Background Survival for cardiac arrest remains poor, and the use of extracorporeal cardiopulmonary resuscitation (ECPR) has been suggested as a potential therapy to improve outcomes. Harefield Hospital has been performing ECPR for over a decade; however outcomes have been historically poor.Methods A retrospective analysis was performed of all ECPR cases at Harefield Hospital between April 2018 and April 2023. A new structured system including a systematic screening process, strict exclusion criteria, assessment of resuscitation adequacy through physiological stop criteria, drilled and standardised cannulation process, post-resuscitation care and neuroprognostication was rolled out on 1 April 2023. A comparison of ECPR performed pre and post was undertaken.Results With the institution of the new system, 13 patients were treated with ECPR over 1 year; of which nine (69.2%) survived with good neurological outcomes at 6 months. In the preceding 3 years, there had been 22 cases; of which, two survived (9.1%). This was a statistically significant higher survival with HR of 4.56 (CI 2.1, 10.2, p<0.05).The average length of hospital stay in the new system was 61.3 days; of which, 32.4 days were in the intensive care unit.Conclusions This study shows that it is possible to significantly improve outcomes in ECPR in a single centre by introducing into standard practice, a high-quality bundle of care.

Introduction Haemodynamic atrioventricular delay (AVD) optimisation has primarily focussed on signals that are not easy to acquire from a pacing system itself, such as invasive left ventricular catheterisation or arterial blood pressure (ABP). In this study, standard clinical central venous pressure (CVP) signals are tested as a potential alternative. Methods Sixteen patients with a temporary pacemaker after cardiac surgery were studied. AV delay optimisation was performed by alternating between a reference AVD of 120ms and tested settings ranging from 40 to 280ms, with 8 replicates for each setting. Alongside (a) the raw data, three methods of correcting for respiration were tested: (b) limiting analysis to a respiratory cycle, (c) asymmetric least squares (ALS) and (d) discrete wavelet transform (DWT). The utility of a quality control step was tested. Results CVP signals were a mirror image of the systolic ABP signals: The four R values were -0.674, -0.692, -0.631, -0.671 respectively (all p<0.001). With quality control, the mirror image was best for DWT (R = -0.76, p<0.001), with the CVP and ABP optima agreeing well (R = 0.78, p<0.001). The automated quality control signal correctly predicted the gap between the AVD optima calculated from ABP and CVP (R = 0.8, p<0.001). Conclusions Central venous pressure signals could be used to optimise AVD, because they have a reliable inverse relationship with ABP when pacemaker settings undergo protocolised testing. However, protocols need careful design to circumvent spontaneous biological variability.

Abstract Funding Acknowledgements Type of funding sources: Private grant(s) and/or Sponsorship. Main funding source(s): British Heart Foundation Introduction Automated classification of arrhythmias in ECGs is becoming increasingly important. Publicly available ECG datasets have been widely used by the research community to create novel artificial intelligence models that improve these detection rates. The development of these models requires access to large volume of labelled data. However, access to such databases is becoming increasingly limited. In addition, the datasets are often unbalanced because abnormal rhythms are far outweighed by normal samples. The unbalanced nature of the datasets can lead to less accurate models. Therefore, generating realistic synthetic signals can augment the real signals found in such databases and provide data that allows sophisticated model development. Purpose In this study, we propose a deep learning-based approach for synthetic ECG signal generation that uses long short-term memory (LSTM) autoencoder and generative adversarial networks (GAN) to generate signals that mimic the distribution of arrhythmia signals (Figure 1). Methods The LSTM autoencoder is composed of two parts: an encoder and a decoder (Figure 1b). The encoder takes original ECG signal as its input and uses LSTM layers to compress the signal into a set of features. The decoder is formed by reversing the encoding process, which uses the encoded features as its input and converts them back into the original signal. To generate synthetic signals, we inserted GANs between the LSTM encoder and the decoder. GANs are composed of a generator and a discriminator (Figure 1c). The generator produces synthetic ECG features based on noise, whereas the discriminator tries to distinguish between real features and results received from the generator. The pathological beats studies were: left bundle branch block (LBBB), right bundle branch block (RBBB), aberrated atrial premature (AA), and normal beats (N) from the MIT-BIH arrhythmia database, using lead II only. To evaluate the quality of our synthetic signals, we trained an LSTM classifier on a combination of our real and synthetic data and compared the testing results with a model trained on real data alone. Results The LSTM encoder, decoder and GAN were trained individually for each beat type, and examples of generated signals are illustrated in Figure 2. The average accuracy of the classification for the original dataset was 90%, with a recall of 98% for N, 36% for AA, 39% for LBBB and 97% for RBBB. Once synthetic signals were added to the training set, the average testing classification accuracy increased to 98%, with a recall of 99% for N, 83% for AA, 99% for LBBB and 99% for RBBB. For fair comparison, the testing set contained only real data and remained unchanged for both groups. Conclusion In this work, we proved the capability of GANs to generate realistic synthetic signals that helped to improve the detection rates of arrhythmias as measured by both increased overall accuracy and recall. model examples_signals

Background: It is not known whether the optimal Atrioventricular delay (AV ) varies between left ventricular (LV) pacing site during endocardial biventricular pacing (BiVP) and may therefore needs consideration. Methods: We assessed the haemodynamic AV in patients with chronic heart failure undergoing endocardial LV lead implantation. AV was assessed during atrio-biventricular pacing (BVP) with a “roving LV lead”. Up to four locations were studied: mid lateral wall, mid septum (or a close alternative), site of greatest haemodynamic improvement and LV lead implant site. The AV was compared to a fixed AV delay of 180ms. Results: Seventeen patients were included (12 male, aged 66.5 +/- 12.8 years, ejection fraction 26 +/- 7%, 16 left bundle branch block or high percentage of right ventricular pacing (RVP), QRS duration 167 +/-27 ms). In most locations (62/63), AV increased systolic blood pressure during BiVP compared with RVP (relative improvement 6 mmHg, IQR 4-9mmHg). Compared to a fixed AV delay the haemodynamic improvement at AV was higher (1mmHg, IQR 0.2-2.6mmHg, p<0.001). Within most patients (16/17), we observed a difference in AV between pacing sites (median paced AV 209 ms, IQR 117-250). Within this range, the haemodynamic impact of these differences was small (median loss 0.6 mmHg, IQR 0.1-2.6mmHg). Conclusion: Within a patient, different endocardial LV lead locations have slightly different haemodynamic AV which are superior to a fixed AV delay. The haemodynamic consequence of applying an optimum from a different lead location is small.

AIMS:For bradycardic patients after cardiac surgery, it is unknown how long to wait before implanting a permanent pacemaker (PPM). Current recommendations vary and are based on observational studies. This study aims to examine why this variation may exist. METHODS AND RESULTS:We conducted first a study of patients in our institution and second a systematic review of studies examining conduction disturbance and pacing after cardiac surgery. Of 5849 operations over a 6-year period, 103 (1.8%) patients required PPM implantation. Only pacing dependence at implant and time from surgery to implant were associated with 30-day pacing dependence. The only predictor of regression of pacing dependence was time from surgery to implant. We then applied the conventional procedure of receiver operating characteristic (ROC) analysis, seeking an optimal time point for decision-making. This suggested the optimal waiting time was 12.5 days for predicting pacing dependence at 30 days for all patients (area under the ROC curve (AUC) 0.620, P = 0.031) and for predicting regression of pacing dependence in patients who were pacing-dependent at implant (AUC 0.769, P < 0.001). However, our systematic review showed that recommended optimal decision-making time points were strongly correlated with the average implant time point of those individual studies (R = 0.96, P < 0.001). We further conducted modelling which revealed that in any such study, the ROC method is strongly biased to indicate a value near to the median time to implant as optimal. CONCLUSION:When commonly used automated statistical methods are applied to observational data with the aim of defining the optimal time to pacing after cardiac surgery, the suggested answer is likely to be similar to the average time to pacing in that cohort.

Permanent pacemaker (PPM) implantation occurs in up to 5 % of patients after cardiac surgery but there is little consensus on how long to wait between surgery and PPM insertion. Predicting the likelihood of a patient being pacing dependent 30 days after implant can aid with this timing decision and avoid unnecessary observation time waiting for intrinsic conduction to recover. In this paper, we introduce a new approach for the prediction of PPM dependency at 30 days after implant in patients who have undergone recent cardiac surgery. The aim is to create an automatic detection model able to support clinicians in the decision-making process. We first applied Synthetic Minority Oversampling Technique (SMOTE) and Bayesian Networks (BN) to the dataset, to balance the inherently imbalanced data and create additional synthetic data respectively. The six resultant datasets were then used to train four different classifiers to predict pacing dependence at 30 days, all using the same testing set. The Bagged Trees classifier achieved the best results, reaching an area under the receiver operating curve (AUC) of 90 % in the train phase, and 83 % in the test phase. The overall classification performance was clearly enhanced when using SMOTE and synthetic data created with BN to create a combined and balanced dataset. This technique could be of great use in answering clinical questions where the original dataset is imbalanced.

Objective The study evaluated the feasibility of mindfulness-based cognitive therapy (MBCT) in patients with non-cardiac chest pain by assessing their willingness to participate and adhere to the programme, and for these data to help further refine the content of MBCT for chest pain. Patients and methods This prospective 2:1 randomised controlled trial compared the intervention of adapted MBCT as an addition to usual care with just usual care in controls. Among 573 patients who attended the rapid access chest pain clinic over the previous 12 months and were not diagnosed with a cardiac cause but had persistent chest pain were invited. The intervention was a 2-hour, weekly, online guided 8-week MBCT course. Compliance with attendance and the home practice was recorded. Enrolled patients completed the Seattle angina questionnaire (SAQ), Hospital Anxiety and Depression Scale, Cardiac Anxiety Questionnaire, Five-Facet Mindfulness Questionnaire, and Euro Quality of Life–5 Dimensions–5 Level at baseline assessment and after 8-week period. Results Persistent chest pain was reported by 114 patients. Of these, 33 (29%) patients with a mean age of 54.2 (±12.2) years and 68% women, consented to the study. Baseline questionnaires revealed mild physical limitation (mean SAQ, 76.8±25), high levels of anxiety (76%) and depression (53%), modest cardiac anxiety (CAQ,1.78±0.61) and mindfulness score (FFMQ, 45.5±7.3). Six patients subsequently withdrew due to bereavement, caring responsibilities and ill health. Of the remaining 27 participants, 18 in the intervention arm attended an average of 5 sessions with 61% attending ≥6 sessions. Although not statistically powered, the study revealed a significant reduction in general anxiety, improved mindfulness and a trend towards improvement in SAQ scores in the intervention arm. Conclusion One-third of patients with persistent non-cardiac chest pain were willing to participate in mindfulness-based therapy. An improvement in anxiety and mindfulness was detected in this feasibility study. A larger trial is required to demonstrate improvement in chest pain symptoms.

Abstract Funding Acknowledgements Type of funding sources: None. Background and Purpose Post-operative bradycardia is common but intrinsic rhythm often recovers. There is little consensus on the optimum time to wait between cardiac surgery and the implantation of a permanent pacemaker (PPM). Earlier device implantation may facilitate shorter length of stay. However, it may expose some patients who have no long-term pacing requirement to the risks associated with device therapy. This study aimed to understand how the number of days between cardiac surgery and PPM implantation is associated with pacing dependence and recovery of intrinsic conduction by 30 days. Methods We examined healthcare records of consecutive patients who underwent cardiac surgery at our centre between 01/01/2015 to 01/01/2021. The primary outcome measures were pacing dependence (PD) at 30 days and recovery of intrinsic conduction at 30 days. Recovery was defined as showing evidence of intrinsic rhythm at the 30 day check after being pacing dependent at the time of implant. Patient demographics, baseline ECG characteristics and surgical procedure were recorded. Time to pacemaker implantation and pacing indication were identified. Pacing checks at 30 days post implant were reviewed and PD defined as no intrinsic rhythm seen over a 30 second period with base rate set at 40bpm. Univariate analysis and binary logistic regression were used to determine factors significantly associated with the primary outcome measures. Subsequent receiver-operator characteristic (ROC) analysis was used to determine the optimal timing of pacemaker implantation as defined by the Youden Index. This aims to maximise sensitivity and specificity of days to implant in predicting PD and conduction recovery at 30 days. Results Following 5849 operations, 103 (1.8%) patients underwent PPM implantation for a new bradycardic indication. The baseline characteristics of those paced are summarised in table 1. Numerous factors were associated with pacing dependence at 30 days on univariate analysis (table 2). However, multivariate analysis showed that only PD at implant and days to implant (DTI) were significant predictors of PD at 30 days. The only significant association with conduction recovery was DTI. ROC analysis showed that the optimal DTI is 12 days for a variety of analyses: 1.) Predicting PD at 30 days for all patients (AUC 0.620, SE 0.056, p=0.031, 95% CI 0.511-0.730) 2.) Predicting PD in patients whose PPM indication was AV nodal dysfunction (AUC 0.706, p=0.001, Youden Index (YI) 1.34). 3.) Predicting recovery of intrinsic rhythm in patients who were pacing dependent at implant (AUC 0.80, p= 0.000, YI 1.515). Conclusions The number of days between surgery and pacemaker implantation is the only factor significantly associated with both pacing dependence and recovery of intrinsic conduction at 30 days. The optimum time to wait is 12 days to allow time for intrinsic conduction to recover.

Abstract Funding Acknowledgements Type of funding sources: Private grant(s) and/or Sponsorship. Main funding source(s): Boston Scientific Background A third of patients that receive cardiac resynchronisation (CRT) are non-responders. Predictors of positive response include broader QRS duration, non-ischaemic aetiology and sinus rhythm, but it is still unclear whether lead placement site determines a positive responder. Purpose We assessed the acute haemodynamic response of endocardial biventricular pacing in patients with intrinsic left bundle brunch block (LBBB) versus LBBB due to pre-existing right ventricular pacing (RVP). Methods Patients who fulfilled standard criteria for CRT implantation but had failed conventional (coronary sinus) left ventricular (LV) lead placement (primary or revision) or were deemed clinical ‘non-responders after > 6 months of conventional CRT were enrolled. The acute haemodynamic response during endocardial biventricular pacing was assessed with a roving LV lead at 9 different locations (basal and mid: septal, anterior, posterior and lateral walls and apex). Acute changes in beat-to-beat systolic blood pressure (SBP) in the left ventricle were recorded and analysed. Results We recruited 23 patients across 10 UK centres: 14 intrinsic LBBB and 9 dependent on RVP. Patient characteristics were comparable: age (mean 67 + 10.6 years vs. 62 + 15.4 years), ischaemic (63% vs 50%), QRS (160 + 18ms vs. 190 + 36ms, p =0.07). Of the RVP group 5/9 had septal RV leads (the remainder apically positioned). There was no difference in the SBP improvement between the groups: change in SBP ranged from -5.25 – 19.91mmHg (median 3mmHg) in RVP patients vs -5.92 – 23.03mmHg (median 3mmHg) for intrinsic LBBB. However, the improvement in SBP was more consistent across the different segments in the patients with RVP (group A), as compared to intrinsic LBBB (group B), where the lateral wall and then non-septal walls provided the greatest haemodynamic improvement. Figure 1: depicts SBP improvement (in mmHg) during endocardial biventricular pacing in different positions within the LV for patients with RVP (A) vs underlying intrinsic LBBB (B): 9 segment model of the LV: Ant (anterior wall), Lat (lateral wall), Post (posterior wall), Sept (septum). Outer ring represents the four basal LV locations, middle ring the mid LV locations and centre ring the apex. Scale depicts mmHg improvement in SBP. Conclusion When implanting an LV lead for patients who are RV pacing dependent any position within the LV provides an acute haemodynamic improvement, compared with those with intrinsic LBBB, where a targeted lateral wall approach is more important. This finding corroborates the key differences in LV activation patterns for induced versus intrinsic LBBB.

Cardiac implantable electronic devices (CIEDs) are the mainstay of care in patients with life-threatening ventricular arrhythmia, severe left ventricular impairment or symptomatic pauses. Magnetic resonance imaging (MRI) is the gold standard diagnostic tool for a wide variety of conditions however,

OBJECTIVES BRAVO (British Randomized Controlled Trial of AV and VV Optimization) is a multicenter, randomized, crossover, noninferiority trial comparing echocardiographic optimization of atrioventricular (AV) and interventricular delay with a noninvasive blood pressure method. BACKGROUND Cardiac resynchronization therapy including AV delay optimization confers clinical benefit, but the optimization requires time and expertise to perform. METHODS This study randomized patients to echocardiographic optimization or hemodynamic optimization using multiple-replicate beat-by-beat noninvasive blood pressure at baseline; after 6 months, participants were crossed over to the other optimization arm of the trial. The primary outcome was exercise capacity, quantified as peak exercise oxygen uptake. Secondary outcome measures were echocardiographic left ventricular (LV) remodeling, quality-of-life scores, and N-terminal pro-B-type natriuretic peptide. RESULTS A total of 401 patients were enrolled, the median age was 69 years, 78% of patients were men, and the New York Heart Association functional class was II in 84% and III in 16%. The primary endpoint, peak oxygen uptake, met the criterion for noninferiority (p(noninferiority) =0.0001), with no significant difference between the hemodynamically optimized arm and echocardiographically optimized arm of the trial (mean difference 0.1 ml/kg/min). Secondary endpoints for noninferiority were also met for symptoms (mean difference in Minnesota score 1; p(noninferiority) = 0.002) and hormonal changes (mean change in N-terminal pro-B-type natriuretic peptide -10 pg/ml; p(noninferiority) = 0.002). There was no significant difference in LV size (mean change in LV systolic dimension 1 mm; p(noninferiority) < 0.001; LV diastolic dimension 0 mm; p(noninferiority) < 0.001). In 30% of patients the AV delay identified as optimal was more than 20 ms from the nominal setting of 120 ms. CONCLUSIONS Optimization of cardiac resynchronization therapy devices by using noninvasive blood pressure is noninferior to echocardiographic optimization. Therefore, noninvasive hemodynamic optimization is an acceptable alternative that has the potential to be automated and thus more easily implemented. (C) 2019 The Authors. Published by Elsevier on behalf of the American College of Cardiology Foundation.

ImportanceThere is a need for better arrhythmic risk stratification in nonischemic dilated cardiomyopathy (DCM). Titin-truncating variants (TTNtvs) in theTTNgene are the most common genetic cause of DCM and may be associated with higher risk of arrhythmias in patients with DCM.ObjectiveTo determine if TTNtv status is associated with the development of life-threatening ventricular arrhythmia and new persistent atrial fibrillation in patients with DCM and implanted cardioverter defibrillator (ICD) or cardiac resynchronization therapy defibrillator (CRT-D) devices.Design, Setting, and ParticipantsThis retrospective, multicenter cohort study recruited 148 patients with or without TTNtvs who had nonischemic DCM and ICD or CRT-D devices from secondary and tertiary cardiology clinics in the United Kingdom from February 1, 2011, to June 30, 2016, with a median (interquartile range) follow-up of 4.2 (2.1-6.5) years. Exclusion criteria were ischemic cardiomyopathy, primary valve disease, congenital heart disease, or a known or likely pathogenic variant in the lamin A/C gene. Analyses were performed February 1, 2017, to May 31, 2017.Main Outcome and MeasuresThe primary outcome was time to first device-treated ventricular tachycardia of more than 200 beats/min or first device-treated ventricular fibrillation. Secondary outcome measures included time to first development of persistent atrial fibrillation.ResultsOf 148 patients recruited, 117 adult patients with nonischemic DCM and an ICD or CRT-D device (mean [SD] age, 56.9 [12.5] years; 76 [65.0%] men; 106 patients [90.6%] with primary prevention indications) were included. Having a TTNtv was associated with a higher risk of receiving appropriate ICD therapy (shock or antitachycardia pacing) for ventricular tachycardia or fibrillation (hazard ratio [HR], 4.9; 95% CI, 2.2-10.7;P < .001). This association was independent of all covariates, including midwall fibrosis measured by late gadolinium enhancement on cardiac magnetic resonance images (adjusted HR, 8.3; 95% CI, 1.8-37.6;P = .006). Having a TTNtv was also associated with the risk of receiving a shock (HR, 3.6; 95% CI, 1.1-11.6;P = .03). Individuals with a TTNtv and fibrosis had a greater rate of receiving appropriate device therapy than those with neither (HR, 16.6; 95% CI, 3.5-79.3;P < .001). Having a TTNtv was also a risk factor for developing new persistent atrial fibrillation (HR, 3.9; 95% CI, 1.3-12.0;P = .01).Conclusions and RelevanceHaving a TTNtv was an important risk factor for clinically significant arrhythmia in patients with DCM and ICD or CRT-D devices. Having a TTNtv, especially in combination with midwall fibrosis confirmed with cardiovascular magnetic resonance imaging, may provide a risk stratification approach for evaluating the need for ICD therapy in patients with DCM. This hypothesis should be tested in larger studies.

Introduction: In patients with a broad QRS and heart failure, the effects on acute systolic blood pressure (SBP) augmentation during biventricular pacing (BVP) when pacing at different LV endocardi...

OBJECTIVES: The intermediate-term all-cause mortality rate of real-world patients with multivessel disease (MVD) treated with percutaneous coronary intervention (PCI) with new-generation drug-eluting stents or coronary artery bypass grafting (CABG) remains unknown. We sought to compare the intermediate-term all-cause mortality rates of real-world patients with MVD including left main stem disease, treated with CABG or PCI. METHODS: All consecutive all-comer patients with MVD undergoing CABG or PCI with second/third generation drug-eluting stents from 2007 to 2015 in Harefield Hospital, UK were included in this study. The revascularization modality was based on heart team discussions. Primary outcome was all-cause mortality. Mean follow-up of the study was 3.3 years. Cox regression analysis and propensity matching were used. RESULTS: Of 6383 patients with MVD, 4230 underwent CABG, whereas 2153 had PCI with new-generation stents. In the CABG group, the mean age was 66.4 +/- 10 years, whereas in the PCI group it was 65.3 +/- 12.1 years (P < 0.001). Fewer female patients with MVD were treated with CABG than were treated with PCI (18.5% vs 20.5%; P = 0.026). There was a higher 5-year estimated survival rate among patients having CABG (88% vs 78.3%; Plog-rank < 0.001). The adjusted hazard ratio (HR) for PCI over CABG was 1.74 [95% confidence interval (CI) 1.41-2.16; P < 0.001]. A total of 653 patients having CABG and 653 having PCI were included in the propensity-matched groups. At mean follow-up, PCI was associated with a higher adjusted HR for all-cause mortality (2.18, 95% CI 1.54-3.1; P < 0.001). CONCLUSIONS: In this contemporary cohort of real-world patients with MVD, CABG was associated with increased intermediate-term survival compared to PCI with new-generation drug-eluting stents.

Balloon aortic valvuloplasty has a role in a select group of patients with severe aortic stenosis. Identifying those appropriate patients who will benefit most is key. Given previous evidence demonstrating that histologically the intervention involves a physical disrupting of the cusp's calcium we hypothesized that the quantity of calcium seen at CT will influence outcome. We examined our cohort of patients who had undergone balloon aortic valvuloplasty and CT-quantified aortic valve calcium (AVC) between July 2011 and April 2014. All patients underwent echocardiography pre- and post-procedure and for those patients managed medically, again at 6 months. A potential predictive AVC value for mortality was calculated using Youden's index. A total of 240 aortic valvuloplasties were performed in 206 patients (male = 124). Valvuloplasty caused a significant (pre 0.63 +/- 0.21 vs post 0.77 +/- 0.27 cm(2), p < 0.01, n = 240), but temporary (post 0.80 +/- 0.27 vs 6 months: 0.64 +/- 0.18 cm(2), p < 0.01, n = 88) increase in valve area. Those patients with a non-severe AVC (<1853.5 AU) had a larger increase in valve area after valvuloplasty compared with those with more calcium (0.10 [95% confidence interval {CI} 0.05 to 0.10] vs 0.15 [95% CI 0.10 to 0.22] cm(2), p = 0.049). Multivariate analysis revealed severe AVC (Hazard ratio 2.79, 95% CI 1.18 to 6.63, p = 0.02) along with pulmonary artery pressure post-valvuloplasty (Hazard ratio 1.02, 95% CI 1.00 to 1.03, p = 0.03) to be predictive of survival. In conclusion, in patients with severe aortic stenosis the degree of AVC impacts on the success of valvuloplasty. (C) 2018 Elsevier Inc. All rights reserved.

Studies comparing coronary artery bypass graft (CABG) and percutaneous coronary intervention (PCI) have largely been performed in the bare-metal stent (BMS) and first-generation drug eluting stent (F-DES) era. Second-generation DES (S-DES) have shown improved outcomes when compared to F-DES, but data comparing CABG with PCI using S-DES is limited. We compared mortality following CABG versus PCI for patients with multivessel disease and analyzed different stent types.A total of 6,682 patients underwent multivessel revascularization at Harefield Hospital, UK. We stratified CABG patients into single arterial graft (SAG) or multiple arterial grafts (MAG); and PCI patients into BMS, F-DES or S-DES groups. We analyzed all-cause mortality at 5 years.4,388 patients had CABG (n[SAG] = 3,358; n[MAG] = 1,030) and 2,294 patients had PCI (n[BMS] = 416; n[F-DES] = 752; n[S-DES] = 1,126). PCI had higher 5-year mortality with BMS (HR = 2.27, 95% CI:1.70-3.05, p<0.001); F-DES (HR = 1.52, 95% CI:1.14-2.01, p = 0.003); and S-DES (HR = 1.84, 95% CI:1.42-2.38, p<0.001). This was confirmed in inverse probability treatment weighted analyses. When adjusting for both measured and unmeasured factors using instrumental variable analyses, PCI had higher 5-year mortality with BMS (Δ = 15.5, 95% CI:3.6,27.5, p = 0.011) and FDES (Δ = 16.5, 95% CI:6.6,26.4, p<0.001), but had comparable mortality with CABG for PCI with SDES (Δ = 0.9, 95% CI: -9.6,7.9, p = 0.844), and when exclusively compared to CABG patients with SAG (Δ = 0.4, 95% CI: -8.0,8.7, p = 0.931) or MAG (Δ = 4.6, 95% CI: -0.4,9.6, p = 0.931).In this real-world analysis, when adjusting for measured and unmeasured confounding, PCI with SDES had comparable 5-year mortality when compared to CABG. This warrants evaluation in adequately-powered randomized controlled trials.

BACKGROUND:More than half of the patients undergoing percutaneous coronary intervention (PCI) have multivessel disease. Whether complete revascularization impacts long-term mortality or whether selected patients or those with specific coronary anatomy benefit from complete revascularization is unclear. METHODS:A total of 14,452 patients underwent PCI between 2004 and 2015 at Harefield Hospital, UK. Of these, 7,076 patients had multivessel disease. We excluded 321 patients with left main-stem stenosis ≥50%, with 6,755 patients included in the analysis (936 patients had complete revascularization). RESULTS:The unadjusted 3-year mortality rates were lower with complete revascularization (10.8% vs 13.1%, P = 0.047). However, multivariable-adjusted analyses indicated that complete revascularization was not independently associated with mortality (HR = 1.01, 95% CI: 0.78-1.31, P = 0.939). These findings were unchanged when addressing measured confounding using propensity-matched analyses (HR = 1.16, 95% CI: 0.81-1.65, P = 0.417) and inverse probability treatment weighted analyses (HR = 1.01, 95% CI: 0.77-1.33, P = 0.950); and unmeasured confounding using instrumental variable analyses (Δ = 0.9%, 95% CI: -2.5%, 4.3%, P = 0.958). There was no association with mortality and untreated LAD disease (HR = 0.92, 95% CI: 0.72-1.17, P = 0.482) and LCx disease (HR = 0.90, 95% CI: 0.74-1.10, P = 0.999). However, untreated proximal LAD disease (HR = 1.23, 95% CI: 1.06-1.51, P = 0.045) and RCA disease (HR = 1.36, 95% CI: 1.08-1.65, P = 0.007) was associated with increased mortality, particularly in patients with ST-elevation acute coronary syndrome (STEACS). CONCLUSIONS:In this study of unselected patients undergoing PCI, complete revascularization did not confer a mortality benefit. However, the presence of untreated proximal LAD and RCA disease was prognostic in patients with STEACS. Thus, complete revascularization may be considered in select patient groups with anatomical subsets of coronary disease.