Effect of urgent treatment for transient ischaemic attack and minor stroke on disability and hospital costs (EXPRESS study): a prospective population-based sequential comparison

BACKGROUND
Evidence is available on the effectiveness and costs of treatments to reduce stroke risk in long-term secondary prevention. However, there are few data on the costs and outcomes of urgent assessment and treatment after the onset of transient ischaemic attack (TIA) or minor stroke. The Early use of eXisting PREventive Strategies for Stroke (EXPRESS) study showed that urgent assessment and treatment reduced the 90-day risk of recurrent stroke by about 80%. We now report the effect of the EXPRESS intervention on admissions to hospital, costs, and disability.


METHODS
EXPRESS was a prospective population-based before (phase 1: April 1, 2002, to Sept 30, 2004) versus after (phase 2: Oct 1, 2004, to March 31, 2007) study of the effect of early assessment and treatment of TIA or minor stroke on the risk of early recurrent stroke. This report assesses the effect of the introduction of the phase 2 clinic on admissions to hospital within 90 days, hospital bed-days, hospital costs, and 6-month new disability (progression from no disability before event [modified Rankin scale score < or =2 points] to disability at 6 months [modified Rankin scale score >2 points]) or death, compared with the phase 1 clinic. To assess the main predictors of these outcomes, multivariate regression analyses were done.


FINDINGS
The 90-day risk of fatal or disabling stroke was reduced in phase 2 (1 of 281 vs 16 of 310; p=0.0005). Hospital admissions for recurrent stroke were also lower in phase 2 than in phase 1 (5 vs 25; p=0.001), which reduced the overall number of hospital bed-days compared with phase 1 (672 vs 1957 days; p=0.017). Hospital bed-days for admissions to hospital due to vascular causes were also lower in phase 2 (427 vs 1365 days; p=0.016), which generated savings of 624 pounds per patient referred to the phase 2 clinic (p=0.028). Results from the multivariate analyses showed that assessment in phase 2 was an independent predictor of reduced disability, days in hospital, and costs.


INTERPRETATION
Urgent assessment and treatment of patients with TIA or minor stroke who were referred to a specialist outpatient clinic reduced subsequent hospital bed-days, acute costs, and 6-month disability.


Interventions
Two modes of access to the EXPRESS TIA and minor stroke out-patient clinic for urgent assessment and treatment were evaluated.
Phase one consisted of an appointment based TIA clinic, to which collaborating primary-care physicians referred any patient with suspected TIA or minor stroke. The clinic then contacted the patient directly to arrange an appointment as soon as possible. After the assessment and investigation, the clinic returned their treatment recommendations to the referring primary-care physicians.
In phase two, patients with suspected TIA or minor stroke were sent directly to the clinic, with no prearranged referral, and treatment was initiated at the clinic, on confirmation of the diagnosis.

Location/setting
UK/primary care.

Analytical approach:
This economic evaluation was based on a single clinical study. The time horizon was six months and the authors did not clearly report the perspective.

Effectiveness data:
The evidence came from a single clinical study, which was a prospective, population-based before (phase one) and after (phase two) study (EXPRESS). Phase one ran from 1st April 2002 to 30th September 2004, and included 310 patients. Phase two ran from 1st October 2004 to 31st March 2007, and included 281 patients. The protocols for the investigation and recommended treatments were identical in both phases of the study, except that treatment was initiated in the clinic in phase two. The patients were followed-up at one and six months after the initial event.
Multivariable modelling was used to assess the main predictors of hospital vascular admission, days in hospital, total costs, the prognostic indicators of new disability or death, and overall disability or death, at six months. Further details of the study can be found in Rothwell, et al. 2005 and2007 (see 'Other Publications of Related Interest' below for bibliographic details).

Monetary benefit and utility valuations:
Not relevant.

Measure of benefit:
The primary clinical outcome in the EXPRESS study was the 90-day risk of recurrent stroke. In addition, re-admission after 90 days and new patient disability or death were also measured.

Cost data:
The costs were those for admissions to hospital and day cases during the 90 days after the initial event, split by cause. The costs of setting up the out-patient clinics for the urgent assessment and treatment of TIA and minor stroke were excluded. The resource use data were obtained from central administrative sources, patients' records, and direct questioning at follow-up. The unit costs were obtained from national reference costs, and standardised to 2005 to 2006 prices using the UK National Health Service hospital and community health services inflation index. The currency was UK pounds sterling (£).

Analysis of uncertainty:
To account for the skewed resource use and cost data, 95% confidence interval (CI)s were calculated nonparametrically from 10,000 bootstrap estimates. A scenario analysis was also carried out for all patients with TIA or minor stroke (National Institute of Health Stroke Scale score of three points or less) in the whole Oxford Vascular Study population, of which the EXPRESS trial population was a sub-sample, irrespective of whether they were referred to the EXPRESS clinic or to other services. A scenario assuming that the efficacy of phase two was halved was also evaluated.

Results
Attendance at the phase two clinic reduced the number of 90-day recurrent strokes compared with attendance at the phase one clinic. In phase two, 6 out of 281 (2%) had another stroke, whilst, in phase one, 32 out of 310 (10%) did (p=0.0001).
Attendance at the phase two clinic also reduced the recurrent fatal strokes, disabling strokes, and overall number of fatal or disabling strokes, which were 1 out of 281 (<1%) for phase two, compared with 16 out of 310 (5%) for phase one (p=0.0005).
The total mean costs per patient for admissions for TIA, stroke, or other vascular disease, were £1,056 (standard deviation, SD: 4,879) for phase one, and £432 (SD: 2,277) for phase two, which equates to a reduction of £624 (95% CI: -1,370 to -104) per patient.
The results were robust to the different scenarios evaluated.