly patients (aged ⬎65 years) re-
ceived ECMO.
Figure 1. Flow Diagram of Patients Receiving Mechanical Ventilation for Suspected 2009
Influenza A(H1N1) Infection at ECMO Centers Of the 68 patients who received
ECMO for influenza-associated ARDS,
201 Patients given mechanical ventilation 66 (97%) fulfilled criteria for pneumo-
for confirmed or suspected influenza
nia and 64 (94%) fulfilled criteria for a
preceding influenzalike illness. Fifty-
68 Received ECMO 133 Did not receive ECMO
three patients (78%) had 2009 influ-
enza A(H1N1) detected by polymer-
ase chain reaction or viral culture and
61 Confirmed 2009 7 Had suspected but 133 Confirmed 2009 8 patients (12%) had serological evi-
influenza A(H1N1) unconfirmed influenza influenza A(H1N1)
or influenza A not or influenza A not dence of recent influenza A that was not
subtyped subtyped subtyped and was regarded as sus-
pected to be 2009 influenza A(H1N1)
(FIGURE 1). The remaining 7 patients
53 Confirmed 2009 8 Confirmed influenza A 6 Alive 116 Alive
influenza A(H1N1) not subtyped 1 Still in ICU 11 Still in ICU (10%) had preceding symptoms of in-
42 Alive 6 Alive 1 Died 17 Died fluenzalike illness and were also re-
4 Still in ICU 1 Still in ICU
11 Died 2 Died garded as having suspected 2009 in-
fluenza A(H1N1). Seasonal subtypes of
ECMO indicates extracorporeal membrane oxygenation; ICU, intensive care unit. influenza A were not detected in any
patient. Nineteen (28%) patients also
had a secondary organism isolated from
Table 1. Comparison of Patients With Influenza A Who Received ECMO and Those Who
Received Mechanical Ventilation But Without ECMO at ECMO Centers a a respiratory tract specimen or blood
Mechanical Ventilation sample at the time of hospital presen-
ECMO But Without ECMO P tation, the most common being Strep-
Parameter (n = 61) (n = 133) Value tococcus pneumoniae (n = 10) and
Age, median (IQR), y 36 (27-45) 44 (31-54) .02 Staphylococcus aureus (n=4).
Male sex 29 (48) 63 (47) .54
During the study period, 252 pa-
BMI, median (IQR) 29 (23-36) 29 (24-37) .92
tients were admitted to the 15 partici-
Chronic lung disease 18 (30) 35 (26) .64
pating ICUs with influenza. Of these pa-
APACHE III comorbidity b 5 (8) 30 (23) .02
tients, 201 received mechanical
Pregnancy or postpartum 10 (16) 12 (9) .21
ventilation. Of the 194 patients with
Diabetes mellitus 9 (15) 23 (17) .64
either confirmed 2009 influenza
H1N1 positive 56 (92) 107 (80) .05
A(H1N1) or influenza A not sub-
At ICU admission
Mechanical ventilation 53 (87) 117 (88) .80 typed, 61 treated with ECMO were
Vasopressor 35 (57) 46 (34) .02 compared with the 133 treated with me-
Renal replacement therapy 5 (8) 9 (7) .95 chanical ventilation but without ECMO
Duration or length of stay, in TABLE 1.
median (IQR), d The estimated incidence of ECMO
Mechanical ventilation 18 (9-27) 8 (4-14) .001
use for the combination of confirmed
ICU 22 (13-32) 12 (7-18) .001
and suspected 2009 influenza A(H1N1)
Hospital 28 (15-43) 20 (13-31) .07
during the winter influenza season was
Mortality
in ICU 14 (23) 12 (9) .01 2.6 (95% confidence interval [CI], 2.0-
in hospital 14 (23) 17 (13) .06 3.2) cases per million people. When
Abbreviations: APACHE, Acute Physiology and Chronic Health Evaluation; BMI, body mass index (calculated as weight in only confirmed cases were consid-
kilograms divided by height in meters squared); ECMO, extracorporeal membrane oxygenation; ICU, intensive care unit;
IQR, interquartile range.
ered, the estimated incidence was 2.0
a Data are presented as No. (%) unless otherwise specified.
b The presence or not of at least 1 comorbidity.
(95% CI, 1.4-2.6) cases per million. In
the jurisdictions where patients were
1890 JAMA, November 4, 2009—Vol 302, No. 17 (Reprinted) ©2009 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ by Vitor Monteiro on 05/17/2021

 EXTRACORPOREAL MEMBRANE OXYGENATION FOR INFLUENZA A(H1N1)

treated, the estimated incidence of ceived ECMO required retrieval and in- sure before ECMO commencement
ECMO use varied from 1.6 (95% CI, ter-hospital transfer to the ECMO- were 1.0 (1.0-1.0), 18 (15-20) cm H2O,
1.1-2.1) to 5.3 (95% CI, 4.3-6.3) cases providing site; of these, 38 (78%) were 5.6 (4.6-6.7) mL/kg, and 36 (33-38) cm
per million for confirmed and sus- started on ECMO at the referring site H2O, respectively. All but 2 patients had
pected 2009 influenza A(H1N1). The and successfully transferred while re- a PaO2/FIO2 ratio of 83 or less, and both
total ECMO burden for the cohort was ceiving ECMO. of these had a PaCO2 of 98 or more and
828 days of ECMO (32; 95% CI, 30-34 a pH of 7.07 or less. All patients had
ECMO days per million). The number Severity of Illness and Treatment either a modified acute lung injury score
of patients treated concurrently with Before Commencement of of 3.0 or more, or the combination of
ECMO in Australia and New Zealand Mechanical Ventilation and ECMO hypercapnia and a pH of less than 7.2.
peaked 8 weeks after the first patient Median (IQR) duration of mechanical Representative images of a chest radio-
was treated and then decreased dur- ventilation before commencement of graph and a computed tomogram for
ing the next 4 weeks, with the maxi- ECMO was 2 (1-5) days. Before me- these patients are shown in FIGURE 3.
mum number of 23 patients (34%) on chanical ventilation, the median (IQR) In cases with available data before
3 consecutive days in early August respiratory rate, arterial oxygen satu- commencement of ECMO, clinicians
(FIGURE 2). In the previous winter ration (SaO2), and PaO2 were 44 (31- used rescue ARDS therapies such as re-
(June 1-August 31, 2008), only 4 pa- 48)/min, 83% (77%-88%), and 53 (47- cruitment maneuvers in 38 patients
tients (estimated incidence of 0.15 cases 60) mm Hg, respectively. Details of (67%), prone positioning in 12 pa-
per million people) received ECMO for severity of illness in the 6 hours be- tients (20%), high-frequency oscilla-
ARDS in participating sites. fore ECMO commencement are shown tory ventilation in 3 patients (5%), in-
The median (IQR) interval between in TABLE 2. Overall, patients had a me- haled nitric oxide in 20 patients (32%),
the onset of influenzalike symptoms dian (IQR) lowest PaO2/fraction of in- or prostacyclin in 14 patients (22%).
and hospital admission, ICU admis- spired oxygen (FIO2) ratio of 56 (48- Overall, 55 patients (81%) received at
sion, and ECMO was 5 (3-6) days, 5 63), a lowest pH of 7.2 (7.1-7.3), a least 1 of these therapies. Further-
(3-7) days, and 9 (5-13) days, respec- highest PaCO2 of 69 (54-83) mm Hg, more, 46 patients (68%) received va-
tively. Oseltamivir (administered en- and a modified acute lung injury score soactive drugs and 16 patients (24%)
terally) was used as initial antiviral treat- of 3.8 (3.5-4.0). The median (IQR) received renal replacement therapy be-
ment in 64 patients (94%) for a median highest recorded FIO2, positive end- fore commencement of ECMO. Pa-
(IQR) duration of 8 (7-11) days. Forty- expiratory pressure, tidal volume (per tients with secondary bacterial infec-
nine of 68 patients (72%) who re- kg body weight), and peak airway pres- tion at the time of hospital presentation

Figure 2. Histogram of Number of Concurrent Patients Receiving ECMO Across Australia and New Zealand in 2009

25

20

15
No. of Patients

10

0
9 11 13 15 17 19 21 23 25 27 29 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 1 3 5 7
Jun Jul Aug Sep

Date

ECMO indicates extracorporeal membrane oxygenation.

©2009 American Medical Association. All rights reserved. (Reprinted) JAMA, November 4, 2009—Vol 302, No. 17 1891

Downloaded From: https://jamanetwork.com/ by Vitor Monteiro on 05/17/2021

 EXTRACORPOREAL MEMBRANE OXYGENATION FOR INFLUENZA A(H1N1)

(n=19) were more likely to receive va- tance oxygenators. The initial mode of (4.0-5.9) and 4.9 (3.9-6.0) L/min,
soactive drugs (90% vs 59%, respec- ECMO was veno-venous in 63 pa- respectively.
tively; P=.01). tients (93%) and veno-arterial in 5 pa- All adult patients had vascular can-
tients (7%). No arteriovenous (pump- nulae inserted through a peripheral ap-
Technical Details of ECMO Support less) support was used. The median proach into the femoral, jugular, or both
All centers provided ECMO with cen- (IQR) duration of ECMO support was vessels, and 1 child had central cannu-
trifugal blood pump driven circuit flow 10 (7-15) days. Median (IQR) circuit lae. In 33 patients (49%), a second ac-
and polymethylpentene low-resis- blood flow at 4 and 24 hours was 4.9 cess cannula was needed to augment
ECMO support. Hemorrhagic compli-
Table 2. Severity of ARDS Before Commencement of ECMO cations occurred in 37 patients (54%)
2009 Influenza A(H1N1) during ECMO therapy, with the most
common sources being ECMO cannu-
Confirmed Suspected
Infection Infection All Infections lation sites in 15 patients (22%), gas-
Characteristics (n = 53) (n = 15) (N = 68) trointestinal tract in 7 patients (10%),
Ventilation parameters, median (IQR) respiratory tract in 7 patients (10%),
Lowest PaO2/FIO2 ratio 55 (48-65) 57 (45-62) 56 (48-63) vaginal bleeding in 6 patients (9%), and
Highest FIO2 1.0 (1.0-1.0) 1.0 (1.0-1.0) 1.0 (1.0-1.0)
intracranial hemorrhage in 6 patients
Highest PEEP, cm H2O 18 (15-20) 15 (14-18) 18 (15-20)
(9%).
Highest peak airway pressure, 36 (34-40) 34 (29-36) 36 (33-38)
cm H2O The median (IQR) amount of blood
Lowest pH 7.2 (7.1-7.3) 7.2 (7.1-7.3) 7.2 (7.1-7.3) administered per patient was 1880
Highest PaCO2, mm Hg 69 (54-86) 67 (61-73) 69 (54-83) (904-3750) mL. Infective complica-
Highest tidal volume, mL/kg 5.6 (4.8-6.6) 5.7 (4.4-6.7) 5.6 (4.6-6.7) tions occurred in 42 patients (62%) dur-
Quadrants of radiograph 4 (4-4) 4 (4-4) 4 (4-4) ing ECMO therapy, with the most com-
infiltrate, No. mon sites being respiratory tract in 30
Acute lung injury score a 3.8 (3.3-4.0) 3.5 (3.3-3.8) 3.8 (3.5-4.0) patients (44%), bloodstream in 14 pa-
Pneumothorax pre-ECMO, No. (%) 9 (17) 1 (7) 10 (15) tients (21%), non-ECMO catheter-
Rescue ARDS therapies used, No. (%) related in 13 patients (19%), and ECMO
Recruitment maneuver 30 (66) 8 (66) 38 (67)
cannulae-related in 7 patients (10%).
Prone positioning 11 (22) 1 (8) 12 (20)
High-frequency oscillation 3 (6) 0 3 (5) Details of ICU Support
Nitric oxide 19 (38) 1 (8) 20 (32) and Outcomes for Patients
Prostacyclin 12 (23) 2 (15) 14 (22)
Requiring ECMO
Abbreviations: ARDS, acute respiratory distress syndrome; ECMO, extracorporeal membrane oxygenation; FIO2, frac-
tion of inspired oxygen; IQR, interquartile range; PEEP, positive end-expiratory pressure.
a Data were missing in 4 cases for PaO /FIO ratio, in 4 cases for PEEP, in 17 cases for lung compliance, and in 5 cases
The median (IQR) duration of mechani-
2
for quadrants of radiograph infiltrate.
2
cal ventilation was 25 (13-34) days (26
[14-34] and 14 [7-29] days for survi-
Figure 3. Chest Radiograph and Computed Tomogram of 2 Patients Successfully Treated vors and nonsurvivors, respectively)
With ECMO for Confirmed 2009 Influenza A(H1N1) (T ABLE 3). Tracheostomy was per-
formed to assist weaning from mechani-
Chest radiograph Computed tomogram cal ventilation in 39 patients (57%). The
median (IQR) durations of ICU admis-
R L R L
sion and hospitalization were 27 (16-
37) and 39 (23-47) days, respectively.
Of the 68 patients, 53 (78%; 95%
CI, 68%-88%) had been weaned from
ECMO, 13 had died while receiving
ECMO, and the other 2 were still
receiving ECMO as of September 7,
2009. Of the 53 patients weaned from
ECMO, 1 had died and 52 (76%)
were still alive. Of the 52 patients still
alive and weaned from ECMO, 4 were
still in the ICU and 48 (71%; 95% CI,
60%-82%) had survived to ICU dis-
ECMO indicates extracorporeal membrane oxygenation. The images demonstrate severe bilateral airspace dis- charge. Of the 48 ICU survivors, 16
ease with massive loss of normal aerated lung tissue.
patients (24%; 95% CI, 13%-34%)
1892 JAMA, November 4, 2009—Vol 302, No. 17 (Reprinted) ©2009 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ by Vitor Monteiro on 05/17/2021

 EXTRACORPOREAL MEMBRANE OXYGENATION FOR INFLUENZA A(H1N1)

were still in the hospital and 32 ECMO, and received prolonged me- Several factors may have contrib-
patients (47%; 95% CI, 35%-59%) chanical ventilation and ECMO sup- uted to the observed mortality rate.
had survived to hospital discharge. port. Children and elderly persons were First, our patients were young and had
Of the 32 hospital survivors, 31 pa- infrequently treated with ECMO. The ARDS secondary to viral pneumonia,
tients (97%) were ambulant. In 20 of majority of patients underwent re- which when managed with ECMO has
32 hospital survivors, pulse oximetry trieval to a specialist center for ECMO. been associated with higher survival
data on room air were available and all Despite the disease severity and the in- rates than other causes of ARDS.12-14
patients had recordings of 92% or more tensity of treatment, the mortality rate Second, improvements in ECMO tech-
(median [IQR], 97% [95%-98%]). In was low. nology (eg, heparin-bonded cannulae,
the 14 patients who died (21%; 95% CI, rotary pumps, and small efficient long-
11%-30%), intracranial hemorrhage Comparison With Previous Studies lasting oxygenators) and staff training
(n=6), other hemorrhage (n = 4), and To our knowledge, this is the first mul- have occurred since previous publica-
intractable respiratory failure (n = 4) ticenter study on the use of ECMO for tions, leading to safer and more effec-
were the most common conditions con- 2009 influenza A(H1N1)–associated tive ECMO application. All of the pa-
tributing to death. Of the 10 pregnant ARDS. Publications from an interna- tients fulfilled the ARDS severity criteria
or postpartum patients, 7 (70%) were tional ECMO registry12 and from cen- for enrollment in a recently reported
alive. Of the 3 children treated with ters experienced in the use of ECMO for randomized controlled trial (the CESAR
ECMO, all 3 were alive; however, 1 ARDS of heterogeneous etiology have re- study16) of ECMO treatment.
child remained in the ICU. ported mortality rates between 30% and
48%.13-15 Although our patients had a Implications for Policy Makers
Details of Outcomes for Patients mortality rate of 21% (95% CI, 11%- and Clinicians
With and Without ECMO 30%), several patients remained in the Our findings have implications for
From the group of 194 mechanically ICU at the time of reporting. health care planning and the clinical
ventilated patients with confirmed 2009
influenza A(H1N1) or influenza A not
Table 3. Patient Outcomes a
subtyped (not all of whom had ARDS),
2009 Influenza A(H1N1)
patients treated with ECMO (n = 61)
were compared with those without Confirmed Suspected
(n=133). The patients who were treated Infection Infection All Infections
Outcome Measure (n = 53) (n = 15) (N = 68)
with ECMO had longer duration of me- Length of stay, median (IQR), d
chanical ventilation (median [IQR], 18 ICU 26 (16-35) 31 (15-38) 27 (16-37)
[9-27] vs 8 [4-14] days; P = .001), ICU Hospital 35 (24-45) 40 (27-54) 39 (23-47)
stay (median [IQR], 22 [13-32] vs 12 Duration, median (IQR), d
[7-18] days; P =.001), and greater ICU Mechanical ventilation 24 (13-31) 28 (13-34) 25 (13-34)
mortality (14 [23%] vs 12 [9%]; P=.01). ECMO support 10 (7-14) 11 (10-16) 10 (7-15)
Survival at ICU discharge 38 (72) 10 (67) 48 (71)
COMMENT Still in ICU 4 (8) 2 (13) 6 (9)
Summary of Study Findings Survival at hospital discharge 22 (42) 10 (67) 32 (47)
Still in hospital b 14 (26) 2 (13) 16 (24)
We identified all patients who re-
Ambulant at hospital discharge c 21 (95) 10 (100) 31 (97)
ceived ECMO for severe ARDS during
SaO2 on room air at hospital 97 (95-98) 97 (95-98) 97 (95-98)
the 2009 influenza A(H1N1) winter discharge, median (IQR), % c
pandemic in Australia and New Zea- Discharge destination
land. Although there are almost 200 Died 11 (21) 3 (20) 14 (21)
ICUs across these 2 countries, all Home 18 (34) 4 (27) 22 (32)
ECMO was provided at just 15 special- Other hospital 0 1 (7) 1 (1)
ist centers. Within these centers, the Rehabilitation facility 4 (8) 5 (33) 9 (13)
burden was substantial, as high- Cause of death d
Hemorrhage 3 (27) 1 (33) 4 (29)
lighted by the provision of a large num-
Intracranial hemorrhage 4 (36) 2 (66) 6 (43)
ber of total days of ECMO support and
Infection 1 (9) 0 1 (7)
the use of ECMO support in approxi-
Intractable respiratory failure 3 (27) 1 (33) 4 (29)
mately one-third of all cases requiring Abbreviations: ECMO, extracorporeal membrane oxygenation; ICU, intensive care unit; IQR, interquartile range; SaO2,
mechanical ventilation at these cen- arterial oxygen saturation.
a Data are presented as No. (%) unless otherwise specified.
ters. Affected patients were often young b Not including patients still in the ICU.
c For survivors only.
adults, pregnant or postpartum, obese, d Data are shown as No. (% of deaths) and patients could have more than 1 cause contributing to death.
had severe respiratory failure before
©2009 American Medical Association. All rights reserved. (Reprinted) JAMA, November 4, 2009—Vol 302, No. 17 1893

Downloaded From: https://jamanetwork.com/ by Vitor Monteiro on 05/17/2021

 EXTRACORPOREAL MEMBRANE OXYGENATION FOR INFLUENZA A(H1N1)

management of patients with 2009 in- tients tested positive for 2009 influ- lowing ICU discharge was uncom-
fluenza A(H1N1) during the 2009- enza A(H1N1), the remainder had mon. In addition, we are unable to com-
2010 northern hemisphere winter. Our confirmed influenza A during an out- ment on the long-term outcome of our
results indicate that the incidence of break in which the dominant strain of patients, particularly in relation to the
ARDS sufficient to warrant consider- laboratory-confirmed influenza A has degree of pulmonary dysfunction and
ation of ECMO, based on the criteria been 2009 influenza A(H1N1)19 or had quality of life. Finally, our estimates of
used for the CESAR study,16 exceeds 2.6 features of a preceding influenzalike ill- ECMO use may be affected by changes
per million inhabitants. Given the out- ness complicated by pneumonia. In ad- in virulence of the virus or the devel-
comes reported in the CESAR study and dition, their clinical characteristics were opment and deployment of an effec-
in our study, other clinicians may also similar to those with confirmed 2009 tive and safe vaccine.
choose to treat these patients with influenza A(H1N1). As the diagnostic
ECMO. Approximately 15% of our pa- sensitivity of microbiological tests for CONCLUSION
tients were pregnant or postpartum, the 2009 influenza A(H1N1) is unknown, In Australia and New Zealand, during
largest case series of such patients in the many of these patients are likely to have the 2009 influenza A(H1N1) winter pan-
literature.17,18 Most of these patients been infected with the virus. demic, there was a large increase in the
survived. We are unable to report on the pos- use of ECMO for ARDS in patients com-
Despite the additional disease bur- sible outcome of our patients if ECMO pared with the winter of 2008. Despite
den, ECMO capacity was never ex- had not been used, because allocation their illness severity and the prolonged
ceeded; however, information on the re- to receive ECMO was not conducted in use of life support, most of these pa-
source utilization should facilitate the context of a randomized con- tients survived. This information should
planning in the northern hemisphere. trolled trial. In our study, approxi- facilitate health care planning and clini-
With a similar incidence of ECMO mately 30% of patients who were me- cal management for these complex pa-
use for 2009 influenza A(H1N1)– chanically ventilated with 2009 tients during the ongoing pandemic.
associated ARDS, rough estimates are influenza A(H1N1) were treated with
that the United States and the Euro- ECMO. This compares to an ECMO Published Online: October 12, 2009 (doi:10.1001
pean Union might expect to provide treatment rate for patients who were /jama.2009.1535). This article was corrected on
November 3, 2009.
ECMO to approximately 800 and 1300 mechanically ventilated with 2009 in- Authors/Management and Writing Committee: An-
patients during the 2009-2010 winter, fluenza A(H1N1) of only 10% from all drew Davies, MBBS, FRACP, FJFICM (chair), Austra-
lian and New Zealand Intensive Care Research Cen-
respectively. ICUs in 1 Australian state.20 Of the 187 ter, Monash University, and Alfred Hospital,
ICUs in Australia and New Zealand, Melbourne, Australia; Daryl Jones, MD, BSc(Hons),
Study Strengths and Limitations FRACP, FJFICM, Australian and New Zealand Inten-
only 15 provided ECMO services; how- sive Care Research Center, Monash University, and
Our study is the first to report, to our ever, these centers were often referred Austin Hospital, Melbourne, Australia; Michael Bailey,
knowledge, the ECMO experience for patients with severe respiratory fail- PhD, MSc(statistics), BSc(Hons), Australian and New
Zealand Intensive Care Research Center, Monash Uni-
2009 influenza A(H1N1)–related ARDS ure despite advanced mechanical ven- versity, Melbourne, Australia; John Beca, MBChB,
using a population-based method in 2 tilatory support through semiformal re- FRACP, FJFICM, Auckland City Hospital, Auckland,
New Zealand; Rinaldo Bellomo, MD, FRACP, FJFICM,
developed countries, with well- ferral networks. Australian and New Zealand Intensive Care Research
established and nationally coordinated Of the approximately 4950 patients Center, Monash University, and Austin Hospital, Mel-
bourne, Australia; Nikki Blackwell, BSc, FRCP, FRACP,
critical care systems. To our knowl- requiring hospitalization for 2009 in- DTMH, FAChPM, FJFICM, Prince Charles Hospital,
edge, this is the complete experience of fluenza A(H1N1) in Australia and New Brisbane, Australia; Paul Forrest, MBChB, FANZCA,
Royal Prince Alfred Hospital, Sydney, Australia; David
ECMO in our region during winter. We Zealand as of September 7, 2009 (4561 Gattas, MBBS, MMed(ClinEpi), FRACP, FJFICM, Royal
report important aspects of the epide- in Australia21 and approximately 400 in Prince Alfred Hospital, Sydney, Australia; Emily Granger,
miology, disease burden, and resource New Zealand based on a similar pro- FRACS, St Vincent’s Hospital, Sydney, Australia; Rob-
ert Herkes, MBBS, FRACP, FJFICM, Royal Prince Al-
utilization for ECMO. We confirm pre- portion of confirmed cases22), the ICUs fred Hospital, Sydney, Australia; Andrew Jackson,
vious findings of severe respiratory fail- at the 15 ECMO centers received 252 MBBS, FANZCA, St Vincent’s Hospital, Sydney, Aus-
tralia; Shay McGuinness, MBChB, FRCA, FANZCA,
ure in a subset of patients with 2009 in- patients, 68 of whom received ECMO. Auckland City Hospital, Auckland, New Zealand; Priya
fluenza A(H1N1),3 and also demonstrate Of the 252 patients, 31 died, represent- Nair, MD, FJFICM, St Vincent’s Hospital, Sydney, Aus-
tralia; Vincent Pellegrino, MBBS, FRACP, FJFICM, Al-
that most patients survived. ing 17% of all 2009 influenza A(H1N1) fred Hospital, Melbourne, Australia; Ville Pettilä, MD,
Our study has the inherent limita- deaths in Australia21 and New Zealand.22 PhD, Australian and New Zealand Intensive Care Re-
tions of a case series. To improve ac- With the requirement to inform the search Center, Monash University, Melbourne, Aus-
tralia; Brian Plunkett, MBChB, Royal Prince Alfred Hos-
curacy, we used systematic methods of northern hemisphere for the upcom- pital, Sydney, Australia; Roger Pye, FRACP, FJFICM,
data collection, such as a case report ing winter, we censored our data col- FANZCA, St Vincent’s Hospital, Sydney, Australia; Paul
Torzillo, MBBS, FRACP, FJFICM, Royal Prince Alfred
form, trained research coordinators, lection on September 7, 2009. Accord- Hospital, Sydney, Australia; Steve Webb, MPH, PhD,
predefined data field definitions, and a ingly, final hospital outcomes were not FRACP, FJFICM, Royal Perth Hospital, and School of
Population Health and School of Medicine and Phar-
prospectively constructed data analy- available for some patients. However, macology, University of Western Australia, Perth, Aus-
sis plan. Although only 78% of pa- death after weaning from ECMO or fol- tralia; Michael Wilson, BScMed, FRACS, Royal Prince

1894 JAMA, November 4, 2009—Vol 302, No. 17 (Reprinted) ©2009 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ by Vitor Monteiro on 05/17/2021

 EXTRACORPOREAL MEMBRANE OXYGENATION FOR INFLUENZA A(H1N1)

Alfred Hospital, Sydney, Australia; Marc Ziegenfuss, Bird, Simon Finfer, Carole Foot, Richard Piper, Ray- 10. Australian Bureau of Statistics. Australian demo-
MBBCh, BSc, Dip(PEC), FRCS, FJFICM, Prince Charles mond Raper, Elizabeth Steel); Monash Medical Cen- graphic statistics. http://www.abs.gov.au/ausstats
Hospital, Brisbane, Australia. tre, Melbourne, Australia (Pauline Galt, Craig Walker); /[email protected]/mf/3101.0/. Accessed September 10,
Author Contributions: Dr Davies had full access to all Royal Perth Hospital, Perth, Australia (Andree Gould, 2009.
of the data in the study and takes responsibility for Geraldine McEntaggart, Steve Webb). 11. Statistics New Zealand. Population Indicators.
the integrity of the data and the accuracy of the data Project Support: Siouxzy Morrison, Belinda Howe http://www.stats.govt.nz/methods_and_services
analysis. (Australian and New Zealand Intensive Care Research /access-data/tables/pop-indicators.aspx. Accessed Sep-
Study concept and design: Davies, Jones, Beca, Centre). tember 10, 2009.
Bellomo, Gattas, Granger, Jackson, Nair, Pellegrino, Additional Contributions: We thank Simon Finfer, 12. Conrad SA, Rycus PT, Dalton H. Extracorporeal
Plunkett, Pye, Torzillo, Webb, Wilson, Ziegenfuss. FRACP, FJFICM (George Institute for International Life Support Registry Report 2004. ASAIO J. 2005;
Acquisition of data: Davies, Jones, Beca, Bellomo, Health, University of Sydney, Sydney, Australia), for his 51(1):4-10.
Blackwell, Forrest, Gattas, Granger, Herkes, Jackson, advice and comments during the preparation of the 13. Peek GJ, Moore HM, Moore N, Sosnowski AW,
McGuinness, Nair, Pellegrino, Plunkett, Pye, Torzillo, manuscript; and Ian Seppelt, FANZCA, FJFICM (Nepean Firmin RK. Extracorporeal membrane oxygenation for
Webb, Ziegenfuss. Hospital, Sydney, Australia), for assisting with the eth- adult respiratory failure. Chest. 1997;112(3):759-
Analysis and interpretation of data: Davies, Jones, ics committee approval process. We also thank all the 764.
Bailey, Bellomo, Blackwell, Gattas, Jackson, physicians, nurses, and perfusionists who cared for these 14. Hemmila MR, Rowe SA, Boules TN, et al. Extra-
McGuinness, Pettilä, Torzillo, Webb. complex patients. Drs Finfer and Seppelt did not re- corporeal life support for severe acute respiratory dis-
Drafting of the manuscript: Davies, Jones, Bellomo, ceive any compensation for their contributions. tress syndrome in adults. Ann Surg. 2004;240(4):
McGuinness, Nair, Pye, Torzillo, Webb, Ziegenfuss. 595-605.
Critical revision of the manuscript for important in- REFERENCES 15. Lindén VB, Lidegran MK, Frisén G, Dahlgren P,
tellectual content: Davies, Jones, Bailey, Beca, Bellomo, Frenckner BP, Larsen F. ECMO in ARDS: a long-term
Blackwell, Forrest, Gattas, Granger, Herkes, Jackson, 1. World Health Organization. Influenza-like illness follow-up study regarding pulmonary morphology and
McGuinness, Nair, Pellegrino, Pettilä, Plunkett, Pye, in the United States and Mexico. http://www.who function and health-related quality of life. Acta An-
Torzillo, Webb, Wilson, Ziegenfuss. .int/csr/don/2009_04_24/en/index.html. Accessed aesthesiol Scand. 2009;53(4):489-495.
Statistical analysis: Davies, Bailey. September 10, 2009. 16. Peek GJ, Mugford M, Tiruvoipati R, et al. Effi-
Obtained funding: Bellomo, Herkes. 2. Dawood FS, Jain S, Finelli L, et al; Novel Swine- cacy and economic assessment of conventional ven-
Administrative, technical, or material support: Jones, Origin Influenza A (H1N1) Virus Investigation Team. tilatory support versus extracorporeal membrane oxy-
Beca, Bellomo, Forrest, Gattas, Granger, Jackson, Emergence of a novel swine-origin influenza A (H1N1) genation for severe adult respiratory failure (CESAR):
McGuinness, Nair, Pellegrino, Pettilä, Plunkett, Pye, virus in humans. N Engl J Med. 2009;360(25): a multicentre randomised controlled trial [published
Webb, Wilson. 2605-2615. online September 16, 2009]. Lancet. doi:10.1016
Study supervision: Davies, Beca, Bellomo, Forrest, 3. Perez-Padilla R, de la Rosa-Zamboni D, Ponce de /S0140-6736(09)61069-2.
Gattas, Nair, Pye, Torzillo, Ziegenfuss. Leon S, et al; INER Working Group on Influenza. Pneu- 17. King PT, Rosalion A, McMillan J, Buist M, Holmes
Financial Disclosures: None reported. monia and respiratory failure from swine-origin influ- PW. Extracorporeal membrane oxygenation in
Participating Sites and Investigators (sites in order enza A (H1N1) in Mexico. N Engl J Med. 2009; pregnancy. Lancet. 2000;356(9223):45-46.
of largest number of patients): Royal Prince Alfred Hos- 361(7):680-689. 18. Cunningham JA, Devine PC, Jelic S. Extracorpo-
pital, Sydney, Australia (Paul Forrest, David Gattas, 4. Worldwide H1N1 (Swine Flu) Infection Data. http: real membrane oxygenation in pregnancy. Obstet
Robert Herkes, Brian Plunkett, Dorrilyn Rajbhandari, //www.flucount.org/. Accessed September 10, Gynecol. 2006;108(3 pt 2):792-795.
Caitlin Rees, Paul Torzillo, Michael Wilson); St Vin- 2009. 19. Australian Department of Health and Aging. Aus-
cent’s Hospital, Sydney, Australia (Emily Granger, An- 5. Webb SAR, Seppelt IM; ANZIC Influenza tralian National Influenza Surveillance 2009. http:
drew Jackson, Priya Nair, Roger Pye, Claire Rey- Investigators. Pandemic H1N1 2009 influenza (“swine //www.healthemergency.gov.au/internet
nolds); Auckland City Hospital, Auckland, New Zealand flu”) in Australian and New Zealand intensive care. /healthemergency/publishing.nsf/Content/ozflu2009
( John Beca, Nicola Gini, Shay McGuinness, Rachael Crit Care Resusc. 2009;11(3):170-172. .htm. Accessed September 10, 2009.
Parke); Prince Charles Hospital, Brisbane, Australia 6. Australian Department of Health and Ageing. Pan- 20. Lum ME, McMillan AJ, Brook CW, et al. Impact
(Nikki Blackwell, Angela McCosker, Dan Mullany, Marc demic (H1N1) 2009 update bulletins. http://www of pandemic (H1N1) 2009 influenza on critical care
Ziegenfuss); Alfred Hospital, Melbourne, Australia ( Jas- .healthemergency.gov.au/internet/healthemergency capacity in Victoria. Med J Aust. 2009;191:1-5. http:
min Board, Andrew Davies, Andrew Hilton, Vincent /publishing.nsf/Content/bulletins-13-19-july09. //www.mja.com.au/public/issues/191_09_021109
Pellegrino, Carlos Scheinkestel, Shirley Vallance); Sir Accessed September 10, 2009. /lum10916_fm.pdf. Accessed September 29,
Charles Gairdner Hospital, Perth, Australia (Stuart 7. Call SA, Vollenweider MA, Hornung CA, Simel DL, 2009.
Baker, Brigit Roberts, Paul Woods); Westmead Hos- McKinney WP. Does this patient have influenza? 21. Australian Department of Health and
pital, Sydney, Australia (Rowena Boyd, Peter Clark, JAMA. 2005;293(8):987-997. Ageing. Pandemic (H1N1) 2009 update bulletins.
Vineet Nayyar, Christina Skelly, Eddie Stachowski); St 8. Chow AW, Hall CB, Klein JO, Kammer RB, Meyer http://www.healthemergency.gov.au/internet
George Hospital, Sydney, Australia (Deborah Inskip, RD, Remington JS; Infectious Diseases Society of /healthemergency/publishing.nsf/Content
Doris Lam, John Myburgh, Rebecca Sidoli); Austin Hos- America and the Food and Drug Administration. Evalu- /2FBF38115CD98BB7CA25762A00121475
pital, Melbourne, Australia (Rinaldo Bellomo, Glenn ation of new anti-infective drugs for the treatment of /$File/h1n1-update-20090907.doc. Accessed Sep-
Eastwood, Daryl Jones); Liverpool Hospital, Sydney, respiratory tract infections. Clin Infect Dis. 1992; tember 29, 2009.
Australia (Sharon Micallef, Michael Parr); Princess Al- 15(suppl 1):S62-S88. 22. New Zealand Ministry of Health. Pandemic in-
exandra Hospital, Brisbane, Australia (Meg Harward, 9. Murray JF, Matthay MA, Luce JM, Flick MR. An fluenza (H1N1) 09 swine flu: update 148. http://www
Chris Joyce, Peter Kruger); Royal Children’s Hospital, expanded definition of the adult respiratory distress .moh.govt.nz/moh.nsf/indexmh/influenza-a-h1n1
Melbourne, Australia (Warwick Butt, Melissa Culka); syndrome. Am Rev Respir Dis. 1988;138(3):720- -update-148-070909. Accessed September 29,
Royal North Shore Hospital, Sydney, Australia (Simon 723. 2009.

©2009 American Medical Association. All rights reserved. (Reprinted) JAMA, November 4, 2009—Vol 302, No. 17 1895

Downloaded From: https://jamanetwork.com/ by Vitor Monteiro on 05/17/2021