Continuing Medical Education Article

Noninvasive positive-pressure ventilation avoids recannulation andfacilitates early weaning from tracheotomy in children*

Brigitte Fauroux, MD, PhD; Nicolas Leboulanger, MD; Gilles Roger, MD; Francoise Denoyelle, MD, PhD;Arnaud Picard, MD, PhD; Erea-Noel Garabedian, MD; Guillaume Aubertin, MD; Annick Clement, MD, PhD

Objective: To show that noninvasive positive-pressure ventilationby means of a nasal mask may avoid recannulation after decannu-lation and facilitate early decannulation.

Design: Retrospective cohort study.Setting: Ear-nose-and-throat and pulmonary department of a pe-

diatric university hospital.Patients: The data from 15 patients (age � 2–12 yrs) who needed

a tracheotomy for upper airway obstruction (n � 13), congenitaldiaphragmatic hypoplasia (n � 1), or lung disease (n � 1) wereanalyzed. Four patients received also nocturnal invasive ventilatorysupport for associated lung disease (n � 3) or congenital diaphrag-matic hypoplasia (n � 1). Decannulation was proposed in all patientsbecause endoscopic evaluation showed sufficient upper airway pa-tency and normal nocturnal gas exchange with a small size closedtracheal tube, but obstructive airway symptoms occurred either im-mediately or with delay after decannulation without noninvasivepositive-pressure ventilation.

Interventions: In nine patients, noninvasive positive-pressure ven-tilation was started after recurrence of obstructive symptoms after adelay of 1 to 48 mos after a successful immediate decannulation.Noninvasive positive-pressure ventilation was anticipated in six pa-

tients who failed repeated decannulation trials because of poorclinical tolerance of tracheal tube removal or tube closure duringsleep.

Measurements and Main Results: After noninvasive positive-pres-sure ventilation acclimatization, decannulation was performed withsuccess in all patients. Noninvasive positive-pressure ventilation wasassociated with an improvement in nocturnal gas exchange andmarked clinical improvement in their obstructive sleep apnea symp-toms. None of the 15 patients needed tracheal recannulation. Nonin-vasive positive-pressure ventilation could be withdrawn in six pa-tients after 2 yrs to 8.5 yrs. The other nine patients still receivenoninvasive positive-pressure ventilation after 1 yr to 6 yrs.

Conclusions: In selected patients with upper airway obstruction orlung disease, noninvasive positive-pressure ventilation may repre-sent a valuable tool to treat the recurrence of obstructive symptomsafter decannulation and may facilitate early weaning from tracheot-omy in children who failed repeated decannulation trials. (Pediatr CritCare Med 2010; 11:31–37)

KEY WORDS: upper airway; decannulation; congenital airway abnor-malities; noninvasive positive-pressure ventilation; child

LEARNING OBJECTIVESAfter participating in this educational activity, the participant should be better able to:1. Understand the use of the technique of noninvasive positive pressure ventilation in children with tracheotomy.2. Recognize factors favorably influenced by noninvasive positive pressure ventilation in children with tracheotomy.3. Understand the factors associated with the successful transition to noninvasive positive pressure ventilation in childrenwith tracheotomy.Unless otherwise noted below, each faculty or staff’s spouse/life partner (if any) has nothing to disclose.The authors have disclosed that they have no financial relationships with or interests in any commercial companiespertaining to this educational activity.All faculty and staff in a position to control the content of this CME activity have disclosed that they have no financialrelationship with, or financial interests in, any commercial companies pertaining to this educational activity.Visit the Pediatric Critical Care Medicine Web site (www.pccmjournal.org) for information on obtaining continuing medicaleducation credit.

*See also p. 146.Professor (BF, GA, AC), Pediatric Pulmonary,

Hopital Armand Trousseau, Paris, France; AssociateSurgeon (NL, GR), Otolaryagology-Head & Neck Sur-gery Department, Hopital Armand-Trousseau, Univer-sitie Pierre et Marie Curie, Paris, France; AssociateProfessor (FD), Otolaryagology-Head & Neck SurgeryDepartment, Hopital Armand Trousseau, UniversitiePierre et Marie Curie, Paris, France; Professor (AP),

Plastic Surgery and Maxillo-Facial Department, HopitalArmand-Trousseau, Paris, France; and Professor(E-NG), Chief of Department (E-NG), Hopital Armand-Trousseau, Universitie Pierre et Marie Curie, Paris,France; and Professor (GA), Pediatric Pulmonary, Ho-pital Armand Trousseau, Paris, France.The research is supported, in part, by the Associ-

ation Francaise contre les Myopathies (AFM), Assis-tance Publique-Hopitaux de Paris, INSERM, Legs Poix,

ADEP Assistance, and Universite Pierre et Marie Curie-Paris 6 (BF).For information regarding this article, E-mail:

brigitte.fauroux@trs.aphp.frCopyright © 2010 by the Society of Critical Care

Medicine and the World Federation of Pediatric Inten-sive and Critical Care Societies

DOI: 10.1097/PCC.0b013e3181b80ab4

31Pediatr Crit Care Med 2010 Vol. 11, No. 1

Congenital or acquired upperairway abnormalities are com-mon in children. Laryngealabnormalities (such as laryn-

gomalacia, subglottic stenosis, or vocalcord paralysis), tracheal abnormalities(such as tracheomalacia or tracheal ste-nosis) as well as Pierre Robin syndromemay be responsible for severe upper air-way obstruction that may persist despitemedical and surgical treatment. In thesecases, tracheotomy is indicated to pre-vent potentially serious complications,such as airway obstruction and suddendeath (1, 2), pulmonary hypertension andcor pulmonale (2, 3), failure to thrive (4,5), and neurocognitive dysfunction withthe risk of intellectual impairment (6). Incase of severe lung disease, tracheotomymay be indicated to allow invasive venti-lation in order to improve alveolar venti-lation (7).

However, tracheotomy is associatedwith a significant morbidity and discom-fort and may impair normal developmentand, particularly, language development(8, 9). Discomfort and social life and fam-ily disruption are common in patientswith a tracheotomy (10). A recent studyhas shown that parents caring for chil-dren with tracheotomy tubes experiencesignificant caregiver burden and that themental health status for an adult caringfor a child with a tracheotomy tube issignificantly reduced (10). Although tra-cheotomized children may be safely dis-charged home after careful family educa-tion and training, home treatment maybe difficult or even unfeasible for somefamilies (7, 11). Thus, whenever possible,decannulation should be proposed assoon as possible. But decannulation fail-ure is not uncommon, and apart from themedical consequences, the psychologicalconsequences on the child and the familyare important to consider. Nursing staff’sobservations of restlessness, anxiety, anddepression appeared more frequently inchildren who failed decannulation (12).

Noninvasive positive-pressure ventila-tion (NPPV), which consists of the deliv-ery of positive airway pressure by meansof a nasal mask, has been shown to re-duce the work of breathing in childrenwith upper airway obstruction associatedwith alveolar hypoventilation (13–16).NPPV, by maintaining the patency of theupper airways during the breathing cycle,allows an increase in tidal volume andminute ventilation, which translates intoan improvement in gas exchange (15, 16).NPPV has also demonstrated its benefit in

children with chronic lung disease, suchas cystic fibrosis (17–19). In recent years,we have used NPPV in patients who havehad their tracheotomy removed and sub-sequently developed recurrent airway ob-struction. We then used NPPV to facili-tate early decannulation in patients whofailed repeated decannulation trials. Inthe present study, we relate our experi-ence over the last 12 yrs.

MATERIALS AND METHODS

A tracheotomy was performed in 171 pa-tients during the 12-yr study period (1996–2008) (Fig. 1). Seventy-one patients (42%)have been decannulated successfully and re-mained asymptomatic on systematic follow-upexaminations. Sixty patients are still tracheot-omized and undergo regular evaluations.Nineteen of these 60 patients could be futurecandidates for decannulation because a suffi-cient improvement in upper airway obstruc-tion may be expected in the future. Theother patients have too severe upper airwayobstruction or associated morbid condi-tions. Nineteen patients returned to theirprimary hospital and have been lost to fol-low-up. Six patients with severe underlyingconditions died.

Fifteen tracheotomized patients (8.8%)have been treated with NPPV during the studyperiod and are analyzed in the present report.Written approval for care was provided by allparents and the analysis of the data was ap-proved by the local ethical committee.

In a first approach, NPPV was used to treatthe recurrence of obstructive symptoms afteran immediate successful decannulation (de-layed NPPV group). Patients with tracheoto-mies undergo regular video-endoscopic evalu-ation of the upper airways under generalanesthesia (spontaneous breathing and as-sisted ventilation). Decannulation is proposedwhen the following criteria are fulfilled:

1. Sufficient airway patency during spontane-ous breathing in the operating room wasevaluated by endoscopy and reflected bynormal breathing and normal gas exchangein room air.

2. Adequate airway patency during sleep wasassessed by the absence of obstructivesymptoms, such as stridor, agitation,arousals, night sweats, and the absence ofnocturnal hypoxemia (�5 consecutivemins with a pulse oximetry [SpO2] �90%)with hypercapnia (transcutaneous carbondioxide [PtcCO2] �50 torr, �6,7 kPa) whilesleeping in room air with a closed trachealtube. Before this test, the tracheal tube ischanged for a smaller model to favor easierbreathing during nocturnal sleep.

All patients who fulfilled these criteriawere decannulated without NPPV. A regular3-mo follow-up was systematically performedwith clinical examination, endoscopic evalua-tions when indicated, and nocturnal gas ex-change recordings at least every 6 mos.

In a second approach, we used NPPV tofacilitate early weaning from tracheostomy(immediate NPPV group). In these patients,airway patency with a closed smaller tube wasnot sufficient during nocturnal sleep as re-flected by obstructive symptoms (describedabove) and abnormal nocturnal gas exchangein room air with at least 5 consecutive minswith a SpO2 �90% and/or a PtcCO2 �50 torr(�6,7 kPa). In this situation, or after previousseveral failed decannulation attempts, NPPVwas proposed before decannulation.

In this immediate NPPV group, the decan-nulation procedure was as follows. The firststep consisted of the acclimatization of thepatient to the nasal mask only, without theventilator. This step took between 2 to 15 days,depending on the patient’s age, and his/hermedical and psychological history. Then,when the patient accepted to wear the nasalmask with the headgear, NPPV without seda-tion was tried for short periods, lasting 2 to 5mins, which were repeated during daytime.When daytime tolerance exceeded 15 contin-uous mins, NPPV was tried during the night as

Figure 1. Flow chart of diaphragm of patients.NPPV, noninvasive positive-pressure ventilation.

32 Pediatr Crit Care Med 2010 Vol. 11, No. 1

the tracheal tube was closed. Because of thepartial airway obstruction due to the presenceof the tube, the inspiratory and expiratoryNPPV pressures were set 2 cm H2O above thelevels that would have been chosen withoutthe presence of the tube. Overnight SpO2,PtcO2, and PtcCO2 recordings were systemati-cally performed. At initiation, NPPV was per-formed at night and during daytime with threeto four breaks of 2 to 3 hrs. But within 1 wk,all the patients were able to use NPPV exclu-sively during sleep, at night in all the patientsand during daytime naps in the youngest pa-tients. The tracheal tube was then removedand NPPV settings were adjusted to obtain anormal breathing pattern without stridor andadequate gas exchange, as defined by the ab-sence of nocturnal hypoxemia (�5 consecu-tive mins with an SpO2 �90%) and hypercap-nia (no periods with a PtcCO2 �50 torr [�6,7kPa]). The acceptance and optimal setting ofNPPV lasted 3 to 15 more days, depending onthe age of patient, his/her medical history,anxiety, and psychological stress.

NPPV was always performed by pressure-controlled ventilators (Harmony or Synchrony(Respironics, Craquefou, France), VPAP 3STor STA (Resmed, Saint Priest, France), Vivo 40(Breas Medical, Saint Priest, France), orKnightstar (Tyco Healthcare, Elancourt,France), delivering bilevel positive airwaypressure by means of a commercially available(Respironics, Resmed or Fisher Paykel nasalmasks), or custom-made nasal mask. Thesecustom-made masks were composed of a ther-moformable plastic frame (VT Plastics, Gen-nevilliers, France) with an interior coverage ofeither self-sticking foam (Adhesia Laboratoire,Mulhouse, France) or a protection and com-fort gel (Adhesia Laboratoire, Mulhouse,France). The nasal mask was connected to anexpiratory valve and a nonrebreathing circuitby a plastic tube of an inner diameter of 22mm, which was fixed on the mask by an au-topolymerizable resin (Orthoresin, Dentsply,Weybridge, United Kingdom). The masks weremodeled on plaster phantoms correspondingto the age and the physiognomy of the patient.Bedside adjustments were then realized, ifnecessary, by thermoforming the plastic frameto obtain the best comfort with minimal leaks(20). Custom-made masks were used in case ofage �2 yrs, facial deformity, and/or intoler-ance of an industrial mask (20). Inspiratorypressures of 6 to 10 cm H2O and expiratorypressures of 4 to 6 cm H2O were used, with aramp when available and a back-up rate of 2 to5 breaths below the patient’s spontaneousbreathing rate.

Nocturnal gas exchange was routinely eval-uated by oxygen and CO2 recording, either bythe Tina monitor (Tina, Radiometer, Copen-hague, Denmark), or by the SenTec DigitalMonitor (SenTec, Therwil, Switzerland) dur-

ing the acclimatization to NPPV. Sleep studieslooked for obstructive apnea, which was de-fined as the absence of air flow with continuedchest-wall and abdominal movement for atleast two breaths (21, 22). Hypopnea was de-fined as a decrease in nasal flow of �50% witha corresponding decrease in SpO2 �4% and/orwith associated arousal. The apnea index andthe hypopnea index were defined as the num-bers of apnea and hypopneas per hour of totalsleep time. A desaturation was defined as adecrease of SpO2 �4% below baseline and thedesaturation index was calculated as the num-ber of desaturations per hour of total sleeptime (21, 22). Evaluation of diurnal gas ex-change was assessed by arterialized capillaryblood gases in the morning, after a night ofNPPV (23). Initially, the tracheal stoma wasoccluded by a sticking plaster. The number ofpatients who needed a surgical closure of thetracheal stoma was recorded.

Discharge to home with NPPV was allowedwhen the following criteria were fulfilled:

● Ability to sleep at least 5 hrs with NPPV;● Absence of nocturnal hypoxemia or hyper-

capnia while on NPPV without supplemen-tal oxygen;

● Parents and family educated to NPPV.

RESULTS

Delayed NPPV Group

The delayed NPPV group comprisednine patients (Table 1). In seven patients(patients 1 to 7), tracheotomy was per-formed before 6 mos of age, because ofvocal cord paralysis (which was alwaysassociated with another cause of airwayobstruction) in four patients (patients 2,3, 6, and 7), Treacher-Collins syndromein one patient (patient 1), congenital pa-ralysis of the diaphragm in one patient(patient 4), and cystic lymphangioma inone patient (patient 5). Patients 4 and 6received nocturnal NPPV on the trachealtube before decannulation because of di-aphragmatic paralysis and associatedbronchopulmonary dysplasia, respec-tively. Immediate decannulation was welltolerated by all patients, without clinicalsymptoms of upper airway obstruction,nocturnal desaturations, or hypercapniaas previously defined. However, after adelay of 1 to 48 mos, all patients devel-oped symptoms of obstructive sleep ap-nea with stridor, night sweats and arous-als, daytime fatigue, and change in moodand attention. Nocturnal SpO2, PtCO2, andPtcCO2 recordings in room air showed thepresence of apneas and hypopneas, asso-ciated with periods of desaturation and

hypercapnia (Table 2). These obstructivesymptoms were explained by an increaseof pharyngolaryngeal hypotonia or the re-currence of the primary disorder forwhich no satisfactory surgical interven-tion could be proposed. NPPV was initi-ated to avoid recannulation. Nocturnalbilevel positive-pressure ventilation wasperformed with inspiratory pressures of 6to 10 cm H2O and expiratory pressures of4 to 6 cm H2O. The five youngest patientswere equipped with a custom-mademask, whereas the four older patientsused a commercially available mask. Thedaytime and nocturnal gas exchange in-dices and sleep parameters exhibited atrend toward improvement but becauseof the small number of patients, the dif-ferences did not reach statistical signifi-cance (Table 2). All the patients were dis-charged home. A surgical closure of thetracheal stoma was required in four pa-tients (patients 1, 2, 8, and 9). NPPVcould be definitely withdrawn in threepatients, after 2 yrs in patients 2 and 5,and after 8.5 yrs in patient 8. At follow-up, none of the patients needed a recan-nulation and none of the patients died.

Immediate NPPV Group

The immediate NPPV group com-prised six patients (Table 1). In four ofthese patients, tracheotomy was per-formed before 3 mos of age because ofsevere upper airway obstruction in theneonatal period. The two other patientswere tracheotomized at ages 1.5 yrs and2.7 yrs because of mandibular hypoplasia(patient 14) and chronic respiratory in-sufficiency related to an acute respiratorydistress syndrome of unknown origin (pa-tient 15). This latter patient received noc-turnal pressure-controlled ventilation be-cause of his associated lung disease aswell as patient 10 because of associatedbronchopulmonary dysplasia. In thesetwo patients, NPPV was started beforedecannulation because of the need tocontinue nocturnal ventilatory support.Once the decision had been made to at-tempt tracheotomy removal, a smallertube was placed to facilitate spontaneousbreathing. In the four patients who werenot on long-term invasive ventilation, de-cannulation without NPPV was not pos-sible because of nocturnal hypoventila-tion due to upper airway obstruction. Allpatients developed clinical symptoms ofupper airway obstruction with stridor,chest retractions, and night sweats, asso-ciated with periods exceeding 5 continu-

33Pediatr Crit Care Med 2010 Vol. 11, No. 1

ous mins with a SpO2 �90% and a PtcCO2

�50 torr (�6,7 kPa) during tracheal tubeclosure and/or removal trials. In thesepatients, no surgical option was availableto facilitate decannulation without seri-ous drawbacks, such as an increase in therisk of false passages. Previous attemptsof decannulation without NPPV had beenundertaken at least two times in all sixpatients over long periods (mean pe-riod � 14.3 mos; range � 6–32 mos)without success. Mean age at decannula-tion in these patients was 6.7 yrs(range � 2.5–12 yrs). The daytime andnocturnal gas exchange indices and sleepparameters with nocturnal NPPV werewithin the normal range (Table 2).

The three youngest patients wereequipped with custom-made masks andthe three older patients with commer-cially available nasal masks. All the pa-tients were discharged home and thethree patients who were in a transitional

care facility were able to return to theirfamilies. After decannulation, all the pa-tients experienced improvements in lan-guage and development. Initially, the tra-cheal stoma was occluded by a plasticsticker, but a secondary surgical closureof the tracheal stoma was necessary in allpatients. Three patients could be weanedfrom NPPV after 2 yrs to 3 yrs and threeothers were still on NPPV at ages 9, 14,and 17 yrs. None of the patients died.

DISCUSSION

This study shows that NPPV is able totreat successfully the recurrence of ob-structive airway disorders after tracheot-omy weaning in children. But, most im-portantly, we show for the first time thatNPPV may facilitate decannulation inchildren who failed repeated decannula-tion trials. Tracheotomy weaning couldonly be achieved with immediate NPPV

support. The use of bilevel positive end-expiratory pressure by means of a nasalmask was associated with an improve-ment of nocturnal gas exchange in allpatients and during follow-up, none ofthe patients needed recannulation.

NPPV has been shown to be an alter-native to invasive ventilation in selectedpatients with neuromuscular disease,managed by highly skilled teams. In ad-olescents and young adults with Duch-enne muscular dystrophy, NPPV bymeans of a nasal mask during the nightand a mouthpiece during the day, associ-ated with cough-assisted techniques, mayallow extubation or decannulation andprolong survival in well-trained andhighly qualified teams (24, 25). Even inyoung infants with spinal muscular atro-phy Type I, a noninvasive respiratorymanagement approach may be successfulin selected cases, with an improvement in

Table 1. Description of the patients

Patient Gender DiagnosisAge at

TracheotomyAge at

Detubation (yrs)NPPV Delay

(mos) Outcome

Delayed NPPVgroup

1 female Treacher-Collinssyndrome

1 mo 2.5 6 on NPPV since 1 mo

2 female Vocal cord paralysis �tracheomalacia

1 mo 2.5 4 successful NPPV withdrawal atage 5, now 7.5 yrs old

3 male Vocal cord paralysis �polymalformation

3 mos 11 48 still on NPPV at age 18

4 male Congenitaldiaphragmatichypoplasia #

3 mos 5.7 12 still on NPPV at age 12

5 female Cystic lymphangioma 6 mos 2 1 successful NPPV withdrawal atage 4, now 11 yrs old

6 male Vocal cord paralysis �BPD #

6 mos 2 1 still on NPPV at age 4

7 male Vocal cord paralysis �multiple congenitalanomalies

6 mos 10 6 still on NPPV at age 11

8 male Laryngeal cleft 1 yr 3 6 successful NPPV withdrawal atage 12, now 13 yrs old

9 female Vocal cord paralysis �cerebral tumor

6.5 yrs 10.5 9 still on NPPV at age 12

Immediate NPPVgroup

10 male Pierre Robin sequence� BPD #

1 mo 3.5 0 successful NPPV withdrawal atage 7, now 8 yrs old

11 female Cystic lymphangioma� mandibularhypoplasia

1 mo 12 0 still on NPPV at age 17

12 male Laryngeal cleft gradeIV

2 mos 2.5 0 successful NPPV withdrawal atage 5, now 8 yrs old

13 male Vocal cord paralysis �tracheomalacia

3 mos 6 0 successful NPPV withdrawal atage 8, now 8.5 yrs old

14 female Mandibular hypoplasia 1.5 yrs 9 0 still on NPPV at age 1415 male ARDS sequelae # 2.7 yrs 7 0 still on NPPV at age 9

NPPV, noninvasive positive-pressure ventilation; BPD, bronchopulmonary dysplasia; ARDS, acute respiratory distress syndrome; #, patients receivingnocturnal positive-pressure ventilation on the tracheal tube before decannulation.

34 Pediatr Crit Care Med 2010 Vol. 11, No. 1

medical condition and quality of life forthe child and his/her family (26, 27).

To our knowledge, information re-garding the use of NPPV to facilitate ex-tubation or tracheotomy weaning inother pediatric diseases is lacking. Indi-cations for tracheotomies in children in-clude airway obstruction, inadequate air-way protection, chronic lung disease,neuromuscular weakness, and central hy-poventilation (7). All these conditionsmay be managed by NPPV, if the patienthas an adequate respiratory autonomy,allowing NPPV to be used preferentiallyat night. Interestingly, regular use ofNPPV at night was associated with animprovement in daytime spontaneousbreathing and gas exchange, as observedin children with other conditions, such asneuromuscular disease (28–30). As such,NPPV could be proposed to a selectedgroup of tracheotomized patients to facil-itate decannulation.

Our protocol is based on the assump-tion that long-term use of NPPV is asso-ciated with a better quality of life for thepatient and his/her family and fewer sideeffects than a prolonged tracheotomy. Toour knowledge, NPPV has not been com-pared with tracheotomy in a prospective,

randomized, controlled trial but the fea-sibility of such a study may be extremelydifficult because of practical and ethicalissues. Even if large series have reporteda low occurrence rate of tracheotomy-related mortality and morbidity, severecomplications may occur during the can-nulation period, such as tube obstructionor dislocation, accidental decannulation,or pneumothorax (7, 31–34). Long-termcomplications include tracheal stenosis,stomal narrowing, and recurrent lung in-fections occurring in 10% to 40% of pa-tients (7, 33, 34). There is also evidencethat children with tracheotomies are atrisk for delays in receptive and expressivelanguage development as well in deficitsin oral/vocal speech and voice production(35–37). NPPV, as a noninvasive tech-nique, is not associated with these nu-merous side effects. Importantly, unex-pected readmission rate may reach 63%in some series of children with tracheot-omy (7), whereas none of the 15 patientsreported in the present study requiredreadmission to the hospital for an acuteupper or lower respiratory tract problemduring the follow-up. Importantly, all thefamilies and patients, when able to ex-

press their will, refused the eventuality ofa recannulation.

We present here a series of 14 patientswith upper airway disease and one patientwith lung disease in whom NPPV was ableto facilitate decannulation or treat ob-struction recurrence. Several requisitesmust be fulfilled for NPPV to be success-ful. First, endoscopic evaluation shouldshow a sufficient airway patency, allow-ing acceptable tolerance of spontaneousbreathing without a tracheal tube. Be-cause of lack of objective measurable cri-teria to estimate airway caliber in youngchildren, this decision relies on the sub-jective estimation of the airway patencyby an experienced pediatric ear-nose-throat surgeon.

Second, NPPV in this age group re-quires some technical requisites. Thechoice of the interface is problematic inyoung children (20). To our knowledge,no adequate commercial nasal masks areavailable for children weighing �10 kg.These young children need thus custom-made masks, which require an experi-enced and motivated pediatric maxillofa-cial team. Because of facial growth andpotential side effects, such as skin injuryand facial deformity, a close follow-up ofthese masks is absolutely necessary (20).With this maxillofacial monitoring, nosignificant skin or facial side effects wereobserved in the patients included in thepresent study.

Finally, the psychological aspect is ofparamount importance. A tracheotomy isan invasive procedure, associated with re-current invasive maneuvers, such as as-piration, tube removal, and change, and,as a consequence, hospital visits and hos-pitalizations. All these children have se-vere, and often multiple medical disor-ders, which contribute to anxiety, pain,and psychological stress, both for thechild and his/her family. This may explainwhy some children resist any procedureinvolving the face or the upper airway. Inmore than half of the children in thepresent study, acclimatization to the na-sal mask took �1 wk, which is twice aslong as our experience with children whodid not have a prior tracheotomy. Theactive role of the parents, but also thenurses, psychologists, and school teach-ers are of great help in such situations.Encouragements and positive rewardswere used by the medical team and familyto help the child to accept NPPV. Onechild, who underwent tracheotomy in theneonatal period for congenital myasthe-nia which was diagnosed only at the age

Table 2. Daytime and nocturnal gas exchange indices

Normal Valuesa

Delayed NPPV Group n � 9Immediate NPPV

Group n � 6Before NPPV With NPPV

Daytimeparameters

PaO2, torr 80 � 10 78 � 8 73 � 2 82 � 7PaCO2, torr 35 � 10 37 � 3 39 � 3 40 � 4

Nocturnalparameters

Apnea index �1 3 � 3 0.0 � 0.0 0.2 � 0.4Hypopnea

index�1 37 � 39 0.6 � 1.3 0.4 � 0.5

% of time withSpO2 �95%

�99% 54 � 33 69 � 29 36 � 29

% of time withSpO2 91%–94%

�1% 29 � 21 29 � 13 63 � 16

% of time withSpO2 �90%

0 % 16 � 34 2 � 6 1 � 1

Desaturationindex

�1 38 � 35 25 � 22 22 � 14

PtcO2 min, torr �70 62 � 27 69 � 11 63 � 11PtcCO2 max,

torr�50 51 � 8 49 � 3 48 � 4

Delta PtcCO2

max, torr�10 15 � 8 11 � 3 9 � 3

NPPV, noninvasive positive-pressure ventilation; PaO2, partial arterial oxygen pressure; PaCO2,

partial arterial carbon dioxide pressure; SpO2, pulse oximetry; PtcO2 min, minimum transcutaneousoxygen pressure; PtcCO2 max, maximum transcutaneous carbon dioxide pressure; delta PtcCO2 max,maximum awake/night change of transcutaneous carbon dioxide pressure.

aAmerican Thoracic Society (21) and Montgomery-Downs et al (22).

35Pediatr Crit Care Med 2010 Vol. 11, No. 1

of 4 yrs, never accepted the nasal maskand NPPV. His spontaneous sleep withouta tracheotomy and ventilatory supportwas acceptable after decannulation butafter 1 yr, he required 5 days of invasiveventilation for an acute viral lower airwayinfection. It is highly possible that, in thisneuromuscular patient, the long-termuse of NPPV would have been able toprevent acute invasive ventilatory sup-port (38, 39). These medical, technical,and psychological requirements may ex-plain why NPPV is not considered as aroutine procedure to facilitate decannu-lation in young children.

We acknowledge that the patients ofthe present study had particularly severeupper airway obstruction that requiredmaintenance of the tracheal tube to a veryadvanced age (�6 yrs of age) whereas de-cannulation for usual upper airway ob-struction is generally achieved before theage of 4 yrs. At that advanced age, ourpatients did not tolerate decannulationwithout NPPV even after several trials, re-flecting the severity of the residual obstruc-tion that NPPV was able to improve.

An important observation from the de-layed NPPV group is that patients withsevere upper airway obstruction need aclose and prolonged medical follow-upafter decannulation. Alveolar hypoventi-lation could recur as long as 48 mos afterdecannulation (patient 3). Apart frommedical and endoscopic examination,careful attention must be paid to sleep-disordered breathing and sleep-relatedproblems. Ideally, all those patientsshould have a complete polysomnogra-phy, which is the gold standard for thediagnosis of sleep-disordered breathing.Such a complete evaluation was not pos-sible in the immediate NPPV group be-cause of the unstable clinical condition ofthe patients. In the delayed NPPV group,some patients had a polysomnographybut others, because of time and practicalconstraints, had only an overnight re-cording of gas exchange. However, theserecordings were always sufficiently ab-normal to institute NPPV without delay.

Finally, the use of NPPV as an alter-native to recannulation and as a weaningtool for tracheotomized patients raisesthe question of the optimal indicationsfor tracheotomy in children. The useful-ness of NPPV as a first-line treatment inselected patients with severe upper air-way obstruction, who have sufficient air-way patency when awake during daytimeand who need positive pressure only dur-ing sleep, is worthy of evaluation.

CONCLUSIONS

To our knowledge, we report here afirst experience of successful use of NPPVto treat recurrent respiratory obstructionafter decannulation and facilitate decan-nulation in young children with severeupper airway obstruction or lung disease.The increasing use of NPPV in children,both in the acute (40) and chronic setting(41, 42), should be extended also to aselected group of tracheotomized pa-tients, to improve the quality of life forthe child and his/her family.

ACKNOWLEDGMENT

We thank Emmanuelle Cohen for herexcellent technical assistance.

REFERENCES

1. Sivan Y, Ben-Ari J, Schonfeld TM: Laryngomala-cia: A cause for early near miss for SIDS. InternJ Pediatr Otorhinolaryngol 1991; 21:59–64

2. Marcus CL, Crockett DM, Davidson Ward SL:Evaluation of epiglottoplasty as treatment for se-vere laryngomalacia. J Pediatr 1990; 117:706–710

3. Naiboglu B, Deveci S, Duman D, et al: Effectof upper airway obstruction on pulmonaryarterial pressure in children. Int J PediatrOtorhinolaryngol 2008; 72:1425–1429

4. Marcus CL, Carroll JL, Koerner CB, et al:Determinants of growth in children with theobstructive sleep apnea syndrome. J Pediatr1994; 125:556–562

5. Roger G, Denoyelle F, Triglia JM, et al: Se-vere laryngomalacia: surgical indications andresults in 115 patients. Laryngoscope 1995;105:1111–1117

6. Kheirandish L, Gozal D: Neurocognitive dys-function in children with sleep disorders.Dev Sci 2006; 9:388–399

7. Graf JM, Montagnino BA, Hueckel R, et al:Pediatric tracheostomies: A recent experi-ence from one academic center. Pediatr CritCare Med 2008; 9:96–100

8. Dubey SP, Garap JP: Pediatric tracheostomy:An analysis of 40 cases. J Laryngol Otol 1999;113:645–651

9. Wetmore RF, Marsh RR, Thompson ME,et al: Pediatric tracheostomy: A changingprocedure? Ann Otol Rhinol Laryngol 1999;108:695–699

10. Hartnick CJ, Bissell C, Parsons SK: The impactof pediatric tracheotomy on parental caregiverburden and health status. Arch OtolaryngolHead Neck Surg 2003; 129:1065–1069

11. Ruben RJ, Newton L, Jornsay D, et al: Home careof the pediatric patient with a tracheotomy. AnnOtol Rhinol Laryngol 1982; 91:633–640

12. Waddell A, Appleford R, Dunning C, et al: TheGreat Ormond Street protocol for ward decan-nulation of children with tracheostomy: In-

creasing safety and decreasing cost. Int J Pedi-atr Otorhinolaryngol 1997; 39:111–118

13. Waters WA, Everett FM, Bruderer JW, et al:Obstructive sleep apnea: The use of nasalCPAP in 80 children. Am J Respir Crit CareMed 1995; 152:780–785

14. Guilleminault C, Pelayo R, Clerk A, et al:Home nasal continuous positive airway pres-sure in infants with sleep-disordered breath-ing. J Pediatr 1995; 127:905–912

15. Fauroux B, Pigeot J, Polkey MI, et al: Chronicstridor caused by laryngomalacia in children.Work of breathing and effects of noninvasiveventilatory assistance. Am J Respir Crit CareMed 2001; 164:1874–1878

16. Essouri S, Nicot F, Clement A, et al: Nonin-vasive positive pressure ventilation in infantswith upper airway obstruction: Comparisonof continuous and bilevel positive pressure.Intensive Care Med 2005; 31:574–580

17. Fauroux B, Pigeot J, Isabey D, et al: In vivophysiological comparison of two ventilatorsused for domiciliary ventilation in childrenwith cystic fibrosis. Crit Care Med 2001; 29:2097–2105

18. Fauroux B, Nicot F, Essouri S, et al: Settingof pressure support in young patients withcystic fibrosis. Eur Resp J 2004; 24:624–630

19. Fauroux B, Louis B, Hart N, et al: The effectof back-up rate during non-invasive ventila-tion in young patients with cystic fibrosis.Intensive Care Med 2004; 30:673–681

20. Fauroux B, Lavis JF, Nicot F, et al: Facial sideeffects during noninvasive positive pressureventilation in children. Intensive Care Med2005; 31:965–969

21. American Thoracic Society: Standards andindications for cardiopulmonary sleep stud-ies in children. Am J Resp Crit Care Med1996; 153:866–878

22. Montgomery-Downs HE, O’Brien LM, Gul-liver TE, et al: Polysomnographic character-istics in normal preschool and early school-aged children. Pediatrics 2006; 117:741–753

23. Gaultier C, Boule M, Allaire Y, et al: Deter-mination of capillary oxygen tension in in-fants and children: Assessment of methodol-ogy and normal values during growth. BullEurop Physiopath Resp 1978; 14:287–294

24. Gomez-Merino E, Bach JR: Duchenne mus-cular dystrophy: Prolongation of life by non-invasive ventilation and mechanically as-sisted coughing. Am J Phys Med Rehabil2002; 81:411–415

25. Toussaint M, Steens M, Wasteels G, et al:Diurnal ventilation via mouthpiece: Survivalin end-stage Duchenne patients. Eur Resp J2006; 28:549–555

26. Bach JR, Niranjan V: Spinal muscular atrophytype I: A noninvasive respiratory managementapproach. Chest 2000; 117:1100–1105

27. Bach JR: The use of mechanical ventilation isappropriate in children with genetically provenspinal muscular atrophy type 1: The motion for.Paediatr Respir Rev 2008; 9:45–50; quiz 50; dis-cussion 55–56

28. Simonds AK, Ward S, Heather S, et al: Out-come of paediatric domiciliary mask ventila-

36 Pediatr Crit Care Med 2010 Vol. 11, No. 1

tion in neuromuscular and skeletal disease.Eur Respir J 2000; 16:476–481

29. Mellies U, Ragette R, Dohna Schwake C, etal: Long-term noninvasive ventilation inchildren and adolescents with neuromus-cular disorders. Eur Respir J 2003; 22:631–636

30. Annane D, Orlikowski D, Chevret S, et al:Nocturnal mechanical ventilation forchronic hypoventilation in patients withneuromuscular and chest wall disorders. Co-chrane Database Syst Rev 2007; 4:CD001941

31. Simma B, Spehler D, Burger R, et al: Trache-ostomy in children. Eur J Pediatr 1994; 153:291–296

32. Carr MM, Poje CP, Kingston L, et al: Com-plications in pediatric tracheostomies. La-ryngoscope 2001; 111:1925–1928

33. Alladi A, Rao S, Das K, et al: Pediatric tra-

cheostomy: A 13-year experience. PediatrSurg Int 2004; 20:695–698

34. Mahadevan M, Barber C, Salkeld L, et al: Pedi-atric tracheotomy: 17 year review. Int J PediatrOtorhinolaryngol 2007; 71:1829–1835

35. Simon B, Handler SD: The speech pathologistand management of children with tracheosto-mies. J Otolaryngol 1981; 10:440–448

36. Kaslon KW, Stein RE: Chronic pediatric tra-cheotomy: Assessment and implications for ha-bilitation of voice, speech and language inyoung children. Int J Pediatr Otorhinolaryngol1985; 9:165–171

37. Arvedson JC, Brodsky L: Pediatric tracheot-omy referrals to speech-language pathologyin a children’s hospital. Int J Pediatr Otorhi-nolaryngol 1992; 23:237–243

38. Young HK, Lowe A, Fitzgerald DA, et al:Outcome of noninvasive ventilation in chil-

dren with neuromuscular disease. Neurology2007; 68:198–201

39. Dohna-Schwake C, Podlewski P, Voit T, et al:Non-invasive ventilation reduces respiratorytract infections in children with neuromus-cular disorders. Pediatr Pulmonol 2008; 43:67–71

40. Essouri S, Chevret L, Durand P, et al: Nonin-vasive positive pressure ventilation: Five yearsof experience in a pediatric intensive care unit.Pediatr Crit Care Med 2006; 7:329–334

41. Fauroux B, Sardet A, Foret D: Home treat-ment for chronic respiratory failure in chil-dren: A prospective study. Eur Resp J 1995;8:2062–2066

42. Fauroux B, Boffa C, Desguerre I, et al: Long-term noninvasive mechanical ventilation forchildren at home: A national survey. PediatrPulmonol 2003; 35:119–125

37Pediatr Crit Care Med 2010 Vol. 11, No. 1