Bronchial intubation may produce significant hypoxemia. Current methods of detecting this condition include radiography, auscultation, and direct visualization. While radiography and direct visualization may involve delays and be unavailable outside the hospital, auscultation may have limited accuracy. The purpose of this study is to measure breath sound asymmetry caused by bronchial intubation, and assess the utility of that asymmetry for bronchial intubation detection.
After IRB approval and informed consent, breath sounds were recorded in 19 healthy subjects undergoing general surgery. While patients were supine, 2 electronic stethoscopes were placed at the right and left intersections of the axillary and nipple lines. After anesthesia induction, breath sounds were recorded for tracheal and bronchial intubation, which were confirmed fiberoptically. The acoustic signals were converted into digital form using a PC. The breath sound signal energy before and after filtering out certain acoustic frequencies (300-600 Hz) was calculated, along with the ratio of the acoustic energy between the left and right stethoscopes to assess breath sound asymmetry. Energy ratios for the tracheal and bronchial intubations were compared using the Wilcoxon signed-rank sum test.
Accuracy for separating tracheal and bronchial intubation was 100% when the acoustic signals were filtered (p<0.00001). The 100% separation using computer-assisted breath sound measurements suggest a high sensitivity and specificity for bronchial intubation detection.
These preliminary results suggest that devices implementing this technology may be a reliable, accurate, portable, and inexpensive. Such devices can be used for both online monitoring of ETT position and during initial intubation, and may be most useful when radiographs are unavailable, unpractical or unreliable. Further studies will determine the applicability of this device to a wider range of patients with more diverse medical conditions and different body weights and sizes.
Improved bronchial intubation detection may assist clinicians in accurate and inexpensive assessment of patient status, thereby lowering morbidity, mortality and financial costs.
R.A. Balk, None.