Canadian TIA Score Versus ABCD2 Score in Risk Stratification of Transient Ischemic Attack (TIA)

Rationale: Patients with transient ischemic attack (TIA) are at high risk of developing a subsequent stroke, especially in the short term. Identifying patients at greater risk allows for optimized management to provide aggressive treatment in the early days following the event, a strategy that is not feasible for the entire patient cohort. Over time, several prognostic scores have been studied, among which the most widely used is the ABCD2 score; however, this score has not proven particularly reliable in distinguishing between low-risk and high-risk patients in prospective validation studies. More recently, the Canadian TIA score has been proposed, but it has only been validated in the Canadian population.

Study design: Patients with TIA attending the Emergency Department of IRCCS AOU di Bologna Policlinico di S.Orsola will be included, and they will be admitted to the Emergency Medicine department according to the currently implemented protocol. At the time of admission, the various prognostic scores relevant to the study (Canadian TIA score, ABCD2, and ABCD3) will be completed by the physician during the standard medical assessment. Patients will also be contacted by phone at 7 and 90 days.

The study is observational and non-interventional. Patients will be treated according to standard clinical practice based on the physician's judgment and the information provided in the Technical Data Sheet of each individual product for any concomitant therapies administered according to clinical practice.

Objectives: The study aims to evaluate the superiority of the Canadian TIA score compared to the ABCD2 score in stratifying the risk of stroke or carotid revascularization following TIA at 7 days, and to assess its superiority at 90 days in a European population. Additionally, secondary objectives will include evaluating the superiority of the Canadian score over the ABCD2 score in stratifying the risk of major cerebrovascular events following TIA at 7 and 90 days across different subpopulations (age \< or ≥ 65 years; TIA etiology according to the TOAST classification) and assessing the superiority of the Canadian score compared to the ABCD2 score in stratifying the risk of major cerebrovascular events following TIA at both 7 and 90 days.

Outcomes: The primary outcome will be composite: the occurrence of stroke or carotid revascularization (endarterectomy or stenting) within 7 days of emergency department admission. Stroke will be defined as the sudden onset of neurological symptoms that persist for more than 24 hours or until death, in the absence of apparent non-vascular causes. A secondary outcome will be the occurrence of a stroke within 7 days of emergency department admission, regardless of carotid revascularization, and at 90 days. Outcomes will be assessed through electronic medical records and emergency department reports. All patients will be contacted at 7 and 90 days to evaluate any subsequent TIA or stroke using the validated questionnaire "Questionnaire for Verifying Stroke-Free Status."

Population: Adult patients with TIA attending the Emergency Department of IRCCS AOU di Bologna Policlinico di S.Orsola will be included and admitted to the Emergency Medicine department according to the currently implemented protocol. Recruitment is expected to last 24 months, followed by 12 months for data processing.

Statistical Analysis: The characteristics of patients enrolled in the study will be summarized in tables. Descriptive statistical analyses will be conducted to describe the collected data: continuous variables will be described using means, standard deviations, minimum and maximum values, and percentiles; discrete or nominal variables will be summarized using relative frequency percentages. The classification performance of the Canadian TIA Score will be calculated for each risk category (low, medium, and high) using interval odds ratios with 95% confidence intervals. The interval odds ratio is the multilevel extension of positive and negative odds ratios, which are applicable only when a test/risk score is dichotomized into two levels (positive vs. negative or high risk vs. low risk). We pre-specified risk thresholds (low risk \<1%, medium risk 1-5%, and high risk >5%) based on literature reports. Sensitivity and specificity will be calculated for dichotomized scores for each integer value from -3 to 23. The classification of ABCD2 and ABCD3 scores will also be calculated for both primary and secondary outcomes based on their ability to classify patients as low, medium, and high risk, using cutoff points based on the same risk thresholds (low risk \<1%, medium risk 1-5%, and high risk >5%). The C statistic (area under the curve) will then be compared using discrete values for each score, employing the DeLong method to verify the significance of these differences. Finally, net reclassification improvement indices will be calculated using three levels (low, medium, and high risk), comparing the Canadian TIA Score with the ABCD2 and ABCD3 scores.