Journal of Pediatric Surgery 55 (2020) 39–44

Contents lists available at ScienceDirect

Journal of Pediatric Surgery

journal homepage: www.elsevier.com/locate/jpedsurg

Low postnatal CRI values are associated with the need for ECMO in
newborns with CDH☆,☆☆,☆☆☆
David K. Leopold a,⁎, Ryan C. Phillips a, Niti Shahi a, Jason Gien b,c, Ahmed I. Marwan a,c, John P. Kinsella b,c,
Jane Mulligan d, Kenneth W. Liechty a,c, Steven L. Moulton a,d
a
Department of Surgery, Division of Pediatric Surgery, University of Colorado School of Medicine, Aurora, CO
b
Department of Pediatrics, Division of Neonatology, University of Colorado School of Medicine, Aurora, CO
c
Colorado Fetal Care Center, Children's Hospital Colorado, Aurora, CO
d
Flashback Technologies Inc., Louisville, CO

a r t i c l e i n f o a b s t r a c t

Article history: Introduction: Accurate, real-time technology is needed to predict which newborns with congenital diaphragmatic
Received 8 September 2019 hernia (CDH) will require ECMO. The Compensatory Reserve Index (CRI) is a noninvasive monitoring technology
Accepted 29 September 2019 that continuously trends an individual's capacity to compensate from normovolemia (CRI = 1) to decompensa-
tion (CRI = 0). We hypothesized that postnatal CRI values would be lower in CDH newborns that required ECMO
Key words: than those who did not require ECMO.
Compensatory reserve index
Methods: Newborns with a CDH were prospectively monitored with a CipherOx® CRI M1 device. We compared
Congenital diaphragmatic hernia
Extracorporeal membrane oxygenation
CRI values from delivery to ECMO (ECMO group) versus delivery to clinical stabilization (non-ECMO group).
Photoplethysmography Results: Postnatal CRI values were available from 26 newborns. Eight underwent ECMO within 33 h of delivery,
and median CRI prior to ECMO was 0.068 (IQR: 0.057, 0.078). Eighteen did not require ECMO. Median CRI
from birth to 48 h was 0.112 (IQR: 0.082, 0.15). CRI values were significantly lower in newborns that required
ECMO versus those who did not (p = 0.0035). Postnatal CRI had the highest AUC (0.85) compared to other
prenatal prognostic measures.
Conclusion: Humans from newborns to adults share elemental features of the pulsatile waveform that are
associated with progression to decompensation. CRI may be helpful when deciding when to initiate ECMO.
Level of evidence: Level III.
Type of study: Diagnostic test.
© 2019 Elsevier Inc. All rights reserved.

Congenital diaphragmatic hernia (CDH) is a complex syndrome with
a wide spectrum of severity, depending on the degree of pulmonary hy-
poplasia and pulmonary hypertension [1]. Improvements in standard-
ized protocols, advancements in neonatal management, and the
implementation of various prognostic models have decreased morbid-
ity and mortality in this population. Approximately 30% of newborns
with a CDH require extracorporeal membrane oxygenation (ECMO)
for postnatal cardiopulmonary support [2]. Various prognostic and risk
stratification schemes, which rely on prenatal imaging, genetics, or
postnatal factors, have improved the ability of clinicians to predict the
likelihood of a newborn requiring ECMO and its overall mortality
[1–9]. Once the newborn is delivered, however, the utility of these

Abbreviations: CDH, Congenital Diaphragmatic Hernia; CRI, Compensatory Reserve Index; ECMO, extracorporeal membrane oxygenation; IQR, Interquartile Range; SBP, Systolic blood
pressure; HR, Heart Rate; O/E, Observed to Expected; LHR, Lung to Head Ratio; PPLV, Percent Predicted Lung Volume; TLV, Total Lung Volume; ROC, Receiver Operator Curve; AUC, Area
Under the Curve.
☆ Disclosures: Drs. Mulligan and Moulton, together with Greg Grudic, developed the Compensatory Reserve Index (CRI) algorithm utilized in this study. The underlying intellectual
property is assigned to Flashback Technologies, Inc. and the Regents of the University of Colorado. Drs. Mulligan and Moulton have an equity interest in the company and receive royalty
payments through the University of Colorado. Drs. Leopold, Phillips, Shahi, Gien, Liechty, Marwan and Kinsella have declared no conflicts of interest.
☆☆ Disclaimer: Development of the CRI algorithm was supported in part by funding from the US Army Medical Research and Material Command (USAMRMC) under Grant Nos. DM09027,
W81XWH-15-2-0007, W81XWH-09-1-0750, W81XWH-09-C-0160, W81XWH-11-2-0091, W81XWH-11-2-0085, W81XWH-12-2-0112, and W81XWH-13-C-0121. The views, opinions, and/
or findings contained in this report are those of the authors and should not be construed as an official position, policy or decision of the Department of the Army unless designated by other
documentation. Supported by Children’s Hospital Colorado Foundation and NIH/NCATS Colorado CTSA Grant Number UL1 TR002535. Contents are the authors’ sole responsibility and do
not necessarily represent official NIH views.
☆☆☆ How this paper will improve care: Humans, from newborns to adults, share features of the pulsatile waveform that are associated with progression to decompensation. CRI trends
these features and may aid in earlier initiation of ECMO, before newborns with a CDH experience prolonged hypoxemia and hypotension.
⁎ Corresponding author at: Department of Surgery, University of Colorado School of Medicine, 1640 Oneida Street, Denver, CO, 80220. Tel.: +1 303 818 4749.
E-mail address: David.Leopold@ucdenver.edu (D.K. Leopold).

https://doi.org/10.1016/j.jpedsurg.2019.09.050
0022-3468/© 2019 Elsevier Inc. All rights reserved.
40 D.K. Leopold et al. / Journal of Pediatric Surgery 55 (2020) 39–44
indices diminishes as the clinical team adapts to the rapidly changing
scenario of stabilization and resuscitation [10]. To date, no current tech-
nology accurately predicts which newborns will require ECMO after
delivery.
The Compensatory Reserve Index (CRI) is the first FDA-cleared tech-
nology in the “adjunctive cardiovascular status indicator” class. CRI uses
signals from continuous noninvasive sensors to estimate the proportion
of additional volume loss a patient can tolerate before reaching the
point of hemodynamic decompensation (systolic blood pressure
(SBP) b 80 mmHg) in adults [11]. The technology is based on advanced
signal processing and data analytics applied to an extensive raw signal da-
tabase from a simulated hemorrhage lower body negative pressure
(LBNP) model, in which subjects were taken in a stepwise fashion from
normovolemia to cardiovascular collapse [11]. The intent was to develop
a noninvasive, real-time technology to quickly identify and monitor those
at risk for cardiovascular collapse due to acute blood loss. CRI values range
between 1 and 0, where CRI = 1 indicates supine normovolemia in adults,
and CRI = 0 indicates the point of decompensation (SBP b 80 mmHg in
adults, loss of consciousness, or neurologic symptoms associated with de-
creased perfusion). CRI values between 1 and 0 indicate the proportion of
remaining volume reserve before the onset of decompensation.
CRI has proven to be a sensitive and specific, noninvasive method for
monitoring acute volume loss in adults [11–14]. Although the algorithm
was developed and FDA-cleared based on sensor data from adults
[11,12,14–22], it has been shown to correlate with fluid requirements
and the extent of burn injury in children [23]. CRI has also been
shown to identify hypovolemia and trend fluid resuscitation in children
with dengue shock syndrome [24]. Recent work by Choi et al. [25]
showed that CRI is able to detect the difference in physiologic reserve
due to hypovolemia and peritonitis in children with perforated versus
non-perforated appendicitis. This study showed that this difference
diminished after surgical intervention and fluid administration.
The physiology and hemodynamics of newborns are entirely
different from those of adults or children. While a heart rate of 120 to
160 is normal in newborns, the same values would be a clinical concern
in adults and children [26]. All humans, however, newborns to adults,
share elemental features of the pulsatile waveforms that are associated
with progression to decompensation. Due to the infrequency of
cardiovascular collapse in newborns and the unethical nature of
subjecting newborns to controlled models to simulate decompensation,
CRI has not been validated in this population. To validate if CRI would ef-
fectively trend the ability of a newborn to compensate up until cardio-
vascular decompensation, we chose to investigate the use of CRI in the
setting of newborns with CDH. We hypothesized that postnatal CRI
values would be lower in newborns with a CDH who required ECMO
than those who did not require ECMO. Thus, the aims of this study
were: 1) determine whether CRI can discriminate the range of physio-
logical compromise in newborns with CDH, and 2) evaluate the associ-
ation between post-natal CRI values and the need for ECMO in
newborns with CDH.
1. Methods
1.1. Study population
The Colorado Institutional Review Board approved this prospective
observational study. Informed consent was obtained from all subjects'
parents before data collection, when able. If consent was not able to
be obtained at the time of delivery or transfer, a parent of each subject
consented as soon as possible, given the clinical situation. Newborns
delivered at Children's Hospital Colorado or transferred within the
first 24 h of life with a prenatal or postnatal confirmed diagnosis of
CDH were eligible for the study. Newborns with a CDH, with planned
ECMO immediately after delivery or ECMO initiation within one hour
of birth were excluded.
1.2. Data collection
Continuous CRI measurements were collected using CipherOx® CRI
M1 monitors (Flashback Technologies Inc., Louisville, CO), from the time
of birth or enrollment until after initial stabilization and resuscitation
were performed and subjects were considered reasonably stable by
the clinical team, or the subject was placed on ECMO. Caregivers and
clinicians were unaware of the CRI values since they were not visible
on the investigational device's screen. A CipherOx® CRI M1 device
was placed on the middle fingers or palm of the left hand which is the
best place for determining postductal oxygen saturation. Clinical
information, such as laboratory values and the timing of significant
events, was retrospectively collected from electronic medical records.
1.3. Statistical methods
Demographic variables were summarized using medians and inter-
quartile ranges (IQR). Counts and proportions were used for categorical
variables. Comparisons of the ECMO and non-ECMO groups were
performed using Wilcoxon rank sum, Chi-square, and Fisher's exact
tests. The average CRI for each subject was calculated from (i) the
time of birth/ enrollment to the time of (ii) ECMO flow initiation or
clinical stabilization. The “low CRI” for the ECMO group was the mini-
mum of the hourly CRI averages from the time of birth/ enrollment to
the time of ECMO flow initiation. The “low CRI” for the non-ECMO
group was the minimum of the hourly CRI averages from the time of
birth/ enrollment until clinical stabilization. CRI values were also
separated into sequential ten-minute segments from birth to one hour
of life and the average of these segments was used for comparative
analysis.
Receiver operator curves (ROCs) were calculated using the low CRI
and prenatal prognostic factors as the predictor and ECMO initiation
as the outcome. Area under the curve (AUC) and corresponding CI
were used to compare the fit of the different models. The threshold
was determined by overall accuracy, meaning the combination of
sensitivity and specificity. The AUC was considered significant when
the confidence interval (CI) was greater than 0.5. R version 3.5.1
software (R Foundation for Statistical Computing, Vienna, Austria,
http://www.R-project.org) was utilized. Significance level was set at
0.05.
2. Results
2.1. Demographics and clinical information
A total of 28 subjects were enrolled over a 30-month period between
2016 and 2019. Two subjects, who were enrolled and placed on ECMO
at 46 and 48 min after delivery, were not included in the analysis.
Data were analyzed for the remaining 26 subjects, who had a confirmed
diagnosis of CDH, of which 8 eventually required ECMO support.
Demographics and prenatal information are presented in Table 1.
Comparing the ECMO and non-ECMO newborns, there were no signifi-
cant differences in baseline demographics such as gender distribution,
gestational age, birth weight, delivery method, or prenatal diagnosis of
CDH prior to delivery. Comparing prenatal imaging MRI and ultrasound
findings between the two groups, lung–head ratio (LHR) and the
observed-to-expected (O/E) LHR were not significantly different.
There was a nonsignificant higher incidence of left-sided diaphragmatic
hernias in the non-ECMO group, as the ECMO group had two right-sided
defects and the non-ECMO group had none (p = 0.16). The median
percent predicted lung volume (PPLV) and total lung volumes (TLV)
were significantly higher in the non-ECMO group, measuring 20.3
(IQR: 18, 26) and 26.9 (IQR: 21.6, 34) versus 15.2 (IQR: 13, 16.8) and
16.0 (IQR: 14.2, 21.5), respectively (p = 0.007, p = 0.021).
 D.K. Leopold et al. / Journal of Pediatric Surgery 55 (2020) 39–44 41

Table 1
Demographics and clinical information. Numbers reported are medians and interquartile
ranges for continuous variables and counts and proportions for categorical variables, un-
less otherwise specified.

Non-ECMO (n = 18) ECMO (n = 8) p=

Gender (male) 10 (55%) 5 (63%) 1

Left Diaphragmatic 18 (100%) 6 (75%) 0.16
Hernia
Estimated Gestational 38.00 (IQR: 36.47, 37.43 (IQR: 36.5, 0.16
Age 38.14) 38.00)
Birth weight (kg) 3.04 (IQR: 2.5, 3.5) 3.00 (IQR: 2.88, 3.11) 0.68
Vaginal Delivery 12 (66%) 3 (38%) 0.38
Prenatal Diagnosis 17 (94%) 8 (100%) 1
O/E LHR 0.49 (IQR: 0.41, 0.53) 0.41 (IQR: 0.28, 0.47) 0.14
LHR 1.3 (IQR: 1, 1.5) 1.1 (IQR: 0.88, 1.35) 0.3
PPLV 20.3 (IQR: 18, 26) 15.15 (IQR: 12.95, 0.007
16.75)
TLV 26.9 (IQR: 21.6, 34) 16.00 (IQR: 14.11, 0.021
21.45)

2.2. CRI values

Average CRI values were significantly lower in neonates who

required ECMO, with a median of 0.068 (IQR: 0.057, 0.078) compared
to average CRI values in the non-ECMO group with a median of 0.112
(IQR: 0.082, 0.15) (p = 0.0035). The “low CRI” was significantly lower
for the neonates who required ECMO with a median of 0.035 (IQR:
0.024, 0.046) compared to the “low CRI” values in the non-ECMO
group with a median of 0.084 (IQR: 0.049, 0.106) (p = 0.005). The 10-
minute averages in the first hour were not significantly different
between the ECMO and non-ECMO group, except for the CRI averages
from 50 min to 60 min of life (p = 0.049). The CRI average from birth
to the first hour of life was not significantly different between the
groups, with a median CRI for the non-ECMO group of 0.102 (IQR:
0.081, 0.118) and a median of 0.083 (IQR: 0.048, 0.121) in the ECMO
group (p = 0.438). CRI medians and IQRs are listed in Table 2. Fig. 1 il-
lustrates the difference in distribution of average CRI values between
the two groups.
2.3. Missing values
There were missing CRI values at various time points among subjects
from either birth to ECMO or from birth through resuscitation due to
accidental sensor dislodgement or removal of the sensor by the clinical
team. Values were missing within one hour of birth or transfer in four
subjects: three subjects in the non-ECMO group and one subject in the
ECMO group. Median recording duration was 160 min (IQR: 83, 697)
for the ECMO group versus 348 min (IQR:118, 700) for the non-ECMO
group. The overall range of recording time was 39 to 2880 min. The
differences in recording time between the two groups were not statisti-
cally significant (p = 0.49). There was a wide range of time from birth to
ECMO, with a mean time of 665 min (SD: 702) and a median time of
362 min (IQR: 130, 1021).
Fig. 1. Boxplot of average CRI values of CDH newborns.
2.4. CRI and ROC
Fig. 2 shows the ROCs of the “low CRI”, O/E LHR, LHR, TLV, and PPLV
with the corresponding AUC and CI. AUC values from the “low CRI”,
PPLV, and TLV were considered significant. The “low CRI” had the
highest AUC (0.85, 95% CI: 0.6856–1) compared to O/E LHR (AUC =
0.69, 95% CI: 0.48–0.90), LHR (AUC = 0.63, 95% CI: 0.38–0.88), PPLV
(AUC = 0.84, 95% CI: 0.62–1) and TLV (AUC = 0.79, 95% CI:
0.61–0.98). For the “low CRI” metric, we explored different thresholds
to understand potential application of this metric in the clinical setting.
A threshold of 0.054 was associated with the highest level of accuracy
among potential thresholds, resulting in a sensitivity of 0.875, specificity
of 0.72, positive predictive value of 0.583, and negative predictive value
of 0.929 (95% CI: 0.69–1).
3. Discussion
The management of newborns with a congenital diaphragmatic
hernia, especially those with poor prognostic features, remains
challenging. Advancements in prognostic indices have provided insight
and understanding into the likelihood of physiological decompensation
after delivery. These prognostic models are based on retrospective data,

Table 2
CRI values over time. Numbers reported are medians and interquartile ranges.

Non-ECMO (n = 18) ECMO (n = 8) p=

a
Average CRI 0.112 (IQR: 0.082, 0.15) 0.068 (IQR: 0.057, 0.078) 0.0035
Low CRI 0.084 (IQR: 0.049, 0.106) 0.035 (IQR: 0.024, 0.046) 0.005
10 min CRI average 0.148 (IQR: 0.123, 0.205) 0.342 (IQR: 0.272, 0.351) 0.136
10–20 min CRI average 0.113 (IQR: 0.087, 0.165) 0.149 (IQR: 0.122, 2.12) 0.692
20–30 min CRI average 0.106 (IQR: 0.09, 0.161) 0.105 (IQR: 0.082, 0.128) 0.844
30–40 min CRI Average 0.097 (IQR: 0.062, 0.113) 0.044 (IQR: 0.035, 0.104) 0.225
40–50 min CRI Average 0.099 (IQR: 0.069, 0.121) 0.039 (IQR: 0.03, 0.166) 0.431
50–60 min CRI Average 0.093 (IQR: 0.089, 0.135) 0.068 (IQR: 0.05, 0.092) 0.049
1st Hour CRI Average 0.102 (IQR: 0.081, 0.118) 0.083 (IQR: 0.048, 0.121) 0.438
a
Average CRI was calculated from (i) the time of birth/ enrollment to the time of (ii) ECMO flow initiation or clinical stabilization.
42 D.K. Leopold et al. / Journal of Pediatric Surgery 55 (2020) 39–44
Fig. 2. AUC ROC of CRI and prenatal prognostic factors.
however, which can limit their utility, especially when prenatal factors
are not readily available. Currently, there is limited technology to
continuously trend a newborn's capacity to compensate for hemody-
namic insults or measure the future risk of deterioration. The physiology
of a newborn with a CDH is dynamic and can rapidly change; other
diagnostic measures such as arterial pH, base deficit, and lactate take
time to result. CRI is immediately and continuously available, and it
requires no baseline or calibration measurement. CRI was developed
based on adult physiology, but our study suggests that humans, from
newborns to adults, share elemental features of the pulsatile waveform
that are associated with progression to decompensation. This prospec-
tive pilot study showed that these features, and CRI's ability to detect
them, may be utilized as an adjunct to aid in earlier initiation of
ECMO, potentially before newborns with a CDH experience prolonged
periods of hypoxemia and hypotension. In the present study, we
observed lower postnatal CRI values among newborns with a CDH
that required ECMO, compared to those that did not require ECMO.
We believe these lower CRI values reflect the reduced physiologic
reserve of these newborn to compensate and respond to resuscitative
interventions, which ultimately necessitates ECMO.
The individual-specific nature of CRI may be particularly helpful in
the newborn CDH population, whose response to resuscitative inter-
ventions can vary greatly between individuals. The present study
shows that CRI values were significantly lower in newborns with a
CDH who required ECMO, compared to those that did not require
ECMO. More striking was how the difference became more apparent
between the two groups over time, as those that went on to ECMO
showed progressively more muted responses in their CRI values with
resuscitative measures. Moreover, the overall trend in CRI in neonates
that required ECMO was a decline in CRI values after the first few
hours of birth and the persistence of low CRI values subsequently. A sin-
gle low CRI value, which represents a five second time interval, was not
predictive in this study. The presence of a single hourly average value,
“low CRI”, below the threshold of 0.054 had the highest level of accuracy
to identify those newborns with a CDH who required ECMO. A “low CRI”
below this threshold was predictive whether it occurred hours before or
immediately preceding ECMO initiation. The trend towards a decline in
CRI values in the ECMO group is illustrated by the finding of this study
that both the average CRI and “low CRI” values were significantly
lower in the ECMO group than in the non-ECMO group. This trend is
likely indicative of clinical deterioration of the newborn's ability to com-
pensate in response to resuscitative interventions.
Due to the wide range of abnormalities in the CDH population,
certain abnormalities will need to be taken into consideration in the
analysis of CRI data. In this study, CRI, which was developed to monitor
hypovolemia in adults, detected the significant difference between
newborns with CDH that required ECMO and those that did not require
ECMO. This likely indicates that CRI can identify hemodynamic deterio-
ration in newborns with CDH, and has the beneficial feature of being
based on postnatal real-time data versus prenatal prognostic factors.
Furthermore, our study suggests that CRI may be comparable if not
superior to prognostic models based on prenatal imaging. This study
 D.K. Leopold et al. / Journal of Pediatric Surgery 55 (2020) 39–44
does not, however, address CRI's relevance for newborns with other
congenital diseases.
CRI has several clear advantages for its use as an adjunctive car-
diovascular status monitor. The CipherOx® CRI M1 rapidly pro-
duces initial values and provides real-time monitoring at the
bedside in any acute setting. It is noninvasive and intuitive to use,
as it only requires applying a finger sensor and reading the CRI de-
vice display. Lastly, it generates data that are easy to comprehend
for healthcare providers and may allow clinicians to evaluate the
individual's responsiveness to certain clinical interventions, such
a pressor administration or fluid boluses.
There are several limitations in this study. First, there were missing
CRI values at various time points when the device was accidentally
dislodged or removed by the study participants. These episodes were
random, however, and the average recording length was not statistically
different between the two groups. Second, we simplified the CRI data by
using averages of CRI values over time, which reduces the amount of
power to detect meaningful differences. Third, the time from birth or
enrollment to stabilization or ECMO in the study was variable among
the study subjects, resulting in some subjects receiving various inter-
ventions, such as pressor administration or fluid boluses, prior to enroll-
ment. Nonetheless, we found a significant difference in average CRI
values between subjects that needed ECMO and those that did not. For
similar reasons, the type and timing of resuscitative measures were
not used for this analysis because documentation of timing of may
have been inaccurate, and there were significant differences between
subjects dictated by the clinical situation. Fourth, a priori power analysis
was not performed as this was an exploratory study, and the small sam-
ple size limited our analysis of interventions and complications. Fifth,
we were not able to analyze the relationship between CRI and the vary-
ing degrees to which children with CDH compensated because there is
no gold standard to quantify compensation, unlike volume loss in con-
trolled blood loss experiments. Lastly, and most importantly, the CRI al-
gorithm was developed based on data collected from healthy adults and
therefore is not specifically calibrated to the physiology of children or
newborns. However, as one of the primary aims of this study, we
show that CRI can differentiate newborns with a CDH that progress to
ECMO, versus those that are able to compensate with appropriate
resuscitation.
4. Conclusion
This is the first study to evaluate the Compensatory Reserve Index
(CRI) in the setting of newborns or newborns with CDH. CRI is a nonin-
vasive, real-time, individual-specific, hemodynamic parameter that
proved to be useful in assessing newborns with reduced compensatory
capacity and evaluating the need for ECMO. Further studies are needed
to determine if the post-natal use of CRI can improve CDH-related mor-
bidity and mortality.
Appendix A. Discussions
Presenter Ryan Phillips.
Q. Matt Harding, University of Texas in Houston Very nicely presented.
Interesting work. Be great to be able to predict this earlier. My
question is, why do you think this is the case? Certainly there
is emerging evidence that left ventricular dysfunction plays a
much larger role than we thought previously. But I'm inter-
ested, did you correlate this with echocardiography? Did
you correlate it with what your fluid resuscitation status
was? It seems like there's a lot of factors that might play
into this. Again, thank you.
Ryan Phillips Thank you for the question. It's a great question. We have
not actually done that yet, but we're in the process of doing
43
that because we agree that that's an important thing to
look at.
Q. Stan Adkins, Columbia, South Carolina Were you guys blinded to the
CRI when you were making the decisions to go on ECMO?
Ryan Phillips Yes, we were.
Q. Jack Langer from Toronto, Canada As a person that comes from a
place that hardly ever uses ECMO and still has the same
kind of survivals that ECMO centers do, the term ECMO is a
relative term. You're really saying those are the patients that
you chose to put on ECMO based on clinical criteria. So you're
now correlating that with another measurement, which re-
ally just echoes the decision making that you made based
on the other stuff. So I'm not sure how it's actually going to
help you. You're still going to put those patients on ECMO
based on the clinical criteria that you chose in the first place.
Ryan Phillips Thank you for your comments and question. I think the
main thing is that it would help potentially to get you on ear-
lier before the labs change. Because we've seen the labs
change after the CRI values.
References
[1] Losty PD. Congenital diaphragmatic hernia: where and what is the evidence? Sem-
inars in pediatric surgery. Elsevier; 2014.
[2] Jancelewicz T, Brindle ME, Harting MT, Tolley EA, Langham Jr MR, Lally PA, et al. Ex-
tracorporeal membrane oxygenation (ECMO) risk stratification in newborns with
congenital diaphragmatic hernia (CDH). J Pediatr Surg 2018;53(10):1890–5.
[3] Akinkuotu AC, Cruz SM, Abbas PI, et al. Risk-stratification of severity for infants with
CDH: prenatal versus postnatal predictors of outcome. J Pediatr Surg 2016;51(1):44–8.
[4] Russo FM, De Coppi P, Allegaert K, et al. Current and future antenatal management of
isolated congenital diaphragmatic hernia. Seminars in fetal and neonatal medicine.
Elsevier; 2017.
[5] Jani J, Keller RL, Benachi A, et al. Antenatal-CDH-Registry Group. Prenatal prediction
of survival in isolated left-sided diaphragmatic hernia. Ultrasound Obstet Gynecol
2006;27(1):18–22.
[6] Jani J, Nicolaides K, Benachi A, et al. Timing of lung size assessment in the prediction
of survival in fetuses with diaphragmatic hernia. Ultrasound in Obstetrics and Gyne-
cology 2008;31(1):37–40.
[7] Triebwasser JE, Treadwell MC. Prenatal prediction of pulmonary hypoplasia. Semi-
nars in fetal and neonatal medicine. Elsevier; 2017.
[8] Kehl S, Becker L, Eckert S, et al. Prediction of mortality and the need for neonatal ex-
tracorporeal membrane oxygenation therapy by 3-dimensional sonography and
magnetic resonance imaging in fetuses with congenital diaphragmatic hernias. J Ul-
trasound Med 2013;32(6):981–8.
[9] Oluyomi-Obi T, Kuret V, Puligandla P, et al. Antenatal predictors of outcome in pre-
natally diagnosed congenital diaphragmatic hernia (CDH). J Pediatr Surg 2017;52
(5):881–8.
[10] Datin-Dorriere V, Rouzies S, Taupin P, et al. Prenatal prognosis in isolated congenital
diaphragmatic hernia. Am J Obstet Gynecol 2008;198:80e1–5.
[11] Moulton SL, Mulligan J, Grudic GZ, et al. Running on empty? The compensatory re-
serve index. J Trauma Acute Care Surg 2013;75:1053–9.
[12] Stewart CL, Mulligan J, Grudic GZ, et al. Detection of low-volume blood loss: compensa-
tory reserve versus traditional vital signs. J Trauma Acute Care Surg 2014;77:892–7.
[13] Stewart CL, Mulligan J, Grudic GZ, et al. The compensatory reserve index following
injury: results of a prospective clinical trial. Shock 2016;46:61–7.
[14] Convertino VA, Howard JT, Hinojosa-Laborde C, et al. Individual-specific, beat-to-
beat trending of significant human blood loss: the compensatory reserve. Shock
2015;44:27–32.
[15] Hinojosa-Laborde C, Howard JT, Mulligan J, et al. Comparison of compensatory re-
serve during lower-body negative pressure and hemorrhage in nonhuman primates.
Am J Physiol Regul Integr Comp Physiol 2016;310:R1154–9.
[16] Janak JC, Howard JT, Goei KA, et al. Predictors of the onset of hemodynamic decom-
pensation during progressive central hypovolemia: comparison of the peripheral
perfusion index, pulse pressure variability, and compensatory reserve index. Shock
2015;44:548–53.
[17] Moulton SL, Mulligan J, Santoro MA, et al. Validation of a non-invasive monitor to
continuously trend individual responses to hypovolemia. J Trauma Acute Care
Surg 2017;83:S104–11.
[18] Poh PY, Carter 3rd R, Hinojosa-Laborde C, et al. Respiratory pump contributes to in-
creased physiological reserve for compensation during simulated haemorrhage. Exp
Physiol 2014;99:1421–6.
[19] Van Sickle C, Schafer K, Mulligan J, et al. A sensitive shock index for real-time patient as-
sessment during simulated hemorrhage. Aviat Space Environ Med 2013;84:907–12.
[20] Carter 3rd R, Hinojosa-Laborde C, Convertino VA. Variability in integration of mech-
anisms associated with high tolerance to progressive reductions in central blood vol-
ume: the compensatory reserve. Physiol Rep 2016;4:e12705.
[21] Convertino VA, Grudic G, Mulligan J, et al. Estimation of individual-specific progres-
sion to impending cardiovascular instability using arterial waveforms. J Appl Physiol
2013;115:1196–202.
44 D.K. Leopold et al. / Journal of Pediatric Surgery 55 (2020) 39–44

[22] Nadler R, Convertino VA, Gendler S, et al. The value of noninvasive measurement of
the compensatory reserve index in monitoring and triage of patients experiencing
minimal blood loss. Shock 2014;42:93–8.
[23] Stewart CL, Mulligan J, Grudic GZ, et al. A noninvasive computational method for
fluid resuscitation monitoring in pediatric burns: a preliminary report. J Burn Care
Res 2015;36:145–50.
[24] Moulton SL, Mulligan J, Srikiatkhachorn A, et al. State-of-the-art monitoring in treat-
ment of dengue shock syndrome: a case series. J Med Case Reports 2016;10:233.
[25] Choi YM, Leopold D, Campbell K, et al. Noninvasive monitoring of physiologic compromise
in acute appendicitis: new insight into an old disease. J Pediatr Surg 2018;53(2):241–6.
[26] Dawson JA, Schmolzer GM, Wyllie J. Monitoring heart rate in the delivery room.
Semin Fetal Neonatal Med 201;23327–32