Impact of absence of critical respiratory rate change on oxygen desaturation following tracheal extubation after general anaesthesia: a propensity score-matched analysis.

Abstract

It has been suggested that oxygen desaturation may be paradoxically related to the absence of an abnormal respiratory rate (RR) during acoustic respiratory rate (RRa) monitoring in a postoperative setting. We retrospectively compared the incidence of desaruration in patients without an abnormal RR with that in patients with an abnormal RR using propensity scorer matching. We also explored the factors contributing to oxygen desaturation without an RR monitoring alert. We used ≤ 8 h postoperative data of the first 935 patients. Outcomes of patients with and without critical RR changes (RR > 30 or < 8 beats per min for > 2 min) (critical RR change vs. noncritical RR change) were first compared according to oxygen desaturation levels (SpO₂< 90% for > 10 s). Multivariate analysis was used to determine oxygen desaturation-associated explanatory factors. Propensity score matching yielded 259 patients without critical RR changes and 259 patients with critical RR changes, respectively. Oxygen desaturation rates were higher in patients without critical RR changes [noncritical RR change vs. critical RR change: 39/220 (15.1%) vs. 16/243 (6.2%)]. An odds ratio and 95% CI for the noncritical RR change was 2.56 (1.38-4.55, P = 0.002). A critical change in the RRa was not observed in 576 patients; of these, oxygen desaturation was observed in 76 (13.2%) patients. Surgery duration (OR, 1.018 per 10 min increase; 95% CI, 1.002 to 1.035) was independently associated with oxygen desaturation without critical RR change. Postoperative oxygen desaturation paradoxically occurred more frequently in patients without critical RR changes, whose RR was monitored by the RRa under oxygen therapy. The duration of surgery may explain the possibility of postoperative oxygen desaturation without an RRa device alert.