Performance of pre-hospital evaluations in ruling out invasive chest stab wounds

Diagnostic accuracy of clinical examination to identify life- and limb-threatening injuries in trauma patients

BACKGROUND

Timely and accurate identification of life- and limb-threatening injuries (LLTIs) is a fundamental objective of trauma care that directly informs triage and treatment decisions. However, the diagnostic accuracy of clinical examination to detect LLTIs is largely unknown, due to the risk of contamination from in-hospital diagnostics in existing studies. Our aim was to assess the diagnostic accuracy of initial clinical examination for detecting life- and limb-threatening injuries (LLTIs). Secondary aims were to identify factors associated with missed injury and overdiagnosis, and determine the impact of clinician uncertainty on diagnostic accuracy.

METHODS

Retrospective diagnostic accuracy study of consecutive adult (≥ 16 years) patients examined at the scene of injury by experienced trauma clinicians, and admitted to a Major Trauma Center between 01/01/2019 and 31/12/2020. Diagnoses of LLTIs made on contemporaneous clinical records were compared to hospital coded diagnoses. Diagnostic performance measures were calculated overall, and based on clinician uncertainty. Multivariate logistic regression analyses identified factors affecting missed injury and overdiagnosis.

RESULTS

Among 947 trauma patients, 821 were male (86.7%), median age was 31 years (range 16-89), 569 suffered blunt mechanisms (60.1%), and 522 (55.1%) sustained LLTIs. Overall, clinical examination had a moderate ability to detect LLTIs, which varied by body region: head (sensitivity 69.7%, positive predictive value (PPV) 59.1%), chest (sensitivity 58.7%, PPV 53.3%), abdomen (sensitivity 51.9%, PPV 30.7%), pelvis (sensitivity 23.5%, PPV 50.0%), and long bone fracture (sensitivity 69.9%, PPV 74.3%). Clinical examination poorly detected life-threatening thoracic (sensitivity 48.1%, PPV 13.0%) and abdominal (sensitivity 43.6%, PPV 20.0%) bleeding. Missed injury was more common in patients with polytrauma (OR 1.83, 95% CI 1.62-2.07) or shock (systolic blood pressure OR 0.993, 95% CI 0.988-0.998). Overdiagnosis was more common in shock (OR 0.991, 95% CI 0.986-0.995) or when clinicians were uncertain (OR 6.42, 95% CI 4.63-8.99). Uncertainty improved sensitivity but reduced PPV, impeding diagnostic precision.

CONCLUSIONS

Clinical examination performed by experienced trauma clinicians has only a moderate ability to detect LLTIs. Clinicians must appreciate the limitations of clinical examination, and the impact of uncertainty, when making clinical decisions in trauma. This study provides impetus for diagnostic adjuncts and decision support systems in trauma.

Complications associated with pre-hospital open thoracostomies: a rapid review

Background

Open thoracostomies have become the standard of care in pre-hospital critical care in patients with chest injuries receiving positive pressure ventilation. The procedure has embedded itself as a rapid method to decompress air or fluid in the chest cavity since its original description in 1995, with a complication rate equal to or better than the out-of-hospital insertion of indwelling pleural catheters. A literature review was performed to explore potential negative implications of open thoracostomies and discuss its role in mechanically ventilated patients without clinical features of pneumothorax.

Main findings

A rapid review of key healthcare databases showed a significant rate of complications associated with pre-hospital open thoracostomies. Of 352 thoracostomies included in the final analysis, 10.6% (n = 38) led to complications of which most were related to operator error or infection (n = 26). Pneumothoraces were missed in 2.2% (n = 8) of all cases.

Conclusion

There is an appreciable complication rate associated with pre-hospital open thoracostomy. Based on a risk/benefit decision for individual patients, it may be appropriate to withhold intervention in the absence of clinical features, but consideration must be given to the environment where the patient will be monitored during care and transfer. Chest ultrasound can be an effective assessment adjunct to rule in pneumothorax, and may have a role in mitigating the rate of missed cases.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13049-021-00976-1.

The contribution of helicopter emergency medical services in the pre-hospital care of penetrating torso injuries in a semi-rural setting

BACKGROUND

Although the merit of pre-hospital critical care teams such as Helicopter Emergency Medical Services (HEMS) has been universally recognized for patients with penetrating torso injuries who present with unstable physiology, the potential merit in patients initially presenting with stable physiology is largely undetermined. The ability to predict the required pre-hospital interventions patients may have important implications for HEMS tasking, especially when transport times to definitive care are prolonged.

METHODS

We performed a retrospective cohort study of patients who sustained a penetrating torso injury and were attended by the Air Ambulance Kent Surrey Sussex (AAKSS) over a 6-year period. Primary outcome was defined as the percentage of patients with penetrating torso injuries requiring HEMS-specific interventions anytime between HEMS arrival and arrival at hospital. Secondary outcomes were the association of individual patient- and injury characteristics with the requirement for HEMS interventions.

RESULTS

During the study period 363 patients met inclusion criteria. 90% of patients were male with a median age of 30 years. 99% of penetrating trauma incident occurred more than 10-min drive from a Major Trauma Centre (MTC). Presenting GCS was > 13 in 83% of patients. Significant hemodynamic- or ventilatory compromise was present in more than 25% of the patients. Traumatic cardiac arrest was present in 34 patients (9.4%), profound hypotension with SBP < 80 mmHg in 30 (8.3%) and oxygen saturations < 92% in 30 (8.3%). A total of 121 HEMS-specific interventions were performed. Although HEMS-specific interventions were associated with presenting physiology (TCA OR 1.75 [1.41-2.16], SBP < 80 mmHg (OR 1.40 [1.18-1.67] and SpO₂ < 92% (OR 1.39 [1.17-1.65], a minority of the patients presented initially with stable physiology but deteriorated on route to hospital and required HEMS interventions (n = 9, 3.3%).

CONCLUSION

HEMS teams provide potentially important contribution to the pre-hospital treatment of patients with penetrating torso injuries in rural and semi-rural areas, especially when they present with unstable physiology. A certain degree of over-triage is inevitable in these patients, as it is hard to predict which patients will deteriorate on route to hospital and will need HEMS interventions. The results of this study showing a potentially predictable geographical dispersion of penetrating trauma could inform multi-agency knife crime prevention strategy.

Development of a field artificial intelligence triage tool: Confidence in the prediction of shock, transfusion, and definitive surgical therapy in patients with truncal gunshot wounds

BACKGROUND

In-field triage tools for trauma patients are limited by availability of information, linear risk classification, and a lack of confidence reporting. We therefore set out to develop and test a machine learning algorithm that can overcome these limitations by accurately and confidently making predictions to support in-field triage in the first hours after traumatic injury.

METHODS

Using an American College of Surgeons Trauma Quality Improvement Program-derived database of truncal and junctional gunshot wound (GSW) patients (aged 16-60 years), we trained an information-aware Dirichlet deep neural network (field artificial intelligence triage). Using supervised training, field artificial intelligence triage was trained to predict shock and the need for major hemorrhage control procedures or early massive transfusion (MT) using GSW anatomical locations, vital signs, and patient information available in the field. In parallel, a confidence model was developed to predict the true-class probability (scale of 0-1), indicating the likelihood that the prediction made was correct, based on the values and interconnectivity of input variables.

RESULTS

A total of 29,816 patients met all the inclusion criteria. Shock, major surgery, and early MT were identified in 13.0%, 22.4%, and 6.3% of the included patients, respectively. Field artificial intelligence triage achieved mean areas under the receiver operating characteristic curve of 0.89, 0.86, and 0.82 for prediction of shock, early MT, and major surgery, respectively, for 80/20 train-test splits over 1,000 epochs. Mean predicted true-class probability for errors/correct predictions was 0.25/0.87 for shock, 0.30/0.81 for MT, and 0.24/0.69 for major surgery.

CONCLUSION

Field artificial intelligence triage accurately identifies potential shock in truncal GSW patients and predicts their need for MT and major surgery, with a high degree of certainty. The presented model is an important proof of concept. Future iterations will use an expansion of databases to refine and validate the model, further adding to its potential to improve triage in the field, both in civilian and military settings.

LEVEL OF EVIDENCE

Prognostic, Level III.