Assessing the prediction of arterial CO₂ from end tidal CO₂ in adult blunt trauma patients.

Abstract

The control of PaCO₂ in ventilated patients is known to be of particular importance in the management and prognosis of trauma patients. Although EtCO₂ is often used as a continuous, non-invasive, surrogate marker for PaCO₂ in ventilated trauma patients in the emergency department (ED), previous studies suggest a poor correlation in this cohort. However, previous data has predominantly been collected retrospectively, raising the possibility that the elapsed time between PaCO₂ sampling and EtCO₂ recording may contribute to the poor correlation. As such this study aimed to analyse the correlation of PaCO₂ to EtCO₂ in the ventilated blunt trauma patient presenting to the ED through contemporaneous sampling. This study was conducted as a prospective observational study analysing the near simultaneous recording of EtCO₂ and Arterial Blood Gas sampling of ventilated adult trauma patients in the ED of a Level 1 trauma centre over a 12-month period. Data was analysed using linear regression and subgroup analysis by Injury Severity Score (ISS) and Abbreviated Injury Score (AIS) of the Chest. Linear regression of EtCO₂ vs PaCO₂ demonstrated a moderate correlation with r = 0.54 (p < 0.01, n = 51, 95 % CI 0.31-0.71). Subgroup analysis by ISS, revealed a stronger correlation in those with minor ISS (0-11) (r = 0.76, p < 0.01, n = 13, 95 % CI 0.36-0.92) compared to those more severely injured patients (ISS > 15) (r = 0.44, P < 0.01, n = 38, 95 % CI 0.14-0.67). Analysis by AIS Chest demonstrated similar correlation between patients without chest injuries (AIS 0) (r = 0.55, n = 29, p < 0.01, 95 % CI 0.23-0.76) and those with an AIS >1 (r = 0.51, n = 22, p = 0.02, 95 % CI 0.11-0.77). In patients with traumatic head injuries who had an EtCO₂ between 30 and 39 mmHg, only 57 % had a measured PaCO₂ within 5 mmHg. As patients transition from minor to seriously injured, a decreasing strength of PaCO₂ to EtCO₂ correlation is observed, decreasing the reliability of EtCO₂ as a surrogate marker of PaCO₂ in this patient group. This inconsistency cannot be accounted for by the presence of chest injuries and worryingly is frequently seen in those with traumatic brain injuries.