Level of agreement between prehospital and emergency department vital signs in trauma patients.

OBJECTIVES

Describe the level of agreement between prehospital (emergency medical service [EMS]) and ED vital signs in a group of trauma patients transported to an inner city Major Trauma Centre. We also sought to determine factors associated with differences in recorded vital sign measurements.

METHODS

All adult patients meeting trauma triage criteria and transported directly from scene of injury by New South Wales Ambulance to our institution were included. The primary outcome was the difference in vital signs: heart rate (HR), systolic blood pressure (SBP), respiratory rate (RR) and Glasgow Coma Scale (GCS), between ED and EMS recorded measurements. Agreement was assessed using intraclass correlation coefficients and enhanced Bland-Altman plots. Multivariable linear regression models were used to determine factors associated with vital sign differences.

RESULTS

The 1181 trauma patients met inclusion criteria. Intraclass correlation coefficients were as follows: GCS 0.74 (95% confidence interval [CI], 0.37, 1.12); HR 0.41 (95% CI, 0.30, 0.53); SBP 0.37 (95% CI, 0.27, 0.46); and RR 0.29 (95% CI, 0.06, 0.51). Bland-Altman derived 95% limits of agreement lay outside a priori limits of clinical agreement for SBP and RR and were within limits of clinical agreement for GCS and HR. SBP and HR differences were associated with prehospital airway and fluid intervention.

CONCLUSIONS

Agreement was demonstrated between EMS and ED GCS scores but not RR and SBP recordings. Discrepancies appeared to reflect physiological changes in response to EMS initiated interventions. Trauma triage algorithms and risk models might need to take these measurement differences, and factors associated with them, into account.