ABSTRACT
BACKGROUND: The detection of haemorrhage in trauma casualties may be delayed owing to compensatory mechanisms. This study aimed to evaluate whether the cardiovascular reserve index (CVRI) on arrival detects massive haemorrhage and predicts haemorrhage development in trauma casualties. METHODS: This was an observational prospective cohort study of adult casualties (≥18 years) who were brought to a single level-1 trauma centre, enroled upon arrival and followed until discharge. Vital signs were monitored on arrival, from which the CVRI and shock index were retrospectively calculated (blinded to the caregivers). The outcome measure was the eventual haemorrhage classification group: massive haemorrhage on arrival (MHOA) (defined by massive transfusion on arrival of ≥6 [O+] packed cells units), developing haemorrhage (DH) (defined by a decrease in haemoglobin >1 g/dl in consecutive tests), and no significant haemorrhage noted throughout the hospital stay. The means of each variable on arrival by haemorrhage group were evaluated using the analysis of variance. The authors evaluated the detection of MHOA in the entire population and the prediction of DH in the remainders (given that MHOA had already been detected and treated) by C-statistic predefined strong prediction by area under the curve (AUC) greater than or equal to 0.8, P less than or equal to 0.05. RESULTS: The study included 71 patients (after exclusion): males, 82%; average age 37.7 years. The leading cause of injuries was road accident (61%). Thirty-nine (54%) patients required hospital admission; distribution by haemorrhage classification: 5 (7%) MHOA, 5 (7%) DH, and 61 (86%) no significant haemorrhage. Detection of MHOA found a strong predictive model by CVRI and most variables (AUC 0.85-1.0). The prediction of DH on arrival showed that only lactate (AUC=0.88) and CVRI (0.82) showed strong predictive model. CONCLUSIONS: CVRI showed a strong predictive model for detection of MHOA (AUC>0.8) as were most other variables. CVRI also showed a strong predictive model for detection of DH (AUC=0.82), only serum lactate predicted DH (AUC=0.88), while all other variables were not found predictive. CVRI has advantages over lactate in that it is feasible in pre-hospital and mass casualty settings. Moreover, its repeatability enables detection of deteriorating trend. The authors conclude that CVRI may be a useful additional tool in the evaluation of haemorrhage.
