Automatic heart rate complexity monitoring in a swine model of cardiac arrest due to massive exsanguination

Abstract

checking postimplant results is usually obtained by mean of clinic (New York Heart Association [NYHA] class improvement) and instrumental (ejection fraction [EF] and left ventricular volumes by means of echocardiogram) evaluation. Objective of the present study was to evaluate the association between the novel nonlinear index, calculated from electrocardiogram (ECG) preprocedure and postprocedure, compared with traditional linear indices to assess the reliability and effectiveness in evaluating results of resynchronization therapy. Methods: We have used OntoCare System (OntoMed LLC) tomeasure CRT impact among patients who received CRT via pacemaker or defibrillator in combination with optimal pharmacological therapy, according to guidelines. Particularly, we measured the impact of stimulation analyzing the ECG performed before and after implant. Electrocardiogram complexity values have been considered significant (the difference between basal and stimulated ECG) when ≥25%. The OntoCare system measures complexity analyzing the structure of information, in particular the multidimensional structure of the ECG, taking into account the interactions between ECG channels and establishing an innovative means of characterizing generic dynamical systems and biologic signals. Results: We retrospectively analyzed 87 patient ECGs implantedwith CRT in our center. New York Heart Association class, left ventricular echocardiography (ECHO) measurements, and EF preimplant and postimplant have been analyzed and compared with complexity values. The OntoCare system found that 69 (79%) of 87 patients had a significant impact of therapy (complexity value threshold, 25%), whereas in the remaining 18 (21%) of 87 patients, there is no variation in the complexity. We then compared the 2 groups, called responder (R) and nonresponder (NR), respectively, with clinical and instrumental data in our possession. If we compare the values before and after the implant, we find that left ventricular end-diastolic volume (LVEDV) changes from 190 to 159 mL (−31 mL; P= .039) in R group and from 214 to 201 mL (−13 mL; P = .62) in NR group. The EF increases from 28.1% to 37.2% (+9.1%; P= .0001) in R group and from 27.3% to 33.6% (+6.3%; P = .0005) in NR group; NYHA class changes from 2.84 to 1.69 (−1.15, P = .01) in R group and from 2.55 to 2.65 (+0.1; P = .5) in NR group. Conclusions: Even if results are drawn from a small sample of patients with retrospective analysis, this study shows that the complexity index is reliable and effective in providing prognostic data in patients implanted with CRT. In addition, complexity index shows a strong association with other parameters (EF, ECHO measurements, and NYHA class) traditionally used to evaluate efficacy of CRT.