Early chronotropic response analysis in head up tilt table test

Abstract

Syncope affects up to a quarter of population [1] and reflex etiology is themost prevalent. The head up tilt test (HUTT)was initially used to evaluate the responses to orthostatic stress in terms of research, and thereafter to investigate syncope [2]. Different protocols have been used showing sensitivity ranging from 30% to 50% and specificity around 90%.We studied the heart rate just after tilting to predict HUTT results in a population under investigation of unexplained syncope referred to our center from March 2005 to December 2009. Subjects were enrolled if there was previous unexplained syncope, no structural heart disease or use of drugs that interfere with cardiac chronotropism, any condition that affects the autonomic system such as diabetes or parkinsonism, and no contraindications to the test. All patients signed a consent form. Positive tests during the first 10 min were excluded from analysis. HUTT was performed after 6 h of fasting and an initial 10 minutes rest period was observed. Patients were tilted to 70°, up to 40 min during which HR and BP were monitored [3–5]. If necessary, 1.25 mg sublingual isosorbide was administered after 20 min. A drop in HR, BP or both followed by symptoms defined test positivity, according to VASIS classification. The HRstart was that in the end of the rest period. Likewise, the HRmax 10 min was the highest seen at monitor during this period. An area under the ROCwas calculated and used to determine the ability of the early chronotropic response-ECR to discriminate HUTT results. AUC greater than 0.7 mean good statistical relation, and not a performance by chance. We calculated sensitivity, specificity, positive likelihood ratio and negative likelihood ratiowith respective 95% CI for all possible cutoff points and determined the optimal using Youden index [6,7]. The quantitative variables with normal distribution were described as mean and standard deviation. The software SAS 9.2 was used in statistical analysis. We enrolled 390 patients with a mean age of 47 y old and 89 had positive HUTT— 39% hadmixed response, 20% vasodepressor, 16% cardioinhibitory and the remainder 25% a POTS or dysautonomic response [Table 1]. ECR was analyzed as the absolute HR increase and as a percentage increase. The absolute ECR in each pattern of response is shown in Table 2. In ROC analysis, a cutoff 21% increase in HRstart discriminated those with negative and positive HUTT with sensitivity of 66.11%, specificity of 75.28%, and AUC of 0.739 (95% CI: 0.692 to 0.782). An absolute cutoff of 15 bpm increase also discriminated groups with sensitivity of 71.43, specificity of 77.53, and AUC of 0.755 (95% CI: 0.709 to 0.796). Comparing both variables using Bootstrap analysis the difference was 0.01615, not statistically significant p=0.114 [Fig. 1]. Statistical analysis demonstrated that a 10% HRstart increase shows specificity of 93.2. A positive HUTT with smaller ECR is a very unusual finding. Spearman correlation coefficient (CI=95%) showed inverse relation between HUTT duration and ECR.We also studied ECR in different age and sex groups [Tables 3 and 4]. Previous studies showed different results on early hemodynamic parameters in HUTT [8–12]. Our paper has a comparatively greater population and we considered the HRstart that immediately before