Early specialist assessment of transient ischaemic attack (TIA) reduces the risk of stroke and death (Giles and Rothwell, 2007; Johnson et al, 2007; Lavellée et al, 2007; Rothwell et al, 2007). The Royal College of Physicians (RCP) and National Institute for Health and Care Excellence (NICE) have guidelines for the management of TIA (RCP, 2010; 2016; NICE, 2017). This includes guidance related to the use of the Face Arms Speech Test (FAST) to exclude stroke (Dombowski et al, 2015).
TIA is defined as:
‘a transient episode of neurologic dysfunction caused by focal brain, spinal cord or retinal ischemia without acute infarction’ (Easton et al, 2009)
The RCP (2016) recommends that ‘referrers’ discontinue the triage of patients suspected to be experiencing a TIA with risk-stratification tools, and treat all patients as high-risk.
Practice in the UK is to convey patients who call 999 with suspected TIA (i.e. their neurological symptoms have resolved by the time of attendance by paramedics) to an emergency department (ED). These patients are subsequently referred by hospital clinicians to TIA clinics, and ABCD2 scores from the ED can guide the urgency of review.
Conveyance to EDs may be an unnecessary step for some patients with uncomplicated TIA (or for those displaying mimic symptoms) in terms of time, patient anxiety, and unnecessary cost the NHS. There is a need to evaluate alternative pathways via which patients can be directly referred by paramedics to TIA clinics.
The authors aimed to describe referral pathways available in the UK for the pre-hospital care of patients with suspected TIA in terms of their clinical and operational details, usage, and clinical and cost-effectiveness.
Methods
The survey was presented to UK ambulance services at a National Ambulance Research Steering Group meeting. A follow-up email invitation was sent, inviting participation in the survey. xsxsThe authors requested details of initiatives with relevant documentation, if available, and for the contact details of someone with good knowledge of the initiatives who could be contacted for more information. Ethical approval was not required as only health professionals took part; patient care was not altered, nor were patients contacted.
Findings
Of the 13 UK ambulance services, 12 (92%) responded to the survey over, approximately, a 2-month period in 2016. Nine (69%) reported that they had no current TIA referral pathway. Three (23%) reported currently using a pathway (Ambulance Service (AS) 1, 2, and 3). Only one (8%) previously used a pathway which was discontinued owing to reconfiguration within the service (AS4). Although protocol compliance was audited in an initial pilot in one service (AS1), no formal evaluation of usage, safety or effectiveness was reported by any AS. AS4 has been included in the results, though it should be noted that it had been discontinued when this survey took place.
Assessment of suspected TIA patients
All four pathways used:
Patients were classified as low-risk if their ABCD2 score was 3 or below. AS2 and AS4 stated that if a patient presented late (more than 1 week after symptoms), they should be categorised as low-risk. AS2 advised consideration of the patient's social environment—whether they had support from family, friends or neighbours who could recognise a further event and summon help. AS3 conducted the Miami Emergency Neurologic Deficit (MEND) exam (Box 1) (Brotons et al, 2012) if the paramedic had received appropriate training and recorded each element as free-text on the patient's clinical record. AS3 also provided a list of stroke mimics which the paramedics were advised to consider while assessing the patient—including migraine, seizures, anxiety and concussion. AS4 suggested that if oxygen saturation was <94%, consideration should be given to other causes of symptoms.
Eligibility of suspected TIA patients
Exclusion criteria for the use of a TIA referral pathway can be seen in Table 1. Contraindications to aspirin administration differed between ambulance services and can be seen in Table 2.
| Exclusion criteria to TIA pathway | Ambulance service | |||
|---|---|---|---|---|
| 1 | 2 | 3 | 4 | |
| Crescendo TIA (two or more TIAs in a week) | ✓ | ✓ | ✓ | ✓ |
| Use of anticoagulants | ✓ | ✓ | ✓ | ✓ |
| Contraindication to aspirin | ✓ | ✓ | ✓ | ✓ |
| Atrial fibrillation | ✓ | ✓ | ||
| Fever | ✓ | |||
| Any positive response to MEND examination | ✓ | |||
| Haemophilia or other coagulopathy | ✓ | ✓ | ||
| Patient cannot understand information given or cannot self-medicate | ✓ | |||
| Contraindication to aspirin | Ambulance service | |||
|---|---|---|---|---|
| 1 | 2 | 3 | 4 | |
| Peptic ulceration | ✓ | ✓ | ✓ | ✓ |
| Haemophilia or other bleeding disorder | ✓ | ✓ | ✓ | ✓ |
| Allergy or hypersensitivity to aspirin | ✓ | ✓ | ✓ | ✓ |
| Aged under 16 years old | ✓ | ✓ | ✓ | ✓ |
| Pregnant/breastfeeding | ✓ | |||
| Severe hepatic or renal impairment | ✓ | |||
| Existing use of anti-platelet drugs, e.g. 300 mg aspirin or clopidogrel | ✓ | ✓ | ✓ | |
| Extreme hypertension (blood pressure >200/105) | ✓ | |||
| Acutely uncontrolled diabetes | ✓ | |||
| Unstable vital signs (haemodynamically unstable) | ✓ | |||
| Headache | ✓ | ✓ | ||
| Allergy/hypersensitivity to non-steroidal anti-inflammatory drugs (NSAIDs) | ✓ | ✓ | ✓ | |
Method of referral
AS1 and AS4 made referrals by telephone. AS1 used a computer to confirm eligibility. The clinic contacted the patient regarding their appointment within 3 days. In AS4, a clinician conducted a repeat ABCD2 score to confirm eligibility and appointment details were left with the patient. AS2 used fax and AS3 used fax or telephone depending on the receiving hospital—patients received a call to arrange their appointment within 24 hours. AS2 provided the patient with a contact number for the clinic in case they did not receive an appointment.
All pathways provided an information leaflet to the patient, advised them not to drive, and administered a 300 mg aspirin loading dose, followed by 300 mg per day. AS1 stated that paramedics must seek advice from the patient's GP if the patient was already taking aspirin and convey them to hospital if this was not possible. For patients already taking 75 mg aspirin per day in AS3, the dose was changed to 300 mg per day. AS1 and AS3 stated that they sent a letter to the patient's GP if the pathway was used.
Discussion
A minority of UK ambulance services have introduced TIA referral pathways with variations in practice. In line with NICE (2017) Guidelines for the management of TIA, the existing pathways:
Only one pathway incorporated MEND (Brotons et al, 2012), which has not been rigorously tested. One pathway excluded patients with fever in case symptoms were caused by central nervous system (CNS) infection, though CNS infection did not appear on another pathway's list of stroke mimics. AS4's suggestion to consider differential diagnoses if oxygen saturations were below 94% seemed appropriate as TIA would not usually significantly affect a patient's breathing. The exclusion of patients with atrial fibrillation (AF) from some referral pathways acknowledges that these patients have a higher risk of stroke.
Though all patients referred by one of these pathways should have an appointment scheduled within 3 days, it is not known whether they are then assessed in a TIA clinic within the advised timeframe (RCP, 2010; 2016; NICE, 2017). It is also unknown whether all paramedics consider the patient's social situation when using a pathway (except for AS2) or whether all patients attended their appointments. All pathways used telephone or fax to refer patients to the TIA clinic; no electronic referral methods were used. This may not be practical in the pre-hospital environment.
None of the pathways have been evaluated in terms of uptake, safety or effectiveness. It is not known how often the pathways are used or whether they have caused harm to patients, whether they save the NHS money, or indeed whether they cost more compared with when no pathway is established. There is currently no evidence regarding which of these pathways includes or excludes patients appropriately, whether detailed neurological examination by paramedics is appropriate, and whether or not these pathways are acceptable to patients. Of note, the most recent RCP (2016) guidelines states:
‘there was insufficient precision in the current evidence … to make recommendations concerning risk stratification by community-based clinicians.’
The Guideline suggests patients be given 300 mg of aspirin immediately and assessed by a specialist within 24 hours, regardless of ABCD2 score—not in keeping with recommendations from NICE.
Limitations
The current survey has limitations in that one AS did not respond; there may therefore be a further TIA pathway in use that the authors are not aware of. In addition, as AS in the UK are dynamic and may undergo reconfiguration, the pathways presented here may have been altered or may no longer be in use at time of publication.
The TIER (Transient Ischaemic Attack 999 Emergency Referral) feasibility study, not yet published, examines the potential for undertaking a fully-powered randomised controlled trial (RCT) to evaluate the clinical and cost-effectiveness of a pathway for patients assessed as low-risk by paramedics, without immediate conveyance to the ED. A specialist clinical panel has used the results of the current survey alongside a scoping review of the literature to define the complex intervention for testing in the TIER feasibility study.
Conclusion
A minority of UK ambulance services have introduced referral pathways for patients with TIA, avoiding the EDs, although none have evaluated safety in terms of subsequent stroke or serious misdiagnosis, cost-effectiveness of the pathway, or acceptability to patients.