Background: An increase in mortality has been documented in association with paramedic rapid sequence intubation (RSI) of severely head-injured patients. This analysis explores the impact of hypoxia and hyperventilation on outcome.
Methods: Adult severely head-injured patients (Glasgow Coma Scale score of 3-8) unable to be intubated without neuromuscular blockade underwent paramedic RSI using midazolam and succinylcholine; rocuronium was administered after confirmation of tube position. Standard ventilation parameters were used for most patients; however, one agency instituted use of digital end-tidal carbon dioxide (ETCO2) and oxygen saturation (Spo2) monitoring during the trial. Each patient undergoing digital ETCO2/Spo2 monitoring was matched to three historical nonintubated controls on the basis of age, gender, mechanism, and Abbreviated Injury Scale scores for each of six body regions. Logistic regression was used to explore the impact of oxygen desaturation during laryngoscopy and postintubation hypocapnia and hypoxia on outcome. The relationship between hypocapnia and ventilatory rate was explored using linear regression and univariate analysis. In addition, trial patients and controls were compared with regard to mortality and the incidence of "good outcomes" using an odds ratio analysis.
Results: Of the 426 trial patients, a total of 59 had complete ETCO2/Spo2 monitoring data; these were matched to 177 controls. Logistic regression revealed an association between the lowest ETCO2 value and final ETCO2 value and mortality. Matched-controls analysis confirmed an association between hypocapnia and mortality. A statistically significant association between ventilatory rate and ETCO2 value was observed (r = -0.13, p < 0.0001); the median ventilatory rate associated with the lowest recorded ETCO2 value was significantly higher than for all other ETCO2 values (27 mm Hg vs. 19 mm Hg, p < 0.0001). In addition, profound desaturations during RSI and hypoxia after intubation were associated with higher mortality than matched controls. Overall mortality was 41% for trial patients versus 22% for matched controls (odds ratio, 2.51; 95% confidence interval, 1.33-4.72; p = 0.004).
Conclusions: Hyperventilation and severe hypoxia during paramedic RSI are associated with an increase in mortality.