Hydrofluoroalkane Pressurized

Transcription

1 Salbutamol Delivery From a Hydrofluoroalkane Pressurized Metered-Dose Inhaler in Pediatric Ventilator Circuits* An In Vitro Study Johannes H. Wildhaber, MD; Mark J. Hayden, MD; Nigel D. Dore, MRRS; Sunalene G. Devadason, PhD; and Peter N. LeSouef, MD Study objectives: The aim ofour study was to determine the in vitro delivery of salbutamol from a pressurized metered-dose inhaler (pmdi) containing hydrofluoroalkane (HFA) propellant through various delivery devices to four models of a pediatric lung. Design: To determine the effect of electrostatic charge, delivery of salbutamol was initially assessed with a multistage liquid impinger (MSLI) through an inline nonchamber device (Baxter MDI Adapter) and a small (Aerochamber MV) and a large (Nebuhaler) inline chamber device. Following this, the delivery was assessed to four lung models appropriate for a child of 70 kg, 50 kg, 15 kg, and 4 kg, with the same three reduced static devices inserted directly into a pediatric ventilator circuit. Measurements and results: Reduction of electrostatic charge improved small particle delivery through holding chambers to the MSLI by 12 to 14%. In the ventilator model, the mean delivery was between 1.9% and 5.4% for the nonchamber device, between 14.3% and 27.2% for the small holding chamber, and between 7.2% and 25.7% for the large holding chamber. Deliverywas the least efficient in the 4-kg model compared to the 70-kg, 50-kg, and 15-kg models. Conclusions: Salbutamol from an HFA pmdi is delivered efficiently through inline holding chambers with reduced static in pediatric ventilator settings. A large holding chamber has no advantage over a small holding chamber. In addition, salbutamol delivery is more efficient through a holding chamber than through a nonchamber device. (CHEST 1998; 113:186-91) Key words: aerosol therapy; electrostatic charge; inhalation devices; mechanical ventilation Abbreviations: CFC = chlorofluorocarbon; HFA=hydrofluoroalkane; MSLI=multistage liquid impinger; PIP=peak inspiratory pressure; pmdi=pressurized metered-dose inhaler; SVN=small-volume nebulizer; =tracheal tube; Vt.tidal volume "D ronchodilators for the therapy of reversible air- *-* way obstruction are often delivered via inhala tion to the lower respiratory tract in both spontane ously breathing and mechanically ventilated patients. The devices usually used for delivery of aerosols to ventilated patients are small-volume nebulizers (SVNs) and pressurized metered-dose inhalers *From the Perth Medical Aerosol Research Group, Departments of Respiratory Medicine (Drs. Wildhaber, Dore, Devadason, and LeSouef) and Intensive Care (Dr. Hayden), Princess Margeret Hospital for Children, Subiaco, Western, Australia. Manuscript received January 6, 1997; revision accepted June 25, Reprint requests: Johannes Wildhaber, MD, Department of Re spiratory Medicine, Princess Margeret Hospital for Children, Roberts Road, Subiaco 6008, WA, Australia; hannesw@ cyllene.uwa.edu.au (pmdis). There is still a controversy about which may be the optimal system for inhalation therapy ventilated patients in in ICUs from the point ofview of both economic and therapeutic effectiveness.1 Several in vitro and in vivo studies have shown that pmdis are more efficient in aerosol delivery than SVNs.2-5 In addition, itwas also shown that both SVNs and pmdis can effectively deliver aerosols in a ventilator model and that aerosol delivery can be significantly improved when a proper technique of administration is followed.68 To enhance lung deposition of aerosols from pmdis, several accessory devices have been devel oped. The accessory devices most commonly used to deliver aerosols from pmdis into ventilator circuits 186 Laboratory and Animal Investigations

2 are inline nonchamber devices and inline holding chambers. The advantage of a holding chamber inserted in the ventilator circuit is that the actuated aerosol cloud is retained within the chamber and hence impaction of the drug within the ventilator circuit is reduced. Holding chamber devices have been shown to be superior vices.9'10 to nonchamber de Until recently, pmdis contained chlorofluorocarbon (CFC) propellant, which is damaging to the ozone layer.911 These are now being replaced by pmdis containing hydrofluoroalkane (HFA) propel lant.12 As the aerosol and delivery characteristics of HFA pmdis may be different from CFC pmdis, it is important to test them in in vitro and in vivo studies.13 The aim of our study was to compare the in vitro delivery of salbutamol from an HFA pmdi through an inline nonchamber device and a small and a inline large holding chamber inserted in a pediatric ven tilator circuit under various ventilator settings. Materials and Methods In vitro salbutamol delivery was measured through the follow ing devices using a high-performance multistage liquid impinger (MSLI) (Copley; Nottingham, UK): an inline nonchamber device (Rigid 22mm pmdi Adapter; Baxter; Irvine, Calif); a small inline holding chamber (Aerochamber MV; volume 165 ml; Trudell; Ontario, Canada); and a large holding chamber (Nebuhaler; 750 ml; Astra; Lund, Sweden) modified to function as an inline device for a ventilator circuit. Modification of the large chamber was performed by sealing off air leaks, removing the valve, and inserting the inline nonchamber device in the actuator orifice. The salbutamol HFA pmdi (20 canisters, Lot 96E01K of Airomir; 3M Pharmaceuticals; St. Paul, Minn) was actuated 10 times into the study devices and shaken vigorously between actuations. Air was drawn through the study devices that were attached to the MSLI at a continuous flow of 60 L/min. Aerosol droplets were deposited on the device, the glass throat of the MSLI, or one of four stages. The MSLI was calibrated by the manufacturer so that particles >13, 6.8 to 13, 3.1 to 6.8, and <3.1 ixm were deposited on stage 1, 2, 3, and 4, respectively. To determine the influence of electrostatic charge on drug delivery, salbutamol delivery was measured through new and detergent-coated devices. Electrostatic charge on the surface of the devices was measured prior to the in vitro test using modified electrometer and a was classified as none (negligible charge, 0 to 1.2 ixc/m2), low charge (1.2 to 3.3 fic/m2), and high charge (3.3 to 6.7 ixc/m2). All devices were then detergent coated to reduce electrostatic charge. To coat the devices with deter gent, they were soaked in diluted (1:250) cationic detergent (Cetrimide 40%; Princess Margaret Hospital Pharmacy; Perth, Australia) for 1 h and drip dried for <24 h.14 In addition, in vitro salbutamol delivery was measured through the three devices, an inline nonchamber device (Rigid 22-mm pmdi Adapter; Baxter), a small inline holding chamber (Aero chamber MV; Trudell), and a large holding chamber (Nebuhaler; Astra) modified as an inline device for a ventilator circuit inserted in a dry unheated pediatric ventilator circuit (Servo Ventilator 900C; Siemens; Elema, Sweden). Four different models were constructed to mimic appropriate ventilator settings for children weighing (a) 70 kg, (b) 50 kg, (c) 15 kg, and (d) 4 kg. All models used the same ventilator (Servo) and circuit, but tracheal tube () size and the volume/compliance of the test lung were altered to mimic that of children of different sizes (test lung for infants: small volume and low compliance; test lung for older children: large volume and high compliance). Additionally the compliance of the test lung in model b was altered by constricting "band" a to mimic normal and "stiff lungs. In keeping with standard clinical practice, the ventilator was set on volume control mode for models a, b, and c, and on pressure control mode for model d. Thus models a, b, and c, had a set tidal volume (Vt) and resultant peak inspiratory pressure (PIP), whereas model d had a set PIP and resultant Vt. Positive end-expiratory pressure was set at 4cm H20 in all models as was inspiratory time (25%) and pause time (10%). On this type of ventilator, inspira tory flow is not set and is dependent on PIP, inspiratory time, pause time, and lung model compliance. The models used were as follows: (a) (Portex; Kent, England) internal diameter 9 mm,vt 750 ml, rate 10 breaths/min, PIP 20cm H20; (b) 7.5 mm, Vt 450 ml, breaths/min 10, PIP 20 cm H20 and 40 cm H20; (c) 5 mm, Vt 165 ml, breaths/min 18, PIP 20 cm H20; and (d) 4 mm, Vt 40 ml, breaths/min 30, PIP 20 cm H20, respectively (Fig 1). In addition, in vitro salbutamol delivery through all three devices was measured in a humidified and heated (37 C) circuit with ventilator settings for a 70-kg child (model a). All devices were detergent coated to reduce electro static charge. Theywere inserted into the ventilator circuit at the opening of the. Salbutamol was collected on an inspiratory filter (Anesthesia Filter, Curity; Kendall; Mansfield, Mass) in serted at the tip ofthe. An identical filter was inserted in the expiratory line. The HFApMDI was actuated 10 times just prior to initiation of inspiration and shaken vigorously inbetween actuations. Between actuations, a sufficient number of inspiratory and expiratory cycles were allowed, to totally clear the dead space of the delivery device and the ventilator circuit. Each device, throat,, inspiratory and expiratory filters, and each ofthe four stages ofthe MSLI were washed separately with 45 ml of methanol immediately after the tests. Five milliliters of 0.1 mmol/l NaOH was added to each wash and the volume was made up to 50mL with methanol. The absorbance (wave length 246 nm) of each sample was measured in duplicate spectrophotometer (Hitachi U-2000; on a San Jose, Calif). The con centration of salbutamol in each sample was obtained by using the absorbance of a standard solution containing a known concentration ofsalbutamol. The total actuated dose (10 puffs) the MSLI in was determined by adding the amount of drug in the device, the glass throat, and the four stages. The total actuated dose (10 puffs) in the ventilator circuit was determined by adding the amount of drug in the device, the filters, and the. The standard curve for salbutamol was linear (^=1.00) for concen trations between 0 and 2,100 [ig/ml. Each experiment from actuation ofthepmdi to the measurements ofthe drug concen tration was repeated four times. Validation of the method, to ensure that no other contents of the pmdi interfere with the spectrophotometric measurement of salbutamol at the same wave length, was done, using a placebo pmdi containing only the propellant HFA but no salbutamol (Airomir placebo, 3M Phar maceuticals). All measurements were undertaken under the following atmospheric conditions: mean temperature was 21.7 C (range, 21 to 24 C) and mean barometric pressure was 760mm Hg (758 to 766 mm Hg). Statistical analysis was carried out using analysis of variance (StatView 512 +; Abacus Concepts Inc; Berkeley, Calif) with a significance level of 95% (p<0.05). CHEST/113/1 /JANUARY,

3 pmdi/inline non-chamber device test lung ^ Y-ventilator connector / expiratory filter Y-ventilator connector/ expiratory filter Y-ventilator connector/ expiratory filter Figure 1. Illustration ofthe three devices, an inline nonchamber device (top, a), a small inline holding chamber (center, b), and a large holding chamber modified as an inline device for a venti lator circuit (bottom, c) inserted in a pediatric ventilator circuit under different ventilator settings. Results Measurements of electrostatic charge and drug delivery using the MSLI are shown in Table 1. The electrostatic charge on the new devices ranged from low to high. Reducing electrostatic charge by coating the devices with cationic detergent significantly in creased the delivery of small particles (<3.1 jmm) from the inline holding chambers (p<0.001). Elec trostatic charge did not affect delivery dirough the nonchamber device. Delivery of small particles (<3.1 xm) is increased by 12% from the small holding chamber (Aerochamber MV) and by 14% from the large holding chamber (Nebuhaler). Delivery of small particles (<3.1 jmm) was significantiy higher from the small holding chamber (Aerochamber MV) device compared to the inline nonchamber device (Raxter MDI Adapter) (p<0.001). However, delivery of small particles (<3.1 [xm) was less from the small holding chamber (Aerochamber MV) compared to the large holding chamber (Nebuhaler) (p<0.001). Drug deposition for the four lung models using the pediatric ventilator circuit is shown in Table 2. Drug deposition on the inspiratory filter as a proportion of drug delivery to the patient was between 1.9% and 5.4% through the inline nonchamber device (Raxter MDI Adapter) and was significantly less than through the two inline holding chambers (Aerochamber MV and Nebuhaler) (p<0.001). Drug deposition on the inspiratory filter was between 14.3% and 26.0% through the small holding chamber (Aerochamber MV) andwas significandyhigher thanthrough the large holding chamber (Nebuhaler) for ventilator settings appropriate for a child of50 kg (b), 15 kg (c), and 4 kg (d) (p<0.001). However, there was no significant dif ference for a child of 70 kg (a). The amount of salbutamol through the small holding chamber (Aero chamber MV) deposited on the inspiratory filter was significantly higher (p<0.001) in ventilator settings appropriate for a child of 70 kg (a) and 50 kg (b) than in ventilator settings for a child of15 kg (c) and 4 kg (d), whereas drug deposition on the expiratoiy filter was significantly lower (p<0.001). Drug deposition on the inspiratory filter was between 7.2% and 25.6% for the large inline holding chamber (Nebuhaler) and de creased with lower Vts from the ventilator setting for a child of 70 kg (a) to the ventilator setting for a child of 4 kg (d) (Fig 2).A change in the compliance ofthe lung model while maintaining the same ventilator setting appropriate for a child of50 kg (b) did not alter drug delivery through the small holding chamber (Aero chamber MV). However, humidification of die ventila tor circuit for a child of 70 kg resulted in a significant (p<0.001) decrease of salbutamol deposition on the inspiratory filter by 24% for the large and 18% for the small inline holding chambers (Table 2). The mean (range) of total dose (10 puffs) of salbutamol from HFA pmdis was 1,040 p,g (890 to 1,180 n.g). Discussion Inhalation therapy with bronchodilators to me chanically ventilated patients is usually performed 188 Laboratory and Animal Investigations

4 New static Spacer Throat Stage Stage Stage 4 Reduced static Spacer Throat Stage Stage Stage 4 Table 1.Measurements ofelectrostatic Charge and Drug Delivery* Nonchamber (Baxter pmdi Adapter) Low 4.2% (2.8) 64.5% (2.3) 0.2% (0.2) 7.8% (0.5) 23.3% (1.9) None 5.4% (1.4) 63.0% (1.6) 0.2% (0.2) 7.3% (0.4) 24.2% (1.2) Small Chamber (Aerochamber MV) Low 42.0% (0.3) 1.2% (0.1) 2.7% (0.3) 14.2% (0.3) 39.8% (0.5) None 34.5% (1.7) 1.2% (0.2) 4.0% (1.8) 15.8% (0.7) 44.5% (0.8) Large Chamber (Nebuhaler) Low-high 38.3% (0.9) 0% 0.8% (0.1) 17.5% (0.4) 43.3% (1.3) None 20.2% (0.5) 2.0% (0.7) 5.9% (0.9) 22.5% (1.9) 49.5% (2.5) *Mean (SD) in vitro delivery of salbutamol from HFA pmdi through different inline devices under static and reduced static conditions measured using an MSLI with cutoffs of particles >13, 6.8 to 13, 3.1 to 6.8, and <3.1 xm for stage 1, 2, 3,and 4, respectively, expressed as percentage of the total actuated dose. using SVNs or pmdis containing CFC propellant. Recent studies have shown that CFC pmdis are an effective alternative to SVNs for aerosol therapy the mechanically ventilated patient.24 However, be to cause of its harmful effect on the ozone layer, the use of CFC pmdis will be discontinued. Thus, CFCfree products, such as HFA pmdis, have been introduced for inhalation therapy. The performance of salbutamol HFA pmdis in drug delivery is dif ferent fromcfc pmdis in some situations.13 How ever, another study has shown that salbutamol HFA pmdis did not perform differently from salbutamol CFC pmdis in drug delivery through spacer de vices.15we have shown in our study that salbutamol in an in from an HFApMDI is delivered efficiently vitro model of a pediatric ventilator circuit. Our results with a maximal deposition of 26.0% of salbu tamol from an HFA pmdi on the inspiratory filter are comparable to another study measuring albuterol delivery from a CFC pmdi in an adult ventilator circuit model with a maximal deposition of 25.1%.6 The technique of administration and the devices Ventilator settings (a: 70 kg) Inspiratory filter (b: 50 kg) Inspiratory filter (c: 15 kg) Inspiratory filter (d: 4 kg) Inspiratoiy filter Table 2.Drug Deposition for the Four Lung Models Using the Pediatric Ventilator Circuit* Nonchamber (Baxter pmdi Adapter) Dry Humidified 29.2% (1.5)/30.5% (2.0) 65.0% (3.0)/63.0% (3.4) 0.5% (0.1)/0.5% (0.3) 5.4% (2.1)/6.3% (1.6) 27.8% (3.0) 66.1% (3.4) 1.4% (1.4) 4.6% (0.8) 26.0% (3.1) 67.8% (3.2) 2.5% (0.5) 3.7% (0.5) 31.1% (3.0) 62.5% (1.2) 4.4% (1.1) 1.9% (0.7) Small Chamber (Aerochamber MV) Dry Humidified 12.4% (1.2)/16.5% (2.2) 58.9% (3.0)/61.8% (2.0) 2.9% (0.6)/0.8% (0.1) 25.9% (1.5)/21.0% (1.5) PIP, 20cm H2O/40 cm H % (2.1)/10.4% (1.7) 60.2% (2.9)/60.3% (4.2) 3.5% (0.7)/3.7% (0.9) 26.0% (2.5)/25.6% (4.6) 13.4% (2.6) 59.3% (3.5) 7.5% (1.4) 19.8% (0.6) 36.9% (1.5) 25.8% (2.7) 23.0% (3.6) 14.3% (1.3) Large Chamber (Nebuhaler) Dry Humidified 6.9% (1.0)/12.8% (1.2) 59.9% (1.9)/60.7% (2.2) 7.5% (0.7)/7.1% (1.2) 25.6% (1.1)/19.6% (1.6) 7.4% (0.9) 61.1% (1.2) 11.8% (2.1) 19.7% (2.1) 22.1% (3.5) 36.0% (4.7) 25.7% (0.6) 16.2% (0.6) 59.8% (1.7) 25.7% (3.2) 7.2% (2.1) 7.2% (2.1) *Mean (SD) in vitro delivery of salbutamol from HFA pmdi through an inline nonchamber device and two ventilator circuit under different ventilator settings appropriate for a child of 70 kg (a), 50 kg (b), 15 kg inline holding chambers in pediatric (c), and 4 kg (d) with and without humidification, expressed as percentage of the total actuated dose. CHEST/113/1 /JANUARY, 1S 189

5 Baxter pmdi Adapter?Aerochamber WI.V. HNebuhaler Figure 2. Mean (SD) drug deposition on the inspiratory filter of salbutamol from HFA pmdi through an inline nonchamber device and two inline holding chambers in pediatric ventilator circuit under different ventilator settings appropriate for a child of 70 kg, 50 kg, 15 kg, and 4 kg, respectively, expressed percentage of the total actuated dose. as used for administration ofcfc pmdis play a major role in the efficacy of aerosol delivery.6 We have shown in the present study the importance of the choice of devicefor HFA pmdi administration. We showed in a previous study that electrostatic charge on the surface of a plastic spacer device is the most important factor influencing salbutamol delivery from a CFC pmdi.16 We also found in that study that the shape and the volume of the spacer device play a major role in delivery of salbutamol CFC pmdi.16 In the present study, we have shown that the delivery of salbutamol from an HFA pmdi through a plastic holding chamber is also dependent on electrostatic surface charge, and hence, is in creased when the holding chambers are coated with ionic detergent to reduce static. In the MSLI model, salbutamol delivery from the HFA pmdi was higher through the large-volume holding chamber than the small. One might there fore expect that a large holding chamber may be superior to a small holding chamber in delivery of salbutamol from an HFA pmdi in a pediatric ventilator circuit. However, we have shown that a large chamber has no advantage over a small cham ber in this setting. This might be explained by the low Vts used in pediatric ventilator settings and the clearance of a holding chamber, inserted between the Y-ventilator connector and the in a ventilator circuit, in both the inspiratory and expiratory phase. Thus, a large inline chamber might not be totally cleared in inspiration but will be cleared in both inspiration and expiration, hence, wasting drug dur ing the expiratory phase. We have shown that inline holding chambers are only efficiently cleared in inspiration when the Vt exceeds the volume of the inline holding chamber. This problem may be over come by inserting the spacer in the inspiratory limb of the ventilator circuit. We have also shown that a small chamber is more efficient in pediatric ventila tor settings where lowvts are used. A change in the compliance of the lung model while maintaining the same ventilator settings did not alter drug delivery. In addition, we found that salbutamol delivery is less efficient through a nonchamber device than through a holding chamber. A change in the compliance ofthe lung model did not alter salbutamol delivery. However, salbutamol delivery through the holding chambers was reduced by 18% through the small chamber and by 24% through the large chamber in a heated and humidi fied ventilator circuit compared to a dry ventilator circuit. Humidification may lead to hygroscopic growth of the aerosols and increased impaction of aerosols on the chamber surface, and hence, de creased delivery of salbutamol to the inspiratory filter.17 Our in vitro study with measurements of drug deposited on a filter gives only an idea about the dose most likely reaching the lungs in a clinical setting. The distribution of the aerosol and the clinical response remain unknown. However, the importance of the in vitro assessment of different devices, under different conditions, proceeding in vivo studies, is shown by the results of three previous studies. One study, using a suboptimal delivery device (pmdi adapter), did not show a clinical response of bronchodilators administered by CFC pmdi,18 whereas the other two studies, using a delivery device (inline holding chamber), shown in vitro to be optimal, found a significant clinical response.1920 In conclusion, we have shown that salbutamol from an HFApMDI is efficiently delivered through inline holding chambers with reduced static in pedi atric ventilator circuits and that a large holding chamber has no advantage over a small holding chamber. However, delivery through device a nonchamber was distinctly inefficient. References 1 Newhouse MT, Fuller HD. Rose is a rose is a rose?am Rev Respir Dis 1993; 148: Fuller HD, Dolovich MB, Chambers C, et al. Aerosol delivery during mechanical ventilation: a predictive in-vitro lung model. J Aerosol Med 1992; 5: Gay PC, Patel HG, Nelson SB, et al. Metered dose inhalers for bronchodilator delivery in intubated, mechanically venti lated patients. Chest 1991; 99: Fuller HD, Dolovich MB, Posmituck G, et al. Pressurized aerosol versus jet aerosol delivery to mechanically ventilated patients: comparison ofdose to the lungs. Am Rev Respir Dis 1990; 141: Arnon S, Grigg J, Nikander K, et al. Delivery of micronized 190 Laboratory and Animal Investigations

6 budesonide suspension by metered dose inhaler and jet nebulizer into a neonatal ventilator circuit. Pediatr Pulmonol 1992; 13: Diot P, Morra L, Smaldone GC. Albuterol delivery in a model of mechanical ventilation: comparison of metered-dose in haler and nebulizer efficiency. Am J Respir Crit Care Med 1995; 152: Everard ML, Stammers J, Hardy JG, et al. New aerosol delivery system for neonatal ventilator circuits. Arch Dis Child 1992; 67: Fink JB, Dhand R, Duarte AG, et al. Aerosol delivery from a metered-dose inhaler during mechanical ventilation: an in vitro model. Am J Respir Crit Care Med 1996; 154: Newman SP. Metered dose pressurised aerosols and the ozone layer. Eur Respir J 1990; 3: Rau JL, Harwood RJ, Groff JL. Evaluation of a reservoir device for metered-dose bronchodilator delivery to intubated adults: an in vitro study. Chest 1992; 102: Manzer LE. The CFC-ozone issue: progress on the develop ment of alternatives to CFCs. Science 1990; 249: Kleerup EC, Tashkin DP, Cline AC, et al. Cumulative dose-response study of non-cfc propellant HFA 134a sal butamol sulfate metered-dose inhaler in patients with asthma. Chest 1996; 109: June D, Carlson S, Ross D. Reduced effect oftemperature drug delivery characteristics of CFC-free metered dose on in halers (MDIs) compared to current CFC metered dose inhalers. Eur Respir J 1996; 9(suppl 23):255s 14 Wildhaber JH, Devadason SG, Hayden MJ, et al. Electro static charge on a plastic spacer device influences the delivery of salbutamol. Eur Respir J 1996; 9: Schultz D, Carlson S, Ross D. In vitro performance charac teristics of two CFC-free metered dose inhalers (MDIs) with large and small volume spacers. Eur Respir J 1996; 9(suppl 23):255s 16 Wildhaber JH, Devadason SG, Eber E, et al. Effect of electrostatic charge, flow, delay and multiple actuations on the in vitro delivery of salbutamol from different small volume spacers for infants. Thorax 1996; 51: Scherrer PW, Haselton FR, Hanna LM, et al. Growth of hygroscopic aerosols in a model of bronchial airways. J Appl Physiol 1979; 47: Manthous CA, Hall JB, Schmidt GA, et al. Metered-dose inhaler versus nebulized albuterol in mechanically ventilated patients. Am Rev Respir Dis 1993; 148: Dhand R, Jubran A, Tobin MJ. Bronchodilator delivery by metered-dose inhaler in ventilator-supported patients. Am J Respir Crit Care Med 1995; 151: Dhand R, Duarte AG, Jubran A, et al. Dose-response bronchodilator delivered by metered-dose inhaler to in ventila tor-supported patients. Am J Respir Crit Care Med 1996; 154: CHEST/ 113/1 /JANUARY,