Update on Extracorporeal Life Support 2004
Heidi J. Dalton, MD,* Peter T. Rycus, MPH,† and Steven A. Conrad, MD, PhD‡
Since its beginnings in 1989, the Extracorporeal Life Support Organization (ELSO) Registry
has collated and reported data on over 30,000 patients. The majority of patients entered
into the Registry have been neonates with respiratory failure from meconium aspiration,
persistent pulmonary hypertension, or congenital diaphragmatic hernia. These patients
suffer from refractory hypoxemia; thus, this supportive technique came to be called “Extracorporeal Membrane Oxygenation (ECMO)” for its ability to provide excellent gas
exchange. With advances in prevention, diagnosis, and treatment measures for neonatal
respiratory failure, need for ECMO support has fallen from the peak of 1500 cases in the
early 1990s to 800 cases annually. Sixty-six percent (over 19,000) of patients in the
Registry are under the category of neonatal respiratory failure, with a 77% overall survival
reported to discharge. The success of neonatal ECMO has led to expansion of the field to
pediatric, cardiac, and adult patients. An average of 200 pediatric patients receive ECMO
for respiratory failure per year with an overall survival of 55%. Adult respiratory failure
patients form a smaller group, with less than 100 cases reported to the ELSO registry per
year. Survival mirrors that noted in the pediatric ECMO population. The application of
ECMO or related techniques continues to increase for cardiac failure across all age groups.
Overall survival in cardiac patients ranges from 33% to 43%. A novel form of extracorporeal
support is “ECPR” or ECMO during cardiac arrest. Bypass circuits and equipment can be
set up and instituted within a very short period of time in this circumstance, thus the name
“rapid deployment ECMO” has become associated with this form of support. Overall
survival in the near-600 patients placed on ECMO during resuscitation is 40%.
Semin Perinatol 29:24-33 © 2005 Elsevier Inc. All rights reserved.
KEYWORDS extracorporeal membrane oxygenation (ECMO), extracorporeal life support organization (ELSO), extracorporeal life support (ECLS)
I
n September of 2004, the Extracorporeal Life Support Organization (ELSO) celebrated its 15th anniversary by honoring Dr. Bob Bartlett, the founding father of extracorporeal
membrane oxygenation. One of the highlights of this anniversary gathering was the personal testimony offered by individuals who have benefited from extracorporeal life support. Chief among these was a woman named Esperanza.
Esperanza, who is now a grown woman with children of her
own, was the first neonate to receive extracorporeal membrane oxygenation for respiratory failure. Esperanza, which
means “hope,” was an abandoned infant with newborn respiratory failure failing conventional therapy. Dr. Bob Bartlett
*Pediatric Intensive Care Unit, Children’s National Medical Center, Professor of Pediatrics, The George Washington University School of Medicine,
Washington, DC.
†Extracorporeal Life Support Organization, Ann Arbor, MI.
‡Health Sciences Center, Louisiana State University, Shreveport, LA.
Address reprint requests to Heidi J. Dalton, MD, Children’s National Medical
Center, 111 Michigan Ave NW, Suite 100; 3 W, Washington, DC 20010.
E-mail: hdalton@cnmc.org
24
0146-0005/05/$-see front matter © 2005 Elsevier Inc. All rights reserved.
doi:10.1053/j.semperi.2005.02.005
and others were involved with laboratory modification of
intraoperative cardiopulmonary bypass into a mode that
could be used to support children with severe respiratory
failure for several days while their underlying disease was
resolving. This technique was called “Extracorporeal Membrane Oxygenation” and was abbreviated as “ECMO.”1 When
consulted about this neonate who was dying from respiratory
failure, work from the laboratory was rapidly transformed
into a “bench to bedside” effort. The rest, as they say, is
history.
The first few successes with ECMO support were tempered
by equal numbers of failures and complications. As a forum
to discuss these events and the early experiences with extracorporeal support, informal meetings were held with surgeons, neonatologists, intensivists, biomedical engineers,
nurses, and perfusion personnel with interest in the field. To
facilitate inclusion of all patients who received extracorporeal
support, the term “ECMO” became interchangeable with
“ECLS,” which stands for extracorporeal life support. From
these informal gatherings, a database for tracking in whom
Update on extracorporeal life support
25
Table 1 Total Numbers of ECLS Cases Reported by the ELSO
Registry International Summary, July 2004
Group
Neonatal
Respiratory
Cardiac
ECPR
Pediatric
Respiratory
Cardiac
ECPR
Adult
Respiratory
Cardiac
ECPR
Total
Total
Cases
Survive to
DC (no.)
Transfer
(%)
19,061
2,215
151
14,681
841
65
77
38
43
2,762
2,936
282
1,536
1,256
111
56
43
39
972
474
132
28,985
515
156
50
19,211
53
33
38
66
and how extracorporeal life support was delivered was developed. From its beginning forum of about 30 individuals,
ELSO has grown to several hundred participants that represent over 100 international centers. Members get together at
the annual ELSO conference to share experiences and provide networking opportunities. Vendors and representatives
from research and development companies related to bypass
and extracorporeal support also interact directly with participants to hear concerns over equipment and plan future research endeavors. Through interactions between business
personnel and practitioners in the field, advances in the circuitry for ECLS have been made.
One of the most vital activities of ELSO is to maintain and
share data from the Registry. To date, over 30,000 patients
have been entered into the ELSO database. Periodic reports
from the ELSO Registry are provided to member centers
which give both global and center-specific data. Annual fees
from the member centers help maintain the database and
information delivery systems, although much of the leadership roles and other activities of the organization remain
largely volunteer efforts. Members have also worked to create, edit, and distribute educational material such as the Extracorporeal Life Support Specialists manual and the Extracorporeal Life Support textbook. These informative texts are
written, produced, and distributed by ELSO. Guidelines for
training and application of ECMO techniques have also been
developed and published from ELSO and its participants.
Data reported to ELSO include basic patient descriptive
information, perinatal information (for neonates), pre-ECLS
physiologic data, ECLS equipment and implementation data,
complications (mechanical and patient related), and basic
outcome information. With the continuing increase in the
application of ECLS techniques to children with cardiac disease, an addendum that tracks more specific data on cardiac
failure patients was created in 2001. From this addendum,
more specific information regarding experiences with ECLS
support in these complex children will be available. It is
hoped that these data will facilitate determination of outcome
predictors for this patient population. In a similar manner, an
addendum for patients placed on ECMO during cardiac arrest has also been recently created to more specifically track
the occurrence and outcomes associated with this novel use
of extracorporeal resuscitation.
The following overview of ELSO registry data represents a
look at the history of extracorporeal support to date and
illustrates the changing environment that is the current state
of ECLS. The data provided were obtained from the July
2004 ELSO International Summary. The overall experience
with ECLS across age and support category is shown in Table
1 and Fig. 1.
Neonatal Registry Summary
The majority (66%) of patients are neonatal, with an overall
survival of 77% (Table 1). Neonatal ECLS has shown a progressive decline in the number of patients treated per year.
Peaking at about 1500 cases in the early 1990s, recent years
have seen an average of 800 patients treated per year (Fig. 2).
These changes may reflect better prenatal care and perinatal
preventive medicine as well as the availability of alternative
therapies for support of neonatal respiratory failure, such as
high frequency ventilation, inhaled nitric oxide, and surfactant. In particular, randomized studies of inhaled nitric oxide
in neonatal respiratory failure and pulmonary hypertension
noted that response to inhaled nitric oxide obviated the need
for ECMO support in about one-third of patients.2,3 There
has been concern that the availability of alternative therapies
leads to a delay in ECMO institution and may be responsible
in part for the poorer outcome noted in recent years of neonatal ECMO. Survival between 1995 and 2003 has fallen
from 76% in 1995 to 62% in 2003 (P ⫽ 0.0002 with 8 df by
Chi Square). Limited comparison data on severity of illness
between patients treated in 1995 or 2003 are available from
within the Registry, and comparison data to track equally ill
patients treated with other methods than ECMO are unavailable. Thus, there is little “hard evidence” to show that the
Figure 1 ECLS cases reported to the ELSO Registry as of July 2004.
The bar graph represents the number of cases reported on an annual
basis, while the line graph represents the cumulative cases reported
as of the end of each year. A steady decrease in the number of cases
reported annually has been occurring since 1992.
H.J. Dalton, P.T. Rycus, and S.A. Conrad
26
Table 2 Extracorporeal Life Support for Neonatal Respiratory
Failure (July 2004)
Primary
Diagnosis or
Mode
Figure 2 Neonatal respiratory ECLS cases reported to the ELSO registry as of July 2004. The bar graph represents the number of cases
reported on an annual basis, while the line graph represents the
cumulative cases reported as of the end of each year. A steady
decrease in the number of cases reported annually has been occurring since 1992. The figure for 2004 is lower due to delays in
reporting of cases.
decline in neonatal ECMO survival is due to “sicker” patients
receiving ECMO. It is true that the “simple” neonatal patient
with meconium aspiration, an entity with a high rate of survival with ECMO, is found less frequently in the ELSO registry than in earlier years (Fig. 3). The presence of comorbidities and congenital diseases may also affect survival. As an
example, patients with trisomy 21 who receive ECMO support have been shown to have decreased survival to hospital
discharge when compared with infants with similar disease
but without trisomy 21 who received ECMO support (66%
versus 96%, P ⫽ 0.03, risk ratio ⫽ 1.6 with 95% CI 1.042.5).4 Patients with congenital diaphragmatic hernia also
form a larger proportion of neonatal ECMO patients compared with earlier years. These patients also have decreased
Figure 3 Neonatal respiratory ECLS cases reported to the ELSO registry as of July 2004. The bar graph represents the percent of cases
by diagnosis reported on an annual basis.
Neonatal cases by
diagnosis
CDH
MAS
PPHN/PFC
Infant RDS
Sepsis
Other
Neonatal mode of
ECLS
VA
VV
VVDL
VA(ⴙV)
VV ¡ VA
VVDL ⴙ V
Total
Cases
Number
Surviving
%
Surviving
4,491
6,560
2,914
1,380
2,384
1,567
2,367
6,160
2,287
1,161
1,794
1,003
53
94
78
84
75
64
13,301
276
3,537
1,159
544
410
9,882
220
3,053
868
360
346
74
80
86
75
66
84
CDH, congenital diaphragmatic hernia; MAS, meconium aspiration
syndrome; PPHN/PFC, persistent fetal circulation; RDS, respiratory distress syndrome; VA, venoarterial; VV, venovenous;
VVDL, venovenous double lumen.
survival as compared with other groups. Finally, more patients in recent years fall into the category of “other,” which
often means they have unusual disorders outside the traditional diseases which with ECMO has been associated in the
past.
The diagnoses and modes of support are summarized in
Table 2. Venoarterial access remains the most common mode
of support in neonatal respiratory failure, but the number
managed with venovenous access using a double lumen catheter (VVDL) has grown to over 20% of cases. The survival rate
with the VVDL catheter is higher than that reported with VA
support, but whether this reflects a difference in severity of
illness between patients or is a result of the mode of support
is unknown. Traditionally, venovenous ECMO has been
avoided in patients requiring inotropic therapy because of
concerns of inadequate cardiac support with this method of
ECMO. One recent review of neonates requiring inotropic
support before ECMO, however, found that 86% of these
patients were treated with venovenous ECMO and overall
survival was 84%.5 Of 14% of patients who were treated with
venoarterial ECMO, survival was 75%. An inotrope score (a
composite based on the type and dosage of vasoactive agents
required) was calculated for each patient and considered “significant” if a score greater than 10 was required. The majority
of patients had scores ⬎10 which fell to a nonsignificant level
within 24 hours of ECMO regardless of the mode of support.
The report suggested that only patients in whom the preECMO inotrope score was ⬎100 or who could not physically
be cannulated in a venovenous mode should preferentially
receive venoarterial ECMO as the initial cannulation route.
Thus, the application of venovenous ECMO to patients who
have cardiac compromise may be a viable option. Reasons for
Update on extracorporeal life support
27
Figure 4 Pediatric respiratory ECLS cases reported to the ELSO registry as of July 2004. The bar graph represents the
number of cases reported on an annual basis, while the line graph represents the cumulative cases reported as of the end
of each year. The number of annual cases reported continues to grow, but may be reaching a plateau. The figure for
2004 is lower due to delays in reporting of cases.
the “success” of venovenous ECMO in such patients may be
related to the role of ventilator management creating cardiac
compromise in the pre-ECMO period. Once adequate oxygenation and ventilation are established with ECMO and the
degree of ventilator support can be reduced, cardiac function
may rapidly improve. Additionally, venovenous ECMO provides well-oxygenated blood directly to the pulmonary bed,
which may reduce pulmonary hypertension and improve
right ventricular output. Well-oxygenated blood from venovenous ECMO circulates through the pulmonary bed and
returns to be ejected from the left ventricle. Data from microsphere studies have shown that, even during venoarterial
ECMO with a cannula in the ascending aortic arch, the majority of coronary blood flow is provided by native left heart
ejection and not from arterial ECMO return. Thus, coronary
perfusion and improved myocardial oxygen delivery may be
better during venovenous ECMO than with venoarterial
ECMO. All these factors may permit the successful use of
venovenous ECMO in patients with presumed cardiac dysfunction. Another aspect of venovenous cannulation which
makes logical sense is the avoidance of the need to instrument the carotid artery. This may reduce neurologic complications or the risk for stroke later in life, although none of
these theoretical advantages have yet been proven. While
debate over the use of venoarterial or venovenous ECMO still
exists, there is definitely a movement to use venovenous
ECMO in patients whenever possible.
Pediatric Registry Summary
Pediatric ECMO patients include those who are over the age
of 30 days and less than 18 years. The number of pediatric
patients supported for respiratory failure on an annual basis
since inception of the Registry is given in Fig. 4. Since 1994,
about 200 pediatric patients are treated yearly with ECMO
for respiratory failure (Fig. 4). Survival remains relatively
stable at 55% (Table 3). Diagnosis does not appear to have a
major impact on survival in this group. The most common
mode of support in pediatric patients remains venoarterial
(Table 3). Lack of double-lumen single cannulas large
enough to support older children and insufficient size of
femoral vessels for venous cannulation are cited as reasons
for the predominance of venoarterial cannulation. Experience is showing, however, that children can tolerate femoral
venous access at ages as low as 5 or 6 years (or lower in some
situations), and this may shift the support mode in the future.
Venovenous ECMO has accounted for one-third of pediatric
respiratory cases in the past year.
Perhaps the biggest change that has occurred over time in
pediatric ECMO is the expansion to patient groups who
would have been excluded from ECMO support in years past.
Recent reports of successful treatment with ECMO has been
described in patients with trauma, immunosuppression,
burns, underlying bleeding disorders (hemophilia), established multiple organ systems failure.6-9 These patients are a
H.J. Dalton, P.T. Rycus, and S.A. Conrad
28
Table 3 Extracorporeal Life Support for Pediatric Respiratory
Failure (July 2004)
Primary Diagnosis
or Mode
Pediatric cases by
diagnosis
Bacterial pneumonia
Viral pneumonia
Aspiration pneumonia
ARDS
ARF, non-ARDS
Other
Pediatric mode of ECLS
VA
VV
VVDL
VA(ⴙV)
VV ¡ VA
VVDL ⴙ V
Total
Cases
Number
Surviving
%
Surviving
290
728
168
348
605
671
157
457
110
188
286
359
54
63
65
54
47
54
1,663
510
283
89
163
44
851
328
200
42
74
32
51
64
71
47
45
73
ARF, acute respiratory failure; ARDS, acute respiratory distress
syndrome; see Table 2 for mode abbreviations.
far cry from the early days of pediatric ECMO, where healthy
children with an overwhelming pneumonia or viral illness
such as respiratory syncytial disease formed the majority of
patients treated with ECMO. Today, as an outgrowth of years
of experience, better equipment, and innovation in treatment, the range of patients who receive ECMO is extremely
varied. One example of these changes is in the approach to
the patient with sepsis and multiple organ failure. While
these characteristics would likely have excluded patients
from ECMO consideration a few years ago, the use of ECMO
in sepsis is now an accepted therapy. In fact, ECMO is part of
the algorithm in the new PALS and SCCM-AAP guidelines for
hemodynamic support of patients with severe septic shock
for patients with catecholamine-resistant shock.10 Recent
case series and reports which illustrate the successful use of
ECMO in “unusual” circumstances, the availability of newer
circuitry and equipment that reduces the need for systemic
anticoagulation, and the general increase in use of other extracorporeal therapies such as renal replacement, plasma exchange, and plasmapheresis in the pediatric population are
likely factors in the revived interest of ECMO in the pediatric
population. Whether this will relate into increasing numbers
of patients treated with ECMO will only be seen as the future
unfolds. Another factor which may play a role in the use of
ECMO in pediatric respiratory failure is related to the interpretation of data relating outcome to the many other lessinvasive modes of therapy. The mantra in pediatric respiratory failure for the past few years has been the striking
improvement in survival which has been noted. Mortality
with severe respiratory failure is often quoted as 10% to 15%,
reduced from 40% to 50% in the early 1990s. On closer
inspection, however, it is unclear whether this improvement
relates only to healthy children with an acute illness or is
applicable to the pediatric population as a whole. Studies
which have included a wide variety of patients with severe
respiratory failure continue to find mortality rates from 25%
to 45%.11,12 In addition, randomized studies which have
evaluated frequently used modalities such as inhaled nitric
oxide, high frequency ventilation, and prone positioning
have not shown a significant improvement in outcome in
pediatric patients when these techniques are compared with
conventional mechanical ventilation. Whether these results
will influence the use of ECMO in pediatrics in the future is
unknown.
Adult Registry Summary
Adults remain potentially the most underserved group in
terms of ECMO support. Only about 100 adults with respiratory failure are reported to the ELSO Registry each year.
Overall survival is remaining relatively stable at around 53%
(Table 4). The best outcomes appear to be with viral pneumonia, aspiration pneumonia, and acute respiratory failure.
The lingering effects of several adult trials of ECMO which
showed no benefit from ECMO have left many clinicians with
little interest in pursuing ECMO support for patients failing
conventional care. A current-era trial of ECMO in adult patients is underway in Europe, with results expected within
the next year. The lack of available adult ECMO centers is
another factor in the lack of adult ECMO representation. A
newly funded NIH grant to develop and implement large
double-lumen single cannulas for adult ECMO support may
provide an easier route for cannulation in large patients.
Whether these events will lead to an increase in the use of
ECMO in the adult patient will only be seen with time. Expansion of many previously exclusively neonatal or pediatric
ECMO programs to the adult patient is underway in some
ECMO centers.
Cardiac Registry Summary
The largest area of growth in application of ECMO has
undoubtedly occurred in the cardiac population (Fig. 5).
While the majority of patients are those with congenital
heart disease in the postoperative period, patients with
myocarditis, cardiomyopathy, and other forms of cardiovascular collapse have also received ECLS support. The
breakdown of cardiac patients from the Registry is shown
Table 4 Extracorporeal Life Support for Adult Respiratory Failure (July 2004)
Primary Diagnosis
Adult cases by diagnosis
Bacterial pneumonia
Viral pneumonia
Aspiration pneumonia
ARDS, postop/trauma
ARDS, not
postop/trauma
ARF, non-ARDS
Other
Total Number
%
Cases Surviving Surviving
186
87
32
132
196
97
54
18
68
100
52
62
56
52
51
55
317
35
154
64
49
ARDS, acute respiratory distress syndrome.
Update on extracorporeal life support
29
Figure 5 Cardiac ECLS cases per year. Bar is broken into number of deaths (black bars) and number of survivors (white
bars). Adapted from the ECLS Organization International Registry, July 2004.
in Table 5. One of the reasons for the increase in cardiac
ECMO has been the increasing complexity of cardiac repairs now undertaken in small infants. To date, infants and
small children have not had other forms of cardiac assist
devices which could be used, so ECLS has remained the
“default” technique. As ventricular assist devices become
miniaturized, they may also offer another modality for
pediatric cardiac support.
Overall survival for neonatal cardiac cases has been declining slightly, likely as a result of ECMO being applied to more
and more complex cardiac diseases such as hypoplastic left
heart patients. The vast majority of patients are infants
treated after repair of congenital heart disease. Other information regarding surgical types or diagnoses is broken down
by age in Tables 6 and 7. Due to the need to provide full
cardiac support in these patients, the majority are cannulated
via venoarterial access, either cervically through the right
internal jugular vein and the right common carotid artery or
directly through an open mediastinum into the right atrium
and aorta. One alteration in traditional ECMO with cardiac
patients such as hypoplastic left heart patients is that, since
respiratory function may be normal, there is no specific need
for a membrane oxygenator in the system to provide gas
exchange. Removal of the oxygenator simplifies the circuit
and may reduce the amount of anticoagulation needed, but it
also eliminates a site of air bubble trapping that can be an
important safety consideration. This form of extracorporeal
support has become colloquially known as “NOMO,” or “nooxygenator membrane oxygenation.”13 Alternatively, if the
patient is in complete cardiac failure and full heart and lung
support is being provided by the ECLS circuit, native lung
function may be unnecessary. Consideration in these patients, who are often bridging to heart transplant and require
prolonged durations of ECLS support, may be given to extu-
Table 5 Extracorporeal Life Support for Cardiac Failure (July
2004): Cardiac Runs by Diagnosis
Age Group: 0–30 days
Congenital Defect
Cardiac Arrest
Cardiogenic Shock
Cardiomyopathy
Myocarditis
Other
Age Group: 31 days
and < 1 year
Congenital Defect
Cardiac Arrest
Cardiogenic Shock
Cardiomyopathy
Myocarditis
Other
Age Group 1 year and
< 16 years
Congenital Defect
Cardiac Arrest
Cardiogenic Shock
Cardiomyopathy
Myocarditis
Other
Age Group 16 years
and over
Congenital Defect
Cardiac Arrest
Cardiogenic Shock
Cardiomyopathy
Myocarditis
Other
Total
Runs
Survived
%
Survived
2,006
24
22
72
27
168
719
5
11
48
11
74
36
21
50
67
41
44
1,318
25
11
61
33
155
548
6
4
28
18
65
42
24
36
46
55
42
740
49
34
200
96
260
297
19
11
108
60
113
40
39
32
54
63
43
42
43
74
73
14
297
12
9
34
25
9
92
29
21
46
34
64
31
H.J. Dalton, P.T. Rycus, and S.A. Conrad
30
Table 6 Cardiac Runs by Surgical type
Age Group: 0–30 days
Cardiac transplant
Other postop, not bridged
Other postop, bridged to transplant
Not postop, not bridged
Not postop, bridged to transplant
Age Group: 31 days and < 1 year
Cardiac transplant
Other postop, not bridged
Other postop, bridged to transplant
Not postop, not bridged
Not postop, bridged to transplant
Age Group 1 year and < 16 years
Cardiac transplant
Other postop, not bridged
Other postop, bridged to transplant
Not postop, not bridged
Not postop, bridged to transplant
Age Group 16 years and over
Cardiac transplant
Other postop, not bridged
Other postop, bridged to transplant
Not postop, not bridged
Not postop, bridged to transplant
Total Runs
Survived
% Survived
38
1,700
31
536
14
12
611
12
229
4
32
36
39
43
29
79
1,215
26
258
25
40
487
9
122
11
51
40
35
47
44
152
759
44
314
110
80
301
22
144
61
53
40
50
46
55
74
261
21
162
25
25
83
10
54
9
34
32
48
33
36
bating the patient to reduce complications such as ventilatorassociated pneumonia and decrease the need for heavy sedation.
Based on the success of ECMO in supporting cardiac patients, there has been a paradigm shift over the past few years.
While once applied only in the most desperate cases, ECMO
is now applied earlier in cardiac dysfunction and to a greater
variety of patients. One group which represents a novel but
increasingly important segment of ECLS is the category of
“ECPR,” or ECMO during cardiopulmonary resuscitation. Although the successful use of ECMO during cardiac arrest was
described over 10 years ago, it has found renewed enthusiasm as more reports of successful outcomes have appeared in
the literature.14,15 To decrease the time it takes to implement
extracorporeal support in arrest or acutely deteriorating situations, many centers now maintain systems which can be
rapidly deployed for ECLS support. Some centers utilize a
regular roller-head, silicone membrane lung system that is
saline-primed and kept sterile for up to 30 days. Others use a
centrifugal pump and hollow fiber oxygenator circuit which
can be ready for use within minutes. These types of setups
decrease dramatically the priming times required for “traditional” roller-head, silicone membrane systems. Past efforts
in initiating ECLS during arrest frequently noted severe neurologic injury due to prolonged periods required to set up
and implement ECLS support. To date, some 565 patients
have received ECLS with an overall survival of 40%. Longterm neurologic outcomes in these patients are not available
as yet, but preliminary results are encouraging.
Complications
A valuable role of the Registry is to permit benchmarking of
individual centers’ performance. Each ELSO participating
center receives periodic reports of their own experience as
well as those of the ELSO Registry in total. This allows each
center to evaluate outcomes, complications, and trends and
compare them with the national experience. The Registry
provides information not only on demographics but also on
complications. The table reports the incidence of the complications in percent of total for that group and the survival in
percent in patients experiencing those complications. The
complication list is a subset of all of the complications reported, separated into mechanical events which occur within
the ECLS circuit and to patient-related complications (Tables
8 and 9). Hemorrhage (gastrointestinal, cannula site, and
surgical site) remains the major problem associated with
ECLS and appears to be associated with a poorer outcome.
Neurological complications are less frequent but are often
associated with poor outcome.
History of the
ELSO Registry Database
In 1996, the ELSO Steering Committee approved the purchase of computer hardware for implementation of a new
registry database. The new database was implemented using
a relational database system (Microsoft Access 97®). A single
database was constructed, into which the four traditional
Update on extracorporeal life support
31
Table 7 Cardiac Congenital Diagnoses
Age Group: 0–30 days
Left to right shunt (ASD/VSD/PDA/AV canal/AVSD/ECD)
Left-sided obstructive (aortic stenosis/mitral stenosis/coarctation)
Hypoplastic left heart
Right-sided obstructive (pulmonary stenosis/pulmonary or tricuspid
Cyanotic incr. pulmonary flow (truncus arteriosis/TGA/TGV)
Cyanotic incr. pulm. Congestion (TAPVR/PAPVR)
Cyanotic decr. pulmonary flow (TOF/DORV/Ebstein’s)
Other
Age Group: 31 days and < 1 year
Left to right shunt (ASD/VSD/PDA/AV canal/AVSD/ECD)
Left-sided obstructive (aortic stenosis/mitral stenosis/coarctation)
Hypoplastic left heart
Right-sided obstructive (pulmonary stenosis/pulmonary or tricuspid
Cyanotic incr. pulmonary flow (truncus arteriosis/TGA/TGV)
Cyanotic incr. pulm. Congestion (TAPVR/PAPVR)
Cyanotic decr. pulmonary flow (TOF/DORV/Ebstein’s)
Other
Age Group 1 year and < 16 years
Left to right shunt (ASD/VSD/PDA/AV canal/AVSD/ECD)
Left-sided obstructive (aortic stenosis/mitral stenosis/coarctation)
Hypoplastic left heart
Right-sided obstructive (pulmonary stenosis/pulmonary or tricuspid
Cyanotic incr. pulmonary flow (truncus arteriosis/TGA/TGV)
Cyanotic incr. pulm. Congestion (TAPVR/PAPVR)
Cyanotic decr. pulmonary flow (TOF/DORV/Ebstein’s)
Other
Age Group 16 years and over
Left to right shunt (ASD/VSD/PDA/AV canal/AVSD/ECD)
Left-sided obstructive (aortic stenosis/mitral stenosis/coarctation)
Hypoplastic left heart
Right-sided obstructive (pulmonary stenosis/pulmonary or tricuspid
Cyanotic incr. pulmonary flow (truncus arteriosis/TGA/TGV)
Cyanotic decr. pulmonary flow (TOF/DORV/Ebstein’s)
Other
categories of neonatal, pediatric, cardiac, and adult support
were merged. The concept of “reason for going on ECLS” was
eliminated, since it was highly subjective, and has been replaced by physiologic data. The small list of diagnosis fields
has been replaced by standardized coding using ICD-9 (In-
atresia)
atresia)
atresia)
atresia)
Total Runs
Survived
% Survived
89
180
389
85
102
343
235
583
31
50
98
37
34
149
80
240
35
28
25
44
33
43
34
41
326
100
112
54
36
58
102
530
132
36
42
23
17
19
38
241
40
36
38
43
47
33
37
45
151
87
38
56
6
3
97
302
59
40
13
24
1
1
38
121
39
46
34
43
17
33
39
40
4
9
2
2
1
7
17
2
1
0
0
0
4
5
50
11
0
0
0
57
29
ternational Classification of Diseases, 9th Revision) codes.16
A single primary and multiple secondary diagnosis codes are
now allowed. Procedures are now coded separately from diagnoses using CPT (Current Procedural Terminology)
codes.17 Free text entries of equipment and cannula type have
Table 8 Mechanical and Patient-Related Complications for Respiratory Population
Complication
Mechanical
Oxygenator failure
Tubing rupture
Pump malfunction
Cannula problems
Patient-related
GI hemorrhage
Cannula site bleeding
Surgical site bleeding
Hemolysis
Brain death
Seizures: clinically determined
*Table entries are in % reported (% survival).
Neonatal
Respiratory
Pediatric
Respiratory
Adult
Respiratory
5.7 (55)*
0.7 (74)
1.8 (68)
11.1 (70)
13.8 (44)
3.8 (47)
3.1 (47)
14.2 (48)
18.2 (43)
4.0 (30)
4.1 (37)
10.7 (42)
1.7 (46)
6.1 (68)
6.1 (46)
12.2 (68)
1.0 (0)
10.9 (62)
4.0 (25)
9.2 (60)
16.0 (47)
8.8 (42)
6.0 (0)
7.3 (35)
4.3 (26)
11.5 (47)
22.4 (35)
5.3 (28)
3.8 (0)
2.0 (45)
H.J. Dalton, P.T. Rycus, and S.A. Conrad
32
Table 9 Mechanical and Patient-Related Complications for the Cardiac Population
Complication
Mechanical
Oxygenator failure
Tubing rupture
Pump malfunction
Cannula problems
Patient-related
GI hemorrhage
Cannula site bleeding
Surgical site bleeding
Hemolysis
Brain death
Seizures: clinically determined
0–30
days
31 days and
< 1 year
1 year and
< 16 years
16 years and
over
7.2 (23)
0.7 (31)
1.3 (32)
6.7 (33)
7.2 (28)
1.1 (24)
1.9 (26)
5.9 (35)
9.1 (37)
2.0 (30)
2.2 (42)
6.4 (31)
16.4 (27)
0.9 (20)
1.8 (36)
6.8 (32)
0.9 (5)
6.8 (27)
31.0 (29)
10.8 (24)
1.3 (0)
9.7 (29)
1.8 (14)
6.7 (23)
33.9 (36)
9.9 (33)
5.1 (0)
11.0 (24)
2.8 (23)
10.7 (44)
31.3 (42)
8.5 (35)
9.5 (0)
6.8 (21)
2.4 (15)
12.9 (30)
31.9 (27)
8.1 (34)
7.9 (0)
4.8 (12)
been replaced with a standardized listing to assure uniformity. Time intervals (eg, time on ECLS) are no longer reported, as these are now calculated from event times (eg, start
and end times of ECLS). These changes were designed to
improve the robustness of the data.
The new Registry database is designed to support automated data entry. The Case Report Form was implemented in
Microsoft Word 97. The form can be printed and filled in
manually, as has been done in years past. Additionally, however, the document can be filled in directly using Microsoft
Word and then sent electronically to ELSO. The Registry
database can import the field data directly from this Word
form, removing any need for double data entry. In addition,
the form can be printed with its information, allowing for
hard copies of the form to be maintained as records. If used
exclusively, this approach can virtually eliminate data entry
delays, while significantly reducing the potential for errors.
In early 2000, it was decided to move the database for the
ELSO Registry to a new system. Instead of an Access database, the data were moved to Microsoft’s SQL Server. This
change allowed for better security, less of a risk for data
corruption, and future expansions. Also, SQL Server is designed to integrate with Microsoft’s Internet Information
Server which the ELSO Web site is run under.
In the summer of 2001, ELSO introduced a two-page cardiac addendum to supplement the ECLS Case Report Form.
It was apparent that the current information being captured
was not sufficient for cardiac cases. ICD-9 diagnoses and CPT
procedure codes were not describing cardiac cases in enough
detail. The ECLS Case Report Form was also modified to
become HIPAA compliant.
Another recent goal of ELSO is to allow for data entry is
submission over the Internet using World Wide Web-based
data entry forms. A Web server has been established (http://
www.elso.med.umich.edu) which is capable of receiving and
storing form-based data. At this time, ELSO is developing a
Web-based application with plans to provide complete ELSO
Case Report Form data entry. Internet-based data entry is
particularly targeted to facilitate data entry from international
sites. With security features added, it will be possible for
individual centers to execute simple queries directly from the
ELSO registry over the Internet, providing for timely access
to information. Revisions to current database information to
track specific research-related questions or identified areas
which need additions or elimination are ongoing goals for the
Registry. Finally, the ability to utilize the infrastructure and
reliability of ELSO to maintain data on other forms of extracorporeal support such as ventricular assist devices or therapies such as inhaled nitric oxide are active areas of discussion
and interest within ELSO.
Summary
The field of ECLS is currently in a state of flux. Many patients
denied ECMO support in this past are now being considered
for ECMO support and obtaining long-term survival. The
experience and knowledge gained over the past 15 years of
ECMO has resulted in making this therapy more accessible,
safer, and efficient. The revised interest in use of ECLS in
cardiac arrest, adult patients, and other populations may herald an increase in the use of extracorporeal life support in
future days. The ELSO registry continues to adapt to current
needs and provides invaluable information on patients
treated, outcome, and noted complications. It remains a
comprehensive and valuable resource to the ECLS community. With the completion of the re-engineering process currently underway, the Registry will have the capability of more
rapid and error-free data entry via electronic submission,
access from domestic and international centers via the Internet for timely reporting of data, and automation in verification of data. These advances, coupled with efforts to include
patients receiving other forms of mechanical support, such as
ventricular assist devices, will enhance the ability to evaluate
outcomes in a wider variety of patients. Outreach to maintain
database information on patients treated with other therapies
such as inhaled nitric oxide as a basis for comparison between patients who receive such treatment and then require
ECMO or not continues as well. The Extracorporeal Life Support Organization continues to provide a forum for growth
and development in this exciting field of medical care.
Update on extracorporeal life support
33
References
1. Bartlett RH, Gazzaniga AB, Jefferies MR, et al: Extracorporeal membrane oxygenation (ECMO) cardiopulmonary support in infancy.
Trans Am Soc Artif Intern Organs 22:80-93, 1976
2. Clark RH, Kueser TJ, Walker MW, et al: Low-dose nitric oxide therapy
for persistent pulmonary hypertension of the newborn. Clinical Inhaled Nitric Oxide Research Group. N Engl J Med 342:469-474, 2000
3. Christou H, Van Marter LJ, Wessel DL, et al: Inhaled nitric oxide reduces the need for extracorporeal membrane oxygenation in infants
with persistent pulmonary hypertension of the newborn. Crit Care Med
28:3722-3727, 2000
4. Southgate WM, Annibale DJ, Hulsey TC, et al: International experience
with trisomy 21 infants placed on extracorporeal membrane oxygenation. Pediatrics 107:549-552, 2001
5. Roberts N, Westrope C, Pooboni SK, et al: Venovenous extracorporeal
membrane oxygenation for respiratory failure in inotrope dependent
neonates. ASAIO J 49:568-571, 2003
6. Fortenberry JD, Meier AH, Pettignano R, et al: Extracorporeal life support for posttraumatic acute respiratory distress syndrome at a children’s medical center. J Pediatr Surg 38:1221-1226, 2003
7. Szocik J, Rudich S, Csete M: ECMO resuscitation after massive pulmonary embolism during liver transplantation. Anesthesiology 97:763764, 2002
8. Sheridan RL, Schnitzer JJ: Management of the high risk pediatric burn
patient. J Pediatr Surg 36:1308-1312, 2001
9. Thiagarajan RR, Roth SJ, Margossian S, et al: Extracorporeal membrane
oxygenation as a bridge to cardiac transplantation in a patient with
10.
11.
12.
13.
14.
15.
16.
17.
cardiomyopathy and hemophilia A. Intensive Care Med 29:985-988,
2003
Carcillo JA, Fields AI, American College of Critical Care Medicine Task
Force Committee Members: Clinical practice parameters for hemodynamic support of pediatric and neonatal patients in septic shock. Crit
Care Med 30:1365-1378, 2002
Dobyns EL, Cornfield DN, Anas NG, et al: Multicenter randomized
controlled trial of the effects of inhaled nitric oxide therapy on gas
exchange in children with acute hypoxemic respiratory failure. J Pediatr 134:406-412, 1999
Arnold JH, Hanson JH, Toro-Figuero LO, et al: Prospective, randomized comparison of high-frequency oscillatory ventilation and conventional mechanical ventilation in pediatric respiratory failure. Crit Care
Med 22:1530-1539, 1994
Darling EM, Kaemmer D, Lawson DS, et al: Use of ECMO without the
oxygenator to provide ventricular support after Norwood Stage I procedures. Ann Thorac Surg 71:735-736, 2001
Dalton HJ, Siewers RD, Fuhrmana BP, et al: Extracorporeal membrane
oxygenation for cardiac rescue in children with severe myocardial dysfunction. Crit Care Med 21:1020-1028, 1993
Morris MC, Wernovsky G, Nadkarni VM: Survival outcomes after extracorporeal cardiopulmonary resuscitation instituted during active
chest compressions following refractory in- hospital cardiac arrest. Pediatr Crit Care Med 5:440-446, 2004
American Medical Association: International Classification of Diseases
(9th Revision): Clinical Modification (ICD-9-CM 1998). Chicago, IL,
AMA, 1997
American Medical Association: Physicians’ Current Procedural Terminology (CPT 98). Chicago, IL, AMA, 1997