Abstract
We studied the question whether the decision of general practitioners (GPs) to refer patients with transient ischemic attacks (TIAs) to hospital departments of neurology is influenced by certain characteristics of the history or the patient. A postal questionnaire which included 10 TIA cases was sent to a random sample of 10% (n = 635) of all GPs in the Netherlands (response 59%). The cases were constructed as five pairs, in each of which one of the following characteristics was varied: nature of TIA, duration and number of attacks, age and previous history. Eighty percent of the cases were accurately diagnosed. Correct diagnoses were most often made in cases with transient hemispheral (as opposed to retinal) ischaemia, longer attacks, recurrent attacks, age below 65, and in those without a history of non-specific symptoms. Referral to the neurologist was proposed in 45% of the cases, particularly for patients with hemispheral ischaemia, shorter attacks, recurrent attacks and age below 65. Aspirin would have been prescribed in one third of the cases, also of the referred patients. We conclude that clinical trials performed in hospital departments of neurology are affected by referral selection before randomisation. If the characteristics that influence referral are also related to the efficacy of the treatments under study, the results of these trials cannot be extrapolated to general practice.