Diagnostic accuracy of FAST and BEFAST for diagnosing stroke in primary care
Patients with complaints suggestive of neurological deficit often seek medical attention in primary care first. During out-of-hours, this is at the out-of-hours service in primary care (OHS-PC). At the OHS-PC, the (BE-)FAST rule is often used to indicate whether the complaints are a result of a transient ischemic attack (TIA) or stroke.
Aim
De aim of this project was to study the reliability of (BE-)FAST as used during telephonic triage by non-medically trained personnel.
Approach
This project consisted of a cross-sectional study using retrospective data from routine primary care form the SAFETY-FIRST study, a long running study on the efficiency and safety of telephonic triage at the OHS-PC. Data for this study consisted of a sample of recorded telephonic triage calls of patients with complaints suggestive of neurological deficit, selected based on entrance complaint, the ICPC code of the contact and key words in the electronic patient file. To improve the predictive performance of (BE-)FAST we build four logistic regression models:
- model 1 with age and sex
- model 2 with age; sex and a history of cardiovascular disease
- model 3 with age, sex and risk factors for cardiovascular disease
- model 4 with all of the predictors above
We used an interaction term for age and sex. Bootstrapping was used to internally validate the model. The predictive performance of each model was expressed by the c-statistic, the R2 and slope.
Results
We included in total 1381 patients. The median age was 72 years, and 65% was female. All patients contacting the OHS-PC with complaints suggestive of neurological deficit were (BE-)FAST positive, and this test therefore was not able to safely exclude stroke when used by non-medically trained personnel during telephonic triage. In this group of (BE-)FAST positive patients, risk of stroke was 52%, and in total 24% of patients received too low an urgency. Addition of age and sex improved the predicitve value of (BE0-)FAST, this was not the case with the addition of risk factors for or a history of cardiovascular disease.
This project finds that patients calling the OHS-PC with complaints suggestive of neurological deficit are (nearly) all positive on the (BE-)FAST test. Adding age and sex improves the performance of (BE-)FAST. This is not the case for adding cardiovascular risk factors or a history of cardiovascular disease.
Strengths and limitations of this project
The strength of this project was the use of data from routine care, causing for an efficient method of data collection reflecting daily practice. The most important limitation was inherent to this routine care data. There was missing data for which advance statistical methods of multiple imputation were used though risk of some bias could not be avoided. A second limitation was in the outcome verification that could differ between patients. Risk of bias was minimized by requestion all available information up to one month after the OHS-PC contact at the general practitioner of each patient.
Recommendations
The results of this project can be used to improve the Netherlands Triage Standard as used at the OHS-PC, and thereby optimize the care for patients with TIA or stroke.
Features
-
Project number:
10060012210005
-
- Duration: 100%
Duration: 100 %20242024 - Part of programme:
- Related funding round:
-
Project lead and secretary:
dr. S. van Doorn MD PhD
-
Responsible organisation:
Universiteit Utrecht
