Life After Transient Ischemic Attack

Background:

In Denmark approximately 4.000 patients every year are diagnosed with Transient Ischemic Attack (TIA). In Denmark the time-based definition of TIA is used and based on a clinical assessment with acute, often focal neurological symptoms of vascular origin resolving completely within 24 hours. Emerging evidence suggests however, that 1/3 of patients experience lasting symptoms as cognitive impairments, depression, anxiety and/or fatigue, which may impact their ability to return to their normal everyday life. Patients' own perception of a complex and life-altering experience has been identified in the qualitative literature. Some patients experiencing lasting symptoms and alterations in multiple life domains like a feeling of uncertainty, vulnerability, instability and change. According to this review, some patients reported that physical and psychosocial symptoms were not recognized, nor acknowledged by their healthcare professionals. Education for clinicians was recommend to enable them to understand these difficulties, improve interaction and meet the patients' needs.

It is unknown which patients are at increased risk of experiencing the described lasting symptoms when these symptoms arise and how it affects quality of life and return to everyday life. Answers to these questions will allow the design and test of an evidence-based and patient-centered intervention. The ultimate goal will be the creation of an evidence-based guide on how to follow-up and intervene in patients with TIA to ensure the return to their everyday life.

Aim:

The aim of this study is to investigate self-reported lasting symptoms after TIA, timing of these symptoms as well as their potential resolution and characterize patients who have high risk for developing these symptoms.

Furthermore, the socio-economic consequences of TIA will be investigated.

Methods:

The study is designed as a prospective cohort study using mixed methods. As used in the clinical settings in Denmark the time-based definition of TIA will be applied in this study. The diagnosis is defined by acute focal neurological symptoms of vascular origin resolving completely within 24 hours, thereby allowing acute ischemia on imaging.

Recruitment:

All patients diagnosed with TIA from the Stroke Unit, Aalborg University Hospital will be screened for eligibility. Inclusion will follow a twostep approach; 1) using clinical and demographic information from medical records, 2) receiving questionnaires at baseline, 3, 6 and 12 months follow-up.

Baseline characteristics and potential prognostic factors:

A pre-TIA modified Rankin Score (mRS) will be collected to describe the pre-TIA level of function for the included patients.

To investigate which patients are at increased risk of experiencing the lasting symptoms multiple potential prognostic factors will be registered at inclusion. In addition to age and gender these include the following

• Alcohol consumption (<7, 7-14 or >14 units/week) and/or use of narcotics on a weekly basis
• Smoking
• Housing (living in own home, nursing home or other)
• Civil status (living alone, cohabitant (with partner, children, other))
• Cardiovascular risk factors as registered in the Danish Stroke Registry
• Comorbidities in form of other chronic diseases (these will be categorized in neurodegenerative, psychiatric, other)
• Previous stroke
• Previous TIA
• Previous diagnosis or use of medication for depression and/or anxiety
• Previous use of antihypertensive, antiplatelet, oral anticoagulation, statins, antidiabetics
• Acute infarction on imaging
• Highest finished education and corresponding ISCED level.
• Employment status (employed >30 hours/week, employed <30 hours/week, unemployed, retired, other)
• ABCD2 stroke risk stratification.

All interventions initiated after discharge based on the TIA will also be registered.

The diagnosis of TIA in clinical settings is challenging. Therefore, after inclusion of patients the diagnosis of TIA will be verified independently by two stroke neurologists. Disagreement will be resolved by a third senior stroke neurologist. This verification will categorize patients in per protocol confirmed TIA versus non-confirmed TIA allowing for a subgroup analysis of outcome according to this categorization. The verification will be guided by the calculation of The Recognition of Stroke in the Emergency Room (ROSIER) score.

Considerations on sample size:

In the last 5 years, the incidence of TIA registered at the Stroke Unit, Aalborg University Hospital ranged from 369 to 463 patients per year. The content of the project is both to explore non-focal lasting symptoms in patients with TIA as well as prognostic factors, which has not been done previously to this extend. Thus, a specific sample size calculation is therefore not feasible as to many factors are unclear. Based on the average annual admission rate the goal is to include 350-400 patients. There is a time constraint of inclusion for a total of 12 months. This sample-size is deemed sufficient to be able to describe the self-report outcomes across 12 months and provide preliminary data on the proportion of patients in high risk of a poor prognosis.

Data management and statistical analysis:

Data will be handled in REDCap and analyzed in STATA. Baseline descriptive data will be presented as medians and interquartile ranges or frequencies and percentages where appropriate. Missing data will be handled with multiple imputation with chained equations to impute the missing data.

To investigate the development of the included outcomes over time (baseline, 3 months, 6 months and 12 months) a longitudinal regression model will be used. For the primary effect measure, we will implement linear regression, while for supplementary effect measures we will use ordered logistic regression. Because of the repeated measures design the model will be a mixed effects model with a random intercept and fixed treatment variables. After discharge each patient can receive different follow-up and treatment interventions and the implementation of these varies over time. For the primary analysis patients will be assigned to each of the two main treatment tracks at baseline. An additional time-varying variable which indicates the progress of the two treatment tracks categorized as "no further treatment" or "further treatment" will be included in the model. For the subsequent treatment categories, we will conduct similar secondary analyses, but stratified by main treatment arm. For all analyses, the model will include time dependent covariates, thus, potential prognostic factor and baseline characteristic adjustment will be conducted using Marginal Structure models. Study results will be presented with 95% confidence intervals (CI).

To compare results with the existing literature, we will estimate mean and median outcomes with 95% CI for each of the follow-up timepoints

The qualitative semi-structured interviews will be handled in NVivo. An explorative and inductive thematic analysis will be performed by two researchers.