Copyright © American Society of Artificial Internal Organs. Unauthorized reproduction of this article is prohibited.
1
ASAIO Journal 2019 Pediatric Circulatory Support
The number of children receiving extracorporeal mem-
brane oxygenation (ECMO) has increased substantially, and
includes a growing population of children with complex un-
derlying conditions who previously may not have been con-
sidered ECMO candidates. However, it remains unclear to
what extent the underlying disease impacts the risk of death
in these patients, particularly related to malignancy, bone
marrow transplantation, complex congenital heart disease
(CHD), or chromosomal abnormalities. A retrospective study
was performed using the Pediatric Health Information System
database of all children placed on ECMO more than a 10 year
period between 2004 and 2013. Patients with diagnoses of
bone marrow transplant, leukemia, lymphoma, neutropenia,
immune system abnormalities, genetic abnormalities, neo-
plastic disorders, and complex CHD were selected as “high
risk” and their outcomes were compared with overall out-
comes. Extracorporeal membrane oxygenation was used in
9,194 children. Two thousand two hundred (24%) were iden-
tified as high risk. Bone marrow transplant (81% mortality;
OR 3.49), leukemia (66% mortality; OR 1.88), and neutro-
penia (58% mortality; OR 1.62) were associated with higher
odds of mortality. Complex CHD (52% mortality) and ge-
netic syndromes (48%) were not associated with higher mor-
tality. These findings are pertinent for clinicians and families
when considering ECMO candidacy in these children. ASAIO
Journal XXX; XX:00–00.
Key Words: pediatrics, extracorporeal membrane oxygen-
ation; postcardiotomy, stem cell transplant, malignancy, ge-
netic disorders, neutropenia
Over the past decade, extracorporeal membrane oxygena-
tion (ECMO) has become a more frequently employed therapy
in children with cardiac, pulmonary, or cardiopulmonary
failure. Although ECMO is a resource intense and complex
therapy associated with significant morbidity and mortality,
overall mortality rates have improved during this time frame.
1
Alongside these trends are a growing use of ECMO in com-
plex patients, often with a less certain or guarded prognosis
with respect to their underlying disease, particularly with re-
spect to underlying malignancy, hematologic disease, or other
complex medical conditions that carry significant mortality in
the ICU setting. The utility or overall benefit of ECMO in these
high-risk patients, for whom the underlying disease may have a
high mortality or might impact the ECMO course or incidence
of complications, is often questioned. This is largely because of
assumptions and lack of data associating underlying pathology
with survival.
The goal of this study is to compare the outcomes of sub-
groups of pediatric patients that may be considered to be at
a higher risk of mortality, because of the underlying disease.
Materials and Methods
A retrospective review of the Pediatric Health Information
System (PHIS) database was conducted, with institutional re-
view board exemption. The PHIS database is an in-patient chil-
dren’s hospital administrative database that collects data on all
admissions to 43 nonprofit, tertiary care children’s hospitals
in the United States. Patients who were placed on ECMO be-
tween January 1, 2004 and December 31, 2013 were identi-
fied. Patients with diagnoses of bone marrow transplant (BMT),
leukemia, lymphoma, genetic abnormalities, neoplastic disor-
ders, and congenital heart disease (CHD) with a risk-adjusted
congenital heart surgery score (RACHS-1) greater than or equal
to 4 were placed into the “high-risk” cohort and their outcomes
on ECMO compared with the controls
Summary statistics were calculated for patient characteris-
tics and outcomes using frequency with percentage or median
with 25th and 75th percentiles. Summary statistics were also
stratified by survival and compared using χ
2
, Fisher’s exact test,
or Wilcoxon rank sum test as appropriate.
Generalized estimating equations were used to assess the
odds of mortality for factors independently and in multivari-
able models while accounting for the within hospital correla-
tion because of the nested data structure.
Results
From the PHIS database, 9,194 patients were identified for
inclusion in the study, of which 2,200 (24%) were “high risk”
and the remainder (6,994, 76%) were “controls.” Patient char-
acteristics are shown in Tables 1 and 2. Mortality for each
patient group is shown in Table 3. Bone marrow transplant
patients had the highest mortality at 81%. The majority were
From the *Department of Pediatrics, Division of Critical Care Med-
icine, Baylor College of Medicine; †Department of Pharmacy, Texas
Children’s Hospital; and ‡Dan L. Duncan Insitute for Clinical and
Translational Research, Baylor College of Medicine, Houston, Texas.
Disclosure: The authors have no conflicts of interest to report.
Submitted for consideration September 2018; accepted for publica-
tion in revised form March 2019.
Dr. Coleman was supported by NIH (5T32HL007939-08).
Correspondence: Ryan D. Coleman, MD, Pediatric Critical Care
Medicine, Texas Children’s Hospital, 6651 Main St, MC: E1420,
Houston, TX 77030. Email: ryanc@bcm.edu.
Extracorporeal Membrane Oxygenation Mortality in High-Risk
Populations: An Analysis of the Pediatric Health Information
System Database
RYAN D. COLEMAN,* JORDANA GOLDMAN,* BRADY MOFFETT,† DANIELLE GUFFEY,‡ LAURA LOFTIS,*
JAMES THOMAS,* AND LARA S SHEKERDEMIAN*
Copyright © 2019 by the ASAIO
DOI: 10.1097/MAT.0000000000001002