Objective. Masks are an essential interface between valved holding chambers (VHCs), or spacers, and a small child's face for providing aerosol therapy. Clinical experience suggests that many young children do not cooperate with the VHC treatment or tolerate a mask of any kind. This might impair the mask–face seal and reduce the dose delivered to the child. The objective of this study was to evaluate the ability of parents to provide a good mask–face seal in infants and toddlers using 3 masks provided with commonly used pediatric VHCs and compare this with the seal obtained with the Hans Rudolph pediatric anesthesia mask.
Methods. A preliminary in vitro filter study was conducted to validate the assumption that reduced ventilation as a result of increased facemask leak reduces the drug aerosol dose delivered to the mouth. Facemask leak then was studied in vivo for NebuChamber, AeroChamber, BabyHaler, and Hans Rudolph masks by measuring ventilation with an in-line pneumotachograph while the facemask was held in place by experienced parents who were asked to demonstrate how they deliver medication to their children without any additional instruction. Thirty children (mean age: 3.2 ± 1.4 years) performed 4 repeat studies with each mask. The first 10 patients performed the tests once again within 1 month. On the second occasion, the parents were coached continuously and encouraged to hold the mask tightly against the child's face.
Results. The AeroChamber and Hans Rudolph masks provided the best seal as reflected in the magnitude of the ventilation measured through them. The NebuChamber provided the poorest seal, with 45% less ventilation than the AeroChamber and Hans Rudolph masks. There was considerable intraindividual variability for all masks (24% to 48%); however, the variability with the NebuChamber mask was 2-fold greater than the other masks. All ventilatory volumes during the coached session were significantly greater than during the uncoached session. Variability during the coached session was significantly less (except for the BabyHaler, which remained unchanged).
Conclusions. VHCs with masks designed for use with small children may provide a poor seal with the face, leading to reduced or more variable dose delivery. The facemask seal is critical for efficient aerosol delivery to infants and young children, and this should be stressed to parents.
The development of spacers and valved holding chambers (VHCs) as metered-dose inhaler (MDI) accessories has greatly improved inhalation treatment in young children.1,,2 There are 2 main advantages to these add-on devices: 1) aerodynamic filtration: the chamber's dimensions allow the larger particles to decelerate and deposit in the device instead of in the mouth or throat, thus reducing impaction of particles in the upper airway; particularly with inhaled steroids, this greatly decreases local upper respiratory tract side effects (hoarseness, candidiasis) and systemic absorption of medications and improves the therapeutic ratio3; and 2) provision of a reservoir of aerosol from which the infant or child can breathe tidally, thus ensuring aerosol delivery on inspiration without the need to coordinate precisely the aerosol discharge and inhalation.
The addition of a facemask allows VHCs to be used with infants who usually are nosebreathers and are too young to breathe through a mouthpiece.4 In pediatric practice, some VHCs, designed specifically for infants' low tidal volume and inspiratory efforts, have improved our ability to treat adequately even premature neonates.5 This has been achieved by means of a relatively small total volume to achieve a high concentration of small particles together with very low inspiratory valve resistance and minimal dead space.
Although the mask interface undoubtedly is an important determinant of the aerosol dose delivered from the VHC to the infant's or toddler's nose and mouth, its optimal characteristics have gained relatively little attention and comparative studies of various facemasks are lacking. The theoretical factors that determine mask efficiency include mask–face seal, mask dead space, compliance, resistance and seal of inspiration and expiration valves, and the dilutional effect of air entering the mask from exhalation ports. For example, a mask with a large dead space or a leak around the mask will decrease the dose of aerosol obtained from the VHC, particularly in low tidal volume breathing infants. A good seal around the nose and mouth will ensure that inspiration is from the device rather than from the environment.
This study evaluated the ability of parents to provide a good mask–face seal with infants and toddlers. Rather than rigorously, controlled, and supervised laboratory conditions, we encouraged the participating parents to imitate the technique that they usually use at home to provide aerosol medication to their children. We measured the air leak of the 3 commercially available facemasks that are an integral part of 3 widely available pediatric VHCs and compared it with that of a Hans Rudolph anesthesia mask, which was considered the standard.
The study consisted of 2 parts. The first evaluated the in vitro relationship between air leak and the dose delivered to the mouth. The second was a controlled, randomized, crossover clinical study in patients.
In Vitro Study
Previous in vitro and in vivo studies showed a positive correlation between ventilation and the aerosol dose delivered to the mouth.6–8 These studies were conducted with a tight mask-to-face seal, however. To validate this assumed relationship also under the conditions used in the present study, anticipating a less-than-optimal seal, we conducted a preliminary study. This related ventilation through the mask to the dose of albuterol delivered to the mouth with a poor seal. To minimize the influence of the chamber on the delivered dose, we used the NebuChamber VHC, whose chamber (unrelated to the sealing characteristics of the mask) is known to have the most consistent aerosol output.9 The delivered dose was defined as the albuterol dose deposited on a filter inserted between the NebuChamber and the mask. A new albuterol MDI (100 μg per actuation) was used. The canister was shaken vigorously for 10 seconds, discharged 10 times, and then tested. Testing consisted of shaking the inhaler vigorously and then firing 1 puff into the NebuChamber. This was followed immediately by breathing from the mask for 15 seconds while ventilation (through the mask) was measured using a screen pneumotachograph from a portable spirometer (Flowscreen; E. Jaeger, Wuerzburg, Germany) inserted between the mask and the filter holder. The total additional dead space was 26 mL. Twenty-six tests were conducted by 1 fully cooperative healthy adult. Similar respiratory rate and tidal volumes in each test were achieved by means of a metronome. The major independent variable was the leak between the face and the mask. This was accomplished by having the participant hold the mask at various distances (0–10 mm) from the face and by applying various pressures on the mask at 0 distance.
We used electrostatic filters (Pari GmbH, Starnberg, Germany), which were shown previously to collect virtually 100% of the aerosolized medication dose.9 This technique of dose assessment was shown previously to be accurate in determining the dose delivered to the patient.6,8–10 The filters were analyzed for albuterol content by ultraviolet spectrophotometry, and the dose trapped on the filter was related to the corresponding ventilation by simple regression.
Thirty-nine children with asthma or wheezing were enrolled. Inclusion criteria included age younger than 5 years, use of any VHC with a facemask continuously for at least 1 month following appropriate instruction (1-on-1 educational session by a respiratory nurse or pediatric pulmonologist), and absence of symptoms on the day of the study. Excluded were patients who did not allow application of more than 1 mask for more than 1 test period.
The study was conducted after a routine outpatient clinic visit. The first consecutive 10 patients returned within 1 month for an additional assessment. During their first assessment, parents were asked to demonstrate to us how they deliver medications to their children through the masks without any additional instruction. During the second assessment, they were coached continuously and encouraged verbally (by 1 of the authors [IA]) to hold the mask tightly against the child's face during the simulated treatment. The study was approved by the Sieff Hospital ethics committee, and parents gave their written informed consent.
The following silicone masks were studied:
NebuChamber (Astra Draco AB, Lund, Sweden): a small, thin, relatively stiff strait-edged mask, triangularly shaped, and elongated, which must be held in a single vertical position on the face.
BabyHaler (infant size #2; Glaxo GmbH, Germany): a round mask with rolled edge (inside diameter: 5 cm).
AeroChamber (child mask; Trudell Medical, London; Ontario, Canada): a round mask with rolled edge (inside diameter: 5.8 cm), the valve opening from the VHC is located eccentrically with respect to the mask. The mask contains an integral exhalation valve, which was sealed for the purpose of the study (because we were interested only in ventilation through the mask, sealing the exhalation valve has no effect on this variable).
Hans Rudolph (#2, pt #669078–6995; Hans Rudolph, Inc, Kansas City, MO): a mask used mainly for anesthesia. It has a lower flap that helps to stabilize the mask against the chin. This is an anatomically contoured mask that was used as the standard against which the performance of the other masks was compared.
All masks covered both the nose and the mouth.
Assessment of Mask Leakage In Vivo
Figure 1 shows the experimental set-up. We measured the minute ventilation through the 4 masks while parents held the mask on their child's face. During the initial phase of the clinical study, parents were instructed to hold the mask and apply it to the face just the way that they do it at home during actual treatments. No additional instruction, correction of faulty technique, or other assistance was provided.
For the purpose of the study, the inlet portion of the mask, usually 22 mm outer diameter, was attached directly to a screen pneumotachograph of a portable electronic spirometer, calibrated on each of the study days. To quantify ventilation through the mask, we used the maximal voluntary ventilation (MVV) function. Once the mask was applied to the face, a few seconds were allowed to stabilize respiration after which the MVV test was initiated. MVV was measured for 12 seconds. The child rested for a few seconds until the results were recorded. The pneumotachograph was reset, and then the next test was done. For each mask, at least 4 such measurements were obtained. Mask administration was randomized. Approximately 15 minutes were required to test all of the masks in each patient.
Air leakage around the mask during tidal breathing was determined indirectly by comparing the ventilated volume through each mask by means of the pneumotachograph. Thus, the better the seal to the face, the greater the measured ventilation. A leak would decrease measured ventilation. However, increased MVV without any change in the facemask seal could result simply from increased respiratory rate, for example, as a result of agitation. Thus, differences in measured ventilation might not necessarily be caused by a deficient seal. To correct for such a possibility, we compared respiratory rate (which is not affected by the integrity of the seal) between the tests from each patient. Patients whose coefficient of variation (CoV) of respiratory rate was greater than 20% (2 patients) were excluded from the analysis.
Intraindividual variability in measured ventilation was expressed as CoV. Statistical analysis was done with analysis of variance with a significance level of 95% (P < .05).
The in vitro relationship between measured ventilation and the dose delivered to the mouth (onto the filter) using variable degrees of facemask leak is shown in Fig 2. There was a positive linear relationship between the 2 variables (r = 0.55, P = .004). The greater the ventilation, the greater the delivered dose.
Thirty patients (17 boys and 13 girls; mean age: 3.2 ± 1.4 years; asthma duration: 1.1 ± 1.4 years) completed the clinical study. Ventilation measured for 12 seconds, expressed as liters per minute, and the CoV are shown in Table 1.
The AeroChamber and Hans Rudolph masks sealed most effectively as reflected in the magnitude of the leak measured through them. The NebuChamber had the poorest seal, with ventilation of 2.89 L/min, significantly less than the AeroChamber (4.24 L/min; P< .001), Hans Rudolph (4.22 L/min; P < .001), or BabyHaler masks (3.82 L/min; P < .001). There was considerable intraindividual variability for all masks; however, the variability with the NebuChamber mask was 2-fold greater than the other masks (P < .001).
Ten patients performed the tests on 2 separate occasions (before and after additional instruction). The demographics and the baseline (uncoached session) ventilation and CoV data of these 10 patients were not significantly different from the whole group of 30 patients. All ventilatory volumes of these 10 patients during the coached session were significantly (P < .05) greater than during their previous uncoached session (Fig 3). Similarly, variability during the coached session was significantly (P < .05) less, except for the BabyHaler, which remained unchanged (Fig 4).
Of 10 enrolled infants who were younger than 2 years, only 3 cooperated fully and were able to complete at least 3 attempts with each of the masks. The others would not remain still for the 12 to 15 seconds required to complete the measurements. They became agitated and pulled the mask off their face. No patient who was older than 2 years had to be excluded as a result of noncooperation.
This study demonstrated that if there is not a good seal between the mask and the child's face, then ventilation through the mask may be reduced considerably, as shown, in particular, by the less substantial NebuChamber mask. On the basis of the results of our preliminary in vitro study and because clinically relevant drug particles, in particular the fine particle fraction (<3 μm), should follow ventilation through the mask,8 it is highly likely that this also would be the case for the dose of therapeutic aerosol.
The variability of ventilation through the masks was remarkably high. Even with the same VHC mask and in the same patient, there can be considerable variability, which would be likely to translate into marked variability in the aerosol dose delivered from VHCs to the child's mouth. Indeed, relatively large variability was demonstrated previously in studies that were conducted under well-controlled laboratory conditions in which patients were instructed to seal the mask tightly on the face and all efforts were taken to ensure this fit, either by the investigator or by a nurse who was supervising the inhalation.6,,7,11 That is unlike the situation in real life. Even if parents have been instructed and thus are somewhat familiar with aerosol administration using VHC and mask, it is evident from the present study that children often do not achieve an optimally tight fit between the mask and the face. Furthermore, although both ventilation and variability improved markedly when the parents were coached to apply the mask to the face properly, variability still remained between 20% (with the AeroChamber mask) and 28% (with the BabyHaler mask).
In 2 recent studies, considerable variability in the dose of aerosol delivered via VHCs was noted under real-life conditions.12,,13 When a facemask is used in a similar age group, the mean CoV of the doses delivered in these studies using a filter technique was 34% for the NebuChamber metal spacer and 37% for plastic devices, values that approximate the values for CoV of ventilation observed in our study.
Despite the large intraindividual variability, the NebuChamber mask performed significantly worse than the other 3 masks and would be likely to deliver less aerosol to the mouth of the infant. This probably is because the NebuChamber mask has a sharp, flat, and relatively firm edge, whereas the others have an inwardly rounded and very flexible edge that seems to adjust better to the facial contours of the child.
The mask is an integral and vital component of the MDI/VHC delivery system in infants and toddlers, and the facemask seal is the most problematic and least robust link in the chain of events from the MDI to the child's respiratory tract. The preliminary study confirmed the validity of using total ventilation through the VHC masks to reflect the effects of mask–face seal on the aerosol dose presented to the nose and the mouth. It is a simple and practical method for evaluating facemask seal in young children in vivo.
Given the need to measure ventilation, this study was designed to simulate, as closely as possible, aerosol delivery as undertaken at home on a daily basis in infants and toddlers. During the first part of the clinical study, apart from a request to demonstrate the usual manner of applying the facemask to the child's face as parents had been instructed previously, no additional assistance was provided to parents. This was an attempt to simulate real-world conditions at home, although we recognize that in an office setting, with their technique being observed, the parents may have been anxious and so might have performed suboptimally.
Previous studies that assessed the performance of VHCs focused on the chamber portion of the device. For example, it has been demonstrated that the metal NebuChamber eliminates electrostatic charge6,,7 that develops on the walls of most plastic chambers unless the latter are washed from time to time with detergent.14 Factors related to chamber size15 and valve function16,,17 also are important. However, no previous attempt has been made to evaluate the facemask seal for ensuring optimum ventilation and hence aerosol drug delivery from the chamber to the lung. The facemask seal arguably is one of the most important links in the chain of events to ensure optimum aerosol therapy in this age group, and regardless of other known factors, a poor mask seal would result in an unpredictably reduced dose to the lower respiratory tract.
Of particular interest is the fit obtained in infants. Of the 10 infants who were younger than 2 years that we tried to study, reasonable cooperation was obtained in only 3. Lack of cooperation with mask treatment in very young patients was noted previously by Noble et al18, who found that many of their patients did not accept the spacer with mask, even when asleep, and had to be withdrawn from a clinical study of inhaled budesonide. It is a common complaint of parents of infants of this age that it may be very difficult to keep a mask snugly fitted to the infant's face for more than a few seconds. Persisting with a screaming infant, as parents may do, is not a good solution as it has been shown that very little aerosol medication is deposited in the lungs under these conditions.4,,19Similarly, excessive pressure on the mask might encourage crying; a happy medium must be found.
One of the potential limitations of the study relates to the validity of a 12-second measurement of ventilation. The tidal volumes of the children who participated in the study were in the range of 50 to 200 mL.11,,20 Thus, assuming an average rate of 25 breaths/min11,,20 (ie, 1 breath every 2–3 seconds), the resulting 4 to 6 breaths should be enough to empty even the largest of the VHCs whose masks were used in this study. Thus, the period of time during which ventilation was measured in this study should have been adequate, representing approximately the actual time recommended for inhaling the medication. We are aware that MVV testing in young noncooperative infants may not be reproducible, and we therefore excluded noncooperative infants from the study. Thus, both the leak and the variability observed in our study may be underestimated compared with real-life situations.
Another potential concern is that application of the masks in this study was slightly different from the method used during the usual treatment at home, because instead of holding several chambers, they held the same pneumotachograph handle. The reason for the approach used was twofold: 1) we wanted to standardize the tests with the only independent variable being the mask seal, and 2) we found in pilot experiments that measuring ventilation through the mask alone not only was more convenient and practical than measuring the ventilation through the mask attached to the VHC but also allowed exclusion of the potential effect of the VHC on ventilation.
The results of the present study highlight that even with cooperative children who use VHCs with masks, it should not be assumed that poor drug efficacy or an inadequate prescribed dose is the cause when asthma control is difficult to achieve. Rather, it should be assumed until proved otherwise that aerosol delivery is inefficient and suboptimal. Thus, before the medication is changed or the dosage is increased, the caregiver should be asked to demonstrate how he or she uses the VHC with mask. The present study emphasizes the importance of educating parents that a good mask fit is all-important during aerosol treatment when using MDIs and VHCs with masks. This should be stressed not only to parents who use facemasks attached to VHCs but also to those who use facemasks with nebulizers, for which similar findings have been reported.21
With close supervision and encouragement to hold the mask tightly against the child's face, ventilation significantly improved compared with conditions under which no supervision or specific instruction was provided. Parental adherence to proper inhalation technique during treatments at home and over time may, however, diminish,22and the correct technique therefore must be reemphasized and checked frequently during follow-up visits, particularly if asthma control is suboptimal.
- Received August 25, 2000.
- Accepted February 8, 2001.
Reprint requests to (I.A.) Box 1079, Rosh-Pina, Israel 12000. E-mail:
- VHC =
- valved holding chamber •
- MDI =
- metered-dose inhaler •
- MVV =
- maximal voluntary ventilation •
- CoV =
- coefficient of variation
- O'Callaghan C,
- Barry P
- Bisgaard H
- Bisgaard H,
- Anhoj J,
- Klug B,
- Berg E
- Everard ML,
- Clark AR,
- Milner AD
- Berg E,
- Madsen J,
- Bisgaard H
- Agertoft L,
- Pedersen S
- Janssens HM,
- Devadason SG,
- Hop WC,
- LeSouef PN,
- De Jongste JC,
- Tiddens HA
- Janssens HM,
- Heijnen EMEW,
- De Jong VM,
- Hop WC,
- De Jongste JC,
- Tiddens HA
- Pierart F,
- Wildhaber JH,
- Vrancken I,
- Devadason SG,
- Le Souef PN
- Barry PW,
- O'Callaghan C
- Sennhauser FH,
- Sly PD
- Noble V,
- Ruggins NR,
- Everard ML,
- Milner AD
- Everard ML,
- Clark AR,
- Milner AD
- Copyright © 2001 American Academy of Pediatrics