Acute stroke occurs in approximately 13.000 persons every year in Denmark, 10 - 15 % now receives intervenous thrombolytic therapy, which remains the most important acute treatment in ischaemic stroke. For more than a decade there has been an ongoing discussion if Computerized Tomography (CT) or Magnetic Resonance Imaging (MRI) were best before thrombolysis: Magnetic Resonance Imaging is superior in visualising ischaemia, but Computerized Tomography is quicker and more easily applicable.

In the investigators centre primary imaging in work up of acute stroke during working hours will alternate between Computerized Tomography and Magnetic Resonance Imaging days based on a fixed calendar for a 24 months period as a quality development project.

This study is planned to include patients who have acute stroke imagining during this period, a total of 600 patients is expected. The investigators will compare door-to-needle time, patient safety, quality of imaging, patients' experience, physicians' decision certainty, and use of recourses.

Background: Stroke remains the third leading cause of death and the first leading cause of handicap in adult age in high-income countries even now when causal revascularisation therapy (intervenous (i.v) thrombolysis and endovascular therapy (EVT) has become generally available. I.v. thrombolysis has been implemented as routine treatment within 4.5 hours of symptom onset based on evidence from randomised controlled trials. Treatment benefit is highly dependent on time from symptom onset to treatment and is therefore strongly related to organisational issues including selection of imaging modality. I.v thrombolysis registration was based on trials using non-contrast Computerized Tomography (CT). In, general however, Magnetic Resonance Imaging (MRI) is considered to imply a higher standard of care. MRI is considered to give more detailed information and improve patient safety. The issue if CT or MRI is 'best' in i.v. thrombolysis work-up has been discussed by stroke scientists for more than a decade, both nationally and internationally. MRI is superior in detecting acute ischaemia and CT is quicker; but the effect of choice of imaging modality on overall efficacy and safety remains to be determined in spite of completion of more than 74.000 i.v. thrombolysis treatments in Europe according to the SITS-MOST registry. Access to primary work up by MRI is frequently regarded as a token of high treatment quality; nevertheless, few centres worldwide are really primarily MRI based. CT is much faster and has no contraindications. This may be crucial as the efficacy of thrombolysis decreases with time to treatment - numbers needed to treat increases with 1 for every 20 minutes, making the question if MRI is only of academic interest a relevant question.

The aim of this study is to determine if choice of primary imaging modality (CT vs MRI) affects efficacy and safety of i.v. thrombolysis in open quasi-randomised design, where imaging allocation was allocated based on the date of admission. The following items will be compared:

1. Radiological exclusion of other causes of symptoms than acute cerebral ischemia
2. Contraindications to scanning method in terms of contrast and magnetism
3. Delay from admission to thrombolytic treatment in minutes
4. Radiological imaging of diagnostic quality
5. Identification of stroke mechanism according to TOAST classification
6. Visual Analogue Scale of patient experience of the admission radiological examination
7. Visual Analogue Scale of the treating physician's experience of decision support prior to prescription of thrombolytic treatment
8. Deviation from radiological Standard Operational Plan
9. Time consumption for radiologists and radiographers in minutes

Method: Patients admitted with acute stroke during office hours and possibly also during daytime in weekends in Bispebjerg University Hospital will be included into the study during a 24 months study period. Based on activity in 2011, the investigators expect a study population of 600 patients. Door to needle time is the parameter/endpoint strongest associated with prognosis and therefore the most relevant from the patient's point of view. This estimate is well within the planned sample size. However, the sample size is extended in order to allow for analysis of less well-defined endpoints including patient experience.

• Study procedures: Patients are evaluated according to usual standard neurological outcome (National Institute of Health Stroke Scale (NIHSS)) performed on admission and on Day 1. Risk factors and co-morbidities are recorded and the stroke mechanism evaluated. The patients or carers are contacted by telephone after approximately 3 months. Recurrence of stroke, Transient Ischemic Attack (TIA), Myocardial Infarction (MI) and cardiac revascularization procedures are registered. If the patient is diseased, time of death and cause of death will be extracted from patient files. The functional outcome is assessed by the modified Rankin Scale (mRS) (including death).
• Patient experience of scan: The patient is asked day 1 about their experience of the scanning procedure by use of visual analogue scales.
• The physicians' certainty of having made the right decision concerning thrombolysis: The doctor notes by using a visual analogue scale (0-10) the strength of the clinical decision to give or not give tissue Plasminogen Activator (tPA) immediately after the bolus, or after the patient is informed that is not indicated.
• Resource use: Time from arrival until treatment initiation are recorded. Potential deviation of resource consumption when derogating from the SOP will be accounted for.
• Data Analysis: Imaging is evaluated during the acute phase. Further, the diagnostic imaging will be systematically and retrospectively evaluated by two senior neuroradiological consultants that are blinded to clinical information except the side of symptoms. X2-test, t-test/Mann-Whitney-test will be used depending on the distribution of the material, as well as uni and multivariate logistic regression analysis. It will be tested if the age, gender and severity affect the results. Data relating to risk factors and co-morbidity will be used to characterize the patient population. Function after 3 months expressed as mRS will be assessed, but the study was not designed to be able to detect an eventual difference in level of functioning after 3 months.

Ethical considerations: All patients are worked up according to best clinical practice and in accordance with Danish, European and American guidelines. As study procedures are added an interview on patient experience with the radiological procedure and a follow-up MRI, which is not done in case of contraindications to MRI or at patient's request. Further data on risk factors, the stroke event and follow up is recorded. There is no known risk related to MRI in persons with no magnetic metal in the body (e.g. a pacemaker); this is excluded by the same procedures as in clinical practice. The confined space and the noise of the MRI-scanner may cause anxiety in some; this will be prevented and handled by experienced radiographers as in clinical practice. Overall, there is no expected risks to participation in the study and some chance of treatment benefit by additional follow-up MRI. The study may have considerable impact on future treatment in patients with acute stroke as it may provide clarification to which of the methods should be future primary strategy. Patient experience as well as physicians experience is included in the analyses.

Economy: The study is fully financed by a grant from Tryg Fonden. The applicant and research responsible has no other relations to the foundation.

Importance and relevance of the study: If the choice of imaging method significantly affects predictors of good outcome after thrombolytic therapy initial use of the superior method would lead to a benefit, which for the individual patient will correspond to less disability after a stroke. Study results can be implemented directly in clinical practice; thought it will be subject to an eventual need for revised allocation of resources even though the total number of scans will not be affected. Based on the study's registration of resource consumption, the health economic effects of the study can be estimated. Since stroke is a very common and serious disorder, relatively minor improvements in the prognosis may have noticeable effect on group or community level. Further, this study may close a long-standing and sometimes emotional scientific discussion, setting as an example that trials remain the only valid method of answering scientific questions. This set up with a research protocol complementing a quality project enables us to evaluate this issue.

Inclusion Criteria:

• Clinical suspicion of stroke <4.5 hours
• National Institute of Health Stroke Scale (NIHSS) ≥ 1
• Admission to Bispebjerg University Hospital (Copenhagen, Denmark) on even days in the daytime
• Informed consent by patient or proxy

Exclusion Criteria:

• Patients in whom the stroke diagnosis is refuted on arrival
• Patients not providing informed consent