100Do automatic CRT optimisation algorithms work? A comparison of techniques

Abstract

Introduction: At least 30% of heart failure patients who receive cardiac resynchronisation therapy (CRT) are 'non-responders'. Optimisation of pacing intervals, referred to as CRT optimisation, is frequently undertaken in an attempt to reduce non-response and maximise the effectiveness of this therapy. Traditional approaches to CRT optimisation, however, are often expensive and time consuming. Thus, use of automatic, device based algorithms which utilise features of the intracardiac electrogram (IEGM) for CRT optimisation provide an attractive alternative technique. The aim of this study was to assess effectiveness of these algorithms by comparison to manual CRT optimisation using non-invasive cardiac output monitoring (NICOM) bioreactance technology. Method(s): 29 participants with Boston Scientific, St Jude Medical or Medtronic CRT pacemakers or defibrillators were recruited into this single centre, prospective, non-randomised trial. All patients attended routine outpatient CRT optimisation where both manual NICOM optimisation and device based optimisation were performed. Intrinsic cardiac output (CO), CO with baseline settings and CO following each optimisation technique was recorded. Result(s): Figure 1 demonstrates how CO increased from intrinisic rhythm to baseline settings and from baseline to NICOM optimised settings. CO produced as a result of automatic optimisation, however, was significantly less than that produced as a result of manual NICOM optimisation (p<0.000), with CRT settings programmed according to NICOM optimisation in 27 of the 29 participants. There was no significant difference between baseline CO and CO following automatic optimisation (p=0.304). Automatic optimisation produced longer paced atrioventricular delays (pAVDs) than those achieved by NICOM optimisation (p=0.007). There was also no significant difference in CO produced between individual manufacturers' devices (p=0.954). Conclusion(s): Firstly, CO increased with CRT on, thus CRT's positive influence on acute hemodynamics has been demonstrated in this study. Secondly, the additional 11% increase in CO achieved with manual CRT optimisation using NICOM bioreactance technology suggests that optimisation of pacing intervals using this technique is effective in improving hemodynamic performance. Thirdly, in CRT patients, it seems that caution should be taken when attempting to optimise pacing intervals using automatic, device based algorithms. These appear to be ineffective at increasing CO beyond baseline settings. Long pAVDs produced by automated algorithms may allow greater intrinsic contribution to ventricular depolarisation, explaining these results. Finally, none of the individual manufacturers' algorithms appear to produce superior acute hemodynamics. Decisions about device selection should not, therefore, be made on the basis of automatic optimisation algorithm performance. (Table Presented)