Usefulness of clinical factors in predicting outcomes of passive tilt tests in patients with syncope

Pretest patient selection affects the outcome of many diagnostic tests; thi s may be true for tilt-table tests. We assessed the impact of patient age, sex, and symptom burden on the outcome of passive tilt tests. Two hundred o ne patients with idiopathic syncope (87 men, aged 45 +/- 20 years, median 5 fainting spells each) under-went passive, drug-free tilt tests for 45 minu tes. Positive tests were defined as those ending in clinically reminiscent presyncope or syncope. Seventy-eight patients (39%) had a positive tilt tes t. Patients had a wide range of symptom burden, having a median 5 syncopal spells (interquartile range [IQR] 2.5 to 17.5) over a median 52.5 months (I QR 12 to 180) with a median frequency of 0.17 spells/month (IQR 0.042 to 0. 67), None of these measures of symptom burden predicted tilt-test outcome ( p = 0.33 to 0.46). In contrast, the age of the patient strongly predicted t ilt-test outcome. The likelihood of a positive test was 75% in 36 patients <25 years old and 31% in 165 patients greater than or equal to 25 years of age (p <0.0001, chi-square for 2 x 5 table). Younger patients also fainted more quickly: patients <25 years old fainted within 22 minutes of tilt and reached a clearly asymptotic valve, whereas the likelihood of a positive ti lt in patients <25 years old increased linearly with time, and did not reac h an asymptote. Measures of symptom burden do not predict test outcome, and younger patients are more likely to faint during passive tilt testing. (C) 2000 by Excerpta Medica, Inc.