Post‐laryngectomy pulmonary and related symptom changes following adoption of an optimal day‐and‐night heat and moisture exchanger (HME) regimen

This study examined post‐laryngectomy pulmonary and related symptom changes following establishment of an optimal day/night regimen (all day/night wear of devices with improved humidification) using a new generation range of heat and moisture exchanger (HME) devices.


| INTRODUCTION
Anatomical changes following total laryngectomy surgery can have significant negative impacts for pulmonary health. Prior to undergoing laryngectomy surgery, when an individual breathes in, the inhaled air is conditioned (warmed and humidified) and filtered by the nasal airway/ upper respiratory tract prior to reaching the lungs. However, post-surgery the disconnection of the upper and lower airways, and redirection of the airway via a permanent tracheostoma, means that the lower respiratory tract epithelium is repeatedly exposed to unfiltered, colder, dryer air. This results in a decrease in mucociliary function and an increase in mucous production, leading to frequent involuntary coughing and forced expectorations, as well as increased risk of pulmonary and airway infections. 1,2 Furthermore, the combined effects of reduced humidification and impaired mucociliary function can cause tracheobronchial secretions to thicken and dry, forming mucus plugs that can cause airway obstruction. 1 The impacts of these respiratory/pulmonary changes are also wide reaching and can contribute to anxiety and depression as well as disruptions to aspects of daily functioning such as sleep and fatigue. [3][4][5][6][7][8] Recognizing the negative consequences of total laryngectomy on pulmonary function, the use of heat and moisture exchanger (HME) devices are considered best practice care for pulmonary rehabilitation postsurgery. 1,[9][10][11] Attached over the tracheostoma, HMEs typically consist of a plastic housing containing polyurethane foam treated with a hygroscopic salt (e.g., calcium chloride, lithium chloride) to enhance absorption of water vapor and warmth upon inhalation. 1,12 Inhaled air passing through the warm, moist foam of the HME is conditioned and some particulate matter is filtered. 1,12,13 Multiple studies have demonstrated the benefits of HME use post total laryngectomy, including improved mucociliary clearance, 14 a reduction in involuntary coughing and forced expectorations, 3,4,15,16 reduced pulmonary infections 17 and reduced incidence of mucus plugs 1 alongside positive improvements in psychological state and other functions. 3,4,6,7,15,16 Despite these confirmed benefits, patient factors can impact whether the HME is consistently used throughout the day and night, leading to variable HME use 3,15,18 and potential reduced humidification benefits. One issue impacting patient adherence to all day and all night HME use is the patient's response to the breathing resistance created by the HME. Following total laryngectomy surgery, patients can become sensitive to breathing resistance, particularly if already accustomed to breathing through an open tracheostoma, or when activity levels increase. 19,20 To help manage this, patients may choose to remove their HME for periods of time, such as during physical activity when work of breathing is increased. 12 This leads to reduced overall adherence to HME use, and reduced duration that humidification is provided.
To address this and other issues associated with optimizing pulmonary health, a range of new generation HMEs (Provox Life™, Atos Medical, www.atosmedical. com, Sweden) with improved humidification and breathability relative to predecessor products have been developed. The range includes different HMEs designed for specific situational use during the day and night, providing options for patients to switch between HMEs depending on situational needs and individual humidification and breathing resistance requirements. The new devices are accompanied by a full range of attachment options that allow patients to change the type of HME depending on their situation or level of activity, to support greater HME use.
Results of a recently published study by Longobardi et al., 21 which was still being finalized when the current study commenced, have provided initial evidence that this new range of devices has positive pulmonary and related symptom benefits for patients. Longobardi et al. 21 used a randomized cross over study design in which the cohort of 40 laryngectomy patients completed 6 weeks of usual HME use and 6 weeks using the new HME range. Results revealed increased HME adherence and a significant reduction in forced expectorations, dry coughs, improvements in both cough and sputum symptoms and impact, reduced shortness of breath, less skin irritation, as well as reduced anxiety/depression using the new device range.
In the current research, the primary aim was to examine changes in post-laryngectomy pulmonary function following establishment of an optimal day/night regimen (where optimal is defined as HME use all day and night, 24/7, or as maximally tolerated, using the new generation HME devices with improved humidification). Secondary aims were to examine any change in device use and impacts to skin integrity, sleep, shortness of breath, quality of life and patient satisfaction. These aims were explored within a two-phase prospective cohort study design. In the first phase, the patients changed from their usual HME routine to the use of a limited/comparable set of HMEs from the new generation range. It was hypothesized that simply exchanging the participants usual HME with select device/s from the new range, could provide incremental improvement in pulmonary and related symptoms due to the enhanced humidification properties of the new devices. Then in Phase 2, when participants were provided the opportunity to use devices with improved humidification and the choice of HME devices with different levels of breathing resistance to use in different situations, it was hypothesized patients would be able to achieve an optimal HME regime, leading to improved pulmonary and related symptoms.

| METHODS
The study was a multicenter, prospective two-phase trial. The project received ethical and governance approval at the main study site, and governance approval from the other two sites consistent with the "National Mutual Acceptance" policy which supports the single ethical review of multicenter research conducted in public health organizations in Australia (https://www.clinicaltrialsandresearch.vic.gov.au/ national-mutual-acceptance#page_downloads). All participants provided written consent prior to participation.

| Participants
Recruitment occurred in three adult acute services providing HNC care: two from the state of Queensland, and one from New South Wales, Australia. Participants were at least 3 months post total laryngectomy (irrespective of pharynx reconstruction) and 6 weeks post radiotherapy. All were current HME users (no minimum or maximum hours of use was set-participants could be wearing HMEs for any duration, including wearing HMEs 24/7) who had worked with their speech pathologists in the months/years prior to participation in this study to optimize their HME use, with the clinical target being to wear a HME as maximally tolerated day/night. Participants were excluded if: their existing daily HME routine incorporated the Provox Luna HME, their existing HME routine involved daily use of the Provox Micron HME, they had any existing medical problems prohibiting the use of a HME, any active recurrent or metastatic disease (medical deterioration), recent pulmonary infections/ unstable pulmonary condition, inability to independently insert and/or remove a HME, were unable to give informed consent, and/or had insufficient cognitive ability to use HMEs or adhesives. Sample size calculations were based on a 9.8 point estimated change score in the Cough and Sputum Assessment Questionnaire (CASA-Q, 22 determined from post-market surveillance data by Atos Medical) with power of 0.80 and 2-sided p-value of 0.05. This indicated a minimum of 38 participants. Target recruitment was set at n = 46, to allow for 20% attrition.

| Trial equipment and design
The Provox Life™ range (www.atosmedical.com) of HMEs and attachments were used in the trial. This includes five HMEs with different humification and breathing resistance properties to suit specific situational use during the day and night including: the "Home"-when relaxing at home; the "Go"-when doing light to moderate physical activities; the "Energy"-when physically active; the "Protect"when additional protection from bacteria/virus/dust/pollen is needed, and; the "Night"-for sleeping. A "FreeHands" option for hands-free speech is also available. Participants used the HME attachment method (adhesive/LaryTube/ LaryButton) that best suited their needs from the range. For those using adhesives, options included a soft flexible "Standard" adhesive for everyday use, the "Sensitive" adhesive made from hydrocolloid material, the firm acrylic "Stability" adhesive suitable for deep stomas and handsfree speech, and the "Night" adhesive made of hydrogel. All HMEs are interchangeable with any attachment method in the range.
The study involved a two-phase design, in which each participant acted as their own control. This enabled a staged exploration of the functional benefits of using the new device range. In Phase 1 participants used only the Provox Life™ products that were most comparable to their usual care products-that is, "like-for-like." For example, if a participant usually wore a Provox XtraFlow HME during the day/night, and used a Provox Micron HME occasionally, then they were provided with only the Go HME to use during the day/night and then the Protect HME for occasional use. This phase allowed exploration of any benefits patients achieved from the improved humidification properties of the new devices when used in the same manner to the participants prestudy HME device routine.
In Phase 2, participants were asked to use the full range of Provox Life™ HMEs and attachments as needed to achieve an optimal day and night routine, that is, HME use all day and night, 24/7, or as maximally tolerated, using the new device range. Participants were encouraged to use the range of HMEs as designed for the different situations and the attachment/s which best suited their needs. If using adhesives, in Phase 2 participants were also introduced to the Night adhesive, a soft adhesive made with hydrogel, a material found to assist skin recovery overnight. 9 Night adhesive use was determined on an individual basis and could vary from most nights (defined as ≥5 or more nights a week) to not at all, depending on individual need. Phase 2 allowed examination of any benefits achieved through using multiple devices, with enhanced humidification properties, to optimize HME use across the day/night. Each phase began with an adjustment period during which participants were introduced to the devices to use by their speech pathologist. This period provided participants time to adjust to using the new device/s and attachments and to troubleshoot issues with their treating clinician prior to commencing the study Phase. The adjustment period in Phase 1 and 2 was open, with the duration (Phase 1: mode:=7 days, range = 4-21 days; Phase 2: mode = 7 days, range 4-23 days) influenced by individual need and impacted by external factors, for example, COVID-19 appointment rescheduling, clinician availability.
Following the Phase 1 adjustment period, participants commenced Phase 1, and used only the specified "likefor-like" device/s and Provox Life™ attachment/s for a 6 week period. In Phase 2, following the Phase 2 adjustment period, participants were instructed to use the full range of Provox Life™ HMEs and attachments as needed to achieve an optimal day and night regime, that is, use of a HME all day and night, or as maximally tolerated, using the new generation devices. In both phases, participants were instructed to apply the HMEs and attachments consistent with their normal daily routine. All HME's and attachments were provided free of charge to study participants, along with associated products/ equipment to support application (e.g., Provox Skin Barrier and Adhesive Remover) and daily use (e.g., Provox Life Shower for showering). No other financial incentives were provided. Device troubleshooting/support was available throughout the study.

| Data collection
Monitoring and reporting of adverse events was ongoing throughout the study. Data pertaining to device use and functional outcomes was collected at baseline (reflecting "usual care" prior to the study), then at weeks 2 and 6 of Phase 1 and Phase 2 respectively. All participant assessments were scheduled for ±7 days of the 2 and 6 week timepoints in Phase 1 and 2, however, a small proportion of assessments (1.9%) were completed >7 days due to unavoidable COVID-19 lockdowns/restrictions and participant ill health.
Duration and patterns of HME and attachment use was collected at each timepoint using study specific questions reflecting on use "over the past 7 days." A 7-day Diary and Tally Sheet was also collected, detailing the daily type and duration of HME use, and the number of involuntary coughs and forced expectorations experienced (for a 3 day period within the 7 days). Participants also completed outcome measures and study specific questions at each timepoint. Frequency and severity of cough and sputum/mucus production and their impact on everyday life were evaluated using the Cough and Sputum Assessment Questionnaire (CASA-Q), 22 a 20-item assessment, answered on a 5 point scale across a 7 day recall time. The CASA-Q was developed and validated for patients with chronic obstructive pulmonary disease (COPD) and covers four domains (cough symptom, cough impact, sputum symptom and sputum impact) scored from 0 to 100 with lower scores indicating more symptoms/higher impact. Although the CASA-Q was not developed or validated for the laryngectomy population, its four domains assess symptoms experienced by the laryngectomy population. 21 Hence, in the absence of a validated tool for detecting pulmonary symptoms in the laryngectomy population, the CASA-Q was used. The cough symptom and cough impact measures of the CASA-Q were preselected as the primary outcome variables of this trial.
Shortness of breath was assessed using a 3-item questionnaire 8 exploring how often participants felt short of breath when (a) climbing stairs, (b) walking and (c) resting over the past 2 weeks (4-point scale, 1 = not all, 4 = very much). Sleep was assessed using the Jenkins Sleep Evaluation Questionnaire (JSEQ), 23 a validated, 4-item questionnaire based on a recall period of "the past 30 days." Lower total scores indicate better sleep behavior. Regarding skin integrity, participants completed 3 questions (unpublished, used with permission) to describe (a) skin around their stoma (normal, sensitive, red/irritated), (b) how often they experienced skin irritation around their stoma (frequent: every day/second day; sometimes: 1-2 times a week, or; rarely/never), and (c) rate on a scale of 1-10 (10 = a lot of influence) how much skin irritation affects their life. The EQ-5D-5L 24,25 was used to measure quality of life. It examines five health care dimensions (mobility, self-care, daily activities, pain/discomfort and anguish/depression), and participants rate their current health status on a 100 point visual analogue scale (VAS, 100 = best imaginable health state). 25 Overall satisfaction/ patient experience was assessed with study specific questions at the end of Phase 1 and 2 only. These examined: overall experience using device/s; time to adjust to using the device/s; perceived advantages/disadvantages; voicing; and ongoing barriers for not wearing a HME all day/night. At end of Phase 2, additional questions were asked about relative ease of selecting HMEs for different situations, reimbursement considerations, and if they would use the devices in the future.
Due to the impacts of COVID-19, patient reported outcome measures were collected either through paperbased forms, or online using the electronic patient reported outcome (ePRO) platform of Castor EDC (castoredc.com). The 7-day Diary and Tally sheet was completed via paper-based forms and returned/posted back. Study specific questions were completed either independently, or with the study research coordinator either in person, via phone, or via telehealth.

| Analysis
Statistical analysis was conducted using IBM ® SPSS ® Statistics version 19. For the total cohort, the 3 preplanned study endpoints calculated for the primary outcomes and all other variables were: baseline to end of Phase 1; end of Phase 1 to end of Phase 2, and; baseline to end of Phase 2. Following completion of the planned analysis, a secondary sub-analysis was also run on the CASA-Q data from participants who were using HMEs 24/7 at baseline to explore separately the impacts of changing the device type for this cohort. Prior to all analysis, data was first checked for normality, and due to non-normal distribution nonparametric statistics were conducted. Continuous data was analyzed using Friedman tests with post hoc Wilcoxon signed rank tests while categorical data was analyzed using the McNemar test. All analysis was conducted as 2 tailed tests. Acknowledging the multiplicity of tests conducted, the alpha level for significance was reduced from 0.05 to ≤0.01 to reduce risk of Type 1 error. However, as this was an initial study of these devices, and to minimize risk of Type 2 error, results >0.1 and <0.05 are noted as trends.

| Participants
Forty-eight participants were initially recruited. Six later withdrew, 2 after consent but before data collection, 3 during Phase 1, and 1 in Phase 2. Reasons for withdrawal were either equipment related (n = 2, 1 had discomfort with the LaryButton which impacted voicing, 1 developed a skin rash from the adhesive) or non-study related changes in personal health state (n = 4, of these, 3 were deceased by end of the study period). The remaining 42 (95% male; mean age 70.4 years, range 48-82 years) were recruited almost equally from the 3 sites (site 1 = 38%, site 2 = 38%, site 3 = 24%). Participants were a mean of 60.1 months post laryngectomy (median = 49, range 4-210) and the main communication method was tracheoesophageal speech (79%; electrolarynx 14%; AAC/mouthing 7%). Most had a history of smoking (83%; 14% never smoked; 2% currently smoking). Only 24% were in full or part-time employment with most (69%) retired or unemployed, 7% on a disability pension or other. The majority of the cohort was already highly adherent HME users, with 81% of the cohort wearing HMEs ≥20 h/day and 62% wearing a HME 24/7 at baseline. Participants stated that the reasons they were unable to wear a HME 24/7 at baseline (multiple responses allowed) included: shortness of breath (n = 3), unable to use while sleeping (n = 6), taking skin rest/ irritated skin (n = 7), and during specific activities/other, for example, concreting, lawn moving, playing golf, nebulizing, problems with button (n = 5).

| Included data
Partial data from withdrawn participants was not included in analysis. All other data was included except for CASA-Q data for 2 participants, who at two separate timepoints only, had a co-occurring (a) leaking voice prosthesis, and (b) viral illness, which impacted coughing rates, and no data was recorded for these timepoints. Occasional questions were not answered by some individuals and where response rates were less than 42, this is identified in the analysis. Missing data constituted less than 10% of the data set and was determined to be missing at random, hence no data imputations were applied. Data from the 2 week assessment timepoints of Phase 1 and 2 are not reported in this manuscript.

| Pulmonary outcomes
Analysis of the cough symptom and cough impact items from the CASA-Q (primary outcome variables) for the whole cohort revealed a significant improvement in cough symptoms (n = 39, X 2 = 20.464, p = 0.0001) and impact (n = 36, X 2 = 17.91, p = 0.0001). Post hoc analysis confirmed significant improvements between baseline and Phase 1, and baseline to Phase 2 ( Figure 1). CASA-Q scores for sputum symptoms (n = 39, X 2 = 18.242, p = 0.0001) and impact (n = 38, X 2 = 13.956, p = 0.001) for the whole cohort also significantly improved across the phases. Post hoc analysis confirmed significant improvements between baseline and Phase 2 for both variables, and between baseline to Phase 1 for sputum symptoms only (Figure 1).
An additional secondary analysis was also run across all CASA-Q variables for the subset of participants (n = 26) already wearing HMEs 24/7 at baseline. This was conducted to explore any pulmonary benefits identi-

| Adherence and use
There were no serious adverse events. Analysis revealed a significant (n = 41, X 2 = 10.138, p = 0.006) increase in HME use across the phases, from 21.6 h at baseline to 23.1 h at end of Phase 2 ( Figure 2). Post hoc testing revealed a significant (p = 0.0002) increase in HME use between baseline and end of Phase 2. Exploring individual data revealed that 6 participants at baseline could not wear a HME when sleeping, which reduced to 2 participants in Phase 1 and 2. The proportion of adherent HME users (i.e., HME use tally ≥20 h/day) increased from 81% at baseline, to 90% in Phase 1 and 93% in Phase 2. Number of patients who reported achieving an all day/night HME routine (excluding showing, cleaning), increased from 62% at baseline, 76% in Phase 1, to 90% by end of Phase 2. For the 10% who were still unable to wear a HME 24/7 at end of Phase 2, the reasons reported were (multiple responses allowed): issues when sleeping (n = 2), irritated skin (n = 3), shortness of breath (n = 1), and during specific activities/other, for example, nebulizing; problems with tube (n = 2).
As expected with the study design, the number of HMEs used/disposed of every 24 hours increased (n = 42, X 2 = 45.016, p < 0.00001). There was no change from baseline (m = 1.5 devices) to Phase 1 (m = 1.6 devices), however baseline to Phase 2 (n = 42, p = 0.0001) and Phase 1 to Phase 2 (n = 42, p = 0.0001) were significantly different, with 2.4 devices used on average during Phase 2. Individual analysis revealed at baseline 83% used just 1 type of HME daily (17% used 2 e.g., one daytime HME + either an alternate device at night or occasional use of a device with increased filtration). In Phase 1 ("like-forlike"), this pattern was maintained (71% using 1 HME, 29% using 2 types of HMEs). By the end of Phase 2, 0% were using only 1 HME, 19% 2 types, 48% 3 types, 24% 4 types and 9% 5 types as needed per the situation. In Phase 2, different devices were being selected for different situations with the majority using Home when relaxing, Go when out, and the Night when sleeping (Figure 3). By F I G U R E 2 Duration of HME use. end of Phase 2, 60% were using a combination of the Home most hours when awake and the Night >5 nights a week (± other devices within their daily routine)-the two devices that offer the highest humidification. Reasons for HME replacement varied significantly across the phases (n = 42, X 2 = 18.142, p = 0.006). "Routine replacement" increased from 55% at baseline to 74% at Phase 1 and 2, while "blocked by secretions" reduced from 40% at baseline, to 24% Phase 1 and 14% in Phase 2 ( Figure 4).

| Sleep
There was a significant improvement in total sleep score (n = 37, X 2 = 8.574, p = 0.01). Post hoc analysis indicated improvements between baseline and Phase 2, and Phase 1 to Phase 2 ( Figure 5). Examination of the individual questions of the scale revealed the greatest improvements were noted for the items: waking several times a night; having trouble staying asleep, and; waking tired and worn out.

| Participant perceptions and satisfaction
In Phase 1, 34 participants reported positive advantages of the new devices including reduced pulmonary complaints (less coughing, less mucus, less clogging of devices), improved breathability, improved voice, and better adhesive seal. In Phase 2, 33 individuals reported advantages, which were the same as in Phase 1 with the addition of improved sleep. After Phase 1, 9 noted disadvantages including more mucus, issues with adhesive seal, issues with packaging, and issues with the "softer" LaryTube. After Phase 2, 15 reported disadvantages similar to in Phase 1, with an additional issue regarding the Night baseplate adhesion. Across Phase 1 and 2, 15 participants (36%) experienced changes to voicing using the new devices, with 11 reporting positive changes (louder voice, improved clarity, easier to voice, less coughing interruptions) and 4 neutral/negative changes (harder to voice, lower voice, different tone). Half (50%) the participants required 1 day to adjust to the new range of devices (14.3% 2-3 days, 14.3% <7 days, 19.1% >7 days). Ninety percent felt deciding which device to use in different situations was easy/very easy (7% neutral, 2% difficult) and of the 34 who tried the Night adhesive, 65% found determining the best schedule for using the Night adhesive was easy (29% neutral, 6% difficult). At the end of Phase 2, 7 (17%) individuals reported ongoing barriers to wearing a HME all day and night, including difficulty breathing at night, poor baseplate adhesion, sensitive skin/need for skin rest, and costs. Overall, 95% stated they would continue to use the device range in the future. Regarding reimbursement, at baseline the majority used only the HME devices as provided through their equipment subsidy funding (offered by their state health service), with only 10% opting to pay for additional HME equipment as out-of-pocket expenses. At the end of Phase 2, 95% stated they would continue using the day/night range if reimbursement would allow it. A further 41% noted they would pay out-of-pocket if the devices were not reimbursed fully.

| DISCUSSION
The current two-phase study design enabled staged examination of the functional impacts of a new range of HME devices within a cohort of existing HME users post laryngectomy. In comparison to usual HME use (baseline), transitioning to comparable HME devices with improved humidification properties (Phase 1) led to improvements in cough symptoms, cough impact, sputum symptoms, and involuntary coughing, and a reduced need to replace devices due to blocking/secretions. By the end of Phase 2, when using the full range of new HME devices with their different levels of breathing resistance, study participants were better able to achieve an optimal day/night routine, with significant improvements in the duration of device use per day. This was a direct consequence of being able to wear a HME which enabled them to breathe easily in the activity they were engaged in. At the end of Phase 2 participants reported improved cough symptoms, cough impact, sputum symptoms, sputum impact, reduced involuntary coughing, reduced need to change devices due to blocking/secretions, and improved sleep. Despite different methods, and different international study populations, the current data is highly consistent with the findings of Longobardi et al. 21 and supports the use of new generation HME systems to facilitate further optimization of pulmonary function for established HME users post laryngectomy.
Results of the CASA-Q and 7 day tally by end of Phase 2 confirmed improved pulmonary symptoms through establishing an optimal day/night routine. The CASA-Q has not been used as a measure of pulmonary symptoms in the laryngectomy population prior to the Longobardi et al. 21 and current study, as such there is limited understanding of what would be an expected amount of change to observe using this tool. Although the mean scores for all domains at baseline were lower in the current cohort than the cohort studied by Longobardi et al., 21 the extent/percentage change across the domains was highly comparable in the two studies (current study: change between baseline and end of Phase 2 for cough symptoms 15%, cough impact 15%, sputum symptoms 27% and sputum impact 14%; Longobardi et al: 15%, 9%, 24%, and 19% change respectively).
As per prior research 3,9,15 these pulmonary symptom improvements can be attributed to both increased hours of HME use and also to the improved conditioning of inhaled air. Although all participants were already established HME users, the proportion of participants able to achieve a 24/7 day/night regime increased from 62% at baseline to 90% in Phase 2. This enhanced HME adherence was facilitated by the improved breathability of the new devices and the ability to change devices based on the breathing resistance tolerated in different situations. To date, the market has offered laryngectomy patients with a limited choice of HME devices, as confirmed with 83% of the current cohort using only 1 type of HME at baseline. Consequently, to date HME adherence has largely been influenced by an individual's tolerance of the breathing resistance of their chosen device. 19,26 Now, however, clinicians can provide patients with a range of devices with different levels of breathing resistance, which offers greater opportunities to support enhanced HME adherence and improve pulmonary symptom management post laryngectomy. In addition, sub-analysis of the CASA-Q data from the participants who were already 24/7 HME users at baseline, confirmed that changing to devices with improved humidification properties 24/7 supported positive pulmonary benefits.
In Phase 2 participants responded positively to the opportunity to use a range of different devices, adapting quickly to a new routine, with the majority changing their daily routine to using either 2 (Home and Night) or 3 devices (Home, Go and Night) each day, with then occasional use of other devices on some days, depending on situational need. This included adoption of the Night HME by 83% of the cohort, though a small proportion of individuals did not change to using the Night HME and preferred to continue using the same type of HME worn during the day, at night. By the end of Phase 2, more participants had also started to use the Energy HME and the Protect HME in specific situations, when they had short periods of higher physical activity, such as playing golf once a week, or were exposed to more particulate matter in the air, such as when mowing the lawn. These devices were used more sporadically. As both the Energy and Protect offer lower humidification benefits than other devices within the range, their use should be limited to such times of specific need. Where able, patients should be encouraged to maximize wear of the devices that offer the greatest humidification, such as the Home and/or the Go during the day, and the Night at night.
The improved pulmonary function achieved through adopting an optimal day/night routine also had positive impacts for routine HME use. From Phase 1, participants reported significantly improved sputum symptoms with further reductions observed in Phase 2. Potentially related to this change, there was a reduction in the need to change HME's due to blocking/secretions, and a corresponding increase in "routine replacement" as the primary reason to change the HME. However, the impact of changing between multiple devices in Phase 1 may also have influenced this result.
Prior research has demonstrated improved sleep outcomes associated with HME use due to reduced disruptions caused by coughing/mucus management. [3][4][5][6] In Phase 2, 95% of participants were wearing a HME through the night, with 83% using the Night HME which offers the highest humidification properties of the new device range. In Phase 2, significant improvements were observed in total sleep scores, with most change reported for the items relating to waking at night, having trouble staying asleep and waking tired. This result validates the data reported by Longobardi et al. 21 who found a significant reduction in use of sleep medication (surrogate measure for improved sleep) within their cohort using the new device range.
Positive changes in social anxiety, depression and quality of life have previously been observed with HME use. 4,7,15 However, despite the majority of participants noting positive benefits in both pulmonary function and sleep, quality of life scores in the current cohort remained unchanged. Most participants rated their overall quality of life >75/100 on the VAS, and this showed minimal change across the study phases. Similarly, minimal change was observed across the 5 domains of health status. These results were consistent with Longobardi et al. 21 who also failed to find a significant change in VAS or index scores of the EQ-5D.
Skin irritation is a known factor contributing to poor HME adherence. 3,9,19 However, in the current cohort very few participants experienced issues with skin irritation prior to the study trial and this remained unchanged throughout the phases. Accordingly, less than a third of the cohort used the Night adhesive to assist skin repair. Prior research has demonstrated that patients using a soft hydrogel adhesive, as developed for the Provox Luna, experienced an increased frequency of days where skin improvement was observed overnight. 9 Similarly Longobardi et al. 21 also reported reduced skin irritation when their cohort was using the Provox Life range, however the rates at which patients used the Night adhesive was not explicitly reported. Based on such prior research 9 the potential benefits of the Night adhesive should be considered for patients who experience irritation around the stoma when wearing adhesives.
In addition to pulmonary benefits and improved breathability, a quarter of the cohort noted improvements to their tracheoesophageal speech using the new devices. This appeared to be directly related to improved pulmonary symptoms with changes described as louder voice, easier to voice and less coughing interruptions to voicing.
Overall perceptions of the new device range were very positive, with 95% indicating they would use the device range in the future.
The improved clinical outcomes and increased adherence observed in this study came with increased utilization of HMEs, increasing from an average of 1.6 devices/day in Phase 1 to 2.4/day in Phase 2. It is recognized that using more devices each day will come with a corresponding cost, which may impact adoption by some patients. Of note 41% of participants indicated they would be willing to pay additional out-of-pocket costs to continue to use the device range in the future, despite the majority being either retired or on a disability pension. However, as each individual will have a different pattern of device and adhesive use, it is important to note that exact out-of-pocket costs will vary, and need to be discussed with each patient, taking into consideration their individual situation, preferences and needs, and any local models of equipment cost subsidy, if available. For those unable to meet the additional costs, or for those who may find using multiple devices too complex or cognitively demanding, the data from Phase 1 of this study demonstrates that simply exchanging the patients baseline HMEs with comparable HMEs from the new generation range can still achieve some incremental positive pulmonary and other symptom benefits. Prior studies have proven HMEs to be cost-effective. 10,27 A full assessment of the incremental cost-effectiveness of the change in routine using Provox Life is planned to investigate if the added 0.8 HME utilization can be justified by improvements in clinical outcomes and any impacts to quality of life.

| Limitations
As the new generation range of devices differ in size, presentation and range from prior types of HMEs, participants could not be blinded. Therefore, the potential for there to be a positive reporting bias when using the new devices cannot be discounted. Equally, as with all studies that require informed consent to participation, there remains an inherent volunteer bias which must be recognized. In addition, although participants reported that they easily adapted to using multiple devices, it is important to note that the study design excluded participants with insufficient cognitive capacity. As such, the capacity of patients to manage the cognitive load associated with using multiple HMEs in different contexts needs to be considered on an individual basis.
The outcome measures collected in this trial were all subjective/patient reported outcomes. In the future, objective measures such brush biopsies of the tracheal epithelium, 14 and tracheal mucus transport scintigraphy, 14 as well as overnight sleep studies and cough tracking Apps could be incorporated to quantify the current patient reported outcomes. The authors again acknowledge that the CASA-Q tool was not developed for use with the laryngectomy population, and as such there is no reference data available regarding expected values for coughing or sputum function for the laryngectomy population. However, in both the current study, and Longobardi et al. 21 it was sensitive to pulmonary symptom change, and so clinicians may find it a useful tool to monitor pulmonary symptoms over time. This study, and all phases of data collection, was conducted during the COVID pandemic, and data collection occurred in different formats for study participants due to COVID-19 restrictions throughout the study period. However, all opportunities were taken to confirm data results with participants who completed their forms independently where questions arose. It is also noted that the adjustment period extended to 3 weeks for a few individuals due to a combination of troubleshooting and COVID-19 restrictions on clinician appointments which extended the time using the devices for those participants.

| CONCLUSIONS
The current data confirms the potential to improve pulmonary symptoms for existing HME users, even if patients are already 24/7 adherent users, through use of the range of new generation HME's. The current trial design has demonstrated benefits in pulmonary (cough/ sputum) and related functions (sleep) that can be attributed to improved pulmonary rehabilitation achieved through both the enhanced humidification properties of the new devices and the increased hours of use per day facilitated by having a range of devices with different breathing resistance for situational use.

AUTHOR CONTRIBUTIONS
All authors provided contributions consistent with authorship. Elizabeth C. Ward designed the study with the project sponsor Atos Medical. All authors Elizabeth C. Ward, Kelli Hancock, Jenni Boxall, Clare L. Burns, Ann-Louise Spurgin, Belinda Lehn, Juliet Hoey, Rachelle Robinson, and Adele Coleman were involved in recruitment and data collection. Atos Medical had no role in recruitment, data collection or data entry beyond routine auditing of the clinical trial. Elizabeth C. Ward and Adele Coleman provided project management. Kelli Hancock, Clare L. Burns and Rachelle Robinson provided local project leadership at their respective sites. Elizabeth C. Ward and Adele Coleman, with review by Atos Medical, completed data analysis. Elizabeth C. Ward interpreted the data and wrote the manuscript. Kelli Hancock, Jenni Boxall, Clare L. Burns, Ann-Louise Spurgin, Belinda Lehn, Juliet Hoey, Rachelle Robinson, Adele Coleman edited the manuscript. The study sponsor had no role in data interpretation and reviewed the final manuscript for accuracy of company/product name and product description only. The study sponsor placed no restrictions on the dissemination of study findings and was not involved in the journal submission or publication process.