Why breath-by-breath built-in software data should be used to monitor CPAP/NIV in children?

Children receiving home continuous positive airway pressure (CPAP) or noninvasive ventilation (NIV) should be followed on a regular basis to ensure the tolerance and efficiency of the treatment.1 For children on CPAP, international guidelines recommend to periodically perform a polysomnography (PSG) to assess the effectiveness of therapy, while for children under NIV, they suggest that a reassessment with PSG and capnography (CO2) monitoring should be scheduled at least annually. However, access to PSG is limited and not available in all pediatric centers. Consequently, the overnight recording of home nocturnal pulse oximetry (SpO2) and transcutaneous CO2 monitoring together with the analysis of the built-in software data from the CPAP/NIV device may constitute an alternative, at least in selected patients.1 Different types of built-in software data are available according to the manufacturers, the devices, and the ventilatory modality.2 Trend data, overnight trends and statistics are increasingly used to monitor global CPAP/NIV efficiency. Information regarding daily usage (adherence), leaks, ventilation (tidal volume, minute ventilation, respiratory rate), pressure levels, and residual respiratory events may be reviewed. Trend data may be of interest to follow the variations of some parameters to detect changes in daily usage (such as treatment intolerance in case of decreased daily usage, increase of dyspnea or respiratory distress in case of increased daily usage), leaks issues or potential clinical issues, such as respiratory exacerbations or nocturnal hypercapnia/hypoxia.3, 4 However, further studies are required to understand their potential, to investigate their clinical usefulness, and to set strategies for an optimal clinical support following trend data review, in both adults and children. Residual respiratory events and apnea-hypopnea index data should be used carefully in pediatrics, as the algorithms for residual respiratory events on CPAP/NIV devices are based on adult criteria and do not follow the AASM scoring.5 Breath-by-breath airflow and pressure tracings allow an analysis of the nighttime breathing pattern, which can be combined to SpO2 monitoring (and CO2 monitoring for some devices) within the software.5 Moreover, some devices offer the possibility to measure thoraco-abdominal movements and to display these tracings in the built-in software together with the other tracings, which provides valuable complementary data to assess CPAP/NIV efficacy. However, in contrast to trend data, breath-by-breath tracings are still underused in pediatrics. And, to our knowledge, no study has addressed the interest of breath-by-breath tracings for the monitoring of CPAP/NIV in children. Regarding CPAP, the breath-by-breath analysis of the tracings, ideally with simultaneous SpO2, may detect “real” residual respiratory events (Figure 1A,B and Figure S1A) or rule out false events caused by an inappropriate airflow detection (Figure 1C,D). Indeed, Figure 1C,D show the disappearance of the airflow tracing at the end of the night, with concomitant clusters of desaturations in absence of unintentional leaks, due to the displacement of the interface by the 4-year-old child on her forehead. The child was thus not receiving the CPAP, explaining the occurrence of desaturations. On the contrary, Figures 1A,B and Figure S1A display “real” residual respiratory events associated or not with SpO2 desaturations. The overnight review of the breath-by-breath tracings is essential to determine if the number of residual events is significant enough to require an intervention. Accordingly, the detection of significant residual respiratory events by the breath-by-breath analysis should lead to an increase in CPAP level (Figure S1B). A measurement of SpO2 (and eventually transcutaneous CO2) should be performed, after CPAP adjustment, to assess the normalization of nocturnal gas exchange. For children on autoset CPAP, breath-by-breath tracings are also informative to identify residual respiratory events and determine if the settings (minimal and/or maximal CPAP levels) are adapted or should be modified, after potential leaks correction (Figure S2). Online Figure 2B shows the persistence of numerous long residual apneas, in absence of unintentional leaks, in a 11-year-old child. CPAP is plateauing at the maximal pressure (12 cmH2O), as observed during several periods through the night (Figure S2A), highlighting the fact that the maximal CPAP level set at 12 cmH2O is probably not sufficient to correct the residual events. In any case, the correction of unintentional leaks is mandatory before any setting adjustments. Moreover, anyway, leak correction is essential to guarantee a good patient tolerance. Regarding NIV, residual respiratory events and patient-ventilator asynchronies (PVA) may be identified on the breath-by-breath data analysis.2, 6 Figure 2A displays a one-night breath-by-breath tracings of a 15-year-old neuromuscular patient, showing residual respiratory events during NIV. The concomitant measurement of thoraco-abdominal belts connected to the ventilator allowed the characterization of these events as upper airway obstruction with decrease in ventilatory drive, indicating recurrent airway closure with a reduced ventilatory command, and therefore setting adjustments could be done accordingly. Figure 2B also displays a one-night breath-by-breath tracings of an 18-year-old patient with a brainstem tumor, showing long residual respiratory events. Respiratory events were suspected to be caused by a decreased ventilatory drive, which was confirmed by a polygraphy under NIV. Subsequent setting adjustments were made to optimize the ventilation. The identification and characterization of the residual respiratory events are thus crucial to adjust the NIV settings appropriately and therefore optimize NIV efficiency. PVA are a well-known phenomenon in patients under NIV. Different types of PVA exist and include those related to the ventilator trigger (trigger asynchronies), the cycling from inspiration to expiration (cycling asynchronies), the unintentional leaks, the respiratory rate or the patient's assistance.6 The clinical consequences of PVA are not clearly determined, and their effect on nocturnal gas exchange not yet established. Indeed, PVA may decrease ventilator efficiency, patient comfort and quality of sleep, increase respiratory effort, and impair gas exchange but the current literature remains controversial. PVA may be identified on the breath-by-breath tracings by a concomitant analysis of airflow, pressures, tidal volume and leaks (Figure 2C,D and Figure S3A). A correction or at least a reduction of unintentional leaks should be the first step. PVA should then be addressed according to their type, by adjusting the settings appropriately (Figure S3B). Figure 2C shows a one-night breath-by-breath tracings of a 4-year-old patient with a neuromuscular disease. Several attempts to trigger the ventilator were identified as the tracings show ineffective inspiratory efforts, despite the most sensitive inspiratory trigger. These ineffective inspiratory efforts are explained by the weakness of the patient's respiratory muscles which are unable to generate an inspiratory airflow sufficient to trigger the device. The change of the device for a device with a more sensitive inspiratory trigger led to a resolution of this PVA. In patients for whom trigger asynchronies cannot be corrected with the most sensitive trigger of a device, as in young children with weak breathing effort or children with weak respiratory muscles, the device may be changed for a more sensitive one (better airflow detection with level 3 devices, i.e., life supports), otherwise an adaptation of the settings (increase in back-up rate, pressure support…) may be proposed if the device is still not sensitive enough. Figure 2D displays a one-night tracings in an 8-year-old patient with a restrictive disease due to a congenital bone disease. Several double triggering and attempts to prolong inspiration were identified as premature cycling, and were due to a short “back-up” inspiratory time. Conversely, a long “back-up” inspiratory time may generate delayed cycling and therefore hinder the cycling to expiration, as it was the case in a 17-year-old patient with a restrictive disease (Figure S3A). These cycling asynchronies may be corrected by modifying the expiratory trigger and/or inspiratory times (Figure S3B). In conclusion, the analysis of breath-by-breath tracings should be systematically performed to check the efficiency of CPAP/NIV when PSG/polygraphy is not readily available, to detect and correct the residual respiratory events and PVA. This strategy may limit PSG/polygraphy during CPAP or NIV to selected patients who have problems to adapt to CPAP or NIV, or who are insufficiently improved.1 In our experience, residual respiratory events and PVA can generally be managed in the majority of children using CPAP or NIV, by adjusting the ventilator settings and reviewing the built-in software data. Whenever possible, thoraco-abdominal belts connected to the ventilator may be used to characterize the residual respiratory events. The efficacy of setting adjustments to correct residual respiratory events or PVA should be confirmed by the normalization of nocturnal gas exchange, with at least the monitoring of SpO2, at best SpO2 and transcutaneous CO2. Future studies should compare the validity of the combination of built-in software data and nocturnal gas exchange (SpO2 and CO2) versus PSG and transcutaneous CO2 for the follow-up of CPAP/NIV in children. Sonia Khirani: conceptualization; methodology; data curation; supervision; formal analysis; validation; investigation; writing - original draft; writing - review & editing. Marine Dosso: data curation; validation; formal analysis; writing - review & editing. Lorène Gerin: data curation; validation; writing - review & editing. Mihail Basa: data curation; validation; writing - review & editing. Charlotte Collignon: data curation; validation; Writing - review & editing. Meryl Vedrenne-Cloquet: data curation; validation; writing - review & editing. Lucie Griffon: data curation; validation; writing - review & editing. Brigitte Fauroux: data curation; validation; writing - review & editing. The authors declare no conflicts of interest. Ethical approval obtained and written consent obtained from parents. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.