Medication Nebulizer Performance*
`Effects Of Diluent Volume, Nebulizer Flow, and
`Nebulizer Brand
`
`Dean Hess, PhD, RRT; Daniel Fisher, BS, RRT; Purris Williams, BS, RRT;
`Sharon Pooler, RRT; and Robert M. Kacmarek, PhD, RRT
`
`Background: Medication nebulizers are commonly used to delivery aerosolized medications to pa(cid:173)
`tients with respiratory disease. We evaluated output and respirable aerosol available to the patient
`(inbaled mass) for I7 medication nebulizers using a spontaneous breathing lung model.
`Methods: Three nebulizer fill volumes (3, 4, and 5 mL containing 2.5 mg of albuterol) and 3 oxygen
`flows (6, 8, and 10 Umin) were evaluated using the I7 nebulizers. A cotton plug at the nebulizer
`mouthpiece was used to trap aerosol during simulated spontaneous breathing. Following each trial,
`the amount of albuterol remaining in the nebulizer and the amount deposited in the cotton plug were
`determined spectrophotometrically. Aerosol particle size was determined using an ll-stage cascade
`impactor.
`Results: Increasing fill volume decreased the amount of albuterol trapped in the dead volume
`(p<O.OOI) and increased the amount delivered to the patient (p<O.OOI). Increasing flow increased
`the mass output of particles in the respirable range of I to 5 pm (p=0.004), but the respirable mass
`delivered to the patient was affected to a greater extent by nebulizer brand (p<O.OOI) than flow.
`Although 2.5 mg of albuterol was placed into the nebulizers, less than 0.5 mg in the respirable range
`of I to 5 pm was delivered to the mouthpiece.
`Conclusions: The performance of medication nebulizers is affected by fill volume, flow, and nebu(cid:173)
`lizer brand. When they are used for research applications, the nebulizer characteristics must be
`evaluated and reported for the conditions used in the investigation. (CHEST 1996; 110:498-505)
`
`Key words: aerosol therapy; inhaled bronchodilator administration; nebulizers
`
`Abbreviations: GSD =geometric standard deviation; MMAD =mass median aerodynamic diameter
`
`D espite the common use of metered-dose inhalers
`and the availability of dry powder inhalers, aero(cid:173)
`solized medications are still frequently administered by
`nebulizer. Nebulizers are commonly used for inhaled
`
`For editorial comment see page 316
`
`bronchodilator administration to patients with reactive
`airways, including the perioperative and postoperative
`treatment of these patients. Advantages of nebulizers
`include the ability to use tl1em with patients who can(cid:173)
`not coordinate the use of a m etered-dose inhaler1 and the
`ability to conveniently administer a large (or continuous)
`
`*From the Department of Respiratory Care, Massachusetts Gen(cid:173)
`eral HosP.ital , and Harvard M edical School, Boston.
`Presentea, in part, at the annual meeting of the American Associ(cid:173)
`ation for Respiratory Care, Las Vegas, December 1994.
`Supported, in part, by Puritan-Bennett, Hudson RCI, Marquest,
`Professional M edical, SIMS.
`Manuscript received October 20, 1995; revision accepted F ebruary
`26, 1996.
`Reprint requests: Dr. Hess, Respiratory Care, Ellison 401 , Massa(cid:173)
`chusetts General Hospital, Boston, MA 02114
`
`dose into the lungs.2 Important characteristics of nebu(cid:173)
`lizer p erformance include the drug output, the aerosol
`particle size generated, the nebulization time, and the
`amount of drug delivered to the patient. Factors that have
`been shown to affect nebulizer performance include de(cid:173)
`vice construction (ie, manufacturer), fill volume, flow,
`temperature, and humidity of the driving gas.1
`A common feature of nebulizers is dead volume,
`which i s the volume of solution that remains in the
`nebulizer cup after aerosol production ends . Previous
`studies h ave typically evaluated dead volume by serial
`weighing.3-5 This method does not adequately charac(cid:173)
`terize drug output and amount of drug in the dead vol(cid:173)
`ume due to reconcentration in the nebulizer cup.6•7 He(cid:173)
`concentration occurs because of evaporation owning to
`the low relative humidity of the gas powering the nebu(cid:173)
`lizer. Nebulizer output should be determined more ap(cid:173)
`propriately by measuring the amount of medication that
`remains after a erosol production is complete.
`Particle size is an important characteristic of nebu(cid:173)
`lizer performance. Particles too large do not reach the
`
`498
`
`Laboratory and Animal Investigations
`
`IPR2021-00406
`United Therapeutics EX2031
`Page 1 of 8
`
`

`

`A
`
`o,
`
`ring
`stand
`
`c:
`"' E
`t:
`~
`-E
`"'o
`~()
`
`c: .. E
`t: """
`
`c:c.
`·;;: E
`·~ 0
`UU
`
`training test lung
`with lift bar
`
`8
`
`nebulizer
`
`o,
`
`ring
`stand
`
`vacuum
`
`FIGURE l. Top, A: experimental setup used to evaluate nebulizer
`output and mhaled mass. Bottom, B: experimental setup to evalu(cid:173)
`ate aerosol particle size output of the nebulizer.
`
`lower respiratory tract, whereas particles too small are
`exhaled.8 It has been shown that smaller particles are
`produced at higher nebulizer flows.9 Methods that
`determine particle sizes from nebulizers should classify
`them according to aerodynamic diameter and not
`physical diameter. A cascade impactor, unlike the op(cid:173)
`tical laser particle counter, allows quantification of
`drug delivery in terms of aerodynamic size character(cid:173)
`istics of the aerosol. 10 Aerodynamic diameter accounts
`for the density and irregular shape of drug particles and
`more accurately predicts the behavior of the aerosol as
`it is delivered to the patient.
`Loffert et al4 introduced the concept of respirable
`rate, which combines the effects of nebulizer output,
`nebulization time, and percent of particles in the
`respirable range. However, they determined nebulizer
`output by serial weighing rather than measurement of
`the amount of nebulized medication. A more useful
`index of nebulizer function might be respirable mass(cid:173)
`the amount of aerosolized drug in the respirable range.
`In this study, we extended the concept of Loffert
`et al4 by quantifying not only the respirable aerosol
`mass from the nebulizer, but also the respiratory
`aerosol mass made available to the patient at a specific
`ventilatory pattern.
`
`Table !-Nebulizer Brands Evaluated
`
`Nebulizer Brand; Manufacturer; Location
`
`Whisper Jet Nebulizer System; Marquest Medical Products
`Inc; Englewood, Colo
`Ava-Neb Nebulizer; Hudson Respiratory Care Inc;
`Temecula, Calif
`Raindrop Medication Nebulizer; Puritan-Bennett; Lenexa,
`Kan
`Vix One Nebulizer; Westmed Inc; Tucson, Ariz
`Sidestream Nondisposable; Inspired Medical Products;
`Pagham, West Sussex, UK
`T-Updraft II Neb-U-Mist Nebulizer; Hudson Respiratory
`Care Inc; Temecula, Calif
`Fan Jet Nebulizer; Westmed Inc; Tucson, Ariz
`One-No 8900 TG; Salter Labs; Arvin, Calif
`Airlife Misty Neb; Baxter Healthcare Corp; Valencia, Calif
`Hospitak; Lindenhurst, NY
`T Up-Draft; Hudson Respiratory Care Inc; Temecula, Calif
`Sidestream Disposable; Inspired Medical Products; Pagham,
`West Sussex, UK
`Intertech lnspiron; Intertech Resources Inc; Lincolnshire, Ill
`Betamist2 Medication Nebulizer; Professional Medical
`Products Inc; Greenwood, SC
`Micro-Mist; Hudson Respiratory Care Inc; Temecula, Calif
`Ventstream; Inspired Medical Products; Pagham, West
`Sussex, UK
`B & F Medical Products Inc; Toledo, Ohio
`
`A.
`
`B.
`
`C.
`
`D.
`E.
`
`F.
`
`G.
`H.
`I.
`J.
`K.
`L.
`
`M.
`N.
`
`0.
`P.
`
`Q.
`
`It has been suggested that nebulizers be character(cid:173)
`ized by the amount of medication that is delivered to
`the patient. Smaldone11 introduced the term inhaled
`rrwss, which he defined as that mass of drug actually
`delivered by a given nebulizer for a defined breathing
`pattern and period of time. Inhaled mass is affected not
`only by the performance of the nebulizer, but also by
`the breathing pattern chosen. For a given breathing
`pattern, inhaled mass should allow comparison of the
`quantity of drug delivered by different nebulizer sys(cid:173)
`tems and adjustment of the drug dose accordingly. To
`our knowledge, evaluation of inhaled mass has not
`been reported for nebulizers designed primarily for'
`delivery of bronchodilators.
`We conducted this study to evaluate medication
`nebulizer performance addressing those issues de(cid:173)
`scribed above. Dead volume, the amount of drug re(cid:173)
`maining in the dead volume, nebulization time, and
`aerosol available to the patient (inhaled mass) were
`evaluated for 17 nebulizers, 3 fill volumes, and 3 flows.
`We also evaluated particle size for 3 flows with the 17
`nebulizers at a single fill volume.
`
`MATERIALS AND METHODS
`
`Nebulizers Evaluated
`We evaluated 17 commercially available nebulizers (Table l) .
`Nebulizers were provided by their manufacturers. All units were
`from the same lot number and the same packaging case. The neb(cid:173)
`ulizers were provided by the manufacturer from their saleable stock
`none were prototypes or otherwise prepared specifically for thi~
`study, and all were used as supplied from the manufacturer.
`
`CHEST/110/2/AUGUST, 1996
`
`499
`
`IPR2021-00406
`United Therapeutics EX2031
`Page 2 of 8
`
`

`

`e-1 .4
`~1 .2
`n;
`_,[1 .0
`
`"E s 0.8
`~
`ill 0 . 6
`..2
`~ 0.4
`16
`~0.2
`
`0 .0
`
`~
`'5
`~ 1 .5
`
`3 mL 4 mL 5 mL
`6 L/min
`
`3 mL 4 mL 5 mL
`8 L/min
`
`3 mL 4 mL s mL
`10 L/min
`
`0>
`.§.
`E g 1 .0
`E ., ...
`
`E
`:::1
`~0. 5
`-m
`
`"0
`
`0 .0
`
`J K
`I
`A B C D E F G H
`nebulizer brand
`
`l M N 0 P Q
`
`FIGURE 2. Top: effect of volume ( p<0.001) and flow (p=0.02) on
`amount of albuterol trapped in the dead volume. Data are pooled
`from all nebulizers for each flow and volume setting. Bottom: effect
`of nebulizer brand on amount of albuterol trapped in dead volume
`(p <0.001). Data for each nebulizer brand are pooled from all vol(cid:173)
`ume and flow settings.
`
`Evaluation of Dead Volume and Aerosol Available to the Patient
`The experimental system is shown in Figure 1. The nebulizer was
`placed in a clamp and attached to a ting stand in the vetiical posi(cid:173)
`tion. A double-sided t est lung (Michigan Instruments; Grand Rap(cid:173)
`ids, Mich) was used to simulate spontaneous breathing. One side of
`the test lung was attached to a ventilator (Putitan-Bennett 7200
`ventilator; Puritan-Bennett; Carlsbad, Calif) and that side of the test
`lung lifted the contralateral side to simulate spontaneous breathing.
`The lung model was set to simulate breathing at 12 breaths/min,
`fraction of total respiratmy cycle inspiration (Ti!Ttot) approximately
`0.4, peak inspiratory flow approximately 0.3 to 0.4 Us with a sine
`wave pattern, and tidal volume approximately 0.45 to 0.5 L. A pi(cid:173)
`lot evaluation showed that this closely simulated the breathing pat(cid:173)
`tern of a normal volunteer breathing through a mouthpiece with(cid:173)
`out nose clips (imitating mouthpiece breathing witl1 a nebulizer).
`The mouthpiece of the nebulizer was replaced with a step-down
`adapter and a cotton plug was placed into the adapter to trap aero(cid:173)
`sol. Pilot testing using several cotton plugs showed that a single plug
`was 100% effective as a filter of aerosol from the nebulizer. The test
`chamber of the lung model (simulating a spontaneously breathing
`patient) was attached to the adapter containing tl1e cotton plug.
`Sa]jne solution diluent volumes of 2.5, 3.5, and 4.5 mL and 2.5
`mg (0.5 mL of0.5%) of albuterol (Proven til; Schereing; Kenilworth,
`NJ) were placed into tl1e nebulizer cup to produce fill volumes of
`3, 4, and 5 mL. Saline solution and albuterol were precisely mea(cid:173)
`sured using ca]jbrated syringes. Oxygen flows of 6, 8, and 10 Umin
`were used from a ca]jbrated flowmeter (Timeter; Lancaster, Pa)
`connected to the hospital bulk oxygen supply. Three new nebuliz(cid:173)
`ers of each type were evaluated at each combination of diluent
`volume and oxygen flow. All evaluations were conducted at ambi-
`
`ent room temperature (approximately 22°C) and humiruty (approx(cid:173)
`imately 40 to 50% relative humidity).
`ebu(cid:173)
`The nebulizer was tapped periorucally during each trial.
`lization time was determined b y a stopwatch and was considered
`complete w hen there was no visible or auruble evidence of nebu(cid:173)
`lization for a period of 30 s. The nebulizer was weighted empty
`( Ohaus 311 Cent-0-Gram Balance; Carolina Biological; Burlington,
`NC), after it was filled with merucation and diluent, and at tl1e end
`of the trial. The percentage of solution that was nebulized was cal(cid:173)
`culated from these mass values.
`At the end of each trial, the amount of drug remllining in the
`nebulizer cup was determined by washing the inside of the nebu(cid:173)
`lizer cup with 10 mL of sa]jne solution and spectrophotometrically
`detennining the amount of albuterol remllimng in the nebulizer
`cup. The drug present in the cotton plug was extracted u sing 20 mL
`of saline solution and gentle agitation by vortex. The resulting so(cid:173)
`lution was centrifuged at 5,000 g for 10 min to remove all cotton
`fibers from the solution and the amount of albuterol was then d e(cid:173)
`·with other studies using
`termined spectrophotometrically. As
`methods similar to ours, we assumed that all albuterol was extracted
`from the cotton. 12·13
`
`Particle Size Determination
`The ex1)erimental setup used to determine particle size is shown
`in Figure 1. During evaluation, the nebulizer was placed in a clamp
`and attached to a ring stand in the vertical position. Albuterol (0.5
`mL of 0.5%) was placed into the nebulizer cup and ruluted with 2.5
`mL of saline solution. Particle sizes were dete1mined at oxygen
`flows of 6, 8, and 10 Umin. Three new nebulizers of each type were
`evaluated at each oxygen flow.
`Aerosol particle size produced by the nebulizer was determined
`using an 11-stage cascade impactor (Intox; Albuquerque, NM) \vith
`cutoff stages of 12, 9.52, 7.56, 6, 5, 4, 3, 1.8, 1, 0.4, and 0.25 pm.
`Aerosol was sampled 5 em from the outlet of the nebulizer at a flow
`of 2 Umin to the impactor for 2 min. The albuterol deposited on
`each stage of the impactor was collected on plates, washed \vith
`saline solution, and the amount of albuterol was determined spec(cid:173)
`trophotomehically. The cascade impactor was calibrated by the
`manufacturer and used per manufacturer's specifications. Mass
`meruan aerodynamic ruameter (MMAD) and geometric standard
`deviation (GSD) were determined from the calibration curves
`provided b y the manufacturer. Cumulative deposition data were
`plotted agllinst stage cutoff ruameter, and fitted with a logarithmic
`regression curve to determine the particle size at 50% of the accu(cid:173)
`mulated d eposition (MMAD). This relationship wa~ ummodal for
`all nebulizers and R2 for this relationship is typically greater than 0.9
`in our laborato ry. GSD was calculated as the MMAD ruvided by the
`particle size at 16% deposition. In adrution to MMAD and GSD,
`the percentage of particles in the respirable range of 1 to 5 pm was
`determined. Although both larger and smaller particles may have
`clinical benefit, we defined respirable particles as 1 to 5 pm for
`purposes of describing nebulizer performance in the laboratory.
`The same particle size range (1 to 5 pm) has been used by others
`to desctibe nebulizer performance 4
`
`Respirable Mass
`We also calculated respirable mass to combine the effects of drug
`output and the percentage of particles in the respirable range. Drug
`output was calculated b y subtracting the amount of merucation in
`the dead volume from the amount of merucation placed into the
`nebulizer at the beginning of the trial (2.5 mg). Respirable mass
`output of the nebulizer wa~ then calculated b y multiplying the drug
`output times the percentage of particles in the respirable range of
`1 to 5 pm. Respirable mass available to the patient was calculated
`by multiplying the inhaled mass times the percentage of particles
`in the respirable range. Thus, respirable mass output of the nebu-
`
`500
`
`Laboratory and Animal Investigations
`
`IPR2021-00406
`United Therapeutics EX2031
`Page 3 of 8
`
`

`

`0.5 e
`~
`,Q 0.4
`"' C)
`E "-E 0.3
`5
`~ 'E 0.2
`.!!!
`n;
`<>.
`0.1
`
`0.0
`
`eo.e
`~
`.0 n; §:0.6
`c :::>
`~ 0.4
`"' c
`"' 8_o.2
`
`0.0
`
`_15.0
`c:
`§.
`
`i c 10.0
`.2 i
`
`~ 5.0
`
`3 mL 4 mL 5 mL
`6 L/min
`
`3 mL 4 mL 5 mL
`8 L/min
`
`3 mL 4 mL 5 mL
`10 L/min
`
`J K L M N 0 P a
`I
`A B C D E F G H
`nebulizer brand
`
`0 .0
`
`14
`
`12
`
`10
`
`8
`
`6
`
`4
`
`2
`
`0
`
`c
`:§.
`"' E
`=
`
`3 mL 4 mL 5 mL
`6 L/mln
`
`3 mL4 mL 5 mL
`8 L/min
`
`3mL4mL5mL
`10 L/mln
`
`J K L M N 0 P a
`I
`A B C 0 E F G H
`nebulizer brand
`
`FIGURE 3. Top: effect of volume (p<O.OOl) and flow (p=0.02) on
`amount of albuterol delivered to the patient. Data are pooled from
`all nebulizers for each flow and volume setting. Bottom: effect of
`nebulizer brand on amount of albuterol delivered to the patient
`(p<O.OOl). Data or each nebulizer brand are pooled from all volume
`and flow settings.
`
`lizer described the aerosol production of the device, whereas
`respirable mass available to the patient described aerosol available
`at the mouthpiece of the device. Because particle size was deter(cid:173)
`mined only for a single nebulizer volume, these calculations were
`conducted only for a nebulizer volume of 3 mL.
`
`Spectrophotometric Analysis of Albuterol
`
`A stock solution of albuterol (0.05 mglmL) was prepared from
`powdered drug (Sigma; St. Louis) and a standard curve was
`constructed from serial dilutions. An absorbance peak was found at
`278 nm and all absorbance measurements were made at this
`wavelength. Our spectrophotometric scan of the reference drug
`solution agreed well with published spectral properties of al(cid:173)
`buterol.14 The spectrophotometer absorbance was adjusted to zero
`with a saline solution solvent (or saline solution solvent treated with
`cotton as appropriate) before each measurement was made. The
`amount of drug in test solutions was determined from the standard
`curve.
`
`Statistical Analysis
`
`Summary statistics are reported and mean±SE. Differences
`between groups were determined by one-way and multifactorial
`analysis of variance as appropriate. Post hoc analysis was conducted
`using the Scheffe procedure . Statistical significance was set at
`p<0.05.
`
`RESULTS
`
`Dead Volume
`The amount of drug in the dead volume measured
`by serial weighing (gravimetric method) was signifi-
`
`FIGURE 4. Top: effect of volume (p<O.OOl) and flow (p<O.OOl) on
`nebulization time. Data are pooled from all nebulizers for each flow
`and volume setting. Bottom: effect of nebulizer brand on nebuliza(cid:173)
`tion time (p<O.OOl). Data for each nebulizer brand are pooled from
`all volume and flow settings.
`
`cantly less than that determined by measuring the
`amount of drug remaining in the dead volume (spec(cid:173)
`trophotometric method) (0.81±0.01 mg vs 1.14±0.01
`mg; p<0.001). The effects of flow, diluent volume, and
`nebulizer brand on dead volume are shown in Figure
`2. There was a small but significant (p=0.02) decrease
`in dead volume amount with an increase in flow from
`6 to 10 Umin. There was no significant difference
`(p>0.05) between 8 and 10 Umin and 6 and 8 Umin.
`With an increase in diluent volume, there was a
`significant decrease in dead volume amount (p<O.OOl);
`this effect was significant among all levels of diluent
`volume. There were also significant differences in dead
`volume among nebulizer brands (p<0.001).
`
`Aerosol Mass Available to the Patient (Inhaled
`Mass)
`The effects of flow, fill volume, and nebulizer brand
`on the amount of aerosol available at the mouthpiece
`are shown in Figure 3. There was a small but signifi(cid:173)
`cant (p=0.02) decrease in the amount of drug delivered
`to the mouthpiece (and thus available to the patient)
`with an increase in flow from 6 to 10 Umin. There was
`no significant difference (p>0.05) between 8 and 10
`Umin and 6 and 8 Umin. There was a significantly
`greater amount of drug delivered to the mouthpiece
`with an increase in fill volume (p<0.001) from 3 to 4
`
`CHEST /110/2/ AUGUST, 1996
`
`501
`
`IPR2021-00406
`United Therapeutics EX2031
`Page 4 of 8
`
`

`

`50%
`
`~40%
`:::;;
`~30%
`'E
`fl
`iii 20%
`Cl.
`
`10%
`
`0%
`
`60%
`
`0 so%
`< :::;;
`::E 40o/o
`u;>
`
`~ 30o/o
`0
`
`&20%
`
`10%
`
`0%
`
`6
`
`8
`nebulizer flow
`
`10
`
`I J K L M N 0
`H
`A B C D E F G
`nebulizer brand
`
`p a
`
`FIGURE 6. Top: effect of flow on percentage of particles in the re(cid:173)
`spirable range of l to 5 pm (p=0.004). Data are pooled from all
`nebulizers for each flow. Bottom: effect of nebulizer brand on per(cid:173)
`centage of particles in the respirable range of l to 5 pm (p<O.OOl).
`Data for each nebulizer brand are pooled from all flow settings.
`
`Scheffe post hoc analysis, there was a significant
`difference in both MMAD and particles in the respi(cid:173)
`rable range between flows of 6 and 8 Umin and
`between 6 and 10 Umin, but no difference between 8
`and 10 Umin.
`
`Respirable Mass
`Respirable mass output of the nebulizers and respi(cid:173)
`rable mass available to the patient is shown in Figure
`7. With an increase in flow, there was a small but in(cid:173)
`significant increase in respirable mass output (p=0.07)
`and respirable mass available to the patient (p=0.43).
`For nebulizer brands, there were significant differ(cid:173)
`ences in respirable mass output and respirable mass
`available to the patient (p<0.001 in each case). There
`was significantly more respirable aerosol mass output
`from each nebulizer than was available to the patient
`(p<0.001).
`
`DISCUSSION
`In this study, we have demonstrated that medication
`nebulizer function is affected by diluent volume, flow,
`and nebulizer brand. Increasing diluent volume de(cid:173)
`creased the amount of albuterol trapped in the dead
`volume and increased the amount delivered to the
`patient. Increasing diluent volume also increased
`
`6
`
`8
`nebulizer flow
`
`10
`
`6
`
`5
`
`w 4
`e 0
`§. 3
`0 < :::;;
`:::;;
`
`2
`
`0
`
`A B C 0
`
`J K L M N 0 P a
`I
`E F G H
`nebulizer brand
`
`FIGURE 5. Top: effect of flow on MMAD (p<O.OOl). Data are
`pooled from all nebulizers for each flow. Bottom: effect of nebulizer
`brand on MMAD (p<O.OOl). Data for each nebulizer brand are
`pooled from all flow settings.
`
`mL and from 3 to 5 mL. However, there were sig(cid:173)
`nificant differences in albuterol delivered to the mouth(cid:173)
`piece between 4 and 5 mL. There were also significant
`differences among nebulizer brands in the amount of
`drug delivered to the mouthpiece (p<O.OOl).
`
`Nebulization Time
`The effects of flow, fill volume, and nebulizer brand
`on nebulization time are shown in Figure 4. There was
`a significant increase in nebulization time \vith an in(cid:173)
`crease in volume or a decrease in nebulizer flow
`(p<0.001 in each case). These differences were signif(cid:173)
`icant between all levels of volume and nebulizer flow
`(p<0.05 by Scheffe analysis). There were also signifi(cid:173)
`cant differences in nebulization time among nebulizer
`brands (p<0.001).
`
`Particle Size
`The effects of flow and nebulizer brand on particle
`size are shown in Figure 5. There was a significant
`decrease in MMAD \vith an increase in nebulizer flow
`(p<0.001). There were also significant differences in
`MMAD among nebulizer brands (p<0.001). As shown
`in Figure 6, there was a small, but significant (p=0.004)
`increase in the mass of particles in the respirable range
`\vith an increase in nebulizer flow. There were signif(cid:173)
`icant differences among nebulizer brands in the mass
`of particles in the respirable range (p<0.001). By
`
`502
`
`Laboratory and Animal Investigations
`
`IPR2021-00406
`United Therapeutics EX2031
`Page 5 of 8
`
`

`

`nebulization time, the effect of which was offset by
`increased flow. Compared with the effect of diluent
`volume, flow had a lesser effect on nebulizer perfor(cid:173)
`mance. Increasing flow increased the mass of particles
`in the respirable range of l to 5 pm; the respirable mass
`delivered to the patient was affected to a greater extent
`by nebulizer brand than flow. The effect of nebulizer
`brand was relatively greater than the effect of diluent
`volume or flow, with large differences among devices
`of different manufacturers.
`A large fraction of the medication placed into the
`nebulizer cup remains trapped within the device and
`in not made available to the patient. Previous studies
`have evaluated this dead volume by serial weighing.3-5
`However, such gravimetric methods do not account for
`the effect of reconcentration within the nebulizer. We
`found that the amount of albuterol trapped in the
`nebulizer is nearly 50% greater than that estimated by
`gravimetric methods. These results agree with those of
`O'Callaghan et al,7 who also found an overestimation
`of nebulizer output by 50% when serial weighing was
`used rather than direct measurement of nebulizer
`output. This can be explained by the fact that many of
`the particles produced within the nebulizer are baffled
`and return to the nebulizer cup. During this process,
`evaporation occurs so that particles which return to the
`nebulizer solution increase the concentration in the
`dead volume. This effect is accentuated when a cold
`dry gas is used to power the nebulizer, 15 as was the case
`in this study.
`Increasing the diluent volume increased the nebu(cid:173)
`lizer output. Malone et al16 used high-performance
`liquid chromatography to measure the amount of
`albuterol remaining in the nebulizer cup with 3 fill
`volumes (1.5, 2.5, and 3.5 mL). Similar to our findings,
`they found that the amount of albuterol trapped in the
`dead volume increased with smaller fill volumes. It is
`difficult to make direct comparisons between our re(cid:173)
`sults and those of Malone et al. 16 They used a
`compressor (relative humidity approximately 50% and
`flow approximately 6 Umin) rather than dry oxygen to
`power the nebulizer, they used different fill volumes,
`and they used a nebulizer different than any of those
`in our study (DeVilbiss 646). However, they report
`nebulizer outputs at 2.5 mL and 3.5 mL fill volumes
`that are similar to what we found with fill volumes of
`3 mL and 4 mL, respectively. Malone et al16 also
`evaluated the concentration of albuterol in the nebu(cid:173)
`lizer over the nebulization time and found nearly a
`100% increase for smaller diluent volumes, but a
`smaller increase in concentration occurred for the
`larger diluent volume.
`Studies that used gravimetric methods to evaluate
`nebulizer performance have also reported increased
`nebulizer output with increasing fill volume,3·5·17J 8 as
`
`0.6
`
`o;o.s
`E..
`"' ~0.4
`E
`<1>
`~0.3
`
`·c. e 0.2
`
`0.1
`
`0
`
`1 .2
`1 . 1
`
`_0.9
`"" go.s
`iQ 0.7
`E 0.6
`<1>
`~0 .5
`·g-0.4
`~ 0.3
`0.2
`
`0.1
`
`0
`
`6 L/min
`
`nebulizer flow
`
`,I,
`
`II,
`
`II,
`
`f-r-,
`
`f-r-, F,
`
`1-r,
`
`F,
`
`J K L M N 0 P 0
`I
`A B C D E F G H
`nebulizer
`
`FIGURE 7. Top: effect of flow on respirable mass. Solid bars
`represent respirable output of the nebulizer (p=0.07) and open bars
`represent respirable mass available to the patient (p=0.43). Data are
`pooled from all nebulizers for each flow. Bottom: effect of nebulizer
`brand on the respirable mass output of the nebulizer (solid bars,
`p<O.OOl) and respirable mass available to the patient (open bars,
`p<O.OOl ). Data for each nebulizer brand are pooled from all flow
`settings.
`
`well as an increased nebulizer output with increased
`flow.3 In contrast to these findings, we found that flow
`had a relatively small effect on albuterol output from
`the nebulizer. Using three jet nebulizer brands and
`sodium fluoride as a tracer, Dennis et al17 reported that
`nebulizer output was relatively unaffected by flow un(cid:173)
`less very high flows were used. Increased gravimetric
`output with little change in drug output as flow is in(cid:173)
`creased suggests that evaporation and reconcentration
`are greater for higher flows .
`Although flow had little effect on nebulizer output,
`it did affect nebulization time and particle size. A
`higher flow resulted in a shorter nebulization time,
`which could be used clinically to offset the effect of a
`larger diluent volume on nebulization time. Increasing
`the flow decreased the size of the particles generated
`and increased the proportion of particles in the respi(cid:173)
`rable range of l to 5 pm. The effect of flow on parti(cid:173)
`cle size was greater between 6 and 8 Umin than be(cid:173)
`tween 8 and lO Umin.
`It is interesting to note that an increased flow did not
`increase the amount of drug available to the patient.
`When the effects of flow on both particle size and
`amount delivered to the patient are considered (ie ,
`
`CHEST I 110 I 2 I AUGUST, 1996
`
`503
`
`IPR2021-00406
`United Therapeutics EX2031
`Page 6 of 8
`
`

`

`respirable mass), there was little effect of changing
`flow. Increasing flow increases the number of particles
`in the respirable range but also increases the amount
`of waste during the expiratory phase so that respirable
`mass remains relatively constant. This is supported by
`previous studies. Although Clay et al9 reported a
`decrease in aerosol particle size when nebulizers were
`powered with a higher flow, others showed no differ(cid:173)
`ence in bronchodilator response when higher nebu(cid:173)
`lizer flows were used.19·20 The lack of change in clin(cid:173)
`ical response with changes in flow is likely due to our
`findings that respirable mass available to the patient is
`unchanged with changes in flow. It would thus appear
`that flow primarily affects nebulization time and thus
`higher flows are favored for convenience rather than
`improved drug delivery. These results also indicate that
`the amount of aerosol available to the patient should
`not be affected by use of low-flow devices like com(cid:173)
`pressors that are commonly used in the home.
`Differences in performance among nebulizers were
`greater than the differences observed for changes in
`diluent volume or flow. The respirable mass delivered
`by some nebulizers was twice that of others. From our
`results, it is apparent that some nebulizers have been
`engineered for a performance that is superior to oth(cid:173)
`ers. The reasons for this are not readily apparent by
`inspection of the devices and likely relates to a num(cid:173)
`ber of factors, the identification of which was beyond
`the scope of this study. Our in vitro data are consistent
`with the in vivo data of Johnson et al. 21 They compared
`deposition and physiologic response to albuterol in
`patients with stable asthma using a nebulizer that
`produced small particles (MMAD, 3.3 pm) and a
`nebulizer that produced large particles (MMAD, 7.7
`pm). Deposition and physiologic response (ie, FEV1)
`were greater with the nebulizer that produced smaller
`particles. Presumably, the respirable mass was greater
`with the nebulizer that produced smaller particles.
`Nebulizers can be powered only during the inspira(cid:173)
`tory phase to minimize drug waste.22·23 A system has
`also been described that uses a collection bag to trap
`aerosol particles during exhalation and delivers them to
`the patient on the next inspiration.24 These systems are
`not in common use with spontaneously breathing pa(cid:173)
`tients owing to their additional cost and increased
`complexity. Another approach to decrease waste is to
`boost the nebulizer output during inspiration, thus
`decreasing the relative amount of drug lost during the
`expiratory phase, and N ewnham and Lipworth25 found
`that a nebulizer using this feature produced a twofold
`increase in the delivery of salbutamol to the lungs as
`compared with a conventional nebulizer.
`Although a standard dose of 2.5 mg was placed into
`the nebulizer in this study, less than 0.5 mg of respi(cid:173)
`rable mass was delivered to the patient. A number of
`
`the nebulizers that we evaluated delivered only a
`respirable mass of approximately 0.2 mg. It is of inter(cid:173)
`est to note that this is similar to the standard albuterol
`dose delivered by metered-dose inhaler (90 pg per
`actuation; 2 actuations=0.18 mg). This may explain why
`a number of studies have shown that clinical response
`using a nebulizer or metered-dose inhaler are virtually
`equivalent. 1 When the metered-dose inhaler is used,
`patient performance is a major determinant of aerosol
`delivery. With nebulizer use, performance of the neb(cid:173)
`ulizer is a major determinant of aerosol delivery.
`Alvine et al26 found considerable variability in neb(cid:173)
`ulizer function, not only among nebulizer brands, but
`also among nebulizers within specific brands. This is
`not particularly surprising, because nebulizers are
`produced in bulk and are typically short-term single(cid:173)
`patient use devices. Our data do not allow confirmation
`of the results of Alvine et