[Preclinical prediction of prehospital injury severity by emergency physicians : approach to evaluate validity]

The prognosis of polytraumatized patients is basically dependent on the quality of emergency room (ER) management and a smooth transition from prehospital emergency therapy to ER therapy is essential. The accurate prediction of the prehospital injury severity by emergency physicians influences prehospital therapy and level of care of the destination hospital. Furthermore it helps to provide medical resources on time. Overestimation of injury severity wastes resources, underestimation puts patients at risk. Prehospital misjudgement of injury severity is common. The aim of this study was to evaluate reliability of the injury severity estimated by emergency physicians.

For comparison of the prehospital and hospital injury severity the Injury Severity Score (ISS) and Trauma-ISS (TRISS) were calculated. The TRISS consists of the ISS and the Revised Trauma Score (RTS). All diagnoses of the prehospital and admission charts were collected and an injury severity was allocated according to the Abbreviated Injury Scale (AIS). The concordance of prehospital and hospital injury severity at different ranges and according to different body regions was evaluated. A difference of more than 25% between the prehospital injury severity and the injury severity calculated after ER diagnostics was considered as being relevant and judged as overestimation or underestimation. The documented injury severity in the emergency physician protocol was judged as detailed, satisfactory and poor.

Of the patients 73% reached the ER during on-call hours. The mean ER-ISS was 19 (1-50). At a range of ±25% referring to the ER-ISS, 30% overestimation and 36% underestimation of the prehospital injury severity was observed. A concordance of 34% was found. At a range of ±50% the concordance between the prehospital injury severity and the injury severity calculated after ER diagnostics was 57%, at a range of ±75% the concordance was 73%. The mean ER-TRISS was 6.9 points (0.3-98.6) and the mean ER-RTS was 7.569 points (0-7.841). Using the TRISS with a range of ±25% a concordance of 28% was observed. A high concordance of the prehospital and hospital injury severity was found in the region of the face (70%) and external soft tissue injuries (80%). The concordance in the body region of the abdomen was 55%, of the thorax 40%, of the extremities and pelvis 37% and of the head 33%. Underestimation in the region of the abdomen was 32%, of the head 37%, of the thorax 42% and of the extremities and pelvis 47%. Missed injuries were the reason for underestimation in the body region of extremities and pelvis in half of the cases. Of the patients 61% suffered a traffic accident, 25% a fall of less than 3 m and 8% of more than 3 m. In 5% of the cases other mechanisms of injury were observed. Injury severity was documented in a detailed manner in 61% and satisfactory in 26%.

The prediction of prehospital injury severity is difficult and less reliable. Relevant underestimation of injury severity was observed in visceral cavities. In order to evaluate injury severity the use of anatomical trauma scores alone might be not sufficient. In addition, the mechanism of injury and the deduced consequences, such as prehospital therapy, the choice of destination hospital and the need of ER treatment should be taken into account.