Automatic Actuation of a Dry Powder Inhaler into a
Pulmonary Service, Department of Pediatrics, Rigshospitalet, National University Hospital, Copenhagen, Denmark
This article describes a new “automatic spacer” device, which has been developed to improve the de-
livery of inhaled medication to young children. In the device, a dry powder inhaler (DPI) is mechanically
actuated into a nonelectrostatic spacer, producing an aerosol cloud of fine drug particles (aerody-
namic diameter, 4.7 m) with a long half-life. The new device combines the principal advantages
of the conventional spacer and the DPI. It has the potential to provide a high ratio between lung
dose and pharyngeal dose, without need for coordination or forced inhalation, and it avoids expo-
sure of the patient to the additives and propellants used in pressurized metered dose inhalers. Stud-
ies with the protototype device show a high yield of fine drug particles in the aerosol (mass median
aerodynamic diameter, 2.8 m), a high repeatability of drug delivery owing to the mechanical nature
of the actuation (relative standard deviation, 12%), and a prolonged residence time of the fine particle
aerosol (half-life of the fallout of the fine particles, 82 s). These features should prove advantageous
in the treatment of young children with inhaled medication. Bisgaard H. Automatic actuation of
a dry powder inhaler into a nonelectrostatic spacer. AM J RESPIR CRIT CARE MED 1998;157:518–521.
The potential efficacy of inhaled medication in young children The mouthpiece is equipped with a one-way valve to prevent re-
and infants is often compromised by the difficulty of delivery breathing back into the holding chamber, and the face mask is
of therapy (1). With the aim of overcoming the disadvantages equipped with an exhalation valve to permit exhalation, so in both
of existing inhaler systems (Table 1), a novel system is sug- systems the inspiratory and expiratory lines are separated without
common dead space.
gested in which a spacer constructed from nonelectrostatic
material is combined with a dry powder inhaler (DPI; Turbu- Test Drug
haler), and the DPI is activated mechanically into the spacer.
The test drug, budesonide, is administered by Turbuhaler at 400 g/
Suction is provided by the generation of a partial vacuum in
the spacer on release of a spring-driven piston, which can be
reloaded manually (Figure 1). The metered drug powder is Ventilator
drawn into the spacer as an aerosol, where it remains stable,
A Pari Sinus breath simulator was used to withdraw drug aerosol from
ready for inhalation. This concept is likely to improve our abil-
the spacer by simulated tidal breathing. The settings selected to mimic
ity to treat young children with inhaled drug aerosols, as it pro- the breathing of a young child were as follows: tidal volumes of 100
vides a stable aerosol with a long half-life and without additives and 200 mL; inhalation-to-exhalation ratio, 1:2; frequency, 25 breaths/
and propellants. min; duration, 30 s (unless otherwise indicated).
METHODS Dose Ex-spacer, Constant Flow
Automatic Spacer Prototype The spacer was connected to a filter (Vital Signs, model 5098E) 2 s af-
ter actuation, and a constant flow of 30 L/min was drawn through the
The spacer is constructed from nonelectrostatic material and is 275 ml
spacer for the next 5 s. Dose ex-spacer was collected on the filter dur-
in volume. The spacer is combined with a DPI (Turbuhaler), and the
ing this period.
DPI is activated mechanically into the spacer. Suction is provided by
the generation of a partial vacuum in the spacer on release of a spring-
Dose Ex-spacer, Ventilator
driven piston, which can be reloaded manually (Figure 1). The suction
time is 250 ms, with a rise time of 160 ms to reach the peak flow of ap- A filter (Vital Signs, model 5098E) was placed in the inspiratory line
proximately 60 L/min. The metered drug powder is drawn into the between ventilator and the spacer. Inspiratory and expiratory flow
spacer as an aerosol, where it remains stable, ready for inhalation. passed through separate lines, ensuring unidirectional inspiratory flow
Normal tidal breathing by the patient, through either a mouthpiece or through the filter. The ventilator was started 2 s after actuation of the
a specially adapted face mask, delivers aerosolized drug to the patient. Turbuhaler. Dose ex-spacer was collected on the filter.
Particle Size Distribution
(Received in original form May 9, 1997 and in revised form October 15, 1997)
The budesonide dose ex-spacer was collected in an Andersen sampler
This study was funded by ASTRA. with a modified BP twin impinger inlet by applying a constant flow of
Correspondence and requests for reprints should be addressed to Dr. Hans Bis- 28.3 L/min through the impactor for 10 s. The spacer was connected to
gaard, Pulmonary Service, Department of Pediatrics, Rigshospitalet, National the impactor 2 s after actuation of Turbuhaler (unless otherwise indi-
University Hospital, Copenhagen, Denmark. E-mail: Bisgaard@RH.DK cated). The amount of budesonide on each stage of the sampler was
Am J Respir Crit Care Med Vol 157. pp 518–521, 1998 determined.
Bisgaard: Dry-Powder Inhaler with Spacer 519
ADVANTAGES AND DISADVANTAGES OF INHALER DEVICES FOR YOUNG CHILDREN
Nebulizer No coordination Time-consuming
Independent of effort Expensive
No harmful additives Limited portability
Long availability of aerosol Technically demanding
Extensive waste of drug remaining in nebulizer
Flow of aerosol through both inspiration and
expiration leads to:
Drug exposure of skin and eyes
Spacer plus pMDI No coordination required Lubricants and surfactants (airway irritants)
Independent of effort Propellants (ozone depletion or greenhouse gas)
Few coarse particles Separation of drug and propellant (potential
Standing cloud of aerosol for poor repeatability)
Variable electrostatic attraction of particles to
plastic spacer material (reduces repeatability
Dead space between inspiratory and expiratory
valves (reduces cost-effectiveness)
DPI No coordination Effort dependent (potential for poor
No harmful additives repeatability and unsuitable for young
Large fraction of coarse particles (increased
upper airway deposition)
Automatic spacer plus DPI No coordination New concept requiring further study and
Independent of effort clinical trials
Few coarse particles
No harmful additives
High dose-to-dose repeatability
High yield of drug from spacer
Standing cloud of aerosol
Long availability of aerosol
Definition of abbreviations: pMDI pressurized-metered dose inhaler; DPI dry powder inhaler.
Quantification of Budesonide Dose half-life (total dose). Dose ex-spacer was measured with de-
The budesonide on the filters and the stages of the sampler was dis- lays of 2, 10, 20, 40, and 60 s after actuation five times from each of
solved in ethanol containing internal standard (fluocinolone ace- three unused automatic spacers. These measurements were repeated
tonide). The amount of budesonide was quantified by high-pressure after the automatic spacer had been used to withdraw 1,500 doses.
liquid chromatography (relative standard deviation, 2%). The limit Dose half-life (fine and coarse particle dose). The half-life of the
of detection is 0.6 g of budesonide. coarse and fine particles was determined by measuring the amounts
collected on the various stages of the Andersen sampler with delays
after actuation of 2, 10, 20, 40, and 60 s and using three different Tur-
Study Design buhaler inhalers and a single automatic spacer.
Four automatic spacers were used throughout the study, and were
combined with different Turbuhaler inhalers in balanced designs. Be- Statistical Evaluation
fore and after each experiment the delivered dose from each individ- The recovered doses were converted to percentages of the mean de-
ual Turbuhaler was determined by five replicate filter tests. livered dose for each Turbuhaler, as measured in replicates of five, be-
Dose-to-dose repeatability. Dose-to-dose repeatability was deter- fore and after each test. All experiments were complete and balanced.
mined by measuring dose ex-spacer in the 4 automatic spacers with 4 The means given are therefore the observed arithmetic means. Re-
different Turbuhaler inhalers, giving a total of 16 combinations with 5 peatability is expressed as the relative standard deviation (RSD) cal-
replicates of each. culated as the SD per mean. The half-life of the aerosol was estimated
Dose ex-spacer during constant flow and tidal breathing. Doses by fitting a linear line to the logged mean recovered doses versus time.
were withdrawn both with a constant flow of 30 L/min and with a The dose half-life could then be estimated as ln(2)/slope.
breathing simulator (tidal volume, 200 mL) from four automatic spac-
ers in replicates of five. RESULTS
Number of breaths to empty spacer. Doses obtained by the breath-
ing simulator were measured after 1, 2, 3, 4, 5, 6, and 10 breaths with Dose-to-dose repeatability. The RSD of dose delivery from the
tidal volumes of 100 and 200 mL. The study was performed five times four Turbuhaler was 10%. Using the same inhalers the RSD
for each number of breaths, with one automatic spacer and one Tur- of the Turbuhaler–spacer combination was 12%.
buhaler. Dose ex-spacer during constant flow and tidal breathing.
Particle size distribution. Particle size distribution was determined
The dose obtained ex-spacer by constant flow was 40% of the
2 s after actuation by applying a constant flow of 28.3 L/min into an
Andersen sampler. Particles of 4.7 m were termed fine particles,
delivered dose from Turbuhaler as compared with 38% when
and those with larger diameters were termed coarse particles. obtained by tidal breathing, which was not statistically differ-
Three automatic spacers were used, each with three inhalers (i.e., ent (difference, 1.4%; 95% CI, 1.1–4.0).
nine measurements in total). In each measurement, 10 doses were col- Number of breaths to empty spacer. After three breaths of
lected. These measurements were repeated after the spacer had been 200 mL or four breaths of 100 mL no increase in dose ex-
used to withdraw 1,500 doses. spacer was obtainable with further breathing.
520 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 157 1998
Figure 1. A series of scale drawings illustrating the mode of action of the automatic spacer. (A) A section
through the device in the resting state. The drug is delivered from a standard dry powder inhaler (Turbu-
haler), seen on the left. On the right is the spacer chamber in its closed state, beneath which is a spring
under tension. (B) Actuation of the device is accomplished by lifting the spacer chamber on the body of
the device, and then twisting it through 90 . This action primes the Turbuhaler, via a gearing system, and
also releases the spring. (C) Release and expansion of the spring lead to rapid expansion in volume of the
spacer chamber. The mouthpiece is closed by the one-way valve, so the resulting partial vacuum draws air
through the Turbuhaler, into the tower-shaped spacer. The drug aerosol created by this action persists as
a stable cloud in the spacer chamber with a long half-life of 82 s for the fine particle fraction. (D) An exter-
nal view of the device, showing the aerosol path through the mouthpiece. A face mask can be fitted when
the device is used by younger children.
Bisgaard: Dry-Powder Inhaler with Spacer 521
TABLE 2 irritant effects in the airways of the patients as seen from use
PARTICLE SIZES FROM AUTOMATIC of propellant-metered dose inhalers (pMDIs) (2–4), and envi-
SPACER PLUS TURBUHALER ronmental damage including the ozone-depleting effect and
the greenhouse gas effect as caused by available pMDI pro-
MMAD GSD pellants (1).
5.8 m 4.7 m ( m) ( m) There is no requirement for forced inspiration or good co-
New automatic spacer 83% (2) 76% (2) 2.79 1.8 ordination, as opposed to the use of a DPI or pMDI alone. As
After 1,500 doses 79% (4) 73% (5) 2.79 1.8 a consequence of the standardized nature of the mechanical
actuation repeatability of dosing was high in this study (RSD,
Definition of abbreviations: MMAD mass median aerodynamic diameter; GSD
geometric standard deviation. 12%). This is likely to improve on repeatability of drug deliv-
* Values are mean (relative SD). ery to the lungs, as the variation ascribed to changes in inspira-
tory flow is avoided.
The use of nonelectrostatic materials together with elimi-
Particle size distribution. Particle size distribution is shown nation of dead space in the inspiratory line ensure a high yield
in Table 2. of drug from the spacer. The dose ex-spacer was 40% of the
Dose half-life. The half-life of the dose ex-spacer was esti- dose delivered to the spacer from the Turbuhaler. This effi-
mated to be 68 s (95% CI, 62–76 s) with no significant differ- ciency is comparable to the efficiency of a nonelectrostatic
ence between a new and a used automatic spacer. The esti- spacer for use with pMDIs (5, 6).
mated half-life of the fine particle dose was 82 s (95% CI, 66– The residence time of the aerosol is prolonged compared
109 s), whereas the estimated half-life of the coarse particle with traditional spacers. The half-life of 82 s for the fallout of
dose was 36 s (95% CI, 29–47 s) (Figure 2). the aerosol cloud of fine particles (diameter, 4.7 m) com-
pares favorably with a half-life of about 10 s for the fallout of
DISCUSSION fine particles in most standard plastic spacers (7). The slow,
passive fallout is assured by the nonelectrostatic material of
Spacer devices are convenient for the treatment of young chil- the spacer avoiding attraction of the aerosol particles to the
dren as they allow aerosol to be inhaled during tidal breathing, wall of the traditional plastic spacers (1), and by the tower
without the need for coordination or forced inhalation. They shape of the spacer, which maximizes the settling distance for
generally improve the safety of inhaled medication, as they the particles. The long residence time of the aerosol provides
lead to a reduced pharyngeal dose and an increased lung dose an extended period for the young child or infant to inhale the
and clinical effect. required dose. During normal breathing, a 1-yr-old child in-
The mechanical actuation of a DPI (Turbuhaler) into a hales the available dose in three or four breaths, but even in
spacer, the “automatic spacer,” provides a new form of spacer the case of a noncompliant toddler, who may breath-hold or
device for aerosol treatment. The new system can be expected perform periods of shallow breathing, passive fallout of aero-
to retain the advantages of existing spacers. In addition, the sol will have little effect on the dose obtained. The require-
automatic spacer provides the advantage of a drug aerosol de- ments for even passive acceptance are therefore minimal, and
livered without use of any harmful additives. The DPI within the prolonged residence time of aerosol ensures both a high
the device (Turbuhaler) delivers respiratory medication with and repeatable drug delivery, and should permit effective in-
no propellants, lubricants, or surfactants, avoiding unwanted halation, even by children who have previously been unable to
use spacer devices effectively for lack of compliance.
In conclusion, the novel automatic spacer device combines
the advantages of a spacer and DPI without the disadvantages
of either, which should improve our ability to treat young chil-
dren with inhaled drug aerosols.
Acknowledgment : The author acknowledges the technical assistance of
Elna Berg and Sven Andersson of Astra Draco AB in this study.
1. Bisgaard, H. 1997. Delivery of inhaled medication to children. J. Asthma
2. Cochetto, D., R. S. Sykes, and S. Spector. 1991. Paradoxical broncho-
spasm after use of inhalation aerosols: a review of the literature. J.
3. Wilkinson, J. R. W., J. A. Roberts, P. Bradding, S. T. Holgate, and P. H.
Howarth. 1992. Paradoxical bronchoconstriction in asthmatic patients
after salmeterol by metered dose inhaler. B.M.J. 305:931–932.
4. Shaheen, M. Z., J. G. Ayres, and C. Benincasa. 1994. Incidence of acute
decreases in peak expiratory flow following the use of metered-dose in-
halers in asthmatic patients. Eur. Respir. J. 7:2160–2164.
5. Bisgaard, H. 1995. A metal aerosol holding chamber devised for young
Figure 2. Passive fallout of aerosol in spacer. The spacer was emp- children with asthma. Eur. Respir. J. 8:856–860.
tied after increasing delay periods after actuation. The estimated 6. Bisgaard, H., J. Anhøj, B. Klug, and E. Berg. 1995. A non-electrostatic
half-life of the total dose of aerosol was 67 s (95% CI, 56–82 s). spacer for aerosol delivery. Arch. Dis. Child. 73:226–230.
The estimated half-life of the fine particle dose was 82 s (95% CI, 7. Bisgaard, H. 1997. Delivery options for inhaled therapy in children under
66–109 s), whereas the half-life of the coarse particle dose was 36 s the age of 6 years. J. Aerosol Med. 10:37–40.
(95% CI, 29–47 s).