Keywords
pediatric - enteral nutrition - bilevel positive pressure ventilation - noninvasive
ventilation - asthma - status asthmaticus
Introduction
In the United States, asthma is the third leading cause of hospitalization in children
under 15 years of age.[1] Status asthmaticus (SA), defined as failure of conventional medical therapy during
the management of asthma leading to respiratory failure, remains one of the leading
causes for admission to the pediatric intensive care unit (PICU).[2]
[3]
While there is a lack of convincing data demonstrating decrease in intubation or mortality
with the use of noninvasive positive pressure ventilation (NPPV) for critically ill
children admitted with SA,[4] there is increasing evidence of expansive use of NPPV for this condition in recent
years.[5]
[6] However, NPPV has also come with its challenges including being independently associated
with delayed initiation of enteral nutrition (EN).[7]
[8] There is an increased concern for potential complications of feeding intolerance,
aspiration, and subsequently pneumonia that may lead to a reluctance in initiating
EN despite the known benefits of early EN in critically ill children.[9]
[10] Moreover, there is a general concern regarding which patients may potentially fail
NPPV and eventually require endotracheal intubation, adding to the hesitancy in initiating
EN.
There are a few studies detailing the benefits of early EN on infection prevention,
clinical outcomes, and lower hospital charges in the general critically ill pediatric
population.[11]
[12] However, there is limited data on the safety and feasibility of providing EN to
patients admitted with SA receiving respiratory support via NPPV in the form of bilevel
positive pressure ventilation (BIPAP). The need to determine feasibility of EN in
children with SA on BIPAP is important, and differs from other respiratory failure
cases needing the same, because children with SA on BIPAP recover faster and rarely
progress to requiring invasive mechanical ventilation.[5] Therefore, they may be more likely to benefit from early EN and should be studied
separately. The goal of this study was to describe our institutional practice of initiating
EN to critically ill children admitted with SA in the PICU on BIPAP, assess how it
differs from children requiring just simple mask, and to determine the incidence of
adverse events (AE) related to feeding, if any.
Materials and Methods
Study Design
A retrospective chart review was done on all children (2–18 years old) admitted with
the diagnosis of severe asthma to the PICU at St John's Children's Hospital, a tertiary
unit in Springfield, Illinois, from December 2010 to December 2016. Inclusion criteria
included children between the ages of 2 to 18 years admitted with International Statistical
Classification of Diseases, 10th revision (ICD-10) Code of “Status Asthmaticus” or
“Severe Asthma” requiring oxygen supplementation in the form of either simple face
mask or NPPV with BIPAP. Patients who were determined not to have asthma after admission
to the PICU or those who did not have consistently documented clinical asthma scores
(CAS) were excluded. Additionally, children with a known history of developmental
delay, cerebral palsy, cardiac pathology, and tracheostomy or ventilator dependence
were also excluded. The local Institutional Review Board (Springfield Committee on
Research Involving Human Subjects) waived the need for informed consent and approved
the study.
Status Asthmaticus Management Protocol
In July 2010, a protocol was instituted in the PICU at St John's Children's Hospital
where children admitted with SA were managed by respiratory therapists according to
their CAS.[2] CAS, which has been used at our institution for several years, is a modified version
of the validated 5-point Pediatric Asthma Severity Score[13] taking into account respiratory rate, work of breathing/accessory muscle use, air
exchange, wheezing severity, and I:E ratio ([Table 1]). According to the protocol, children were placed on BIPAP if their initial CAS
was 4 on admission to the PICU, or if the score increased despite 2 hours of standard
therapy (continuous albuterol, systemic steroids, and oxygen therapy via simple mask).
If CAS did not decrease after 2 hours of the addition of BIPAP, adjunct therapy, including
magnesium sulfate and helium-oxygen mixture, could be instituted; and finally, invasive
mechanical ventilation, if needed. BIPAP was always initiated at pressure support
of 8 cm H20, titrated to maintain a tidal volume of 6 to 8 mL per kilogram body weight,
and positive end expiratory pressure of 5 cm H20, increased as needed for work of
breathing. Ten to 15 L per minute flow rate was used to drive albuterol when simple
face mask was required. The children who needed BIPAP were first taken off BIPAP 2 hours
prior to being weaned off continuous albuterol.
Table 1
Clinical asthma score
|
Parameters
|
Clinical asthma scores
|
|
Respiratory rate
|
0) Normal
1) Above tachypnea threshold for age
|
|
Accessory muscle use
|
0) Normal/none
1) Subcostal/intercostal
2) Neck and abdominal
|
|
Air exchange
|
0) Normal
1) Localized/decreased
2) Generalized/decreased
|
|
Wheezing
|
0) End expiratory /none
1) Entire expiration
2) Entire inspiration/expiration
|
|
I:E ratio
|
0) ≤1:2 (normal)
1) >1:3 (prolonged exp)
|
Feeding Regimen
As per the protocol, after 2 hours on BIPAP or simple mask, all patients could be
started on EN either per oral or via nasogastric tube, liquid diet advanced as tolerated,
provided patient's respiratory status remained stable or improved as reflected by
the trend in their CAS. Furthermore, it was imperative that children tolerated removal
of mask for the short period of time needed for them to eat or drink, without signs
of increasing respiratory distress. They were then placed back on the masks immediately
after feeding. However, the decision to feed was at the discretion of the attending
physician. Feeding-related AE included aspiration leading to pneumonia, emesis after
feeding, and documented increase in work of breathing after feeding. Full EN was defined
as at least 100% of volume requirement prescribed by the managing team, that is, when
intravenous fluid was either discontinued or decreased to “keep vein open (KVO).”
Data Collection and Outcome Measures
Demographic data, type of respiratory support (BIPAP, or simple mask) time to initiation
of EN, time to full EN, modes of EN (nasogastric or per oral), time on BIPAP or simple
mask, CAS, adjunct treatments, duration of continuous albuterol, and AE related to
feeding were obtained from medical records. Data on AE included escalation of respiratory
support, incidence of pneumonia not present on admission, and feeding intolerance
as evidenced by emesis or need to withhold feeds. Data was extracted and recorded
every 2 hours for the duration the patient was in the PICU. Outcome measures were
time to initiation of EN, time to full EN, and AE related to EN.
Statistical Methods
Categorical variables are reported in counts and proportions with chi-squared analysis
performed. Continuous variables are reported with median. The IQR that is reported
with median stands for the interquartile range, which is the difference between the
25th percentile and the 75th percentile. Wilcoxon (Rank Sum) nonparametric analysis
was conducted for the continuous variables. Analysis of covariances with least square
means follow-up was ran to adjust for gender, weight, CAS at admission, use of adjunct
therapy, and PICU length of stay (LOS), and hospital LOS. Regression analysis was
used to calculate the R2. A 2 × 7 analysis of variance was conducted to compare the
two study groups over the 7 years of the study. SAS 9.4 statistical software (Cary,
North Carolina, United States) was used and statistical significance was determined
using two-sided p-value <0.05.
Results
There were 406 cases admitted with ICD-10 Code of “Status Asthmaticus” or “Severe
Asthma” to the PICU, out of which 43 did not have a diagnosis of asthma and 48 were
not consistently scored for clinical severity during the PICU stay. Therefore, a total
of 315 cases were included in the study, of which 135 required oxygen supplementation
through BIPAP and 180 were placed on simple mask. Demographic data are presented in
[Table 2]. Overall median age was 6 (IQR: 6) years and was similar in both groups. Overall
median weight was 24.8 (IQR: 20.8) kg, greater in the BiPAP group, p = 0.02. There were more females on BIPAP, p = 0.037.
Table 2
Demographic data and severity of status asthmaticus in the PICU
|
Characteristics
|
Overall median (IQR)
n = 315
|
Bilevel positive pressure ventilation group
Median (IQR)
n = 135
|
Simple mask group
Median (IQR)
n = 180
|
p-Value
|
|
Age (y)
|
6 (6)
|
7 (6)
|
6 (6)
|
NS
|
|
Weight (kg)
|
24.8 (20.8)
|
26.8 (23)
|
24.0 (19)
|
0.02
|
|
Females (%)
|
36.6
|
43.7
|
32.3
|
0.037
|
|
Asthma score at admission
|
3 (2)
|
4 (2)
|
3 (2)
|
0.001
|
|
Adjunctive therapy (%)
|
58.4
|
80
|
43.9
|
0.001
|
|
Duration of continuous albuterol (h)
|
26 (39)
|
41 (41)
|
20 (29)
|
0.001
|
|
Hospital length of stay (h)
|
67 (54)
|
87 (57)
|
53 (45)
|
0.001
|
Abbreviations: IQR, interquartile range; NS, not specified; PICU, pediatric intensive
care unit.
Severity of Asthma in the PICU ([Table 2])
Overall CAS at admission to the PICU was median 3 (IQR: 2), compared with simple mask,
patients on BIPAP were sicker, p = 0.001. A total of 58.4% cases required adjunctive therapy including magnesium sulfate
infusion and/or helium-oxygen mixture. Children requiring BIPAP received more adjunctive
therapy, p = 0.001. Overall duration of continuous albuterol therapy and hospital LOS were median
26 (IQR: 39) and 67 (54) hours, respectively; both were significantly longer in children
requiring BIPAP, p = 0.001. None of the patients required any sedation to facilitate BIPAP.
Enteral Nutrition ([Table 3])
Table 3
Enteral nutrition
|
Characteristics
|
Overall median (IQR)
n = 315
|
Bilevel positive pressure ventilation group
Median (IQR)
n = 135
|
Simple mask group
Median (IQR)
n = 180
|
p-Value
|
|
Patients receiving enteral nutrition (%)
|
100
|
100
|
100
|
NS
|
|
Time to initiate feeds (h)
|
6 (11)
|
11 (20)
|
4.3 (7)
|
0.001
|
|
Time to reach full feeds (h)
|
15 (17)
|
23 (26)
|
12 (15)
|
0.001
|
|
Patients started on feeds by 24 hours (%)
|
89.2
|
81.5
|
96.6
|
0.001
|
Abbreviations: IQR, interquartile range; NS, not specified; PICU, pediatric intensive
care unit.
All children studied were on full EN while still on BIPAP and simple mask; 99.3 and
100% were fed per oral, respectively. One child was fed via nasogastric tube and one
via G-tube in the BIPAP group. Overall median time to initiation of EN and full EN
was 6 (IQR: 11) and 15 (IQR: 17) hours, respectively. Both were longer in the BIPAP
group versus simple mask group, p = 0.001. The results remained similar after adjusting for gender, weight, CAS at
admission, use of adjunct therapy, and duration of continuous albuterol. The r2 for
“time to initiation of EN” and “time to full EN” were 0.16 and 0.27, respectively,
while using the parametric data for age, gender, CAS at admission, use of adjunct
therapy, and duration of continuous albuterol as predictors. By 24 hours, 81.54% of
patients on BIPAP and 96.6% on simple mask were started on EN, p = 0.001.
None of the patients required invasive mechanical ventilation. Additionally, there
were no documented AE, including aspiration leading to new pneumonia (pneumonia not
present on admission), emesis, and documented increase in work of breathing related
to feeding itself or removal of mask for feeding, in either group.
Over the years, the duration gap between patients on BIPAP versus simple mask for
time to initiate EN ([Fig. 1]) and reach full EN decreased ([Fig. 2]). In fact, time to initiate feeds was statistically similar in the year 2016, median
6 (IQR: 21) hours in the BIPAP group versus median 5.5 (IQR: 9) hours in the simple
mask group, p > 0.05.
Fig. 1 Time to initiate enteral nutrition. BIPAP, bilevel positive airway pressure ventilation.
Fig. 2 Time to full enteral nutrition. BIPAP, bilevel positive airway pressure ventilation.
Discussion
In this retrospective cohort study, all children admitted to the PICU with a diagnosis
of SA were started on EN while still receiving respiratory support via BIPAP or facemask
with 90% of patients on full EN within 24 hours of admission. Children who are critically
ill are under metabolic stress and EN plays a vital role in recovery. Several recent
studies have highlighted the significance of early EN in the critically ill pediatric
population including its role in shorter PICU and overall LOS.[11]
[14]
[15]
[16] There does not appear to be a consensus, however, on the definition and recommendation
for early EN and has been stated in the past as anywhere from 24 to 48 hours.[8]
While there is plenty of published literature describing EN practices for critically
ill children in general,[12]
[17] there is a paucity of data on early EN in patients on NPPV, specifically those admitted
with SA. To the best of our knowledge, our study is the first to focus on safety of
EN on children with SA admitted to the PICU on BIPAP. One large French observational
study reported that almost 60% of adults with respiratory failure due to various reasons
on NPPV were starved for over 48 hours and only 2.5% received EN.[18] Another large annual cross-sectional, multinational audit conducted over 7 years
demonstrated that among almost 10,000 patients admitted with respiratory failure (out
of which 6.2% were on NPPV and the rest on invasive mechanical ventilation), 40 and
20% were not fed the first and second days of admission, respectively, while on respiratory
support.[19] There are fewer studies in children. In a retrospective cohort study, Leroue et
al described 64% of all children on NPPV were initiated on EN within the first 24 hours
with 54% fed orally. However, only 32% were on BIPAP and only 18% had a diagnosis
of SA.[7] Recently, a retrospective study in four PICUs across four European nations studying
EN practices in children on various forms of NPPV around intubation reported a median
time to initiate feeds of 4 hours. However, only 10.8% fed per oral, only 33.2% were
on BIPAP, and it is unclear if there were any children admitted with SA.[20]
Even though 81.4% of cases on BIPAP in our study were on EN within 24 hours, and 100%
eventually received full EN while still on BIPAP, it was noted that the time to achieve
both was longer in the BIPAP group than the simple mask group. This is intuitive and
can be attributed to increased severity of illness as evidenced by higher asthma scores,
longer requirement of continuous inhaled albuterol, and use of adjunctive therapy
in the former group. Children needed to be able to tolerate removing their BIPAP masks
long enough to eat, as almost 100% of cases received EN by oral route. Additionally,
the hesitancy in initiating EN in the BIPAP group may have been to better determine
the likely clinical course to avoid complications of intubation on a full stomach.[21] Use of NPPV has previously been shown to be independently associated with a delay
in initiating EN predominantly due to fear of feeding-related complications, and difficulty
in anticipating which cohort of patients may need escalation of support and intubation.[22] In one retrospective cross-sectional study looking at barriers to early initiation
of EN in children under 21 years of age in six PICUs, the OR of delayed EN in children
on NPPV compared with no support was 3.37 (95% confidence interval: 1.69–6.72).[17] As providers in our hospital became familiar with NPPV for SA, the time to initiate
EN and reach full EN shortened over the years.
Our study revealed no feeding-related complications namely aspiration pneumonia or
need for mechanical ventilation, contrary to the large French observational study
that reported nosocomial pneumonia and increased mortality rates in adults receiving
EN on NPPV.[18] There are a few studies in children that have reported some AE. Tume et al in their
European study across four PICUs found the most common complications related to EN
on noninvasive ventilation to be gastrointestinal in nature (4–20%) with rare incidence
of aspiration pneumonia.[20] However, as we discussed earlier, only 33% of their patient population was on BIPAP
(n = 108) and their admission diagnosis was varied. Leroue et al demonstrated 12% AE
in the cohort of children on NPPV that they studied, including 10% of patients having
developed a new pneumonia and 3% of patients requiring endotracheal intubation.[7] However, it is not clear what type of noninvasive support was related to AE (high
flow nasal canula [HFNC], continuous positive airway pressure, or BIPAP), and whether
the 18% children admitted with SA actually incurred any AEs. One pediatric study reported
5.8% of patients experienced EN-related AEs when placed on HFNC.[23] However, this study was on children under 24 months of age admitted for bronchiolitis.
It would be remiss not to discuss the use, or lack thereof, of HFNC in this retrospective
review. HFNC therapy had just been introduced as respiratory support during the early
years of this study, mainly in neonates in the neonatal intensive care units and infants
with bronchiolitis.[24] Although it was an emerging mode of respiratory support, there was a lack of clinical
experience and studies in older children.[25] Between the years 2010 and 2016 (the duration of this review), there was no documented
use of HFNC to manage children admitted to the PICU with SA in our hospital.
Our study has several limitations to consider when interpreting these results. Due
to the retrospective nature of the study, data collected was limited to prior documentation.
Data on AEs was also extracted from documentation and some, albeit minor AEs, could
have been missed. Having said that, the data was recorded every 2 hours in the PICU
and feeding data was well documented by nursing staff. We were also limited by our
inability to accurately assess oral intake and protein/calorie requirements. Hence,
we used weaning off intravenous fluids as a surrogate marker for reaching full EN.
Intravenous fluids could have been continued even after full EN was reached, thus
erroneously decreasing the data on overall time to reach full EN. Nonetheless, even
with this limitation, almost 90% children were on full EN at 24 hours. Another limitation
we noted was that even though the protocol for management of SA had guidelines for
initiation for EN, the ultimate decision was at the discretion of the attending physician
and, therefore, not standardized. As “improved PO intake” is subjective, this may
have also been influenced by provider comfort levels and therefore the time to reach
full EN likely represents selection bias. Additionally, we lacked information regarding
provider-specific factors contributing to the clinical decision-making regarding EN.
Finally, this study included a small cohort of patients and it represents a single
institution's experience and unknown accuracy of generalizability to others. As a
result of the nature of the study, the question of EN affecting PICU and hospital
LOS could not be determined. Another important aspect of EN is patient and family
satisfaction that also could not be determined because of the retrospective nature
and lack of data. A larger trial comparing children admitted with SA receiving full
EN versus no EN may be required to answer these questions.
Conclusions
EN is a cornerstone of management and recovery in the PICU. Since most children admitted
for SA recover early and rarely proceed to requiring endotracheal intubation and invasive
mechanical ventilation, early EN may be more feasible than in children admitted with
respiratory failure for other etiologies. Our institutional experience suggests that
EN may be effectively and safely initiated and continued for children admitted with
SA on both simple mask and BIPAP. Although early oral nutrition intuitively results
in a “happy child” and satisfied parents, this warrants further research. Larger studies
are also needed to determine if initiating EN within 24 hours affects time to recovery.
