LAUSR

The use of continuous positive airway pressure (CPAP) for the treatment of obstructive sleep apnoea (OSA) was first described in 1981. What started as a research tool became a clinical treatment that transformed and saved lives and uncovered the public health significance of OSA. CPAP has remained the treatment of choice for OSA over the last 38 years because it is highly efficacious at resolving upper airway obstruction, regardless of the site and severity of collapse. However, adherence to therapy has been continually questioned, perhaps unfairly when considered in the context of the medication adherence challenges encountered in most chronic diseases. Estimates of long-term non-adherence to CPAP vary widely from 29% to 83%. The reasons for poor or non-adherence are diverse and include disease and patient characteristics, treatment titration procedure, device factors and side effects and psychosocial factors. A recent systematic review of 82 clinical trials indicated that the rate of CPAP adherence has remained persistently modest over the last 20 years. However, recent large-scale real-world adherence data paint a rosier picture. Furthermore, continuous technological improvement in mask and device features and the development of digital patient engagement tools are transforming therapy and can be expected to enhance adherence. Among the many factors that influence adherence, a good mask fit is critically important to optimize patient comfort and to minimize air leak. Attempts to address this issue have generated a large variety of mask designs and sizes to cater for the heterogeneity in craniofacial phenotype, including nasal, nasal pillow, oronasal and oral designs. Surprisingly, the evidence base supporting clinical decisions about appropriate mask selection is relatively weak and is based largely on art rather than science. In a recent publication in Respirology, Goh et al. attempted to address this knowledge gap by conducting a randomized crossover study of three mask types in the treatment of OSA. They compared CPAP adherence during 1-month periods of use across three mask types, namely, nasal pillows, nasal and oronasal, in adult patients with moderate to severe OSA. They also sought to elucidate the influence of demographic factors, nasal symptoms [assessed by Nasal Obstruction and Septoplasty Effectiveness (NOSE) Scale] and craniofacial phenotypic characteristics (assessed by digital photographs) on therapy outcomes. Consistent with a recent meta-analysis, they found that adherence was greatest with nasal masks, with an average of 0.7 h per night higher usage relative to oronasal masks and 0.5 h per night higher relative to nasal pillows. Residual apnea-hypopnea index (AHI) (average 7.2/h) and mask leak were also higher with oronasal masks. However, they observed that a quarter of patients had better adherence with oronasal masks, and these patients were characterized by a lower level of subjective nasal obstruction and a larger menton-labrale inferioris/biocular width ratio, the latter representing a proportionally increased lower face height. The study represents an important addition to the literature as it is the largest randomized controlled trial of mask types and the first to examine the influence of craniofacial phenotype using digital photography. This study adds weight to the growing evidence base that oronasal masks appear to produce inferior therapy outcomes compared to nasal masks. It is noteworthy that the original description of CPAP involved the nasal route of delivery on the premise that positive airway pressure creates a seal through apposition of the soft palate and tongue. Although the mechanism by which oronasal masks are associated with a greater degree of residual OSA is yet to be fully elucidated, it is likely that the retrusive forces applied to the mandible by the oronasal mask may worsen upper airway obstruction in some patients. The finding that lower scores of subjective nasal obstruction were associated with better oronasal mask adherence is counterintuitive as such masks were developed largely to overcome the challenge of using a nasal mask in the presence of nasal obstruction and to minimize mouth leak. However, Lebret et al. reported that a NOSE score > 50/100, consistent with severe nasal obstruction symptoms, was independently associated with use of an oronasal mask. Hence, further work using objective assessment of nasal obstruction and assessment of oral breathing is required to clarify this potentially important issue. The craniofacial phenotypic assessment suggests that patients with a proportionally increased lower face height (chin to lower lip distance) may be better suited to oronasal masks. It would have been interesting to specially assess the influence of retrognathia on therapy outcomes with oronasal masks. Further work is required to obtain a deeper understanding of craniofacial phenotype and its influence on mask fit. Although the study has a number of strengths, there are a number of limitations that leave important questions unanswered. First, the studied sample represents a subgroup of eligible patients, and the analysis was restricted to patients who completed the protocol. Hence, there are likely to be biases that limit the generalizability of the study. Furthermore, the sample included Asians from Singapore, and hence, the craniofacial phenotypic influences on mask-related therapy