Patient Behavior Immediately After Transient Ischemic Attack According to Clinical Characteristics, Perception of the Event, and Predicted Risk of Stroke
Background and Purpose— Little research has been done on patients’ behavior after transient ischemic attack (TIA). Recent data on the high early risk of stroke after TIA mean that emergency action after TIA is essential for effective secondary prevention. We therefore studied patients’ behavior immediately after TIA according to their perceptions, clinical characteristics, and predicted stroke risk.
Methods— Consecutive patients with TIA participating in the Oxford Vascular Study or attending dedicated hospital clinics in Oxfordshire, UK, were interviewed. Predicted stroke risk was calculated using 2 validated scores.
Results— Of 241 patients, 107 (44.4%) sought medical attention within hours of the event, although only 24 of these attended the emergency department. A total of 107 (44.4%) delayed seeking medical attention for ≥1 day. Correct recognition of symptoms (42.2% of patients) was not associated with less delay. However, patients with motor symptoms or duration of symptoms ≥1 hour were more likely to seek emergency attention (hazard ratio, 2.1; 95% CI, 1.4 to 3.2; P=0.00005), as were those at higher predicted stroke risk (P=0.001). The other main correlate with delay was the day of the week on which the TIA occurred (P<0.001), with greater delays at the weekend. Delay was unrelated to age, sex, or other vascular risk factors.
Conclusions— Many patients delay seeking medical attention after a TIA irrespective of correct recognition of symptoms, although patients at higher predicted risk of stroke do act more quickly. Public education about both the urgency and nature of TIA is required.
With the advent of thrombolysis and other potential treatments, there has been great interest in the delays between symptom onset and treatment for acute ischemic stroke. Individuals’ behavior after stroke has been studied extensively,1–5 as has knowledge of stroke among the general public and at-risk populations.6–9 In contrast, equivalent studies relating to transient ischemic attack (TIA) are lacking, perhaps because there has been less concern about delays to treatment. However, with the recent emergence of data on the high risk of stroke during the hours and days after a TIA,10–13 the ability to identify very high-risk individuals with simple clinical scores,12,14 and the evidence that early instigation of preventive treatments is highly effective,15,16 it is essential to better understand patient perceptions and behavior after TIA and hence the barriers to rapid presentation. A telephone survey in the United States showed that knowledge of TIA among the general public was poor and that delays to seeing a physician after TIA are common,17 as has also been found in a clinical trial population.18 However, there have been no prospective studies of behavior of patients immediately after TIA in relation to their perceptions of the event, demographics, clinical characteristics and most important, their predicted risk of early recurrent stroke. We therefore studied a population-based cohort and a clinic-based cohort of individuals with TIA.
Data were collected from 2 cohorts of patients with probable or definite TIA. One cohort comprised all patients with TIA in the Oxford Vascular Study (OXVASC), a population-based study of the incidence of TIA and stroke, the methods of which have been described previously.19 Multiple search methods ensured as near complete ascertainment of all TIA cases as possible.20 Inpatients, emergency department (ED) and eye hospital attendees and those who were assessed at home were included in addition to those who attended the OXVASC dedicated daily hospital clinic. Patients who had a TIA but only sought medical attention after an early recurrent stroke were also studied but were excluded where appropriate from some analyses of determinants of delays to seeking medical attention. The other cohort comprised consecutive patients attending local specialist hospital-based TIA outpatient clinics in which the researchers were involved.
All cases were reviewed by consultant neurologists to confirm the diagnosis, which was based on standard criteria.21 Patients with stroke or noncerebrovascular diagnoses or who were unable to recall details of their actions were excluded. The study period was April 1, 2002, to March 31, 2003. OXVASC and related substudies were approved by the local ethics committee.
Patients were interviewed by the study clinician at initial assessment to determine: (1) their initial perception of the cause of their symptoms; (2) their immediate response; and (3) dates and times of symptom onset, when medical attention was sought, and the first contact with medical services.
Demographic data, risk factors, and symptomatology were also recorded, including the main known independent risk factors for early recurrent stroke.12,14
Patients were judged to have acted in an emergency if they attempted to seek medical attention as soon as they were able (usually within minutes to a few hours). For example, a patient who delayed seeking medical attention for 4 hours because of transient incapacity was judged to have acted as in an emergency if they sought attention as soon as they recovered sufficiently, whereas a patient who delayed for a similar time for convenience was not. In patients with multiple TIAs, the most recent event before presentation was studied.
We determined the proportion of patients who had acted as in an emergency and the actual delay to seeking medical attention in relation to whether the patient correctly identified the cause of their symptoms, demographic data, risk factors, symptomatology, and the day on which the presenting TIA occurred. Delay was also related to the predicted early risk of stroke using 2 previously published risk scores based on clinical characteristics (a score developed by Johnston et al12 and the ABCD score14). For this analysis, cells containing <10 individuals were merged with adjacent cells.
Of 151 patients in the OXVASC population who had a TIA during the study period, 4 patients (2.6%) had a major disabling or fatal stroke after seeking medical attention but before assessment so were unable to participate, whereas an additional 9 (6.0%) sought medical attention only after a subsequent stroke. All 138 (91.4%) remaining consecutive OXVASC patients agreed to participate. Of 114 consecutive referrals to the hospital clinic, 5 patients (4.4%) had a major disabling or fatal stroke after referral but before the clinic appointment and therefore did not attend the clinic, and a further 6 (5.3%) did not attend for other reasons and so were excluded. All remaining 103 (90.4%) consecutive hospital clinic attendees agreed to take part in the study. A total of 241 participants were therefore studied.
Table 1 shows patient demographics, behavior, perceptions, risk factors for early stroke, and other vascular risk factors for the hospital clinic and OXVASC cohorts. Clinic patients were younger than patients from OXVASC, had fewer previous risk factors, and were assessed less quickly (median time from TIA 22 days versus 5 days; P<0.001). However, initial patient behavior and perceptions were almost identical in the 2 cohorts, and so OXVASC and clinic patients were combined for further analysis.
Initial Behavior After TIA
Medical attention was first sought via the family practitioner in 209 (86.7%) cases, the ED in 24 (10.0%), and from other sources in 8 (3.3%). No patients telephoned “NHS Direct,” the 24-hour Government Department of Health advice service. A total of 107 (44.4%) acted as in an emergency. Patterns and determinants of emergency action were similar between clinic and OXVASC patients. Table 2 shows the numbers and proportions of individuals who acted as in an emergency and who delayed seeking medical attention according to demographics, perceptions, risk factors for early stroke, and previous vascular risk factors. Patients with symptoms associated with a higher predicted early risk of stroke were more likely to act as in an emergency compared with those without: 53.3% versus 35.5% for motor symptoms (P=0.006) and 56.4% versus 36.1% versus 31.3% for symptom duration ≥60 minutes, 10 to 59 minutes, and <10 minutes, respectively (P for trend 0.001). Overall, 88 (53.3%) of 165 patients with either motor symptoms or duration of symptoms ≥1 hour sought attention as an emergency compared with 19 (25%) of 76 patients with neither characteristic (hazard ratio, 2.1; 95% CI, 1.4 to 3.2; P=0.00005). These associations remained significant when the 9 OXVASC patients who had a stroke before seeking medical attention were included. There was no association between age, sex, vascular territory, or presence or number of any previous vascular risk factors (including previous cerebrovascular disease and diabetes) and emergency action. Importantly, there was no association between correct recognition of TIA and emergency action.
In addition to the 107 patients who acted as in an emergency, an additional 27 (11.2%) also sought medical attention on the same day as their presenting TIA, 43 (17.8%) delayed until the next day, and 64 (26.6%) delayed for ≥2 days. Features associated with delay of ≥2 days included absence of motor symptoms and shorter symptom duration. There was no association between longer delays to seeking medical attention and incorrect recognition or not knowing the cause of the TIA.
Figure 1 shows that the delay in seeking medical attention depended on the day of the week on which the presenting TIA occurred (P<0.001). Occurrence on Friday and at the weekend was associated with the longest delays to seeking medical attention. Twenty-eight (41.2%) of those who had a presenting TIA at the weekend sought medical attention on the same day compared with 105 (60.7%) who had a TIA on a weekday (P=0.006). Thirty-one (53.4%) of those who had a TIA on Friday or Saturday delayed ≥2 days compared with 33 (18.0%) who had a TIA Sunday through Thursday (P<0.001). Of the 5 patients who had a TIA on public holidays during the study period, all delayed until the next day or longer to seek medical attention.
Perceptions of the Cause of the Event
Among those patients who did seek medical attention after a TIA, 231 (95.9%) were able to recall their initial perception of the cause of their symptoms, of whom 98 (42.4%) correctly recognized the cause as TIA or “mini-stroke,” and 82 (35.5%) did not know the cause. The remaining 51 (21.4%) assumed incorrect causes, the most frequent being stress or fatigue (n=12; 5.2%), “eye problems” (n=7; 3.0%), heart attack (n=3; 1.3%), and migraine (n=3; 1.3%). Table 3 shows the numbers and proportions of individuals who correctly recognized the cause of their symptoms according to risk factors and TIA characteristics. Rates and determinants of correct recognition were similar between clinic and OXVASC patients. Those with previous TIA were more likely to be correct (58.3% versus 39.5%; P=0.044), as were those with motor symptoms (49.1% versus 35.7%; P=0.046) and cerebral (versus ocular) symptoms (44.7% versus 24.0%; P=0.055), but there were no other significant associations with correct recognition.
Patient Behavior and Predicted Risk of Early Stroke
Figure 2 shows delays to seeking medical attention stratified by early stroke risk predicted by the score proposed by Johnston et al and the ABCD score.12,14 For both scores, those at higher predicted risks were more likely to act as in an emergency (trends P<0.001 and P=0.001, respectively) and less likely to delay seeking medical attention for ≥2 days (trends P=0.01 and P=0.049, respectively). However, even among those at highest predicted stroke risk (scores of 4 to 5 for the score proposed by Johnston or 5 to 6 for the ABCD score), a delay of ≥1 day was common (7 of 36 [19.4%] and 28 of 90 [31.1%], respectively).
In a telephone survey of randomly sampled US adults, Johnston et al found that only 8.6% could identify a typical symptom of a “TIA.”17 Given that public knowledge of the symptoms of “stroke” is much better,7–9 it is likely that the low levels of knowledge found by Johnston et al were at least partly attributable to poor recognition of the term “TIA” itself. Moreover, actual behavior in the event of a TIA is of more practical importance than public knowledge. In the same study, Johnston et al found that among those with a history of TIA, 64% recalled seeing a physician within 24 hours, and 16% were seen over a week later.17 Castaldo et al reported that only one third of patients who had a TIA during follow-up in the Asymptomatic Carotid Artery Study reported it to medical attention within 3 days. However, to our knowledge, there have been no previous studies in routine clinical practice of behavior after a TIA in relation to perceptions, clinical characteristics, and risk of stroke. We found that despite a high prevalence of vascular risk factors including previous TIA or stroke, only a minority of patients acted as in an emergency, and correct recognition of TIA was poor. However, those at highest predicted stroke risk were more likely to act as in an emergency, although there was no association between recognition of symptoms and urgency of action.
That correct recognition of TIA was not associated with seeking medical attention as an emergency and the low proportion of patients attending the ED (10.0%) indicate a lack of public awareness that a TIA is a medical emergency. Public education should therefore address both the symptoms of TIA and the need to seek urgent medical attention. The widespread lack of perception of urgency is perhaps best illustrated by patient behavior after events occurring on Fridays, weekends, and public holidays, despite universal access to emergency inpatient services at all times. Whether the same patterns of behavior after TIA would be observed in other countries with different healthcare systems is uncertain and requires further research. More encouraging was the observation that patients at higher predicted stroke risk were more likely to act as in an emergency because of the influence of weakness and prolonged symptom duration on behavior. As expected, these symptoms were particularly frequent in those patients who had an early recurrent stroke before seeking medical attention.
Although we think that our observations are valid, our methods did have some potential shortcomings. First, we will undoubtedly have underestimated the extent of the lack of awareness of the need to seek medical attention by studying cohorts all of whom did report their symptoms; an unknown proportion of patients with TIA never seek medical attention. However, the exclusion of patients who experienced a disabling stroke before seeking medical attention may have introduced an opposite bias. Second, it is possible that some patients reporting of initial perceptions and actions were influenced by the knowledge that they were attending a TIA clinic or were participating in a study of TIA. However, such knowledge would have tended to lead to exaggeration of our estimates of correct recognition of symptoms and emergency action. Moreover, we were able to validate much of the reported behavior against what was documented by primary care and the ED, and we found very few inconsistencies. Third, the delay from event to assessment in the hospital clinic cohort was 22 days, which could have reduced the reliability of reporting of perceptions and actions. However, the findings in the OXVASC cohort, in which patients were seen after a median of 5 days, were identical. Fourth, the setting of the study, Oxfordshire, UK, is a county with lower social deprivation and higher educational levels on average than the rest of the United Kingdom, although such differences would lead to higher rates of health awareness and a greater propensity to use emergency services. Finally, there were some differences between the baseline characteristics of patients in the hospital clinic and the OXVASC cohorts. However, these differences were explicable on the basis that the OXVASC study was an inclusive population-based study with high rates of case ascertainment in patients with previous TIA and in older age groups. Moreover, the robustness of our findings across the 2 cohorts suggests that they are likely to be generalizable.
In conclusion, despite the high risk of stroke after TIA, many patients delay seeking medical attention, irrespective of correct recognition of symptoms. However, the impact of such delays is counterbalanced to some extent by TIA patients at higher predicted risk of stroke acting more urgently.
- Received November 24, 2005.
- Revision received January 19, 2006.
- Accepted February 28, 2006.
Harraf F, Sharma AK, Brown MM, Lees KR, Vass RI, Kalra L. A multicentre observational study of presentation and early assessment of acute stroke. BMJ. 2002; 325: 17–22.
Smith MA, Doliszny KM, Shahar E, McGovern PG, Arnett DK, Luepker RV. Delayed hospital arrival for acute stroke: the Minnesota Stroke Survey. Ann Intern Med. 1998; 12: 190–196.
Lacy CR, Suh DC, Bueno M, Kostis JB. Delay in presentation and evaluation for acute stroke: Stroke Time Registry for Outcomes Knowledge and Epidemiology (STROKE). Stroke. 2001; 32: 63–69.
Salisbury HR, Banks BJ, Footitt DR, Winner SJ, Reynolds DJM. Delay in presentation of patients with acute stroke to hospital in Oxford. Q J Med. 1998; 97: 635–640.
Reeves MJ, Hogan JG, Rafferty AP. Knowledge of stroke risk factors and warning signs among Michigan adults. Neurology. 2002; 59: 1547–1552.
Carroll C, Hobart J, Fox C, Teare L, Gibson J. Stroke in Devon: knowledge was good, but action was poor. J Neurol Neurosurg Psychiatry. 2004; 75: 567–571.
Coull A, Lovett JK, Rothwell PM; on behalf of the Oxford Vascular Study. Early risk of stroke after a TIA or minor stroke in a population-based incidence study. BMJ. 2004; 328: 326–328.
Hill MD, Yiannakoulias N, Jeerakathil T, Tu JV, Svenson LW, Schopflocher DP. The high risk of stroke immediately after transient ischemic attack. A population-based study. Neurology. 2004; 62: 2015–2020.
Markus HS, Droste DW, Kaps M, Larrue V, Lees KR, Siebler M, Ringelstein EB. Dual antiplatelet therapy with clopidogrel and aspirin in symptomatic carotid stenosis evaluated using Doppler embolic signal detection: the Clopidogrel and Aspirin for Reduction of Emboli in Symptomatic Carotid Stenosis (CARESS) trial. Circulation. 2005; 111: 2233–2240.
Johnston SC, Fayad PB, Gorelick PB, Hanley DF, Shwayder P, van Husen D, Weiskopf T. Prevalence and knowledge of transient ischemic attack among US adults. Neurology. 2003; 60: 1429–1434.
Castaldo JE, Nelson JJ, Reed JF III, Longenecker JE, Toole JF. The delay in reporting symptoms of carotid artery stenosis in an at-risk population. The Asymptomatic Carotid Atherosclerosis Study experience: a statement of concern regarding watchful waiting. Arch Neurol. 1997; 54: 1267–1271.
Rothwell PM, Coull AJ, Giles MF, Howard SC, Silver LE, Bull LM, Gutnikov SA, Edwards P, Mant D, Sackley CM, Farmer A, Sandercock PAG, Dennis MS, Warlow CP, Bamford JM, Anslow P; Oxford Vascular Study. Change in stroke incidence, mortality, case fatality, severity and risk factors in Oxfordshire, UK from 1981 to 2004 (Oxford Vascular Study). Lancet. 2004; 363: 1925–1933.
Coull AJ, Silver LE, Bull LM, Giles MF, Rothwell PM; Oxford Vascular Study. Direct assessment of completeness of ascertainment in a stroke incidence study. Stroke. 2004; 35: 2041–2045.