Discussion
Recent advances in breath sound analyses have been remarkable10, 11). In clinical practice, attempts have been made to use breath sound analyses for the obstructive evaluation of airway narrowing23, 24), and the technique is improving18, 22). Since the measurement of breath sounds is non-invasive, simple and can be obtained with effortless breath, studies are underway concerning its application for the evaluation of respiratory diseases in children16) and infants14) who cannot perform normal lung function tests.
In the present study, of note, the breath sound spectrogram of expiration in acute bronchiolitis was found to be characteristic. As previously mentioned25, 26), the breath sounds of acute phase of acute bronchiolitis show marked power that is not observed in the breath sound spectrogram of healthy children27) or children with mild exacerbation of asthma16). According to reports of the pathophysiology28), epithelial damage and desquamation that introduce clusters of desquamated epithelial cells and viscous sputum29) are severe in the patients with acute bronchiolitis. Thus, a remarkable increase in the expiratory sound power of infants with acute bronchiolitis may be induced by severe stenosis of airways associated with these causes.
In this report, attending physicians described the presence of rhonchi, which are low-pitched continuous-expiratory sounds30), as a respiratory adventitious sound during expiration in many patients detected on auscultation. Widely, acute bronchiolitis is known to cause wheezes2), which are high-pitched continuous-expiratory sounds30), but our data included fewer patients with wheezes than previous reports on breath sounds25, 26). Typical rhonchi and wheezes are continuous musical sounds31) that are easy to identify as a sinusoidal waveform in the breath sound spectrogram27). However, based on examination of breath sound spectrograms, the expiratory sounds may differ from the waveform sounds of rhonchi and wheezes which commonly found in asthmatic children. We suspect that the non-uniform, band-like sounds in the low-pitched area seen in Figure 1 were heard as rhonchi. The sound spectrum of the expiration showed several repetitive peaks (Figure 2), which we presumed to be a pulse train or “complex repetitive waveforms”, as suggested by Tal, et al.25). We will continue to investigate the cause of such characteristic sound in infants with acute bronchiolitis.
Due to the increased power of the expiration, the E/I values were extremely large, showing a significant decrease in the recovery period. The E/I MF values in the recovery period were almost the same as those in healthy infants (the median value of E/I MF in 26 healthy 1-month-old infants who visited our hospital=0.05, private data). The airway stenosis in the acute bronchiolitis28, 29) may be stronger than that in the acute exacerbation of asthma32), although these results cannot be simply compared, as the airway portion with stenosis may differ between these conditions, and the subjects with acute bronchiolitis were younger than asthmatic children on average9, 33).
Despite these uncertainties, we speculate that our outstanding finding will prove useful for the diagnosis of acute bronchiolitis in infants. Although this study was a single-institutional, short-term study, we were able to limit the target to patients with RSV-induced acute bronchiolitis within the same season in an environment where pediatric respiratory tract infections are almost nonexistent due to the COVID-19 epidemic. Acute bronchiolitis should produce special sounds different from bronchoconstriction induced by smooth muscle constriction33), which may be particularly meaningful in distinguishing this condition from infantile asthma.
According to previous reports of adults, the E/I LF value in healthy adults is 0.36, which is equivalent to that in infants, although the E/I MF value is large at 0.2717). This difference was perplexing, as the I MF value of infants (87.2 dB) was larger than that of adults (57.3 dB) 17). The cause of this result is speculated to be dependent on the difference in airway diameter, respiratory flow rate (L/sec) and respiration pattern between adults and infants.
The values of the four sound spectrum curve indices in the inspiratory sounds13, 16) (A3/AT, B4/AT, RPF75 and RPF50) being significantly increased during the recovery period is attributed to the presence of bronchial constriction during the acute phase, based on previous breath sound analysis results15). In the present study, respiratory symptoms have improved in all patients at the second visit, and there were no adventitious sounds with auscultation. Thus, the changes in breath sounds during the recovery period are considered to indicate direct improvement of the airways.
More interestingly, a significant correlation was found between the E/I MF and SpO2, and even the severity score of bronchiolitis19). Although no correlation was found with the length of hospital stay, the correlation with the severity score used in clinical practice seems to be helpful for creating more objective severity scores. A number of studies have reported that RSV-induced acute bronchiolitis was associated with recurrent wheezing in later childhood3) and that the rate of development of asthma was not small4). Our results seems to be important for establishing more precise severity scores with an objective markers.
One limitation of our report is that we were unable to clearly demonstrate the specificity of the acute bronchiolitis due to the lack of data on the healthy age-matched infants (1-3 months old). Thus, we hope to explore this a topic in the future. Furthermore, we were not able to indicate whether or not the same findings would be obtained in other cases of infantile wheezing diseases with dyspnea33). However, this problem is not easily resolved, as the infants with acute bronchiolitis are mainly <6 months old9), and asthma, which is a representative of childhood wheezing diseases with dyspnea, cannot be diagnosed at this period33, 34). In a study of older children with asthma, it was reported that an intentional loud breathing increases the power of respiration35). There is no doubt that the breathing at rest in infants with acute bronchiolitis is greater than that in normal infants, so the specificity of breath sounds of acute bronchiolitis must be evaluated under such situation. However, the apparent significant increase in the expiration power is very clear, and we hope to evaluate this issue in the future.
Many discussions have been held concerning the effective treatment of acute bronchiolitis. According to the previous reports, inhaled β2 agonists have been shown to be ineffective, while inhaled adrenaline, inhaled saline, systemic steroids and their combination have been reported to be effective5-7). We hope to revisit our evaluation of the advantages of a breath sound analysis13, 16). Objective data over time using breath sound analyses may facilitate a clear evaluation of the effect of pharmacotherapy. As there are no clinical objective lung function tests available for infants33, 34), our findings are considered clinically meaningful. We are now considering a method to objectively evaluate the effects of drugs using a breath sound analysis.
Our results confirmed the specificity of the breath sounds in RSV-induced acute bronchiolitis. It has been reported that the breath sound directly reflects the condition of airways 13, 16). Our results seem to be useful for differentiating the diagnosis of acute bronchiolitis from other wheezing diseases, creating a more accurate severity scores and determining the objective effect of treatment.