The Effect of Independent Humidity Control on Aerosol Drug Delivery During Adult Invasive Mechanical Ventilation

Abstract

Background:

Aerosol therapy is commonly performed during invasive mechanical ventilation. Although heated humidification is standard practice, studies suggest that humidity can reduce the delivered dose. In this study, we aimed to investigate whether the delivered dose can be increased by decoupling humidity control from heating during aerosol drug delivery.

Methods:

In a bench study setup, albuterol sulfate solution was nebulized into an adult invasive mechanical ventilation circuit with a vibrating-mesh nebulizer. An absolute humidifier with decoupled heating and humidification was tested at 0 and 25 mg H₂O/L of added humidity during nebulization and compared with 2 conventional pass-over humidifiers with no independent humidity control. The delivered dose, experimentally determined by the drug deposition on a filter between the endotracheal tube and the test lung, was quantified with a UV–Vis spectrometer. The particle size distribution of the aerosol entering the test lung was measured and used to model the lung deposited dose for healthy and diseased lungs.

Results:

The absolute humidifier at 0 mg/L added humidity led to a significantly higher delivered dose (37.2%) than the pass-over humidifiers (21.2% and 24.1%, P = .01 for both), whereas the absolute humidifier at 25 mg H₂O/L added humidity led to a comparable dose (24.9%) as one of the pass-over humidifiers. All these test scenarios achieved sufficient humidity (>33 mg H₂O/L) in the gas flow during nebulization. Regional lung deposition simulations suggest that <10% of the nebulized dose deposits in healthy lungs for the pass-over humidifiers and around 16% for the absolute humidifier at 0 mg/L added humidity. Simulations also suggest that bronchial obstruction increases the deposited dose, whereas alveolar enlargement decreases the deposited dose.

Conclusions:

The delivered dose and hence the simulated lung deposited dose was improved by allowing independent humidity control during nebulization for adult invasive mechanical ventilation.

Introduction

In the ICU, patients with severe respiratory symptoms are ventilated through invasive mechanical ventilation. The invasive mechanical ventilation breathing circuit is heated and humidified to ensure patient comfort and to prevent damage to the lungs by dry air. ¹ One method to heat and humidify the circuit is to use a heat and moisture exchanger (HME), which uses the patient’s exhaled heat and humidity. However, HMEs do not always provide adequate humidity for the patient. ² Another common method is to use a pass-over humidifier. ³ For these humidifiers, the gas passes over a heated water bath, which is designed to reach and maintain 100% relative humidity (RH) in the gas flow.

Aerosol therapy is often performed during invasive mechanical ventilation, where aerosolized drugs are delivered to the lower respiratory tract via the endotracheal tube (ETT). The most common respiratory drugs include bronchodilators, steroids, and antibiotics. ⁴ Studies have shown that only around 20% of the nebulized drug typically reaches the patient during adult invasive mechanical ventilation, and this delivered dose can sometimes be considerably lower, as it depends on factors including the ventilation pattern, the nebulizer type, and the position of the nebulizer.^5–7 In addition, findings suggest that dry heated circuits result in a higher delivered dose than circuits heated and humidified with pass-over humidifiers.^6,8,9 However, dry gas risks damage to the lungs, and conventional pass-over humidifiers do not control humidity independently. Furthermore, although many clinicians reported turning off the pass-over humidifiers during nebulization in an attempt to improve nebulization efficiency,^4,10 evidence suggests that doing so does not increase the delivered dose during invasive mechanical ventilation ¹¹ and it is not recommended in clinical practice. ³

Recently, it has been shown that independent humidity control during nebulization can be achieved with a new type of humidifier that utilizes a humidification stage decoupled from the heating stage ¹² and thus enables the absolute humidity in the gas flow to be controlled independently from the patient-end temperature. The humidification results from controlled spraying and vaporization of small monodisperse water droplets. Evidence suggests that, for a high-flow nasal cannula (HFNC) setup, by turning off active humidification during nebulization and allowing the drug aerosols to partially evaporate and humidify the gas flow, this type of absolute humidifier can reduce undesired deposition in the breathing circuit, which leads to a substantially higher delivered dose compared with pass-over humidifiers. ¹² It remains unknown whether the absolute humidifier can also enhance the drug delivery efficiency during invasive mechanical ventilation, which generally results in a lower delivered dose than HFNC. We hypothesize that the absolute humidifier can increase the delivered dose when compared with pass-over humidifiers during adult invasive mechanical ventilation.

Therefore, this in vitro study aims to test the hypothesis by quantifying the delivered dose of the absolute humidifier setup at different humidification levels and comparing these results with conventional pass-over humidifier setups. For each setup, the particle size distribution of the aerosol exiting the ETT is also measured, and the lung deposited dose is calculated using software simulations.

QUICK LOOK

Current knowledge

Aerosol therapy during adult invasive mechanical ventilation typically has an efficiency of <25%. One key parameter limiting the delivered dose is the heated humidity. Conventional pass-over humidifiers do not control humidity independently from heating during nebulization, which results in high drug loss in the breathing circuit.

What this paper contributes to our knowledge

This bench study demonstrated that the delivered dose during simulated adult invasive mechanical ventilation could be increased while maintaining sufficient humidity. This was achieved by using an absolute humidifier that allows on-demand real-time humidity modulation, which lowers active humidification during nebulization and allows the aerosol droplets to partially evaporate and humidify the gas flow.

Methods

A ventilator (Dräger Evita XL, Lübeck, Germany) in pressure-regulated volume control mode was used to provide a ventilation pattern with a tidal volume of 500 mL, a PEEP of 5 cm H₂O, a respiratory rate of 12 breaths/min, an inspiratory to expiratory time ratio of 1:2, a peak respiratory flow of 50 L/min, and no measured bias flow. An ETT with a diameter of 8.5 mm was used and connected to a test lung (Elite Test Lung Adult RP20 [FMVC3001], Foxxmed, Taiwan) with a resistance of 20 cm H₂O/L/s, a compliance of 20 mL/cm H₂O and a maximum volume of 600 mL. A minute volume of around 6 L/min was reached. The ETT and test lung were located in a climate chamber set to 37°C to simulate conditions in the lungs. The rest of the circuit components were located outside at room temperature (around 22°C).

Figure 1 shows the ventilation circuit setup with pass-over and absolute humidifiers. Both types of humidifiers were set to the invasive mechanical ventilation setting, aimed to deliver 44 mg/L of water to the test lung. Two conventional pass-over humidifiers, F&P MR850 and F&P 950 (Fisher & Paykel, Auckland, New Zealand), were tested. The former is referred to as PassOverH1 and the latter as PassOverH2. For both pass-over humidifiers, the gas passes over a heated water bath, allowing it to entrain water vapor. To evaluate the effect of humidification in the conventional humidifiers, 2 scenarios were tested: leaving it on during nebulization and turning it off during nebulization. It is not possible to adjust the humidity level while maintaining the same temperature for the pass-over humidifiers. A smooth bore inspiratory tube with a 19 mm inner diameter (Hybernite RT 19, Plastiflex, Beringen, Belgium) was used for PassOverH1, and the inspiratory tube from the Adult Ventilator Dual Heated Circuit (950A81, Fisher & Paykel) with a 17 mm inner diameter was used for the PassOverH2.

Fig. 1.

Invasive mechanical ventilation setups during delivered dose measurements. The nebulizer was placed at position 1 when the pass-over humidifiers were used and at position 2 when the absolute humidifier was used. Two collection filters for measuring the delivered dose and bypass dose are labeled. A part of the setup was located inside the 37°C chamber as indicated.

To demonstrate the effect of independent humidity control, an absolute humidifier described in our previous study ¹² as the decoupled humidifier was used in place of the pass-over humidifiers. The absolute humidifier calculates the amount of water to spray and vaporize in order to reach the user-defined absolute humidity set point, based on the incoming gas conditions. The gas is preheated to supply the enthalpy required for vaporization. Water is sprayed through a Rayleigh nozzle, producing a near-monodisperse fine mist that rapidly vaporizes in the preheated gas. The patient-end temperature is controlled independently of the absolute humidity. The same inspiratory tube used for PassOverH1 (Hybernite RT 19, Plastiflex) was used for the absolute humidifier. During the experiments, the absolute humidifier was first set to a maximum humidification setting of 44 mg/L and left to run until it stabilized at the desired patient-end temperature (37°C). Two humidification settings at 0 and 25 mg/L added absolute humidity were tested during nebulization. The humidification setting was lowered from 44 mg/L to either 0 or 25 mg/L of added humidity immediately before nebulization started. Changing the humidification setting has an instantaneous effect on the absolute humidity delivered in the gas flow, as it directly controls the amount of water sprayed and vaporized inside the humidifier.

To measure the delivered dose, 2.5 mg of albuterol sulfate powder (PHR1053-1G, Sigma-Aldrich, Burlington, Massachusetts) was dissolved in 3 mL of injection-grade 0.9% saline, and the solution was nebulized into the breathing circuit. A vibrating-mesh nebulizer (Aerogen Solo, Aerogen, Galway, Ireland) was used to nebulize the drug for both pass-over humidifier and absolute humidifier setups, but at different positions in the circuit because of physical constraints, as shown in Figure 1. For the pass-over humidifiers, the nebulizer was placed immediately upstream of the pass-over humidifiers as per clinical practice. For the absolute humidifier, the nebulizer was placed immediately downstream of the humidifier to prevent heat damage to the drug and contamination of various sensors and the heating stage. As indicated in Figure 1, 3 absolute filters (Respirgard II 303, Vyaire Medical, Mettawa, Illinois) were placed in the circuit: one between the ETT and the test lung to collect the delivered dose through the ETT, one in the expiratory limb to collect the bypass dose (ie, the dose from the inspiratory limb to the expiratory limb that bypasses the test lung), and another one immediately before the ventilator to protect the ventilator from contamination. The drug deposited on the collection filters was rinsed out and measured with a UV–Vis spectrometer (Cary 60 UV–Vis spectrometer, Agilent Technologies, Santa Clara, California) at 276 nm. The collected doses were then compared against the total amount of drug put in the nebulizer to obtain the delivered dose and the bypass dose, which are fractions of the total dose in the nebulizer. Each test case was repeated 5 times. During the delivered dose measurements, the temperature and RH at the end of the inspiratory tube were recorded with a temperature and humidity sensor (SHT4x, Sensirion, Stäfa, Switzerland).

To measure the particle size distribution of the aerosol delivered to the test lung, a similar setup was used as for the delivered dose measurement. The end of the ETT was connected to an aerodynamic particle sizer (APS 3321, TSI, Minnesota) through a dilution stage (aerosol diluter model 3302, TSI), as shown in Figure 2. The diluter and APS have an inlet flow of 5 L/min. The diluter was used to decrease the total particle concentration of all test scenarios to 1,000–2,000 particles per cubic centimeter to reduce the number of coincidence events. The air exiting the APS exhaust was filtered and fed back into the breathing circuit to maintain the circuit pressure. The APS and the diluter were both stabilized at 37°C inside the climate chamber. Injection-grade 5.85% NaCl solution was used for nebulization for the aerosol particle size measurements to ensure that the particles exiting the circuit are within the measurement range of the APS. The nebulizer positions were identical to those used in the delivered dose measurements (Fig. 1). For the aerosol particle size measurements, each test case had a sampling time of 1 min and was repeated 5 times. The mass median aerodynamic diameter (MMAD) and the geometric standard deviation (GSD) of the individual tests were computed using the APS software (Aerosol Instrument Manager software, TSI).

Fig. 2.

Aerosol particle size measurement setup inside the chamber. The aerosol exiting the ETT first passed through an aerosol diluter, and its particle size distribution was measured by an aerodynamic particle sizer (APS). ETT, endotracheal tube.

The theoretical regional lung deposition of the delivered aerosol particles was computed using Mimetikos Preludium software version 1.2.2. ¹³ Weibel 17 was selected as the lung anatomy model, and the National Council on Radiation Protection model was used for the thoracic deposition calculations. The extrathoracic deposition was fixed at 0, and the extrathoracic volume was set to 0 mL to simulate the invasive mechanical ventilation scenario where the upper airway is bypassed with the ETT. A flow pattern similar to the recorded ventilator flow pattern was used. The tidal volume was set to 500 mL, and the peak inspiratory/expiratory flow was set to 50 L/min with a peak time of 0.2 s. A breath-hold time of 0.5 s, which matches the recorded ventilator flow pattern, was used. Breath-hold times of 0 and 1 s were also tested to demonstrate the effect of breath-hold time on the simulated lung deposited dose. A bolus volume equal to the tidal volume was used as the aerosol was provided continuously during ventilation. Aerosol size distributions with MMADs and GSDs close to the aerosol particle size measurements were used. The total lung deposited dose is the sum of the deposited doses in the tracheobronchial region (excluding the bronchioles), the bronchioles, and the alveoli.

As mechanically ventilated patients in the ICU that require aerosol therapy often suffer from lung diseases, simple models of the most common lung disease conditions, including asthma and 3 phenotypes of COPD (chronic obstruction, emphysema, and chronic obstruction in combination with emphysema), were also considered in the simulations. Patients with asthma or COPD suffer from air flow obstruction. For asthma, the obstruction is generally located in the larger airways, whereas for COPD, it predominantly affects the small airways and distal lung. The emphysematous component of COPD also results in enlarged alveolar spaces. To investigate how these lung disease conditions could affect the lung deposited dose, the airway dimensions in the software were adjusted according to Table 1. Although the severity of these diseases varies greatly, these values are within reason for ICU patients with these conditions.^14–17

Table 1.

Simulated lung disease conditions used in the lung deposited dose software calculations

	Asthma	COPD with chronic obstruction	Emphysema	COPD with chronic obstruction and emphysema
Tracheobronchial diameter (% decrease)	50%	0	0	0
Bronchiolar diameter (% decrease)	0	50%	0	50%
Alveolar diameter due to enlargement (mm)	No enlargement	No enlargement	0.75	0.75

Airway dimensions in the model were modified based on the conditions. Note that the tracheobronchial region specified excludes the bronchioles.

Statistical analysis

For the delivered dose measurements, the Kruskal–Wallis test was first conducted for the groups to check whether the samples came from the same distribution. As the sample size (5) was too small to assess normality accurately, the Mann–Whitney U test was used for pairwise comparisons between the different test cases. Benjamini–Hochberg correction was applied to the P value for all comparisons. A P value of < .05 was considered statistically significant. Confidence intervals (CI) of 95% around the means of the delivered doses were computed to provide clinically meaningful effect size information beyond statistical significance.

Results

The delivered doses for the different humidification scenarios (absolute humidifier at 0 and 25 mg/L added humidity, PassOverH1 and PassOverH2) are shown in Figure 3 Complete data on the delivered dose and bypass dose, including humidifier-off cases for PassOverH1 and PassOverH2, are summarized in Supplementary Table S1. The Kruskal–Wallis test showed that the samples do not all come from the same distribution (P < .001).

Fig. 3.

Delivered dose of the different humidification scenarios (mean ± SD) obtained from the bench study. P values are indicated. AbsH is the absolute humidifier tested at 2 different humidification levels. PassOverH1 and PassOverH2 are the 2 pass-over humidifiers tested. AbsH, absolute humidifier.

PassOverH1 gave the lowest delivered dose (21.2%, CI 19.5–22.8%), whereas PassOverH2 gave a marginally higher delivered dose (24.1%, CI 22.8–25.5%, P = .01). For PassOverH1, turning off the humidifier did not have a significant impact on the delivered dose (P = .11), whereas for the PassOverH2, turning off the humidifier led to a lower delivered dose (20.3%, CI 18.9–21.7%, P = .01). The absolute humidifier at 0 mg/L had a substantially higher dose (37.2%, CI 35.8–38.5%) than PassOverH1 (P = .01) and PassOverH2 (P = .01). The absolute humidifier at 25 mg/L had a somewhat higher delivered dose (24.9%, CI 22.6–27.2%) than PassOverH1 (P = .01) but did not show a significant difference compared with PassOverH2 (P = .55). All humidification scenarios gave a bypass dose of <10% (Supplementary Table S1).

To check whether sufficient humidification was reached during nebulization, the temperature and RH of the respiratory gas were measured at the end of the inspiratory tube. At the end of nebulization, the absolute humidifier at 0 and 25 mg/L of added humidity had an RH of 85.8 ± 2.1% and 94.1 ± 0.2%, respectively. PassOverH1 had an RH of 94.4 ± 1.6%, and PassOverH2 had an RH of 92.1 ± 1.8%. The absolute humidity delivered was calculated using the recorded relative humidities and temperatures. The absolute humidifier at 0 and 25 mg H₂O/L added humidity delivered 38.8 and 46.1 mg H₂O/L of absolute humidity, respectively. When turned on, PassOverH1 delivered 44.3 mg H₂O/L and PassOverH2 delivered 45.1 mg H₂O/L. All these humidification scenarios delivered a sufficient humidity level (>33 mg H₂O/L) during nebulization, fulfilling the performance requirement for invasive mechanical ventilation humidifiers (ISO 80601-2-74:2021). However, when PassOverH1 and PassOverH2 were turned off during nebulization, the absolute humidity delivered was reduced to 29.3 and 22.7 mg H₂O/L, respectively, which are not considered sufficient for invasive mechanical ventilation.

The aerosol particle size distribution (MMAD and GSD) of the aerosol exiting the ETT for the different humidification scenarios is summarized in Table 2. For all test cases, the particle size (MMAD) was around 1.7–2.1 μm. The absolute humidifier at 44 mg/L added humidity gave a similar MMAD (around 0.1 μm difference) as the pass-over humidifiers and slightly higher MMADs at lower added humidity settings.

Table 2.

Particle size distribution of the aerosol exiting the ETT for the different humidification scenarios

	MMAD (μm)	GSD
AbsH 0 mg/L	2.06	1.76
AbsH 25 mg/L	2.00	1.73
AbsH 44 mg/L	1.82	1.70
PassOverH1	1.71	1.61
PassOverH2	1.73	1.58

AbsH, absolute humidifier; ETT, endotracheal tube; GSD, geometric standard deviation; MMAD, mass median aerodynamic diameter.

Figure 4 and Supplementary Table S2 show the modeled regional lung deposition (as a percentage of the delivered dose) for healthy adult lungs when given aerosols with particle size distributions similar to what was measured at the end of the ETT. Multiple MMADs close to the range of the measured values (1.5–2.5 μm) were tested. A GSD of 1.7 was used for all cases. The simulated lung deposited dose decreases quickly with decreasing MMAD. For an MMAD of 2 μm, which is close to the measured size for the absolute humidifier at 0 and 25 mg/L, 41.8% of the delivered dose would deposit in the lungs according to the software. For an MMAD of 1.8 μm, which is close to the measured size for the pass-over humidifiers, 37.7% of the delivered dose would deposit in the lungs. This means that, incorporating the delivered dose measurements, when PassOverH1 and PassOverH2 are used, only 8.0% and 9.1% of the total nebulized dose would deposit in the lungs according to the software, respectively. For the absolute humidifier at 0 mg/L, on the other hand, 15.5% of the nebulized dose would deposit in the lungs. The simulated lung deposited dose also increases with the breath-hold time. For an MMAD of 2 μm, the simulated lung deposited doses with breath-hold times of 0, 0.5, and 1 s were 35.7%, 41.8%, and 46.8% of the delivered dose, respectively.

Fig. 4.

Effect of particle size (MMAD) on the simulated regional lung deposited dose (as percentage of the delivered dose) for healthy lung conditions. Note that the tracheobronchial region specified excludes the bronchioles. GSD = 1.7. Breath-hold time = 0.5 s. GSD, geometric standard deviation; MMAD, mass median aerodynamic diameter.

Table 3 shows the modeled regional lung depositions (as percentages of the delivered dose) for different simulated lung disease conditions. Of all disease conditions, only lungs with emphysema have a deposited dose lower than that of healthy lungs. For the other conditions, the lung deposited dose is higher than for healthy lungs. Lungs with asthma have the highest deposited dose, followed by COPD with chronic obstruction and COPD with both chronic obstruction and emphysema. The lung deposited doses for all disease conditions are increased when the breath-hold time is increased from 0.5 to 1 s.

Table 3.

Total thoracic lung deposition (as percentage of the delivered dose) obtained from the software for different simulated lung disease conditions, with different breath-hold times, an MMAD of 2 μm and GSD of 1.7

	Healthy	Asthma	COPD with chronic obstruction	Emphysema	COPD with chronic obstruction and emphysema
Breath-hold time 0.5 s	41.8%	68.3%	56.3%	36.8%	52.7%
Breath-hold time 1.0 s	46.8%	70.0%	59.5%	40.7%	55.0%

GSD, geometric standard deviation; MMAD, mass median aerodynamic diameter.

Discussion

The main objective of this study was to investigate whether aerosol drug delivery during adult invasive mechanical ventilation can be improved by controlling humidity independently during nebulization. Using an absolute humidifier with a low added humidity during nebulization, the drug droplets were allowed to partially evaporate and humidify the circuit. This results in the delivered dose, and hence the lung deposited dose, being significantly increased when compared with using conventional pass-over humidifiers.

This is the first study to address the detrimental effect of conventional humidification systems on the delivered dose for adult invasive mechanical ventilation, and it acts as a proof-of-concept for enhancing drug delivery for this ventilation scenario. Although recent literature ¹⁸ did not observe substantial differences in aerosol delivery between dry circuits and heated humidified circuits when exhaled humidity was utilized, the delivered dose reported for both scenarios (22–26%) was much lower than when a completely dry circuit was used (up to 37%). Our study showed that we could increase the delivered dose significantly to that achievable with a completely dry circuit while delivering sufficient humidity in the gas flow. In addition, our study provides insights into the lung deposited dose during invasive mechanical ventilation for healthy as well as diseased lungs using simulations with the bench study measurements as inputs. This could aid in better dosing of drugs during aerosol therapy, considering the patient’s conditions, and thereby improve understanding of the patient’s clinical response to the drugs.

The conventional pass-over humidifiers were first tested for validation. The delivered dose measured for PassOverH1 is consistent with a previous study in the literature^5,18 for adult invasive mechanical ventilation using the same type of nebulizer, nebulizer position, and pass-over humidifier. Turning off the pass-over humidifiers during nebulization did not lead to an increase in delivered dose. This result aligns with previous evidence in the literature ¹¹ and is most likely because of the RH in the gas flow remaining high after the humidifier is turned off. ¹²

Several mechanisms of aerosol particle deposition inside the breathing circuit are possible. First, the larger particles could deposit in the tubing by gravitational settling. According to calculations using the terminal velocity of the particles and a settling height similar to the radius of the inspiratory tube, for a freshly nebulized droplet of around 6 μm, it takes around 10 s to settle to the tubing wall from the center of the tube. For a mostly dry particle of around 2 μm, it takes more than a minute. Particles near the bottom of the tubing will take less time to settle. Since the aerosol could spend several seconds in the inspiratory tubing during tidal breathing, there could be sufficient time for the larger particles to settle in the inspiratory limb, considering the breath-hold and expiration time. Deposition by gravitational settling is higher for the larger droplets that are freshly nebulized when there is a high added humidity (ie, the pass-over humidifiers), but it becomes less likely for the scenarios with low added humidity as the droplets evaporate faster. The large particles are also prone to deposition by inertial impaction because of curvatures and angles in the flow path, as they are less able to follow the flow. This is most likely to occur in the pass-over humidifier water chamber, bends in the inspiratory tube, and the nebulizer connector. Using the peak flow of 50 L/min, the Reynolds number of the gas flow is estimated to be around 3,000, so the flow is likely turbulent. The turbulent regions may have increased deposition by impaction and decreased deposition by gravitational settling.

The observed differences in delivered dose can be attributed to the distinct humidification mechanisms of the 2 types of humidifiers. For pass-over humidifiers, the RH contribution comes almost entirely from the water vapor generated by heating the water bath. As a result, the drug aerosol droplets evaporate slowly and remain large for some time in this high-RH environment, resulting in the loss of a large portion of the nebulized dose in the circuit because of impaction and settling. For the absolute humidifier at the 0 mg/L added humidity setting, the RH contribution comes from the evaporation of the continuous stream of drug aerosols. The aerosol droplets evaporate rapidly, producing smaller particles less prone to circuit deposition and resulting in a higher delivered dose. For the absolute humidifier at the 25 mg/L added humidity setting, both active humidification and the evaporation of the drug aerosol contribute to the RH. The added humidity suppresses evaporation of the drug aerosol, resulting in slower evaporation than the 0 mg/L added humidity setting. This likely increased deposition in the tubing, yielding a delivered dose similar to PassOverH2.

The particle size measurements are comparable with the literature^3,11 and seem to support the above explanations. One main observation is that the MMAD of the drug aerosol exiting the ETT decreases with increasing added humidity. This appears counterintuitive, as one expects smaller particles with lower humidity levels because of increased evaporation. The most likely explanation, as mentioned above, is that when there is a high added humidity (ie, pass-over humidifiers and absolute humidifier at 44 mg/L), the drug droplets evaporate slowly, leading to a higher deposition of the larger particles in the tubing before reaching the end of the ETT. Deposition efficiency by impaction and settling increases by the square of the particle diameter and is thus substantially higher for the larger particles. For the pass-over humidifiers, a portion of the larger particles also deposits in the water chamber. ¹² The remaining aerosol particles will thus have a smaller MMAD, likely resulting in a smaller MMAD at the end of the ETT than if the entire nebulized aerosol underwent the evaporation process (ie, absolute humidifier at 0 mg/L). This also explains why the GSD is smaller when there is a higher added humidity, as the loss of larger particles makes the distribution narrower.

In general, the simulated lung deposited doses are low, as most particles in this size range are exhaled. The results show that the lung deposited dose increases with increasing MMAD for MMADs ranging from 1.5 to 2.5 μm, which is in accordance with the literature.^19,20 The effect of the particle size is most evident for the tracheobronchial dose. Increasing the breath-hold time allows more time for the particles to deposit, leading to a small positive effect in increasing the deposited dose, although it might not be a large enough benefit for the physicians to consider a change in the ventilator setting.

The simulations also showed that lung disease conditions lead to changes in the regional deposition in the lungs. Increased alveolar diameter leads to less deposition in the lungs as particle settling distances increase, whereas constriction of the bronchi and bronchioles increases the deposition of particles because of more efficient impaction in the bifurcations and shorter settling distances. These observations align with in vivo studies in the literature.^21–23 Increasing the breath-hold time also increased the deposited dose for all lung disease conditions, as it allows more time for the particles to deposit before exhalation.

Limitations

There are several limitations to this study. First, this study only considered the bronchodilator albuterol sulfate as a respiratory drug. For other drugs such as suspensions, antibiotics, or protein-based drugs, the fluid properties of the drug such as density, viscosity, surface tension, and hygroscopicity, may also influence the evaporation of the aerosol droplets and thus the delivered and deposited dose. To achieve the desired clinical effect, depending on the drug, the humidity setting may need to be adjusted to favor deposition in specific parts of the lungs. Further testing is needed to ensure the functionality of various drugs can be preserved and the aerosols can reach the target areas. Also, only one mesh nebulizer was tested in this study to ensure consistency and fair comparison of the humidification scenarios. Other nebulizers could lead to different delivered doses, but similar trends are expected.

Furthermore, the nebulizer was placed upstream of the pass-over humidifiers but immediately downstream of the absolute humidifier. The absolute humidifier was designed such that the nebulizer must be placed after the humidifier to prevent contamination of the spray chamber and potential degradation of medicine by the heating chamber. For the pass-over humidifiers, according to the manufacturer’s recommendations, the nebulizer should be placed on the dry side of the humidifier. ²⁴ Although this position results in a significant amount of drug lost in the water chamber, our measurements (see Supplementary Table S3) show that placing the nebulizer immediately downstream of the pass-over humidifier would lead to a comparable amount of drug lost in the condensation in the T-piece, which flows back into the water chamber. This is supported by the literature, which suggests that nebulizing at this location, which has 100% RH, would also result in significant drug losses because of condensation and potential particle growth, leading to a lower delivered dose than producing the aerosol at a cool and dry location upstream of the pass-over humidifier. ²⁵ In addition, placing the nebulizer on the wet side of the humidifier also risks interfering with the sensors because of unheated aerosol and condensation for PassOverH1, ⁶ and for PassOverH2 this position is not possible because of a direct electrical connection to the heated inspiratory tube. Moreover, several benefits of placing the nebulizer upstream of the pass-over humidifier in addition to the delivered dose have been identified. ⁵ Therefore, the nebulizer was placed immediately upstream of both pass-over humidifiers, and based on the considerations above, the difference in nebulizer position for the pass-over humidifiers and the absolute humidifier is not expected to affect the main conclusion that the absolute humidifier can give a higher delivered dose because of the effect of humidity control.

The simulated lung deposited doses are only indicative values because of limitations of the particle size measurement and the lung deposition model. The particle size measured likely differs slightly from the actual particle sizes slightly, as the particles travel through additional tubing and a diluter before they reach the APS. Evaporation was, however, limited since the RH for all test cases was quite high, and the RH inside the diluter and APS was equilibrated to the RH inside the circuit. There is also some deposition in this additional tubing and the diluter, with a bias toward higher deposition of the largest particles. Furthermore, the particle size measurements were conducted with 5.85% NaCl solution instead of the albuterol sulfate solution, as the albuterol sulfate solution made with 0.9% NaCl resulted in particles too small to be measured properly by the APS. The 2 salt solutions have different drying properties, so the particle size measurements with NaCl are not fully representative of the drug particle sizes during the delivered dose measurements. To compensate for these uncertainties, a larger range of MMADs was used in the lung deposited dose simulations as a sensitivity analysis. However, the trends observed in the measured particle sizes for the different humidification scenarios are likely representative of what happened and support the trends in the measured delivered dose. In addition, the lung deposition model did not account for intrapulmonary humidity. Although lung humidity should not influence the delivered dose, it could affect the evaporation and hygroscopic growth of the drug particles in the lungs, which affects the simulated lung deposited dose according to Figure 4. However, we do not expect a large effect as the temperature and RH in the ETT were close to those inside the lungs. ³ Nevertheless, we would like to acknowledge that the model is simplified, and the simulated lung deposited dose results might not translate to patient-important outcomes and await in vivo validation.

Lastly, the bypass dose measured was low for all humidification scenarios, as no bias flow was detected for the ventilation setting used. However, different ventilators have different bias flows, which may affect the bypass and delivered doses.^5,18 Future research is needed to assess the effect of bias flow for the absolute humidifier, but based on previous studies on the pass-over humidifier, increasing bias flow is expected to increase the bypass dose and reduce the delivered dose. ⁵

Conclusions

Through an adult invasive mechanical ventilation bench study, it was demonstrated that the delivered dose, for aerosol therapy during invasive mechanical ventilation could be improved by using an absolute humidifier, which decouples active humidification from heating inside the breathing circuit, instead of a conventional pass-over humidifier. Further research and development are needed to confirm the translational relevance of the findings in real-world scenarios and to optimize the absolute humidifier for clinical implementation.