Can short-term heart rate variability be used to monitor fentanyl-midazolam induced changes in ANS preceding respiratory depression?

ABSTRACT

Opioids have an occasional but high-risk side effect of respiratory depression. The detection of critical respiratory depression usually occurs after the event. Earlier detection would be beneficial in preventing increased morbidity and mortality of 0.01 % patients receiving analgesic opioids. Airway patency during inspiration requires vagal modulation. Regulation of the cardiovascular and respiratory centres may be coupled with a central mechanism that is indirectly measurable with heart rate variability (HRV). While opioids tend to increase parasympathetic tone, a decrease in airway stability could be due to a decrease in respiratory parasympathetic activity. Sympathetic arousal generated by apneic events may separately be recognised with short-term HRV. This pilot observational study examined the dynamic sympathovagal changes during fentanyl-midazolam induced respiratory depression on 10 subjects scheduled for minor surgery. A selection of HRV indices, able to work over sub-minute periods on non-stationary signals, were applied including a range of less common indices. Three analyses tested the effects post-fentanyl, preceding the first central depression, and preceding obstruction of the upper airway. Statistical significance was assessed with overlap of bootstrap percentile confidence intervals for the median. A decrease in total variability, Lomb Total using the Lomb-Scargle method, is a positive finding for short-term HRV use in this study. No significant change before critical respiratory events was observed in traditional, spectral power, respiratory or other indices. One index, PolVar20, indicated a burst of sympathetic activity preceding respiratory depression similar to sleep apnoea arousals that restore airway patency. Before its usefulness in early detection of airway tone can be determined, PolVar20 requires further work a statistical method for highly skewed distributions, auto adjustment for baseline variability, and detecting a range of sympathetic responses to apnea.