noninvasive respiratory support in infants and...

Noninvasive Respiratory Support in Infants and ChildrenKatherine L Fedor MBA RRT-NPS CPFT

IntroductionIndications

AsthmaBronchiolitisPediatric ARDSCystic FibrosisObstructive Sleep ApneaNeuromuscular DisordersCardiac DiseasePredictors of Success

ContraindicationsInterfacesDelivery DevicesChallenges

Noninvasive Ventilation-Neurally Adjusted Ventilatory AssistSedationAdherence

Measurements and MonitoringClinical Management

CPAPNoninvasive Ventilation

Alternative Approaches to Noninvasive ModalitiesSafety and ComplicationsSummary

CPAP and noninvasive ventilation (NIV) offer an alternative to intubation and mechanical venti-lation in the treatment of acute and chronic respiratory disorders commonly encountered in infantsand children. There are many distinct challenges associated with the application, management, andsafety of CPAP and NIV in the pediatric population. This review attempts to identify indica-tions, contraindications, management strategies, and safety measures associated with the ap-plication of CPAP or NIV delivery in children. More recently, high-flow nasal cannula (HFNC)has emerged as an alternative to CPAP and NIV. Evidence related to the use of CPAP, NIV, andHFNC is included in this review. Key words: noninvasive ventilation; high-flow nasal cannula;continuous positive airway pressure; respiratory distress syndrome; pediatrics. [Respir Care2017;62(6):699 –717. © 2017 Daedalus Enterprises]

Introduction

Noninvasive respiratory support is used in infants andchildren with acute and chronic conditions associated withimpaired respiratory drive, inadequate lung inflation, andaltered gas exchange resulting in respiratory insufficiencyor failure. Previously, a predominance of the evidence wasdevoted to the neonatal and adult population, but recent

evidence advocating for less invasive ventilation has beena catalyst to increase the study and use of CPAP or non-invasive ventilation (NIV) in infants and pediatric patients.CPAP refers to the administration of continuous positiveairway pressure during all phases of the respiratory cycleof a spontaneously breathing patient, whereas NIV refersto the application of 2 pressure levels (inspiratory andexpiratory) designed to enhance ventilation by creating

RESPIRATORY CARE • JUNE 2017 VOL 62 NO 6 699

pressure gradients that facilitate the movement of air inand out of the lungs. Both CPAP and NIV have a goal thatis consistent with improving or adequately restoring thefunctional residual capacity in the lung while minimizingthe inspiratory work of breathing. The primary goal for theuse of CPAP is to improve oxygenation by improvingfunctional residual capacity and lung inflation in patientswith an adequate respiratory drive. NIV improves the ef-fective minute ventilation enhancing CO2 elimination byaugmenting inspiration in patients with respiratory failureor impending respiratory failure without the use of anartificial airway. Effective ventilation will also result inimproved oxygenation as the ventilation/perfusion ratioimproves. NIV is commonly provided in the form of pos-itive pressure; however, negative pressure delivery devicesare available but are not widely used in the United States.

CPAP was first utilized in the neonatal population totreat respiratory distress syndrome associated with prema-turity in the early 1970s. CPAP is administered by nasalprongs or mask to provide an adequate distending pressurethat subsequently results in expansion and stabilization ofcollapsed alveoli secondary to surfactant deficiency.1-4 Thesuccesses seen in neonatal and adult CPAP and NIV useover the years have generated an expanded interest for itsuse in pediatric patient populations. The recent popularityand widespread use of high-flow nasal cannula (HFNC)have led to its inclusion in this discussion, although HFNCtherapy should not be considered as a form of positivepressure delivery. Restoring or improving the functionalresidual capacity by applying continuous positive pressureto the ventilating lung units reduces the work of breathingby eliminating the need to overcome the opening pressureof the lung, and it prevents collapse and reopening ofalveoli. CPAP does not provide tidal breathing support,and patients receiving CPAP alone must be able to main-tain airway patency and have adequate respiratory musclestrength and neurological drive to breathe. With CPAP andNIV, care must be taken to avoid regional over distentionof the lungs. Reducing the need for an artificial airway isthe primary advantage for the use of noninvasive modal-

ities, whereas secondary advantages include the reducedrisk of nosocomial infections, decreased need for sedation,increased tolerance for enteral feeds, potential for careoutside of the ICU, and an improved ability to ambulate.

Indications

The goals of therapy and the physiological effects willdetermine the indications for use of the various noninva-sive modalities. Common indications for the use of CPAP,NIV, or HFNC (collectively referred to as noninvasiverespiratory support) for specific patient populations areoutlined in Table 1.1-4 The most common use of CPAP isseen in patients with hypoxemia related to alveolar col-lapse or symptoms associated with obstructive sleep apneaor upper-airway collapse. The majority of NIV in pediatricpatients is utilized for the treatment of impending respira-tory failure associated with acute or chronic respiratoryinsufficiency secondary to pulmonary disease, neuromus-cular disease, airway obstruction, infectious processes, orpostextubation management or to avoid intubation or re-intubation. NIV can be effective in treating disorders re-lated to hypercarbia and/or hypoxia and may be adminis-tered as an intermittent or continuous intervention basedon patient need and reversibility of the indication. Pediat-ric studies on the use of noninvasive modalities suggestthat proper patient selection can have a significant im-pact on outcomes.

Asthma

In subjects with asthma, Basnet et al5 performed a pilotstudy to evaluate the safety, efficacy, and tolerability ofearly initiation of noninvasive positive-pressure ventila-tion in pediatric patients admitted with status asthmaticus.This was a small homogeneous study of 20 subjects thatcompared early initiation of NIV plus standard care (short-acting � agonists and systemic steroids) with standard carealone. Subjects received NIV by mask, the inspiratorypositive airway pressure (IPAP) setting was titrated to8 cm H2O to achieve a tidal volume of 6–9 mL/kg, theexpiratory positive airway pressure (EPAP) was set to5 cm H2O, and supplemental oxygen was administered toachieve a saturation of � 92%. Assessments were made atbaseline, 0–2 h, 4–8 h, 12–16 h, and 24 h. The clinicalasthma scores evaluated breathing frequency, accessorymuscle use, air exchange, wheeze, and inhalation/exhala-tion ratios. The scores for the NIV group and the standardgroup indicated no differences in the severity of illness;however, the scores for the NIV group showed a morerapid decline in symptoms than for the standard treatmentgroup at all intervals of assessment.

The use of NIV in asthma exacerbation has been shownto impact the use of mechanical ventilation, hospital stay,

Ms Fedor is affiliated with Pediatric Respiratory Therapy, ClevelandClinic, Cleveland, Ohio.

Ms Fedor discloses no conflicts of interest.

Ms Fedor presented a version of this paper at the 55th RESPIRATORY CARE

Journal Conference, “Pediatric Respiratory Care,” held June 10-11, 2016,in St Petersburg, Florida.

Correspondence: Katherine L Fedor MBA RRT-NPS CPFT, PediatricRespiratory Therapy, Cleveland Clinic, 11100 Euclid Avenue, Cleve-land, OH 44106. E-mail: [email protected].

DOI: 10.4187/respcare.05244

NONINVASIVE RESPIRATORY SUPPORT IN INFANTS AND CHILDREN

700 RESPIRATORY CARE • JUNE 2017 VOL 62 NO 6

mortality rates, and patient costs. Miguel-Dıez et al6 stud-ied children admitted with asthma exacerbation over an8-y period from 2002 to 2010. This study included � 12,000pediatric subjects, and findings related to the use of NIVincluded a decrease in hospital admissions of 1.8/100 pe-diatric subjects and a decrease in the mean length of stay(LOS) of 0.77 d. The presumed decrease in patient costs isattributed to the decreased LOS. During the study period,the use of NIV increased from 0.91% of hospitalized sub-jects in 2002 to 3.30% of the subjects in 2009, whereas asmall decrease occurred in 2010.

Bronchiolitis

The use of NIV continues to gain popularity in bron-chiolitis patients. A study by Ganu et al7 evaluated theimpact of intubation rates over the last decade related tothe increasing use of NIV in infants with severe bronchi-olitis. This study compared outcomes of 520 infants withsevere bronchiolitis with a median age of 2.8 months. Keyoutcomes were the median LOS in the pediatric ICU ofsubjects who were successfully treated with NIV com-pared with those who failed NIV or were treated only with

intubation and mechanical ventilation. The LOS for sub-jects successfully treated with NIV was 2.4 d comparedwith 5.2 d for those who were intubated and mechanicallyventilated and 8.4 d for subjects who failed NIV. Of the520 infants, 121 received no positive pressure support, 237received successful NIV, 114 were intubated and mechan-ically ventilated without a trial of NIV, and 48 failed NIVand required invasive mechanical ventilation. Over the10-y period, the use of NIV rose by 2.8%/y, and invasiveventilation fell by 1.4%/y.

In another study looking at acute respiratory failure dueto bronchiolitis and pneumonia, Abadesso et al8 evaluatedthe effectiveness of NIV in avoiding endotracheal intu-bation and to assess risk factors associated with NIV fail-ure. This study included infants with a mean age of7.2 � 20.3 months (102 subjects had bronchiolitis, 44 withpneumonia, 27 subjects with atelectasis, 20 with apnea,and 43 with upper-airway obstruction). Assessments weredivided into success and failure groups. Both groups showedimprovement in breathing frequency, heart rate, pH, andPCO2

at 2, 6, 12, and 24 h. However, the SpO2/FIO2

wassustained only in the NIV success group. The NIV failuregroup did have an improvement in SpO2

/FIO2, but that im-

Table 1. Indications for Noninvasive Respiratory Support in Infants and Children

Indications Modality

InfantsDelivery-room management CPAP* or NIV†Alveolar recruitment CPAP* or NIV†Post-extubation CPAP*, NIV†, or HFNC*Hypotonia with respiratory insufficiency and depression CPAP* or NIV†Conditions associated with loss of lung volume CPAP or NIV†Obstructive airways diseases (tracheal malesia, bronchial malesia, laryngomalesia) CPAP*

ChildrenRespiratory insufficiency associated with acute lung injury CPAP*, NIV†, or HFNC*Pneumonia CPAP*, NIV†, or HFNC*Asthma CPAP*, NIV†, or HFNC*Bronchiolitis CPAP*, NIV†, or HFNC*Pulmonary edema often associated with renal failure or cardiac failure associated with myopathy or

congenital heart diseaseCPAP* or NIV†

Congenital heart disease is associated with increased work of breathing and pending respiratory failure CPAP*, NIV†, or HFNC*Respiratory insufficiency associated with chronic lung disease CPAP* or NIV†Cystic fibrosis CPAP* or NIV†Neuromuscular diseases CPAP* or NIV†Neurologic disorders associated with respiratory depression CPAP* or NIV†Chest-wall deformities (scoliosis) CPAP* or NIV†Post-extubation CPAP*, NIV†, or HFNC*Obstructive airway diseases (OSA, tracheal malesia, bronchial malesia, laryngomalesia) CPAP*, NIV†, or HFNC*

Data from References 4-6.* Must have adequate respiratory drive to sustain adequate gas exchange.† Indicated when respiratory drive is insufficient to provide adequate gas exchange.NIV � noninvasive ventilationHFNC � high-frequency nasal cannulaOSA � obstructive sleep apnea



provement was only present during the first 4 h followinginitiation of NIV. The overall failure rate in this study was22.5%, with the risk factors being identified as apnea,prematurity, pneumonia, and co-infections. The overall suc-cess rate was 77.5%, with the remaining 22.5% requiringinvasive mechanical ventilation. Subjects with pneumo-nia had the highest failure rate (26%), followed by sub-jects with bacterial co-infections (17%) and apnea (11%).The highest degree of success was for subjects withbronchiolitis (95%), whereas subjects with ARDS hadzero success. Additionally, the LOS for the success groupversus the failure group was 9.7 � 9.5 d versus14.5 � 10.2 d. The results of this study suggest thatpatients with bronchiolitis should be trialed with NIVbefore intubation.

There is some evidence that the use of a heliox (helium-oxygen) gas mixture during NIV may further increase thesuccess of NIV in the treatment of bronchiolitis or otherdiseases associated with upper- or lower-airway obstruc-tion. Martinon-Torres et al9 conducted a single-institutionstudy evaluating subjects with bronchiolitis. Only 1 of 352subjects (0.28%) admitted to the pediatric ICU requiredintubation after being treated with heliox or heliox withNIV. Mixtures can be administered in helium/oxygen con-centrations of 80%/20%, 70%/30%, or 60%/40%, with thegreatest benefit for the 80/20 gas mixture. There are somecommercially available ventilators that are designed to de-liver helium-oxygen gas mixtures and therefore convertmeasurements to accommodate the density of the gas. How-ever, caution must be taken if a clinician modifies a deviceto administer a helium-oxygen gas mixture that does notcorrect for the altered gas mixture. Flows will be higherwhen using a helium gas mixture with a flow meter cali-brated for oxygen. Specifically, the flow through an oxy-gen-calibrated flow meter will be 2.1 times higher whenusing 80% helium, 1.7 times higher when using 70% he-lium, and 1.4 times higher when using 60% helium.

Pediatric ARDS

In 2015, the Pediatric Acute Lung Injury ConsensusConference Group published recommendations on pediat-ric ARDS.10 A key recommendation was that NIV shouldbe delivered in a setting with trained experienced staff thatoffers close monitoring of the patient (100% agreement), itshould be used early in the disease process to improvegas exchange and decrease the work of breathing to avoidinvasive ventilation complications (88% agreement). Thisgroup concluded, with 80% agreement, that immunodefi-cient patients are at a greater risk of complications frominvasive ventilation and may benefit from early interven-tion with NIV. They did not recommend NIV for childrenwith severe disease and recommended intubation for pa-tients who do not show clinical improvement or have signs

of worsening disease (100% agreement). When using NIV,the consensus group recommended using an oronasal ortotal face mask to maximize effective ventilation (84%agreement) and strongly encouraged the use of heated hu-midification (100% agreement). Patient comfort and safetyissues were also addressed. There was strong agreement(100%) on the monitoring of skin breakdown, gastric dis-tention, barotrauma, and conjunctivitis. Sedation was rec-ommended to improve patient ventilator synchronizationand tolerance (88% agreement). The use of pressure sup-port or CPAP was recommended to achieve synchrony orwhen using a nasal mask interface (92% agreement).

Cystic Fibrosis

One of the most debilitating chronic pulmonary diseasesin children is cystic fibrosis. Studies have shown that thereare many benefits to the use of NIV to improve the activ-ities of daily living in this patient population. Flight et al11

studied the rate of decline in lung function and the FEV1

measurements in 47 subjects with cystic fibrosis. Partici-pants in this study had significant lung disease, with anaverage FEV1 of 0.83 L and an FVC of 1.76 L (FEV1/FVCof 0.47) before initiation of NIV. Results indicated that 24of the 33 subjects who had baseline data before and afterinitiation of NIV showed an improvement in FEV1 or areduction in the rate of decline. This study reported thatNIV may improve lung function, prolong and improve thequality of life, and contribute to improved health statusand better outcomes following lung transplantation.

The positive effects on exercise tolerance with the useof NIV on cystic fibrosis subjects is well supported. Onestudy looked at the effects on the walk distance and chest-wall volumes in subjects with cystic fibrosis. In a random-ized crossover clinical trial, Lima et al12 evaluated 13 sub-jects with cystic fibrosis for walk distance and regionalchest wall volumes with and without the use of NIV. Allsubjects were between 7 and 16 y old without evidence ofrespiratory failure within the previous 3 months and hadan FEV1 of � 80% of predicted. Results indicated that thewalk distance had increased by approximately 7%(415.38 � 77.52 m vs 386.92 � 84.89 m) and percent-of-predicted FEV1 increased by approximately 9% (67%vs 61%) with NIV. Fauroux13 suggested that pressure sup-port should be provided in subjects with cystic fibrosis,citing improved patient comfort secondary to decreasedmuscle work and patient the ability of the patient to con-trol the inspiratory and expiratory phase of ventilation.

Obstructive Sleep Apnea

The prevalence of obstructive sleep apnea (OSA) inchildren ranges from 1 to 10% and results in a number ofconsequences, including failure to thrive, attention-deficit



disorder, behavioral problems, poor academic performance,and cardiopulmonary disease.14 OSA in children is definedas an apnea-hypopnea index of � 1 and a minimum oxy-gen saturation of � 92% during polysomnography. Treat-ment of pediatric OSA may involve surgical removal orrepair of the obstructive airway component or the use ofnocturnal CPAP. The use of CPAP in these patients in-volves the application of positive pressure to maintain thepatency of the airway during tidal breathing by creating apneumatic stent preventing airway collapse through theairway cycle. Although CPAP is an effective treatment forOSA, it is challenging in children, with intolerance orfailure to adhere occurring in 20% of subjects. All patientswith suspected apnea should have a sleep study to deter-mine the origin of apnea (central vs obstructive). CPAP isonly indicated in the presence of OSA; patients with cen-tral sleep apnea should be evaluated for placement of apermanent artificial airway versus nocturnal NIV.

Neuromuscular Disorders

Another common indication for the use of NIV is seenin children with neuromuscular disorders. NIV may beused to treat acute respiratory failure or may be used as aprimary intervention in the treatment of chronic hypoven-tilation. In a study conducted by Piastra et al,15 10 pedi-atric subjects with neuromuscular disease and acute respi-ratory failure were treated with NIV, with 8 of the subjectssuccessfully avoiding intubation. Subjects who were suc-cessfully treated with NIV experienced an increase in meanPaO2

/FIO2from 75 (at the onset of NIV) to 240 and 328 at

3 and 12 h following the initiation of therapy, and acuterespiratory failure was resolved within 6 h. Clinical out-come at discharge showed that 3 subjects required trache-ostomy with mechanical ventilation, 3 fully recovered with-out discharge intervention, 2 were discharged receivingNIV, and the 2 failure subjects died: one after 12 monthsand one after 24 months of hospitalization. This suggeststhat NIV can be routinely attempted in pediatric patientswith neuromuscular disease to avoid or delay invasiveventilation with an artificial airway.

Cardiac Disease

In pediatric patients with heart disease, the overall suc-cess rate of NIV following extubation was found to be77.8% when used prophylactically or when acute respira-tory failure was noted.16 Participants were assigned to aprophylactic group or a non-prophylactic group; 201 sub-jects received CPAP, whereas 20 subjects received NIV.The prophylactic group had a shorter LOS in the ICU andin the hospital than the non-prophylactic group (median of49 d vs 88 d in the ICU and median 60 d vs 103 d in thehospital). There were no differences in the LOS between

responders and non-responders, and the ICU mortality ratewas significantly higher in the non-responder group versusthe responders (31% vs 4%), but there was no differencein mortality between the prophylactic and non-prophylac-tic groups (both were noted at 10%). These data suggestthat the successful use of noninvasive modalities follow-ing extubation in children with cardiac disease can lowermortality rates, decrease the LOS in the ICU and the hos-pital, and reduce re-intubation rates.

Predictors of Success

The general success or failure of NIV is difficult topredict and is often attributed to the lack of availability ofpatient-specific interfaces or therapy intolerance. One studyinvolving pediatric subjects with respiratory failure by Ber-net et al17 suggests that the FIO2

level after 1 h of NIV maybe a predictive factor for success. This study showed thatin the first hour following initiation of NIV, non-respond-ers had a median FIO2

requirement of 0.8 compared with0.48 for responders.

Contraindications

Contraindications for the use of noninvasive modalitiesin infants and pediatric patients are shown in Table 2.3,18

Contraindications include medical conditions or treatmentfailures associated with respiratory failure or complica-tions that occur as a direct result of the application ofpositive pressure. In addition to the clinical contraindica-tions, patients may need to be removed from NIV second-ary to serious complications associated with the patientinterface device, such as skin ulcerations or deep-tissueinjury. Although risk of aspiration is not an absolute con-traindication for the use of noninvasive modalities, thesepatients do require additional precautions related to thedelivery interface, feeding, and monitoring to reduce theincidence of aspiration.

Interfaces

With an increased interest and use of noninvasive treat-ment modalities in pediatric patients, manufacturers areincreasing their selection of patient interfaces. Limitationsassociated with successful application and use of CPAPand/or NIV can be attributed to the patient interfaces, thedelivery devices, and the available technology. Interfacesinclude nasal prongs, nasal pillows, nasal masks, oronasalmasks, total face masks, and extended nasal prongs ornasal tubes positioned in the nasal pharyngeal airway. Theprimary patient interfaces used in small infants are nasalprongs and nasal masks. Both nasal prongs and nasal masksare associated with system leaks, which may necessitate



the use of a chin strap to maximize positive pressure de-livery.

The recent popularity of the RAM cannula, which isnot FDA-cleared for use as a CPAP or NIV deliverydevice, poses an added delivery challenge, since this isa non-occlusive nasal cannula interface, and performancewill not be significantly impacted by the use of a chinstrap. The presence of interface leaks on any device willnegatively affect the triggering capabilities of the de-livery device, an effect that is exaggerated by the in-fant’s inability to generate measurable inspiratory flowthrough the system.

Pediatric interfaces are also impacted by design andleaks associated with application. Pediatric interfaces haverecently expanded, but they still present challenges relatedto triggering, monitoring, and volume measuring. Leakcompensation is a mechanism that can determine and quan-tify the presence of an interface leak and thus compensatefor the triggering requirement needed to initiate an inspira-tory effort during NIV, provided the leak is reasonablystable. If the leak is highly variable, compensation willalso be variable and may result in asynchrony. If a sub-stantial volume of gas is contained within a mask inter-face, the ability to effectively trigger breath delivery will

be impacted which can still present a challenge in thepediatric patient. Impaired respiratory muscle strengthwill greatly affect the ability to trigger a breath evenwhen the amount of gas contained within the interfaceis minimized and leak compensation is employed. Un-der these conditions, it may be necessary to utilize aspontaneous timed mode to ensure adequate breath de-livery when NIV is indicated. Other factors to considerwhen selecting the patient interface will include patientheadgear and patient safety risks related to the potentialfor the development of skin ulcerations, gastric insuf-flation, and aspiration. When the risk of aspiration ispresent, it is best to select a nasal mask if possible tominimize exposure to the accumulation of secretionsand/or gastric contents and reduce the potential for suf-focation. The risk of gastric insufflation can be reducedby placement of a vented nasogastric or oral gastrictube, whereas the risk of skin ulcerations can be reducedby alternating patient interface devices, adherence tocareful skin care protocols, and the use of skin barriers.An abbreviated list of commonly used infant and pedi-atric interfaces is shown in Figure 1.

Delivery Devices

There are several stand-alone devices capable of deliv-ering CPAP to infants weighing � 5–7 kg and pediatricpatients � 20 or 30 kg. When NIV is required for patientsweighing between 7 and 20 kg in the in-patient setting,delivery devices are typically limited to the use of univer-sal ICU ventilators or portable home-care ventilators withNIV capabilities. When pediatric patients are dischargedhome receiving positive airway pressure, the choices arefurther limited to the home-care ventilators with noninva-sive capabilities based on the existing FDA-cleared con-straints.

Challenges

The biggest logistical limitations related to the applica-tion of NIV in infants and children are the interfaces anddelivery devices. In recent years, the patient interfacesand headgear for pediatric patients has expanded greatlyand includes a wider variety of total face masks, nasalmasks, nasal pillows, and, most recently, cloth masks. Care-ful attention to the interface device connection is an im-portant component of the circuit; some delivery systemsrequire a vented circuit to allow for leak compensation andexhalation, whereas mechanical ventilators utilize the in-ternal exhalation valve, and a closed circuit is required.Many of the interfaces have interchangeable system con-nections that allow for the exchange of vented and non-vented interface elbows.

Table 2. Contraindications for Noninvasive Respiratory Support

InfantsCPAP or NIV failure

PaCO2� 65 mm Hg

pH � 7.25Frequent and severe apneas

Congenital diaphragmatic hernia - unrepairedGastric insufflation

Tracheoesophageal fistulaCranial facial abnormalities

Preventing interface applicationUntreated pneumothoraxNeuromuscular disorders associated with severe respiratory

depressionChildren

Acute moderate-to-severe ARDS with respiratory failurePaCO2

� 65 mm HgpH � 7.25

Untreated pneumothoraxFacial injuries

Traumatic injuriesBurns

Risk of aspiration (consider interface; nasal mask may beacceptable)

Severe hemodynamic compromise and cardiovascular instabilityPatients undergoing certain upper-gastrointestinal tract surgeriesPatients with increased intracranial pressure (must evaluate the risk

benefit)

NIV � noninvasive ventilation



Fig. 1. Patient interfaces. Courtesy Respironics, Circadiance, Sleepnet, Neotech, ResMed, Teleflex, and CooperSurgical.



Patient triggering presents a significant problem for thepediatric patient population. Challenges associated withtriggering are often due to small tidal volumes, leaks relatedto the type of interface, the volume of gas contained withinthe interface, and poor respiratory muscle strength of thepatient. Triggering capabilities improve as leaks are mini-mized and when total face masks are utilized. Other adjuncts,such as chin straps and nasal prong occluding devices, mayhelp to reduce leaks and improve performance. Some devicesoffer auto-track triggering to accommodate variable leaks inthe system, which will improve effective triggering, providedthe delivery device can detect patient effort.

Mask interfaces are often associated with large residualvolumes within the mask, making it difficult to generateenough flow within the circuit to trigger an inspiratorybreath. The ability to trigger or not trigger will influencehow NIV is managed and must be evaluated on a case-by-case basis. Many infants successfully receive unsyn-chronized NIV using a pressure-limited time-cycled strat-egy, whereas some clinicians attempt to synchronize patientinteraction using pneumatic signals, such as flow or pres-sure. Even when using pneumatic signals to improve pa-tient interaction, rapid breathing frequencies, small tidalvolumes, and leaks can still present a challenge. To com-plicate matters, delivery is generally pressure-limited anddoes not accommodate variability in respiratory demand.

NIV-Neurally Adjusted Ventilatory Assist

Recent advances in technology have given cliniciansother tools to enhance the effectiveness of NIV in all pa-tients. One such tool is the emergence of neurally adjustedventilatory assist (NAVA). In a review of NIV with NAVAin newborns by Stein et al,19 the importance of patientsynchronization was evaluated. NAVA ventilation and thespecially designed NAVA catheter are proprietary to theServo ventilator (Maquet, Solna, Sweden). The NAVAcatheter detects electrical signals received by the diaphragmand is reflective of neural respiratory drive. The diaphragmsignals exist at 2 levels: the maximum diaphragm signal,which represents the inspiratory effort or work of the di-aphragm, and the minimum diaphragm signal, which rep-resents the expiratory effort required to prevent de-recruit-ment. When the diaphragm signal catheter is properlyplaced, signals are based on the muscle activity of thediaphragm. These signals not only provide excellent syn-chronization; they also accommodate variability in the in-fant’s sporadic breathing patterns and are unaffected bysystem leaks.

Successful use of NIV-NAVA is limited by the func-tionality of the NAVA catheter and the neurologic respi-ratory drive of the patient. However, NAVA provides im-portant backup safety features to address these occurrences.If a diaphragm electrical signal cannot be detected, the

ventilator will provide backup noninvasive pressure con-trol ventilation; however, this technology is not availablefor home use.

Baudin et al20 conducted a small crossover study of 11subjects with bronchiolitis to evaluate asynchrony usingnoninvasive NAVA versus noninvasive pressure control.Asynchrony was assessed by the asynchrony index (thenumber of times auto-triggering and/or double triggeringoccurred per minute), and the numbers of ineffective pa-tient breathing efforts were noted per minute. In this study,a NAVA catheter was placed before initiating pressurecontrol, subjects were switched to NAVA after 2 h, andfactors related to patient synchrony were evaluated. Whensubjects were placed into NAVA, the NAVA level wastitrated to a level that produced an equivalent peak pres-sure, PEEP remained stable, and the FIO2

was decreased tothe lowest possible level in both modes. Significant dif-ferences between pressure control and NAVA were notedrelated to auto-triggering (8.2/min vs 0.09/min), ineffec-tive efforts (21.8/min vs 0.54/min) (Fig. 2), asynchronyindex (38% vs 3%), and trigger delays (116.0 ms vs 43.9ms).These authors suggest that these data indicate thatNAVA is associated with a significant improvement inpatient-ventilator interactions. Additional studies shouldbe done to evaluate the clinical impact of this improvedpatient-ventilator interaction on key outcomes measures.Additionally, more randomized controlled trials are neededto determine how to most effectively interpret and adjustmandatory settings. Figure 3 demonstrates the inability totrigger during NIV, Figure 4 demonstrates improved trig-ger capture with the use of NAVA, and Figure 5 demon-strates inadequate support with a high inspiratory work ofbreathing during NAVA.20

Sedation

The use of sedation to improve patient tolerance in pa-tients receiving noninvasive ventilation should be given

Fig. 2. Noninvasive ventilation asynchrony. Pressure-control contin-uous mandatory ventilation (CMV) vs neurally-adjusted ventilatoryassist (NAVA). Data from Reference 20.



careful consideration. Patients may not tolerate appli-cation of the interface appliance or the excessive flowsometimes associated with NIV. If it is necessary toadminister anti-anxiety medications to facilitate patienttolerance, it will be important to choose a pharmaco-logic agent, such as midazolam, that produces minimalrespiratory depression.

Adherence

Adherence in pediatric subjects receiving CPAP wasstudied by Hawkins et al21 They found that overall adher-ence was 49% of the subjects evaluated (69 of 140). Ad-herence was defined as � 70% nightly use and averageusage of � 4 h/night. Specifically, they found that femaleshad better adherence than males (60.9% vs 39.5%), andchildren with developmental delays also had better adher-ence. Poor adherence could contribute to poor outcomesrelated to respiratory insufficiency and may lead providersto perform more invasive alternatives. Interface adherencemay be enhanced by desensitization efforts and role-playing.This process may require creative thinking by the practitionerand assistance from the family or child life department, whichmay be helpful in distracting the patient.

Measurements and Monitoring

Accuracy in volume delivery is difficult to achieve dur-ing NIV. System leaks are the biggest factor in volumemeasurements, but the size of the interface and the amountof residual volume within the interface also contribute tothe volume measurement. Delivered tidal volume is a di-rect reflection of the difference between IPAP and EPAPduring NIV, but due to inaccuracies in measurements, otherclinical indicators should be used in the management ofNIV. This issue is less concerning for patients only re-ceiving CPAP therapy. Clinical indicators of adequate tidalvolume during NIV will include assessment of chest rise,breath sounds, noninvasive forms of O2 and CO2 mon-itoring, and the level of consciousness. Ultimately, ad-equate evaluation of ventilation may require blood gasanalysis.

Monitoring of patients receiving CPAP or NIV shouldinclude pulse oximetry with saturation targets specific forthe patient based on diagnosis and acuity of the diseaseprocess. Pulse oximetry is a good indicator of oxygenationstatus; however, it is of little value related to the ventila-tion status of the patient. The adequacy of ventilation mustinclude some periodic form of CO2 measurement. CO2

measurements can be obtained noninvasively with trans-

Fig. 3. Noninvasive ventilation unable to capture patient effort.



cutaneous monitoring or invasively by arterial, venous, orcapillary blood gas analysis. End-tidal PCO2

monitoringmay be attempted, but it is often inaccurate secondary todilution from the high flow of gas from the delivery de-vice. CO2 targets are disease-specific and should reflect acompensated acid/base status. Other observations relatedto the work of breathing and the level of consciousness canalso provide valuable clinical indicators of the overall ox-ygenation and ventilation status of the patient.

It is important to have a clear understanding of thepatient’s baseline clinical status when determining effec-tiveness of treatment. Many of the tools that are availablein the acute care setting are not utilized in the home;therefore, it will be important to educate caregivers aboutthe early recognition of signs and symptoms of respiratorydistress or other changes in the clinical status. It isessential that all caregivers be familiar with the com-plications associated with the noninvasive modalitiesand the steps necessary to minimize those complica-tions. Chest radiographs can provide useful diagnosticinformation if clinical symptoms are present and will bea useful tool in evaluating the adequacy of lung infla-tion during CPAP or NIV. The radiograph may alsoprovide useful information regarding the appropriateEPAP levels to use during NIV.

Clinical Management

CPAP

The goal of CPAP is to restore adequate functional re-sidual capacity to correct, reverse, or minimize alveolarcollapse. CPAP will effectively reduce the work of breath-ing as it unloads the work of the inspiratory muscles andallows effective inflation in excess of the opening pressureof the lungs.22 Generally, the level of CPAP required toaccomplish this goal in infants is usually 5–8 cm H2O butvaries with disease process; the level necessary for pe-diatric patients will be variable based on the indicationsfor use and the severity of disease. Disorders associatedwith low lung compliance may require CPAP levels inexcess of 10 cm H2O. CPAP should only be used onpatients with an adequate respiratory drive, and pres-sures should be incrementally increased to achieve thetargeted oxygen saturation goal at an FIO2

of � 0.50 –0.60 up to a CPAP level of 10 –12 cm H2O. CPAP de-vices are designed to operate under 2 basic principles;one principle is to provide a constant pressure withvariable flow, whereas other systems provide constantflow resulting in variable pressure delivery to the pa-tient. Pressure-constant delivery devices are preferred

Fig. 4. Noninvasive ventilation-neurally adjusted ventilatory assist (NIV-NAVA) where each patient effort is captured.



to provide a consistent pressure and prevent alveolarcollapse by sustaining lung recruitment. When adequateCPAP levels are achieved, the work of breathing shouldimprove, as evidenced by a decreased breathing fre-quency, less sternal or intercostal retractions, and im-proved oxygen saturations.

When CPAP is indicated to treat acute lung disease, FIO2

and CPAP levels should be weaned as tolerated to main-tain acceptable oxygenation levels. Weaning of CPAP sup-port should commence when resolution or resolving dis-ease processes has been identified. Clinical and diagnosticindicators should be evaluated to determine resolution ofthe disease process and readiness for weaning. In infants,the FIO2

should be weaned to minimally acceptable levelsbefore weaning CPAP levels. If CPAP is applied intermit-tently for nocturnal use for the treatment of OSA, theCPAP level is maintained while the FIO2

is titrated to main-tain predetermined saturation levels. Weaning of CPAPmay involve intermittent trials of HFNC or standard nasalcannula at increasing intervals (sometimes described assprints). CPAP can be trialed off when the patient has beenstable at a level of 5–6 cm H2O with a FIO2

of � 0.40.Patients are often transitioned to HFNC as an intermediatestep following CPAP before transitioning to a standardnasal cannula.

NIV

When NIV is indicated, the clinician must determine theappropriate mode of ventilation to use. The spontaneousmode is a patient-triggered mode that provides an inspira-tory pressure assist without a backup frequency. Whenchoosing this mode, the patient must have a good respi-ratory drive, and the patient must be able to effectivelytrigger breaths. The timed mode offers a set pressure de-livery at a preset mandatory frequency and allows addi-tional patient-triggered breaths as needed. This mode canlead to asynchrony when patient breathing efforts conflictwith the set frequency. The spontaneous/timed mode is apatient-triggered mode that provides inspiratory pressureassist with a backup time-based frequency. This mode mayoffer a combination of patient safety and patient comfort,provided triggering is effective. Each of these modes canbe provided by most noninvasive and invasive deliverydevices.

Initial settings when implementing NIV support will bepatient-specific and require patience on the part of theclinician. It will be important to select the correct pa-tient interface that permits tolerance and maximizes trig-gering capabilities. If a set breathing frequency is uti-lized, the frequency will be age-specific and reflect the

Fig. 5. Noninvasive ventilation-neurally adjusted ventilatory assist (NIV-NAVA) where each patient effort is captured but support isinsufficient (maximum electrical signal of the diaphragm [Edi] is 41, indicating a high inspiratory work of breathing and insufficientNAVA level).



level of support needed to ensure adequate ventilation.The EPAP setting should provide a resting pressure levelthat minimizes alveolar collapse and should be increased ifevidence of lung volume loss is present. The EPAP settinghas a direct relationship with oxygenation and should betitrated based on oxygen saturation. EPAP settings shouldnot exceed 7–8 cm H2O in most infants or 10–12 cm H2Oin most pediatric patients, although this is patient-specificand is not an absolute recommendation.

IPAP is applied to the resting expiratory pressure andtranslates to volume delivery to the patient. IPAP has alinear relationship between pressure and volume. The IPAPsetting should be titrated to achieve adequate chest ex-pansion and CO2 removal. Inspiratory time determinesthe amount of time devoted to inspiration and should besufficient for breath delivery. Excessively long inspira-tory times will be uncomfortable for patients and mayincrease the work of breathing if patients are forced toexhale against inspiratory flow. Other secondary set-tings available are dependent on the device used to pro-vide ventilation and may include the trigger sensitivity,the rise time, the breath transition cycle, and a gradualpressure ramp time.

The trigger sensitivity should be set so that the patientcan easily initiate a spontaneous or assisted breath withoutauto-triggering. This can be difficult to achieve secondaryto leaks that have a large impact on the triggering mech-anism. The rise time reflects the amount of time requiredto reach the preset pressure during pressure support ven-tilation or pressure control ventilation and should be set tomaximize patient comfort. Patients with restrictive lungdisease may prefer a slower rise to improve the distribu-tion of ventilation, whereas patients with increased airwayresistance may prefer a more aggressive rise. The cycletime setting determines the transition period from inhala-tion to exhalation during pressure support ventilation andmay be flow- or time-based. The cycle setting should lowerthe threshold for transition in the presence of large systemleaks. The ramp time, if available, will allow gradual in-crease in the pressure delivery to assist with patient toler-ance and comfort. Invasive mechanical ventilation is some-times indicated to manage patients with respiratory failure;however, there are many complications associated withinvasive mechanical ventilation. This has been a catalystfor the use of NIV, which has become increasingly prev-alent in infants and children.

System leaks have a much bigger impact in the set-up,management, and weaning strategy and must be consid-ered on an individual patient basis. When patients cantrigger the delivery device, the approach is similar to thatof invasive ventilator weaning. When significant leaks arepresent and patients are unable to trigger spontaneous orassisted breaths, spontaneous breaths are usually not sup-ported. In this case, a traditional weaning strategy of in-

crementally reducing the mandatory breathing frequencyshifts more work to the patient and may substantially in-crease the work of breathing because spontaneous breathsare not supported. This can be effective for many patients,but the increased work of breathing for patients with lim-ited reserves may result in them not tolerating this strat-egy. Another approach would be to set the mandatoryfrequency at a physiological level, based on patient ageand disease process, and wean the IPAP until the IPAP andEPAP settings are nearly equal. This approach may offera more gradual shift of the respiratory muscle work. Theoptimal weaning strategy requires more research beforeevidenced-based recommendations can be made. Cliniciansshould always evaluate the efficiency of current strategiesand adjust as needed.

Alternative Approaches to Noninvasive Modalities

The use of HFNC has become very popular in recentyears; however, the data are limited and conflicting. It isimportant to distinguish the intended goals of therapy whenimplementing HFNC.23-27 The high flow used may be suf-ficient to produce a CPAP effect, and may flush a portionof the CO2 contained within the anatomical dead space.The definition of high flow varies with age of the patientand is not universal. Flows � 1 L/min are generally con-sidered high flow in the neonatal patient, whereas flowranges for older children are not well established in theliterature. Typically, flows � 3 L/min meet the criteria forhigh flow in toddlers, whereas flows � 5 L/min are con-sidered high flow in older children. Assumed benefits ofHFNC are realized from maximum mucocilliary clearance,reduced inflammatory reactions, inhibition of bronchoc-onstrictor reflex, reduced airway resistance, and a variabledegree of washout of the nasopharyngeal dead space. Theevidence for positive pressure delivery with HFNC is mixedand highly variable.

Manley et al25 compared treatment failure in infantswithin 7 d of extubation comparing the use of HFNC withCPAP. In this study, 34.2% of the nasal cannula infantsexperienced treatment failure, whereas 25.8% of the CPAPgroup demonstrated failure. Of the failures, 50% were suc-cessfully treated with CPAP without re-intubation. Basedon their predetermined criteria, this led to a conclusion thatthe use and efficacy of HFNC was not inferior and wastherefore similar to that of CPAP as a form of respiratorysupport for very preterm infants following extubation. Theuse of HFNC therapy in pediatric patients is popular andoften creates conflict of point of care (pediatric ICU vsregular nursing floor). In a pilot study conducted by May-field et al,26 33 subjects received standard low-flow oxy-gen while 61 subjects with bronchiolitis were treated withHFNC at 2 L/kg/min at an FIO2

sufficient to maintain anoxygen saturation of 94%. Responders and non-responders



to HFNC were identified within 60 min, with non-responders being 4 times more likely to be admitted to thepediatric ICU than HFNC responders. Pediatric ICU ad-mission was required in 8 of the 61 HFNC group subjects(13%) and in 10 of the 33 standard low-flow nasal cannulagroup subjects (30%). There were no subject demographicor physiological characteristic differences between thegroups. These authors concluded that HFNC responderscould be assessed based on heart rate and breathing fre-quency within the first hour and that responders may besafely cared for on the pediatric wards; they further con-clude that this may result in a cost savings by reducingICU admissions.

In another study conducted by Beggs et al,27 comparingstandard oxygen therapy with HFNC in 19 bronchiolitissubjects, the authors found some differences in SpO2

at 8and 12 h, but no difference was noted at 24 h. There wasalso no difference in duration of oxygen therapy, LOS, ortime to discharge in either group. They did note that HFNCtherapy is feasible and well tolerated.

Safety and Complications

All forms of ventilation require patient safety featuresrelated to alarm settings. If the delivery device allows forhigh tidal volume and high breathing frequency alarms,they should be set at levels that are 20–30% above base-line delivery. High-pressure alarms should be set at5–10 cm H2O above IPAP (or PIP) levels. Low-pressurealarms should be set above the EPAP setting but below theIPAP setting. Traditionally, apnea times are set at 10–15 sfor infants, 15–20 s for pediatric patients, and 20–30 s foradolescent patients. Backup ventilation settings shouldreflect full ventilator support. Other safety measuresthat the clinician should consider are the use of remotemonitoring, including video monitoring, central nursecall monitoring for heart rate, saturation, and discon-nect, especially if the patient is cared for outside of theICU setting.

Complications associated with NIV are most often seenin the most fragile patients; they include gastric distention,aspiration, pneumothorax, and pressure ulcerations. Ofthese, pressure ulcerations are the most preventable. Gas-tric distention can be minimized by ensuring that all pa-tients receiving NIV have vented gastric tubes in place tocircumvent excess air from the stomach. Aspiration pre-cautions involve feeding precautions, close attention tofeeding schedules, elevation of the head of the bed, posi-tioning of the patient, careful suctioning procedures toreduce gagging and emesis, and close patient monitoring.Pneumothoraces can be reduced by minimizing pressuredelivery while observing changes in volume delivery orchest rise, which may indicate a change in lung compli-ance. It is important to observe trends in volume delivery

and to make appropriate changes in patient and alarm set-tings to minimize potential harm from positive pressuredelivery.

Pressure ulcerations can produce serious complica-tions, and consequences can range from temporary su-perficial skin discoloration to deep-tissue injury requir-ing plastic surgery. Pressure ulcerations have beenreported in 4 –27% of children.28 Pressure ulcerationsare classified in stages: Stage 1 is redness at the site andis non-blanchable, state 2 has redness with an openingor blistering at the site of injury (the wound is shallow,and fat is generally not visible), stage 3 has visible fator slough and is slightly deeper than stage 2, and stage4 pressure ulcerations have visible slough and may haveexposed bone or muscle.

Visscher et al29 investigated factors contributing to pres-sure ulcers in 50 subjects. They used color imaging, 3-di-mensional surface imaging, and skin hydration measure-ments to identify early breakdown. They also evaluated 3interventions to reduce trauma: use of a silicone foamdressing, a water/polyethylene oxide hydrogel dressing,and a flexible water-permeable cloth mask. The interven-tions were applied upon identification of skin breakdownat any stage. In summary, their findings were that skinbreakdown was most severe with the oronasal masks, thatthe bridge of the nose represented the area of most severebreakdown, and that maintaining normal skin hydrationare important interventions for protecting skin tissue in-tegrity.

Humidification is recommended to hydrate secretionsand improve patient tolerance when administering CPAPand/or NIV; however, it presents a clinical dilemma re-lated to potential skin breakdown and pressure ulcerations.High humidity has been linked to the development of pa-tient interface pressure points and contributes to the de-velopment of pressure ulcerations as outlined by Visscheret al.29 The incidence of breakdown associated with oro-nasal masks has been confirmed in a study by Schallomet al,30 where 200 subjects were assigned to either anoronasal mask or a total face mask. Findings showedthat 16 subjects in the oronasal group had stage 1 break-down, whereas none were observed in the total facemask group; 4 subjects were noted to have stage 2 break-down in the oronasal group, with one in the total facemask group; and none had deep-tissue injury in theoronasal group, whereas one had deep-tissue injury inthe total face mask group. The authors concluded thattotal face masks offer a greater distribution of pressureand therefore less severe pressure ulcerations than theoronasal mask.

Skin ulcerations can be minimized by careful evalu-ation at regular intervals to identify early signs of po-tential breakdown and liberal policies enabling routineskin care consults when potential injury is identified.



Caregivers should be trained in the stages of skin break-down, and safety measures to minimize risk should beroutinely employed. Safety measures related to the re-duction of skin breakdown should include but are notlimited to the intermittent alteration of patient interfacesto rotate pressure points, intermittent interface breaks iftolerated by the patient, and routine inspection andcleansing of the face and interface device to minimizerubbing friction.

Summary

The key determination of successful implementation ofnoninvasive respiratory support is to change the trajectoryof respiratory distress by alleviating respiratory distressand/or improving gas exchange on a lower FIO2

. The firststep is to select the most appropriate delivery modality,CPAP, NIV, or HFNC. Once the delivery modality is de-termined, the careful selection of the patient interface isimportant. The patient interface must meet the require-ments of the delivery strategy and maximize patient tol-erance criteria. Tolerance is important because it will be akey factor in adherence, which is essential for successfultreatment. Factors that may improve tolerance are theuse of mild sedatives or anti-anxiety agents, a comfort-able interface, appropriate system leak management, pa-tient-specific settings that meet the oxygenation andventilatory needs of the patient, and humidification ofgas delivery. A clear understanding of the limitationsand operations of delivery devices is key. The clinicianmust match the patient goals with the available inter-ventions, and they must be able to operate the deviceseffectively to maximize patient benefits. The clinicianmust also recognize key differences in the operation ofvarious devices to safely set up, control, and manage thepatient. It is important to have clear multidisciplinaryexpectations for the use and limitations of NIV so thatclinical goals can be met with minimal risk to the pa-tient.

REFERENCES

1. DiBlasi R, Courtney SE. Noninvasive respiratory support. In: Gold-smith JP, Karotkin EH, Keszler M, Suresh GK, editors. Assistedventilation of the neonate, 6th edition. Philadelphia, Pennsylvania:Elsevier; 2003:162-179.

2. Cheifetz IM. Invasive and noninvasive pediatric mechanical venti-lation. Respir Care 2003;48(4):442-458; discussion 453-458.

3. Deorari AK, Thukral A. Continuous positive airway pressure. In:Walsh BK, editor. Neonatal and pediatric respiratory care, 4th edition.St Louis, Missouri: Elsevier; 2015:267-286.

4. Teague GW, Thompson-Batt D. Noninvasive mechanical ventilation ofthe infant and children. In: Walsh BK, editor. Neonatal and pediatricrespiratory care, 4th edition. St Louis, Missouri: Elsevier; 2015:287-299.

5. Basnet S, Mander G, Andoh J, Klaska H, Verhulst S, Koirala J. Safety,efficacy, and tolerability of early initiation of non-invasive positive pres-

sure ventilation in pediatric patients admitted with status asthmaticus: apilot study. Pediatr Crit Care Med 2012;13(4):393-398.

6. de Miguel-Dıez J, Jimenez-Garcıa R, Hernndez-Barrera B, Lopez deAndres A, Villa-Asensi JR, Plaza V, Carrasco-Garrido P. Nationaltrends in hospital admissions for asthma exacerbations among pedi-atric and young adult population in Spain (2002-2010). Respir Med2014;108(7):983-991.

7. Ganu SS, Gautam A, Wilkins B, Egan J. Increase in use of non-invasive ventilation for infants with severe bronchiolitis is associatedwith decline in intubation rates over a decade. Intensive Care Med2012;38(7):1177-1183.

8. Abadesso C, Nunes P, Silvestre C, Matias E, Loureiro H, Almeida H.Non-invasive ventilation in acute respiratory failure in children. Pe-diatr Rep 2012;4(2):e16.

9. Marinon-Torres F. Noninvasive ventilation with helium-oxygen inchildren. J Crit Care 2012;27:220.e1-220.e9.

10. Pediatric Acute Lung Injury Consensus Conference Group. PediatrCrit Care Med 2015;16(5):428-439.

11. Flight WG, Shaw J, Johnson S, Webb AK, Jones AM, Bentley AM,Bright-Thomas RJ. Long-term non-invasive ventilation in cystic fi-brosis: experience over two decades. J Cyst Fibros 2012;11(3):187-192.

12. Lima CA, Andrade Ade F, Campos SL, Brandao DC, Fregonezi G,Mourato IP. Effects of noninvasive ventilation on treadmill 6-minwalk distance and regional chest wall volumes in cystic fibrosis:Randomized controlled trial: Respir Med 2014;108(10):1460-1468.

13. Fauroux B. Why, when and how to propose noninvasive ventilationin cystic fibrosis? Minerva Anestesiol 2011;77(11):1108-1114.

14. Chan J, Edman JC, Koltai PJ. Obstructive sleep apnea in children.Am Fam Physician 2004;69(5):1147-1154.

15. Piastra M, Antonelli M, Caresta E, Chiaretti A, Polidori G,Conti G. Noninvasive ventilation in childhood acute neuromus-cular respiratory failure: a pilot study. Respiration 2006;73(6):791-798.

16. Gupta P, Kuperstock JE, Hashmi S, Arnolde V, Gossett JM, ProdhanP, et al. Efficacy and predictors of success of noninvasive ventilationfor prevention of extubation failure in critically ill children withheart disease. Pediatr Cardiol 2013;34(4):964-977.

17. Bernet V, Hug MI, Frey B. Predictive factors for the successof noninvasive mask ventilation in infants and children with acuterespiratory failure. Pediatr Crit Care Med 2005;6(6):660-664.

18. Schibler A, Franklin D. Respiratory support for children in theemergency department. J Paediatr Child Health 2016;52(2):192-196.

19. Stein H, Beck J, Dunn M. Non-invasive ventilation with neutrallyadjusted ventilator assist in newborns. Semin Fetal Neonatal Med2016;21(3):154-161.

20. Baudin F, Pouyau R, Cour-Andlauer F, Berthiller J, Robert D, Ja-vouhey E. Neurally adjusted ventilator assist (NAVA) reduces asyn-chrony during non-invasive ventilation for severe bronchiolitis. Pe-diatr Pulmonol 2015;50(12):1320-1327.

21. Hawkins SM, Jensen EL, Simon SL, Friedman NR. Correlatesof pediatric CPAP adherence. J Clin Sleep Med 2016;12(6):879-884.

22. Loh LE, Chan YH, Chan I. Noninvasive ventilation in children: areview. J Pediatr 2007;83(2 Suppl):S91-S99.

23. Cummings JJ, Polin RA, Committee on Fetus and Newborn, Amer-ican Academy of Pediatrics. Noninvasive respiratory support. Pedi-atrics 2016;137(1). doi: 10.1542/peds.2015-3758.

24. Walsh BK. Oxygen administration. In: Walsh BK, editor. Neonataland pediatric respiratory care, 4th edition. St Louis, Missouri: Elsevier;2015:156-157.



25. Manley BJ, Owen LS, Doyle LW, Anderson CC, CartwrightDW, Pritchard MA, et al. High-flow nasal cannulae in very pre-term infants after extubation. N Engl J Med 2013;369(15):1425-1433.

26. Mayfield S, Bogossian F, O’Malley L, Schibler A. High-flow nasalcannula oxygen therapy for infants with bronchiolitis: pilot study. JPaediatr Child Health 2014;50(5):373-378.

27. Beggs S, Wong ZH, Kaul S, Ogden KJ, Walters JA. High-flow nasalcannula therapy for infants with bronchiolitis. Cochrane DatabaseSyst Rev 2014;(1):CD009609.

28. Bergquist-Beringer S, Cuddigan J, Davidson J. National Database ofNursing Quality Indicators. Pressure injury training. https://members.nursingquality.org/ndnqipressureulcertraining/. Accessed March 20,2017.

29. Visscher MO, White CC, Jones JM, Cahill T, Jones DC, Pan BS.Respir Care 2015;60(11):1536-1547.

30. Schallom M, Cracchiolo L, Falker A, Foster J, Hager J, MorehouseT, et al. Pressure ulcer incidence in patients wearing nasal-oral ver-sus full-face noninvasive ventilation masks. Am J Crit Care 2015;24(4):349-356.

Discussion

Panitch: Thanks for a great talk,Kathy. I have a comment and ques-tion. The comment is when it comesto nomenclature and ventilator ther-apy, I think NIV parlance causesgreater inaccuracies in what we do.You brought that up in terms of howwe reference pressures. But when wetalk specifically about neonatal NIV,the terms that are used to describe var-ious modes of therapy that we thinkare synonymous with NIV modes inolder children have nothing to do witheach other. In fact, there really are notany studies that look at pressure sup-port ventilation (PSV) that is patient-triggered flow-cycled in neonates; theclosest comparison is one study thatused abdominal motion by respiratoryinductance plethysmography for trig-ger and cycle signals, but the machineis no longer available. So NIV is avery different modality in neonates.SiPAP is sort of bi-level CPAP. I thinkit would be great if we could stan-dardize the nomenclature so we allknow what we’re talking about. Myquestion refers to the use of the ICUventilator versus a machine that’sspecifically designed to provide non-invasive support. My experience hasbeen that the ICU ventilator is veryintolerant of leak even when you usea non-vented mask, and often pa-tients do better when we can con-vince our colleagues to switch to amachine designed for the purpose ofNIV, where the flow meets patientdemand better. I would like youropinions.

Fedor: I would agree, we have seena lot of that. Sometimes we’veswitched the patient type to accom-modate increased flow demands modefrom infant to pediatric in order toaccommodate those flows. We’vedone that even in the face of NAVAwith total face shields; we’ve had toincrease our patient age input a (in-fant to peds, or peds to adult depend-ing on the ventilator) so that the ad-ditional flow is necessary. I do thinkthat the standalone devices that haveleak compensation offer enhanced trig-gering capabilities and is an importantpoint that you’re bringing up. The is-sue is that, especially when you getinto total face masks, the amount ofdead space that exists within the maskitself that has to be overcome by thepatient still presents a triggering prob-lem. We have overcome a lot of theICU ventilator issues by incorporat-ing the use of NAVA for better trig-gering and improved flow initiationand termination. But not everybodyhas that capability, and you’re abso-lutely right.

Cheifetz: I am going to challengeyou. When you presented the pediat-ric NIV indication slide, I was sur-prised that ARDS was not on yourlist. Then, on the next slide, you sur-prised me even more when ARDSwas listed as an exclusion criterion.First a comment, and then a question.My comment is if you review the Pe-diatric Acute Lung Injury ConsensusConference (PALICC) publications,1,2

the definition of ARDS actually in-cludes NIV criteria. Based on thePALICC publications, they are in es-sence saying NIV could, and poten-

tially should, be used in ARDS, atleast for mild-to-moderate ARDS.3

Now my question is why do you haveARDS as an exclusion criterion forthe use of NIV?

Fedor: I should have quantified thatfor mild/moderate, I agree. I was think-ing of exclusion in the case of severeARDS where you have such low lungcompliance that you cannot effectivelyventilate noninvasively.

Cheifetz: With that clarification, Iagree.

Berlinski: I would like to commenton 2 issues: first that we need to keepin mind that the use of total face maskincreases the risk of aspiration. It’s abig challenge in choosing the right in-terface for the patient. Second that theuse of NIV and asthma in the emer-gency department is definitely notready for primetime. Even if it takesyou a couple more hours to get to thesame breathing frequency or the sameFIO2

, the expense of $2,500 or morethat the use of a ventilator adds tothe care of the patient without a clearbenefit in mortality or morbidity doesnot seem justified. Maybe it couldbe used in a very select group ofpatients where it might save themfrom intubation. But not as a blankettherapy, because I could see thisspreading around the emergency de-partments of the country adding un-necessary cost.

Fedor: Right, I understand. The pri-mary outcome of that study was safetyand tolerance, so those were the out-comes they reported in terms of clin-



https://members.nursingquality.org/ndnqipressureulcertraining/

https://members.nursingquality.org/ndnqipressureulcertraining/

ical outcomes. I would agree that blan-ket use of NIV for asthma patients inthe emergency department is proba-bly not something we want to embracewithout carefully choosing our pa-tients.

* Myers: Nice job, Kathy. It’s inter-esting as I sit here and reflect, about adecade ago, we had a neonatal/pedi-atrics Journal Conference4,5 where wetalked about a lot of technologies andmedication that were not cleared bythe FDA for kids, that we adapt anduse in the hospital setting to makework. And over a decade later, thetechnology still hasn’t been tested,proven, validated, and cleared by theFDA to do so. We know now that aswe try to shift costs from that acutecare environment to other entities andlocations, that becomes even moreproblematic with using devices andmedications outside of that arena.Again, it’s a challenge to industry tolook at the ability to adapt and getFDA clearance in these areas to ad-dress this pediatric gap that we seemto continue to have for the last 10, 20,30 years.

Walsh: I would actually disagreewith that mainly because I think wehave twice as many devices on themarket now as we did a decade ago.As Kathy alluded to, we have moremasks now than ever. We have thedevices, and many of them go downto 5 kg. Now, they’re expensive. Don’tget me wrong, they are not the $1,500bilevel machine, but they are avail-able now. It used to be that they werenot available and we would do“muzzlePAP,” as I used to call it, witha critical care ventilator and slap amask on and a black strap. So, wehave come a long way; I’m just notsure if we are any clearer as to whenwe should or shouldn’t use it. One ofthe questions I didn’t quite get a solidanswer to is when should we escalateand intubate someone who’s receiv-ing NIV?

Fedor: That’s a really tough ques-tion; in terms of making that decision,I think that is probably a patient-by-patient decision. I think there are someclear guidelines like when your pres-sures exceed a certain point. I don’tknow that the literature out there pres-ents that clear-cut data. I think hospi-tals have their own criteria for goingfrom NIV to intubation, such as in-spiratory pressures of 30 cm H2O, orPEEP, or FIO2

criteria and those kindsof things. That’s a future area where Ithink we need more concrete data interms of what constitutes failure.

Rehder: There are some data inadults as well as pediatrics that inves-tigate NIV and the time frame in whichyou should expect to see improvement.Generally, you should see improve-ment within the first hour or so. I thinkthat’s something we really do need tofocus on and encourage; if you’re notseeing improvement early, then it’stime to escalate, because I think that’swhere many providers get into trou-ble. There was at least one study inadults that showed increased mortal-ity with NIV, but I think that’s exactlywhat it was, that those patients didn’timprove early and they sat on NIV toolong prior to escalation of therapy.

Fedor: Probably some of our biggestchallenges are some of the neuro de-velopmentally delayed patients whocome into our unit. You need to knowwhat the surrogate’s wishes are. It’salways a tough call when deciding tointubate them when they present withsevere respiratory failure and there areno do-not-resuscitate orders. The do-not-intubate advance directive alsocomplicates the question of how far togo.

Stokes: You mentioned briefly thenegative pressure devices. At St Jude,we have a long history of using con-tinuous negative pressure, and recentlyI’ve been interested in the Hayek os-cillator that’s a much more sophisti-cated device than the old converted

iron lung we used in the late 70s andearly 80s. Unfortunately, it’s notwidely available, but I wonderedwhether you’ve had any experiencewith that technology. We have had suc-cess recently with a neuromuscular pa-tient with respiratory failure and wide-spread atelectasis, where we were ableto transition him back to his home ven-tilator by using the Hayek oscillatoras a bridge.

Fedor: I do not have any personalexperience with that; I have some ex-perience with iron lungs and the chestcuirass, but I do not have any personalexperience with the oscillator device.

Cheifetz: I want to go back to a cou-ple points you made during your pre-sentation as well as Tim’s [Myers]comment. We need to be careful notto get too caught up in FDA clearancewhen we know that much of what wedo in pediatrics and neonatology in-volves off-label use of products, de-vices, and medications. Many of theNIV devices and interfaces you men-tioned are not FDA-cleared for the in-dications that we employ in pediat-rics; however, many centers are usingthis approach with good success insmaller and younger patients. I doagree with Tim’s comment that we donot have data to definitely support thisapproach, but we also do not have datato refute its use. Unfortunately, a lotof what we do in pediatrics is basedon anecdotal experience, physiology,and extrapolation from data in neo-nates and adults, but until more defin-itive data are available, we are some-what trapped. I just want to raise thepoint that many centers do employ abroader use of these products than whatthe FDA has given clearance.

Fedor: I agree; in the hospital set-ting, we tend to use those devices off-label more frequently. Where we prob-ably have our biggest challenge is intransitioning these patients home, andthe home care companies may or maynot want to take on that liability.



† Kallstrom: How difficult is it toget these kids home and have a suc-cessful experience once they’re dis-charged from the hospital? I say thatin light of the fact that there aren’t alot of RTs in home care anymore.

Fedor: Yeah, there are not. I alsomanage our rehab facility, which is achronic treatment technology-depen-dent free-standing hospital, and we re-ally have a lot of readmissions fromthose patients because we are relyingon nurses primarily to provide thehands-on care for these patients, andthey don’t necessarily have the knowl-edge base that an RT would have. Thereimbursement for RTs is so low inthat setting that home care companiesaren’t providing RTs, they’re primar-ily doing setup and education. Gettingadherence at home is a problem; onestudy talked about 50% adherence, andthat’s really not very good when you’retalking about something that’s poten-tially lifesaving. It is a really difficultquestion.

Lin: I want to go back to one of Ari-el’s [Berlinski] comments made ear-lier about not necessarily deployingNIV early on in asthma patients as amatter of routine. Certainly, I wouldagree that we shouldn’t do it as a mat-ter of routine, but I do think there is apotential role that should be investi-gated more fully. Early use of NIV topotentially prevent ICU admission, thethought there being that if we can de-crease work of breathing (WOB) byproviding NIV support and making iteasier for them to get that next breathso they spontaneously decrease theirfrequency, then perhaps then theirWOB decreases, and it becomes a sal-utary cycle where they continue to im-prove. I haven’t looked at the litera-ture recently, but in the 2000s, maybeabout 10 years ago, there were somesmaller randomized controlled trialsbased out of emergency departmentsthat demonstrated a positive signal fordecreasing incidence of ICU admis-sion if not even hospitalization. I won-

der if you have any thoughts aboutthat?

Fedor: I think the big question thereis about decreasing ICU admission.We actually would not initiate NIVon our regular pediatric wards. Thatrepresents an escalation of care thatwe would not do in our facility. Wemight do it in our intermediate careunit, but even that would be rare. Ithink it ends up being one of thosesituations where that might be definedby your workflow at your facility.

Lin: I was referring more to NIV ini-tiation in the emergency departmentsetting, not after they’ve already beenhospitalized.

Fedor: I know that in our emergencydepartment, they will frequently ini-tiate NIV pretty commonly. Theyhave a number of devices now; theyeven have them on battery-poweredtransport carts. But they are all ad-mitted to the ICU if they’ve beencommitted to NIV in the emergencydepartment.

Berlinski: We have different viewsabout where that patient needs to go.In my opinion, the use of NIV treat-ment in the emergency department isa surrogate of severity, and I hope weall agree that a patient who requiresNIV has no other place to go than theICU. I have safety concerns if patientsare treated with NIV in the emergencydepartment and a few hours later areadmitted to a regular ward. I think youneed to have protocols to decide ther-apies and disposition. The use of score-based evaluations is crucial because itallows you to talk the same languagebetween the wards and the pediatricICU.

Lin: I believe that we agree that thepatient who presents to the emergencydepartment who is in severe distressand is in the upper 25% of whateveryour score range is, then that patientabsolutely needs to come to the ICU.

Whether they first receive NIV in theemergency department or not is reallya matter of hospital policy. For thosepatients who are in the bottom of yourscoring system, who get one treatmentof albuterol and do great, then we cer-tainly don’t need to be instituting NIVthere. But in those patients who are inthe middle, where the usual course ofemergency department therapy is theyget 3 intermittent nebs and they look alittle bit better, but somebody recom-mends a 2-3-h trial of continuous be-fore making the first call to the ICU… I wonder if that population of pa-tients, if initiated on NIV at the onsetof treatement to help facilitate decreas-ing their WOB would allow them torecover quicker and not turn a 4-hemergency department stay into a 12-hstay. Would that be a potential verynarrow selection of the patient popu-lation where empiric use of NIV earlyon is more effective if they don’t re-spond to that first neb as well as youwant them to, and would that popula-tion of patients be a very narrow in-dication for NIV initiation for asthmain the emergency department?

Berlinski: The Journal published astudy6 in adults using NIV and aero-sols, where the authors looked at in-trapulmonary deposition, FEV1, andoutcomes. And those who were receiv-ing NIV � the aerosol had the samedeposition but better FEV1. I think be-fore going that route, we need a largetrial because of the associated cost.We’re not talking about adding $5 pertherapy, but we are talking about atleast $2,500 to set up the vent for how-ever many moderate-to-severe asthmapatients who come into the emergencydepartment. I don’t think that an in-stitution can support this practice with-out clear data. A randomized con-trolled trial is the only way you cansay for sure when this is a cost-effec-tive treatment modality.

Sweet: Would you do that study inthe emergency department setting orin patients admitted to the ICU, with



the goal of determining whether thetime window of NIV needed to turnthe patient around would be shortenough to make doing it in the emer-gency department practical? If you’regoing to go down that path, you woulddesign your study to put the patient inthe ICU, assess their response receiv-ing NIV, and then ultimately movethe NIV down to the emergency de-partment if you really get a clear sig-nal that doing so would shorten emer-gency department stays and reduceICU admissions.

Lin: I would agree that the RCTshould happen in the ICU environ-ment, but the question then in apply-ing it broadly becomes if everybodythinks you need to come to the ICU inorder to get this started, that sort ofdefeats the purpose of potentially ini-tiating NIV sooner to prevent the ICUadmission in the first place. For safetypurposes, I think absolutely the initialRCT should be done in a higher levelof monitoring environment that has thepotential ability to escalate to higherlevels of support, whether that behigher levels of NIV or invasive ven-tilation. There’s certainly an educa-tional component. Many of our emer-gency department colleagues are lessexperienced with the initiation of NIVin the acute setting; it’s just not some-thing they will typically do. They’lleither intubate or send them up to theICU for additional management. Ithink it’s very interesting that thethought that perhaps early initiationof NIV could lead to a more rapidturnaround by just making it easierfor them to breathe and all of the as-sociated things—the anxiety, in-creased breathing frequency—if youcould get rid of that, it potentially al-lows some of these asthma patients toimprove. As an aside, over the past4 years in St Louis, we’ve seen 7 casesof asthma deaths, which has got to bean epidemic. There were 2 this pastweek alone. One didn’t come out ofthe emergency department, and the

other had brain death and support wasremoved in the ICU.

Walsh: I think the RCT should bedone in the emergency department, ifthe goal is to buy time for them be-cause they are fatiguing. The standardpractice is NIV use in the ICU if theyare fatiguing. In other words, we an-ticipate they will be intubated if weare not careful, and so we proactivelyuse it. I’m wondering if we can headoff fatigue like you’re saying if we areproactive in the emergency depart-ment. If we anticipate they’re fatigu-ing and are in that moderate-to-severecategory, then start it. You’re reallybuying time for steroids and thosetypes of things to kick in, and you’relooking for 6-8 h in those extreme pa-tients. Could you do it and save anICU admission?

Fedor: I would say you would alsohave to include upon admission be-cause so many of our admitted pa-tients come from direct admits to theICU from outside hospitals if you workin an academic setting.

Kuch: I think it’s interesting regard-ing the emergency department andwhere you would conduct this RCT.A couple things. Brian [Walsh], youmentioned that 6-8 h and trying to keepthem in the emergency room and notadmitted to the ICU. I think it becomesmore of a challenge unless you decideto do it in an observation unit, asthroughput is a big goal right now—toget them out and get them throughthe emergency department. In addi-tion, it begs the question, if you startsomeone on NIV because they’remoderately fatigued and starting toshow some signs, a little bit hypoxic,maybe creeping up CO2 where arethey going to end up anyway? I thinkthe default would be admitting themto the ICU even if you capture that;I think there’s a sense of at least aday in the ICU either way. That’s ahard call to make.

Walsh: No, but the outcome wouldbe that it didn’t help, and it would bea negative study. That’s what I’m say-ing, if you can—early intervention iswhat we’re talking about, if you canintervene early—keep them from go-ing to the ICU, then it’s a worthy trial.If not, like Ariel is saying, there is noevidence to say whether it’s the rightthing to do because it adds to cost. So,if you still go to the ICU because yougot started on NIV, you just added tothe cost and not necessarily takenaway.

Kuch: I agree, I feel that finding thatpatient population might be a verysmall window where you might be ableto save them from the ICU versus go-ing to the ICU either way.

Sweet: I’m stirring the pot here, butmy idea was to adjust the current stan-dard of care to then retrospectively iden-tify which patients have a rapid turn-around and might not have needed to goto the ICU. You might be able to retro-spectively get some data from your crit-ical care units to clarify that, but to jus-tify an RCT and start in the emergencydepartment with issues of education andpractice change may be challenging. Itmight be easier to get buy-in if you setforth the idea that we move somebodyquickly, get them started on NIVquickly, and then identify which patientpopulation subset actually got betterquickly and wouldn’t ultimately need tobe in the ICU.

Kuch: My point is that the needed-to-treat number might need to be verylarge to see that, if that makes sense.

Panitch: I would say that in my in-stitution, currently our pulmonaryfloor takes care of some children whorequire NIV and who are very sick.For instance, we admit children withasthma who receive magnesium in theemergency department and are receiv-ing continuous nebulized albuterol,and some of my colleagues wonderwhy we do that and say, “there is



anICU very close by; why don’t weput this patient where he or she be-longs?” So, as we push the envelopein terms of what we do, I think there’snothing wrong with increased obser-vation briefly. The trick is once thosesick patients start to get better, canyou get them back out of the ICU andinto a less expensive, less restrictiveenvironment? But initially, I think thecaregivers all might be more appre-ciative of that ICU environment.

Walsh: I want to go back to Tom’s[Kallstrom] comment because some of

the people sitting around this table ac-tually do this. A lot of times hospitalsnow are providing RTs at their owncost just to try and take better care ofthese kids. I think it’s admirable, Iwished everyone did it, it’s one wayto try and get these medically com-plex kids from the ICU or hospitalsetting to home, and we’re trying tobridge that gap by paying for it our-selves because we sadly don’t get re-imbursed.

REFERENCES

1. Pediatric Acute Lung Injury ConsensusConference Group. Pediatric acute respira-tory distress syndrome: consensus recom-mendations from the Pediatric Acute LungInjury Consensus Conference. Pediatr CritCare Med 2015;16(5):428-439.

2. Khemani RG, Smith LS, Zimmerman JJ,Erickson S, Pediatric Acute Lung InjuryConsensus Conference Group. Pediatricacute respiratory distress syndrome: def-inition, incidence, and epidemiology: pro-ceedings from the Pediatric Acute LungInjury Consensus Conference. Pediatr CritCare Med 2015;16(5 Suppl 1):S23-S40.

3. Essouri S, Carroll C, Pediatric Acute LungInjury Consensus Conference Group. Non-invasive support and ventilation for pedi-atric acute respiratory distress syndrome:proceedings from the Pediatric Acute LungInjury Consensus Conference. Pediatr CritCare Med 2015;16(5 Suppl 1):S102-S110.

4. Current trends in neonatal and pediatric re-spiratory care. Part 1. Respir Care 2003;48(3):192-295.

5. Current trends in neonatal and pediatric re-spiratory care. Part 2. Respir Care2003;48(4):352-464.

6. Galindo-Filho VC, Brandao DC, FerreiraRde C, Menezes MJ, Almeida-Filho P, Par-reira VF, et al. Noninvasive ventilation cou-pled with nebulization during asthma cri-ses: a randomized controlled trial. RespirCare 2013;58(2):241-249.

This article is approved for Continuing Respiratory Care Educationcredit. For information and to obtain your CRCE

(free to AARC members) visitwww.rcjournal.com

* Timothy R Myers MBA RRT-NPS FAARC,invited discussant. Mr Myers is the Chief Busi-ness Officer, American Association for Respi-ratory Care.† Thomas J Kallstrom RRT MBA FAARC, in-vited discussant. Mr Kallstrom is the Executive

Director, American Association for RespiratoryCare.



noninvasive respiratory support in infants and...

Documents