Study population
We analyzed individual values of resting supine blood pressure (BP) and heart rate (HR) collected in the years 2003-2019 in three large databases of patients who had undergone TT for unexplained syncope in the tertiary syncope investigation units of hospitals in Florence (n=805) and Lavagna (n=2798), Italy, and Malmö, Sweden (n=1633). We included patients who had undergone TT for diagnosis of syncope, unexplained by the initial assessment1, with high pre-test probability of a reflex mechanism. We excluded patients with classical orthostatic hypotension (OH), when TT was performed for other than diagnostic reasons (e.g. tilt training), and patients who were investigated for orthostatic intolerance such as postural orthostatic tachycardia syndrome (POTS) but not syncope as the main symptom. Hypertensive patients were included with hypertension being defined by a clinical diagnosis plus hypotensive medication.
Despite the long study period, indications, methodology and interpretation of TT results remained unchanged during this time and were very similar to the recommendations of the current European Society of Cardiology syncope guidelines.1 TT was performed according to the “Italian protocol” which consists of a 20-minute passive phase at a tilt angle of 70°, followed by a 15-minute nitroglycerine-potentiated phase (≈400mcg administered sublingually), if syncope was not induced during the passive phase.18Positive response was defined as reproduction of spontaneous symptoms in the presence of characteristic hemodynamic pattern of bradycardia and hypotension.1 In all centers, baseline hemodynamic data were obtained in supine position after 5-10 minutes of rest prior to TT using validated non-invasive beat-to-beat hemodynamic monitors, Task Force Monitor (CNSystems Medizintechnik GmbH, Graz, Austria) in Florence and Lavagna, and Nexfin (BMEYE, Amsterdam, Netherlands) or Finapres Nova monitors (Finapres Medical Systems, PH Enschede, Netherlands) in Malmö.19,20 The monitors were calibrated before measurement using brachial cuff and oscillometric method according to the manufacturer’s instructions. As the monitors render beat-to-beat data, an average value of a hemodynamically stable period of 10 (Florence and Lavagna) or 30 seconds (Malmö) was recorded in the database. The patient information was de-identified before merging the databases, thus, Ethical approval was not required.