Assessing the prediction of arterial CO 2 from end tidal CO 2 in adult blunt trauma patients.

BACKGROUND: 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.

METHODS: 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.

RESULTS: 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.

CONCLUSIONS: 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.