1. Background 1.1 Investigational Device Custom non-invasive ventilation (NIV) masks will be fabricated based on a facial scan of the participant using sterophotogrammetry (3D photography). The scanner captures and digitizes facial features in a 180o ear-to-ear view in a few seconds. The points in the data set represent the surface of the face. This data set is then converted into a surface representation that can be manipulated using computer-aided design software (MeshMixer, Autodesk, San Rafael, CA). The facial scan data will be imported into computer-aided design (CAD) software (Fusion 360, Autodesk, San Rafael, CA) to create an NIV mask that is customized to the individual face. Once the design is created, a mold will be printed out of polylactic acid filament using a fused filament fabrication 3D printer (Ultimaker II, Ultimaker North America, Cambridge, MA). The printed mold will then be injected with medical grade silicone (Dragonskin, Smooth-On, Macungie, PA), and cured overnight. The coupler, used to attach the mask to the tubing connected to the ventilator, will be directly printed using stereolithography (SL Moai, Peopoly, Los Angeles, CA).
1.2 Previous Studies and Predicate Devices A few reports indicate that quantitative approaches towards NIV mask selection may improve comfort and adherence to treatment. In 2018 Ma et al. reviewed selection and design considerations for continuous positive airway pressure (CPAP) masks and stated "There is potential for optimizing the CPAP mask fit by adapting the interface to the patients' individual characteristics. This holds particularly true for users with abnormal features or simply outside conventional industry sizing standards. There is to date no integrated system that integrates those elements into a standard solution, but several studies have shown its effectiveness for specific cohorts."1
In 2019, Goh et al. reported a quantitative analysis of facial photographs of obstructive sleep apnea (OSA) patients.2 This report suggests that quantifying individual craniofacial features may help guide mask choice for NIV treatment for OSA, although the authors cautioned that the oronasal masks were not as effective as the nasal masks in terms of leakage and in reducing the number of apnea events. Despite the caveats, the report supports that further study of aa more rigorous approach to mask fitting is warranted. In an accompanying commentary, Cistulli and Sullivan state: "Surprisingly, the evidence base supporting clinical decisions about appropriate mask selection is relatively weak and is based largely on art rather than science."3
At least two companies offer custom NIV masks, including Inspirate (Singapore) and ApneaSeal, (Australia). To our knowledge, however, efficacy and compliance of custom versus commercial masks has not been tested rigorously. In addition, the published literature on customized masks for NIV therapy is small. For creating the masks de novo, one report details the fabrication of customized masks using maxillofacial prosthodontic laboratory techniques,4 while another reports a procedure including impression casting, 3D printing and silicone casting.5 Another approach describes modifying standard masks to create 3D-printed interfaces fitted to individual faces.6
1.3 Trial Conduct This study will be conducted in compliance with the protocol approved by the University of Alberta Health Research Ethics Board (HREB), and according to Good Clinical Practice standards.
1.4 Population Children already established on NIV (use for at least 3 month) between the ages of 6-12 years who continue to show sub-optimal adherence are able to be the participants in this clinical trial. Children with craniofacial disease or differences make up a large proportion of children using NIV. Craniofacial dimensions change with growth such that children using NIV may need to change to a different mask to provide the best possible fit. Children with developmental or physical disabilities that would prevent their successful completion of the study protocol (e.g., non-verbal, unable to follow 3-step commands, unable to lie still for 20 min) will be excluded from recruitment.
2. Trial Objectives To compare the efficacy of custom NIV masks to commercial masks as measured by efficacy of treatment delivery, compliance, fit test, and ratings of comfort.
3. Trial Design 3.1 Study Design/Type This is a non-randomized clinical trial using a crossover design comparing a custom NIV mask to the participants current commercial NIV mask.
3.2 Procedures
There will be three parts to this trial to compare the custom mask to the participant's current commercial mask:
i) Fit testing will assess both the unintentional leak and comfort of the custom mask compared to the participant's current commercial mask. This will be completed with the participant sitting in a seated position. For each mask, an in-line flow sensor will be used to assess unintentional leak (i.e., air leak around the mask) at two levels of mask tightness (typical used at home, loose) and two pressures (current mean airway pressure, 4 cmH20 higher). Participants will be asked to rate comfort for each trial using a pictorial visual analogue scale. The total testing time including set-up will be approximately 45 min.
ii) Titration polysomnography will assess the treatment efficacy. Participants will undergo a standard clinical polysomnography in the Pediatric Sleep Laboratory using their current commercial mask and then the custom mask prior with a 2-week measurement of compliance after each polysomnography study. The titration studies will be scored according to the Pediatric Sleep Laboratory protocol and the data extracted for analysis.
iii) NIV compliance \& end of trial questionnaire will assess compliance from NIV machine downloads as well as mask comfort and fit prior to and at the end of each 2-weeks compliance monitoring period, and preference assessed at the end of the study. After each titration study (custom, commercial mask), participants will use the same mask for 2-weeks with NIV machine downloads obtained specific to this time period for each mask. At the end of each 2-week period, participants and their parent/guardian will be asked to complete a questionnaire to assess participantive comfort and fit using a pictorial visual analogue scale as well as to provide narrative data on their experience. Participants and their parent/guardian will also be asked to describe any adverse events. There will be a minimum 2-week wash out period between mask trials where the participants will wear their current commercial ask without monitoring. After each participant has completed both 2-week trials, the participant and their parent/guardian will be asked which mask they prefer and why. The decision as to whether the custom mask is appropriate for on-going use outside the trial will be made by the participant, their parent/guardian, and the NIV clinical team (includes physicians, an NIV nurse practitioner, and an NIV respiratory therapist).
4. Statistical Plan 4.1 Statistical Methods Descriptive statistics will be used to summarize demographic and health information of the participants. Pair-wise t-tests will be used to compare continuous data including unintended air leak, polysomnography variables, and compliance measurements between custom and commercial masks with corrections for multiple comparisons as appropriate. Categorical data, including comfort and mask preference, will be analyzed using Chi-square or Kruskal-Wallis tests.
4.2 Participant Population(s) for Analysis Power calculations show that based on the recruitment of 15 participants, we will have the ability to detect a 1 hour increase in NIV compliance (increase in average use per day from 360 min to 420 min, SD 60 min) with power = 0.86, α = 0.05. Participants will be identified through the NIV clinic and approached for participation as they are identified. As the sample is not random and will focus on those with low compliance, the study results will be relevant to those children most likely to benefit from a custom mask.
