Detection of intracerebral haemorrhage in the prehospital setting

Abstract

Aim: This thesis aimed to improve the prehospital identification of intracerebral haemorrhage (ICH) in patients with suspected stroke.

Methods: A scoping review was conducted to identify early clinical features and portable diagnostic devices capable of distinguishing ICH from other causes of stroke symptoms in prehospital settings. This was followed by a systematic review of existing prehospital prediction models for ICH, assessing their development, predictive variables and performance. Subsequently, prehospital data were used to develop and validate new prediction models for ICH detection. The identified predictors were then evaluated in combination with glial fibrillary acidic protein (GFAP) using clinical data.

Results: The scoping review highlighted several clinical features that may help distinguish ICH from other suspected stroke cases. Promising portable devices, including microwave and ultrasound imaging, were identified; however, their capabilities and application in prehospital settings remain limited. The systematic review revealed substantial methodological limitations in existing prehospital prediction models, rendering them unreliable for practical use. Consequently, new prehospital prediction models were developed using logistic regression and a machine learning approach. These models incorporated nine predictors and demonstrated good internal validity, with optimism-corrected area under the curve (AUC) values of approximately 0.80. Calibration measures showed good agreement between predicted and observed outcomes. The combination of the identified predictors with GFAP achieved better performance, with an optimism-corrected AUC of 0.90.

Conclusions: Prehospital care plays a pivotal role in determining outcomes for patients with ICH, as timely identification, treatment and transfer to an appropriate facility are critical. This thesis demonstrates the potential to distinguish ICH from other suspected stroke cases in the prehospital setting. The findings may ultimately enhance prehospital triage, reduce transport delays and minimise onset-to-treatment times for patients with ICH. Moreover, this work addresses important gaps in prehospital stroke care and provides a foundation for future research to refine and implement these strategies in practice, potentially leading to better outcomes for this high-risk stroke subtype.