Prehospital Identification of Stroke Subtypes in Chinese Rural Areas

Development and internal validation of prehospital prediction models for identifying intracerebral haemorrhage in suspected stroke patients

Introduction

Distinguishing patients with intracerebral haemorrhage (ICH) from other suspected stroke cases in the prehospital setting is crucial for determining the appropriate level of care and minimising the onset-to-treatment time, thereby potentially improving outcomes. Therefore, we developed prehospital prediction models to identify patients with ICH among suspected stroke cases.

Methods

Data were obtained from the Field Administration of Stroke Therapy-Magnesium prehospital stroke trial, where paramedics evaluated multiple variables in suspected stroke cases within the first 2 hours from the last known well time. A total of 19 candidate predictors were included to minimise overfitting and were subsequently refined through the backward exclusion of non-significant predictors. We used logistic regression and eXtreme Gradient Boosting (XGBoost) models to evaluate the performance of the predictors. Model performance was assessed using the area under the receiver operating characteristic curve (AUC), confusion matrix metrics and calibration measures. Additionally, models were internally validated and corrected for optimism through bootstrapping. Furthermore, a nomogram was built to facilitate paramedics in estimating the probability of ICH.

Results

We analysed 1649 suspected stroke cases, of which 373 (23%) were finally diagnosed with ICH. From the 19 candidate predictors, 9 were identified as independently associated with ICH (p<0.05). Male sex, arm weakness, worsening neurological status and high systolic blood pressure were positively associated with ICH. Conversely, a history of hyperlipidaemia, atrial fibrillation, coronary artery disease, ischaemic stroke and improving neurological status were associated with other diagnoses. Both logistic regression and XGBoost demonstrated good calibration and predictive performance, with optimism-corrected sensitivities ranging from 47% to 49%, specificities from 89% to 90% and AUCs from 0.796 to 0.801.

Conclusions

Our models demonstrate good predictive performance in distinguishing patients with ICH from other diagnoses, making them potentially useful tools for prehospital ICH management.

Prehospital identification of intracerebral haemorrhage: a scoping review of early clinical features and portable devices

INTRODUCTION

Prehospital identification of intracerebral haemorrhage (ICH) in suspected stroke cases may enable the initiation of appropriate treatments and facilitate better-informed transport decisions. This scoping review aims to examine the literature to identify early clinical features and portable devices for the detection of ICH in the prehospital setting.

METHODS

Three databases were searched via Ovid (MEDLINE, EMBASE and CENTRAL) from inception to August 2022 using prespecified search strategies. One reviewer screened all titles, abstracts and full-text articles for eligibility, while a second reviewer independently screened 20% of the literature during each screening stage. Data extracted were tabulated to summarise the key findings.

RESULTS

A total of 6803 articles were screened for eligibility, of which 22 studies were included for analysis. Among them, 15 studies reported on early clinical features, while 7 considered portable devices. Associations between age, sex and comorbidities with the presence of ICH varied across studies. However, most studies reported that patients with ICH exhibited more severe neurological deficits (n=6) and higher blood pressure levels (n=11) at onset compared with other stroke and non-stroke diagnoses. Four technologies were identified for ICH detection: microwave imaging technology, volumetric impedance phase shift spectroscopy, transcranial ultrasound and electroencephalography. Microwave and ultrasound imaging techniques showed promise in distinguishing ICH from other diagnoses.

CONCLUSION

This scoping review has identified potential clinical features for the identification of ICH in suspected stroke patients. However, the considerable heterogeneity among the included studies precludes meta-analysis of available data. Moreover, we have explored portable devices to enhance ICH identification. While these devices have shown promise in detecting ICH, further technological development is required to distinguish between stroke subtypes (ICH vs ischaemic stroke) and non-stroke diagnoses.

Clinical features and novel technologies for prehospital detection of intracerebral haemorrhage: a scoping review protocol

INTRODUCTION

The detection of intracerebral haemorrhage (ICH) in the prehospital setting without conventional imaging technology might allow early treatment to reduce haematoma expansion and improve patient outcomes. Although ICH and ischaemic stroke share many clinical features, some may help in distinguishing ICH from other suspected stroke patients. In combination with clinical features, novel technologies may improve diagnosis further. This scoping review aims to first identify the early, distinguishing clinical features of ICH and then identify novel portable technologies that may enhance differentiation of ICH from other suspected strokes. Where appropriate and feasible, meta-analyses will be performed.

METHODS

The scoping review will follow the recommendations of the Joanna Briggs Institute Methodology for Scoping Reviews as well as the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews checklist. A systematic search will be conducted using MEDLINE (Ovid), EMBASE (Ovid) and CENTRAL (Ovid). EndNote reference management software will be used to remove duplicate entries. Two independent reviewers will screen titles, abstracts and full-text reports according to prespecified eligibility criteria using the Rayyan Qatar Computing Research Institute software. One reviewer will screen all titles, abstracts and full-text reports of potentially eligible studies, while the other reviewer will independently screen at least 20% of all titles, abstracts and full-text reports. Conflicts will be resolved through discussion or by consulting a third reviewer. Results will be tabulated in accordance with the scoping review's objectives along with a narrative discussion.

ETHICS AND DISSEMINATION

Ethical approval is not required for this review, as it will only include published literature. The results will be published in an open-access, peer-reviewed journal, presented at scientific conferences and form part of a PhD thesis. We expect the findings to contribute to future research into the early detection of ICH in suspected stroke patients.

Diagnostic accuracy of clinical tools for assessment of acute stroke: a systematic review

INTRODUCTION

Recanalisation therapy in acute ischaemic stroke is highly time-sensitive, and requires early identification of eligible patients to ensure better outcomes. Thus, a number of clinical assessment tools have been developed and this review examines their diagnostic capabilities.

METHODS

Diagnostic performance of currently available clinical tools for identification of acute ischaemic and haemorrhagic strokes and stroke mimicking conditions was reviewed. A systematic search of the literature published in 2015-2018 was conducted using PubMed, EMBASE, Scopus and The Cochrane Library. Prehospital and in-hospital studies with a minimum sample size of 300 patients reporting diagnostic accuracy were selected.

RESULTS

Twenty-five articles were included. Cortical signs (gaze deviation, aphasia and neglect) were shown to be significant indicators of large vessel occlusion (LVO). Sensitivity values for selecting subjects with LVO ranged from 23 to 99% whereas specificity was 24 to 97%. Clinical tools, such as FAST-ED, NIHSS, and RACE incorporating cortical signs as well as motor dysfunction demonstrated the best diagnostic accuracy. Tools for identification of stroke mimics showed sensitivity varying from 44 to 91%, and specificity of 27 to 98% with the best diagnostic performance demonstrated by FABS (90% sensitivity, 91% specificity). Hypertension and younger age predicted intracerebral haemorrhage whereas history of atrial fibrillation and diabetes were associated with ischaemia. There was a variation in approach used to establish the definitive diagnosis. Blinding of the index test assessment was not specified in about 50% of included studies.

CONCLUSIONS

A wide range of clinical assessment tools for selecting subjects with acute stroke has been developed in recent years. Assessment of both cortical and motor function using RACE, FAST-ED and NIHSS showed the best diagnostic accuracy values for selecting subjects with LVO. There were limited data on clinical tools that can be used to differentiate between acute ischaemia and haemorrhage. Diagnostic accuracy appeared to be modest for distinguishing between acute stroke and stroke mimics with optimal diagnostic performance demonstrated by the FABS tool. Further prehospital research is required to improve the diagnostic utility of clinical assessments with possible application of a two-step clinical assessment or involvement of simple brain imaging, such as transcranial ultrasonography.