Received: 29 January 2020 | Accepted: 29 April 2020

DOI: 10.1002/ppul.24811

ORIGINAL ARTICLE: SLEEP & BREATHING

The prevalence of hypercapnia during acute infection in

children on chronic noninvasive ventilation:
A retrospective study

Tessi Beyltjens MD1 | Sophie R. de Leede BSc1 | Ine van Eekelen BSc1 |
Fleur F. W. van Ginneken BSc1 | Evelyn Wyckmans MSc1 | Sophie Installe BSc1 |
Kim Van Hoorenbeeck MD, PhD1,2 | Stijn Verhulst MD, PhD1,2

1
Department of Paediatrics, Antwerp
University Hospital, Edegem, Belgium Abstract
2
Laboratory of Experimental Medicine and
Background/Aim: Children on chronic noninvasive ventilation are at risk for
Pediatrics, Faculty of Medicine and Health
Sciences, University of Antwerp, nonelective hospitalizations, mainly for acute infections. This study examined the
Antwerp, Belgium
prevalence of hypercapnia in children on chronic ventilatory support during an
Correspondence acute admission.
Stijn Verhulst, MD, PhD, Department of
Methods: This retrospective study included children aged 0 to 18 years who reg-
Pediatrics, Antwerp University Hospital,
Wilrijkstraat 10, B‐2650 Antwerp, Belgium. ularly used bilevel positive airway pressure or continuous positive airway pressure
Email: [email protected]
at home, and who were diagnosed with an acute infection, and were hospitalized at
the pediatrics department or pediatric intensive care unit. Capillary blood gas
analysis and parameters of the built‐in software of the home ventilator were
recorded.
Results: Among the 43 cases included, hypercapnia was prevalent in 23% with a
mean partial pressure of carbon dioxide of 51.7 ± 6.4 mm Hg. These children also
had lower oxygen saturation levels. The respiratory rate 48 hours before admission
was significantly higher in the hypercapnic group and the volume guarantee mode
was less frequently used in the hypercapnic group.
Conclusion: Approximately, a quarter of the cases of chronic home ventilation ex-
perience hypercapnia during an acute infection. Our data warrant a prospective
study on the monitoring of respiratory rate in patients with chronic respiratory
insufficiency as an indicator for hospitalizations with hypercapnia; we also re-
commend the use of volume guarantee mode of ventilation to prevent hypercapnia.

KEYWORDS

acute infection, children, home ventilation, hypercapnia

1 | INTRODUCTION innovations that have made CPAP and NIV at home more feasible.1
The increasing prevalence of children on CPAP and NIV poses sev-
Continuous positive airway pressure (CPAP) and chronic noninvasive eral challenges considering that they are an inherently vulnerable
ventilation (NIV) in children has increased in recent years. This is patient group. For instance, Kun et al2 described that children who
partially explained by better medical care, which improves the are ventilator dependent are at risk for nonelective hospitalizations,
prognosis of children with complex conditions, and technical with acute infections as the main cause. The aim of this study is to

Pediatric Pulmonology. 2020;1–5. wileyonlinelibrary.com/journal/ppul © 2020 Wiley Periodicals LLC | 1

2 | BEYLTJENS ET AL.

examine the prevalence of hypercapnia in children with chronic signed‐rank test was used. P < .05 was defined as statistically
ventilatory support during an acute infection requiring hospitaliza- significant.
tion. Second, we wanted to identify factors that are different in the
hypercapnic group from both clinical and ventilation perspectives.
These factors were collected at two time points (48 hours and 3 | RE SU LTS
2 weeks before admission) to examine whether early identification of
respiratory insufficiency would be possible. Based on our findings, A total of 74 patients were included in this retrospective study.
there could be a future opportunity to modify these factors and to Between January 2011 and March 2019, there were 76 hospital
prevent some of these acute admissions. admissions for acute infections in this group, which corresponds to an
incidence of 13% per year. The 74 cases we analyzed 43 cases for
further analysis (Figure 1): 59% of these patients had an underlying
2 | MAT E R I AL S A N D M E TH O DS neuromuscular disease, 18% had cerebral palsy, 14% had an under-
lying thoracic or lung disease including Jeune syndrome, lung hypo-
This retrospective, cross‐sectional study was approved by the Ethics plasia, or tracheomalacia, and 9% had Down syndrome with severe
Committee of the Antwerp University Hospital. Included patients
were part of the chronic CPAP and NIV population, followed
between January 2011 and March 2019 at the Department of
Pediatric Pulmonology at the Antwerp University Hospital. The in-
clusion criteria were as follows: aged between 0 and 18 years, regular
use of bilevel positive airway pressure (BiPAP) or CPAP at home,
hospitalization at the pediatrics department or intensive pediatric
care unit (PICU) with a diagnosis of acute infection. Hospitalizations
for noninfectious causes or scheduled admissions were excluded.
Hospitalizations with missing partial pressure of carbon dioxide
(pCO2) measurements were excluded from the analysis.
Hypercapnia was defined in our study as a pCO2 greater than
45 mm Hg in a capillary or arterial blood gas sample, obtained at
initial presentation. Oxygen saturation (SaO2) measured by a pulse
oximeter was also recorded. Secondary outcome parameters in-
cluded the following ventilator settings at the time of admission:
targeted tidal volume (TTV), back‐up respiratory rate, inspiratory
positive airway pressure (IPAP), and expiratory positive airway
pressure (EPAP) in cases of BiPAP and positive end‐expiratory
pressure (PEEP) in cases of CPAP. Data from the ventilator‐in‐built
software were also obtained (tidal volume [TV], respiratory rate,
achieved pressure during inspiration (IPAP) and expiration (EPAP),
and CPAP‐pressure) from 48 hours before admission and at 2 weeks
before admission. These data were collected as mean values during a
48‐hour observation period. This way, it would be shown whether or
not the ventilation settings were met during this period and whether
or not there was a visible trend in the 2 weeks before admission. We
also collected baseline characteristics, clinical management and in-
formation concerning complications during hospitalization. Data
were collected by searching for medical records. All patients
(including those on CPAP) were ventilated using the Trilogy
100 ventilator and data were extracted using the Direct View pro-
gram (versions 2.4.1 and 2.4.2) (Philips Respironics, Murrysville).
Statistical analyses were performed using IBM SPSS Statistics.
Data are expressed as the mean ± standard deviation, the median
with the interquartile range, and proportions (absolute frequencies)
as appropriate. The Mann‐Whitney test was used to compare con-
F I G U R E 1 Study flow chart. BiPAP, bilevel positive airway
tinuous variables, while the χ² test or Fisher’s exact test were used to pressure; CPAP, continuous positive airway pressure; pCO2, partial
compare proportions. For comparison of paired data, the Wilcoxon pressure of carbon dioxide
BEYLTJENS ET AL. | 3

obstructive sleep apnea. All of our patients underwent regular of acute infection, the total incidence of nonelective hospitalizations
poly(somno)graphy and/or saturation and transcutaneous CO2 is underestimated when compared with Kun et al,2 where children
monitoring to control their CPAP or NIV settings. All the study pa- with other indications, such as tracheostomy obstruction and failure
tients had normal gas exchange studies under stable conditions. None to thrive, were also included. Moreover, our community care is well
of the included patients were using chronic supplemental oxygen. expanded and children are followed‐up closely in specialized centers,
The prevalence of hypercapnia was 23% (10/43) in this study so the threshold for hospital admission is relatively high.
population, with a mean pCO2 of 51.7 ± 6.4 mm Hg. The clinical pCO2 values were available only in approximately 50% of our
characteristics of the groups with and without hypercapnia are hospitalizations. The main reason for this was the absence of a
shown in Table 1. Patients with hypercapnia tended to have lower standard protocol for acute admissions for this specific population.
oxygen saturation levels. The majority of patients received BiPAP In the present study, hypercapnia was present in 23% of patients
ventilation at home, and there was no difference in the proportion of with chronic home ventilation admitted to the hospital. To the best of
CPAP and BiPAP users between the two groups. All patients with our knowledge, this is the first study to report this prevalence in this
hypercapnia at admission were diagnosed with a lower airway in- specific pediatric patient population. Indicators for the severity of
fection. In normocapnic patients, this was the most frequent diag- acute infection, including the need for additional oxygen, length of
nosis. We noted that 12% (5/43) of the patients needed to be oxygen supplementation, need for intensive care and length of hos-
admitted to the PICU from day 1 of hospitalization, another 9.3% pital stay, were not significantly different in patients with hy-
(4/43) had to be transferred during hospitalization. The proportion of percapnia compared to patients with normal pCO2 values. Previous
patients requiring PICU care was not different between the studies in different patient groups described a correlation between
two groups. Patients with hypercapnia had a higher respiratory rate hypercapnia and ICU mortality and morbidity.3,4 In these studies, a
at presentation. This difference remained significant after controlling higher limit for pCO2 of 50 mm Hg was used, and the correlation was
for age. All patients on BiPAP were ventilated on the S/T setting of also the strongest for hypercapnic acidosis. In our study, the mean
the Trilogy 100 ventilator. Cases with hypercapnia were significantly pH value at admission was not significantly lower in the hypercapnic
less frequent on average volume‐assured pressure support (AVAPS) group than in the normocapnic patients. There was no mortality
compared to the normocapnia group. Other ventilation parameters associated with any of the cases described here.
were similar between the groups. After analysis of the ventilator data from 48 hours before ad-
Complete ventilator readings were available for 20 out of 43 cases. mission in the BiPAP group, only the respiratory rate and the
In 25 cases, we analyzed respiratory rate, EPAP, and IPAP in patients percentage of patients on AVAPS were significantly different be-
using BiPAP. Overall, we did not observe a significant change between tween groups. Borel et al5 found similar findings in adult patients
TV, IPAP or EPAP values from 2 weeks before to 2 days before hospital with chronic obstructive pulmonary disease (COPD); the risk of an
admission. Respiratory rate 48 hours before presentation was acute exacerbation increased when the respiratory rate was higher
significantly higher in hypercapnic patients (38.3 ± 11.7; P = .03). The before admission. The study by Blouet et al,6 published in 2018,
Wilcoxon signed‐rank test indicated that the respiratory rate 2 weeks showed the same results in patients with COPD. Further readings of
before presentation (26.0 ± 7.6 bpm) was lower than the respiratory ventilator data showed an increase in the respiratory rate compared
rate 48 hours before presentation. However, this was not significant to the steady‐state value 2 weeks before presentation. An increase in
(27.3 ± 9.9 bpm; P = .3). Similar results were observed when the the respiratory rate is an adequate response to increased ventilation
hypercapnia group was analyzed separately; the TV, IPAP, and EPAP of and excretion of CO2. However, when there is insufficient support
respiratory rate did not change significantly over time, but respiratory for ventilation, tachypnea can lead to an increase in dead space
rate rose between 2 weeks before admission (33.9 ± 11.8 bpm) to ventilation and a subsequent rise in carbon dioxide in the blood. As
48 hours before admission (38.3 ± 11.7 bpm; P = .2). there was no concomitant increase in TV or inspiratory pressure,
which supports this hypothesis. Since TV in our patients before
hospital admission were approximately 100% of the TTV, we suggest
4 | D I S C U S SI O N increasing these settings for TTV when signs of a lower respiratory
infection appear. Another important point is that AVAPS was sig-
The present retrospective study showed that hypercapnia during an nificantly less frequently used in the hypercapnic cases. AVAPS at-
acute infection occurs in approximately a quarter of the children with tempts to deliver a consistent TV by adjusting the inspiratory
chronic NIV. Tachypnea is associated with hypercapnia, and these pressures between predefined ranges. All of our patients with un-
patients are less frequently ventilated using a volume guarantee mode. derlying neuromuscular disease are placed on AVAPS in our center,
The overall annual incidence rate for nonelective hospital but our study suggests that it would be recommended to expand this
admission because of an acute infection was 13% in our study. to all our patients. The usefulness of AVAPS in children with chronic
Compared to Kun et al,2 who reported that children who are venti- NIV warrants further research. Evidence demonstrating its super-
lator dependent are at risk for nonelective hospitalizations, with iority is sparse, but there have been studies that have demonstrated
an incidence of 40% per 12 months, our numbers are substantially comparable efficacy between the AVAPS and traditional modes.
lower. However, since we only included cases with a clear diagnosis Studies in adults with chronic respiratory failure involve small sample
4 | BEYLTJENS ET AL.

T A B L E 1 Comparison between patients with and without hypercapnia

Hypercapnia (pCO2 > 45 mm Hg) Normocapnia (pCO2 ≤ 45 mm Hg)
N = 10 N = 33 P value

Age, y 4.3 (4.7) 7.7 (5.7) .1

Sex—male (n [%]) 7 (70.0) 19 (58.8) .7

Mode of home ventilation (n [%])

BiPAP 10 (100) 30 (90.9)
CPAP 0 (0) 3 (9.1) .0

Parameters on admission
pCO2 (mm Hg) 51.7 (6.4) 34.5 (6.0) <.001
SaO2 (%) 87.1 (7.0) 92.5 (7.9) .06
pH 7.4 (0.06) 7.4 (0.08) .1

Diagnosis (n [%])
LRTI 10 (100) 28 (84.8)
URTI 0 2 (6.1)
Gastrointestinal infection 0 3 (9.1) .7

Department of Admission (n [%])

Pediatrics 7 (70.0) 31 (93.9)
PICU 3 (30.0) 2 (6.1) .07

Hospitalization characteristics
Need for oxygen supplementation (n [%]) 8 (80.0) 16 (48.5) .2
Duration of oxygen supplementation (d) 5.7 (1.9) 6.9 (6.0) .8
Intensive care admission (n [%]) 0 4 (12.1) .6
LOS, d 7.5 (4.9) 8.4 (7.4) .9

Ventilatory settings on admission

TTV (mL/kg) 8.0 (1.1) 7.8 (1.1) .7
AVAPS (n [%]) 3 (30%) 24 (80%) .006
IPAP, cm H2O 17.1 (4,0) 18.1 (4.9) .5
EPAP, cm H2O 6.8 (1.8) 5.8 (1.0) .09
Back‐up respiratory rate, bpm 19.8 (11.1) 18.8 (5) .5
Ventilator readings 48 h prehospitalization in patients using BiPAP
TV, mL/kg 6.0 (1.5) 7.3 (2.4) .14
TV (% TTV) 92.9 (9.0) 91.5 (19.3) .7
IPAP (cm H2O) 15.9 (4.5) 16.2 (4.1) .0
IPAP (% IPAP set) 93.4 (3.8) 84.4 (12.0) .07
EPAP, cm H2O 6.2 (1.7) 5.5 (1.0) .44
EPAP (% EPAP set) 98.3 (4.1) 97.7 (13.8) .9
Respiratory rate, bpm 38.3 (11.7) 27.3 (6.6) .03
Respiratory rate (% of back‐up) 180.1 (83.0) 143.1 (33.5) .6
Leak, L/min 36.4 (9.5) 34.9 (24.6) .8

Ventilator readings 14 d prehospitalization in patients using BiPAP

TV (mL/kg) 6.0 (2.0) 7.0 (2.1) .2
TV (% TTV) 103.9 (4.5) 87.8 (21.1) .4
IPAP (cm H2O) 15.5 (5.0) 15.7 (4.4) .0
IPAP (% IPAP set) 90.7 (7.8) 81.7 (12.9) .2
EPAP (cm H2O) 6.3 (1.8) 5.6 (1.1) .4
EPAP (% EPAP set) 100.4 (7.8) 97.8 (13.9) .7
Respiratory rate, bpm 33.9 (11.8) 26.0 (7.6) .1
Respiratory rate (% of back‐up) 155.9 (70.1) 135.6 (31.5) .8
Leak, L/min 32.8 (10.7) 29.8 (20.0) .8

Note: Values in this table are presented as the mean with the standard deviation in parenthesis, unless indicated otherwise.
Abbreviations: AVAPS, average volume‐assured pressure support; BiPAP, bilevel positive airway pressure; CPAP, continuous positive airway pressure;
EPAP, expiratory positive airway pressure; IPAP, inspiratory positive airway pressure; LOS, length of hospital stay; LRTI, lower respiratory tract infection;
pCO2, partial pressure of carbon dioxide; PICU, pediatric intensive care unit; SaO2, oxygen saturation; TTV, targeted tidal volume per kg; TV, tidal volume
per kg; URTI, upper respiratory tract infection.
BEYLTJENS ET AL. | 5

sizes but have shown some promise. The benefits noted with AVAPS, 2. Kun SS, Edwards JD, Ward SL, Keens TG. Hospital readmissions for
however, did not translate into increased survival, decreased hospi- newly discharged pediatric home mechanical ventilation patients.
Pediatr Pulmonol. 2012;47(4):409‐414.
talizations, or improved quality of life compared to BiPAP.7
3. Nin N, Muriel A, Peñuelas O, et al. Severe hypercapnia and outcome of
The main limitations of this study is its retrospective nature. Clinical mechanically ventilated patients with moderate or severe acute
scores reflecting the severity of acute respiratory infections were not respiratory distress syndrome. Intensive Care Med. 2017;43(2):200‐208.
available. We would also like to stress that there is a relatively high 4. Tiruvoipati R, Pilcher D, Buscher H, Botha J, Bailey M. Effects of
hypercapnia and hypercapnic acidosis on hospital mortality in
threshold for PICU admission in our center and that our pediatric ward
mechanically ventilated patients. Crit Care Med. 2017;45(7):e649‐e656.
has experience in managing children on continuous NIV and tracheost- 5. Borel JC, Pelletier J, Taleux N, et al. Parameters recorded by software
omy. Finally, a prospective study could also assess the change in para- of noninvasive ventilators predict COPD exacerbation: a proof‐of‐
meters between baseline, stable condition and during an exacerbation. concept study. Thorax. 2015;70(3):284‐285.
6. Blouet S, Sutter J, Fresnel E, Kerfourn A, Cuvelier A, Patout M.
In conclusion, our data warrant a prospective study on the mon-
Prediction of severe acute exacerbation using changes in breathing
itoring of the respiratory rate in patients with chronic respiratory pattern of COPD patients on home noninvasive ventilation. Int
insufficiency as an indicator for hospitalizations with hypercapnia and J Chronic Obstruct Pulm Dis. 2018;13:2577‐2586.
on the use of AVAPS as a protective tool against hypercapnia. 7. Pluym M, Kabir AW, Gohar A. The use of volume‐assured pressure
support noninvasive ventilation in acute and chronic respiratory
failure: a practical guide and literature review. Hosp Pract (1995).
CO NFLICT OF I NTERE STS
2015;43(5):299‐307.
The authors declare that there are no conflict of interests.

OR CID
How to cite this article: Beyltjens T, de Leede SR, van Eekelen
Stijn Verhulst http://orcid.org/0000-0002-1922-9716
I, et al. The prevalence of hypercapnia during acute infection
in children on chronic noninvasive ventilation: A retrospective
REFERENC ES
study. Pediatric Pulmonology. 2020;1–5.
1. Amin R, Sayal P, Syed F, Chaves A, Moraes TJ, MacLusky I. Pediatric
long‐term home mechanical ventilation: twenty years of follow‐up https://doi.org/10.1002/ppul.24811
from one Canadian center. Pediatr Pulmonol. 2014;49(8):816‐824.