Sedation strategies for defibrillation threshold testing: safety outcomes with anaesthesiologist compared to proceduralist-directed sedation: an analysis from the SIMPLE study Free

Abstract

Introduction

The implantable cardioverter-defibrillator (ICD) is a well-established therapy for patients who have suffered a life-threatening ventricular arrhythmia or are at high risk for sudden cardiac death.^1–3 Traditionally defibrillation threshold (DT) testing, which is performed to assess device detection and termination of ventricular fibrillation has been performed under general anaesthesia.⁴ While the use of general anaesthesia provides ideal sedation for patient comfort, as DT testing can be painful and traumatic; a dedicated anaesthesiologist can also manage airway and haemodynamic issues that may occur among this patient population, with a high prevalence of comorbidities. However, anaesthesiologist-directed sedation (ADS) may increase procedural time, increase use of procedures such as arterial lines, have higher cardiorespiratory complications and be more costly.^5–7 Given this, DT testing is now occurring under the direction of implant operators using moderate sedation and analgesia.

However, data regarding the safety of proceduralist-directed sedation (PDS) for DT testing is sparse.^8–11 In two retrospective, single-centre studies, PDS using midazolam was related to respiratory depression leading to acute hypoxia in 5.9% of patients, all successfully treated with head and chin tilt and manual ventilation¹⁰ while 10% of patients undergoing PDS with propofol for DT testing had a serious adverse event and 38.7% experienced a non-serious adverse event.⁸ A single observational study comparing ADS with PDS found the use of intravenous etomidate to induce deep sedation for DT testing in the absence of an anaesthesiologist or the use of propofol in the presence an anaesthesiologist evoked no difference in episodes of arterial hypotension, hypoxia, or in-hospital mortality.¹¹

At present, no standard practice exists with regards to who should perform sedation at the time of DT testing. Therefore, we evaluated adverse and safety outcomes among 1242 patients undergoing DT testing with either ADS or PDS in the Shockless Implant Evaluation (SIMPLE) trial.

Methods

Study population

The SIMPLE trial design and results have been previously published.¹² Briefly, this is a single-blind, randomized, clinical trial comparing the efficacy and safety of ICD implantation with and without DT testing at 85 hospitals in 18 countries worldwide. A total of 2500 patients were included in this study and followed for a mean of 3.1 years. For our analysis, we included the 1253 patients who were randomized to DT testing. Data on intra-procedural sedation at time of DT was present for 1242 patients. The DT protocol required induction of ventricular fibrillation to show either one successful arrhythmia cessation at 17 J or two successful cessations at 21 J with reconfiguration and retesting if initial DT testing was unsuccessful. Among the 85 enrolling centres there were 46 sites where DT was performed with PDS and 39 who performed DT with anaesthesia support.

Outcomes

We assessed two adverse composite outcomes (intraoperative and in-hospital) and two 30-day safety composite outcomes from the main trial with use of ADS vs. PDS.¹² Intraoperative adverse events included need for chest compressions or an aortic balloon pump during the ICD implantation procedure, non-elective intubation, complications from arterial line insertion, and intraoperative hypotension necessitating use of vasoconstrictors for more than 15 min. In-hospital adverse events included death, myocardial infarction (MI), stroke (CVA), transient ischaemic attack, systemic or pulmonary embolism, and heart failure requiring intravenous diuretics or inotropes.

The primary composite safety endpoint included death, MI, CVA, systemic or pulmonary embolism, heart failure requiring intravenous diuretics or inotropes, the need for chest compressions or an aortic balloon pump during the ICD implantation procedure, use of intraoperative vasoconstrictors for more than 15 min, non-elective intubation, arterial-line complications, an unplanned stay in the intensive care unit, other anoxic brain injury, pneumothorax, cardiac perforation, ICD infection, or aspiration pneumonia. A secondary composite safety endpoint included everything from the primary endpoint except anoxic brain injury, aspiration pneumonia, pneumothorax, pericarditis, cardiac tamponade, and ICD infection, was pre-specified to examine complications thought to be more directly related to induction of ventricular fibrillation. All adverse and safety outcomes were adjudicated by a committee blinded to randomization group.

Statistical analysis

Baseline and implant characteristics of patients according to sedation provider were summarized as mean ± standard deviation for continuous variables, and frequency/percentage for categorical variables. To test baseline differences in patient demographic and clinical characteristics, a two sample t-test was used for continuous variables and χ² or Fisher exact test was used for categorical variables. Logistic regression was used for baseline characteristics to determine predictors associated with ADS with forward selection and a significant P-value of 0.1 was needed for entry into the model. The associated between sedation approach and risk of clinical outcomes were investigated using logistic regression and odds ratios were calculated [odds ratio (OR), 95% confidence interval (CI)] for the relative chance of an event (non 0) happening with ADS compared to PDS. We also performed non-parsimonious logistic regression analyses and propensity score adjusted analyses for the primary and secondary safety composite endpoints and the intraoperative composite safety outcome. The models were adjusted according to variables that were clinically important and variables that were significant in univariate analysis. The non-parsimonious logistic regression models were adjusted for age, sex, left ventricular ejection fraction, systolic blood pressure, New York Heart Association (NYHA) class, underlying coronary artery disease, history of cardiac arrest, and impaired renal function. The propensity score analysis adjusted for age, sex, left ventricular ejection fraction, systolic and diastolic blood pressure, NYHA class, underlying coronary artery disease, beta blocker use, history of cardiac arrest, and impaired renal function. Additional analyses was not performed for in-hospital adverse composite endpoint because of the small number of events. All tests were two-sided and conducted at the 0.05 level. Analyses were conducted in SAS 9.4 (SAS Institute, Inc., Cary, NC, USA).

Results

Baseline characteristics

The baseline and implant characteristics for DT testing according to ADS and PDS are shown in Table 1. Compared to DT testing performed by PDS, patients who had undergone DT with ADS were younger (62 ± 12 years vs. 64 ± 12 years, P = 0.01), had lower blood pressure (systolic blood pressure 122 ±18 mmHg vs. 126 ± 20 mmHg, P = 0.001 and diastolic blood pressure 72 ±11 mmHg vs. 74 ± 12 mmHg, P = 0.003), received an ICD for a primary prevention indication (78.4% vs. 69.2%, P = 0.0002), had a lower ejection fraction (33% ± 13% vs. 31% ± 12%, P = 0.03), and received more arterial lines for invasive haemodynamic monitoring (43% vs. 8%, P = <0.0001). In the ADS group, the intraoperative sedation was more likely inhaled anaesthetics (15% vs. 0%, P = <0.0001), systemic narcotics i.e. fentanyl (62% vs. 31%, P = <0.001), and propofol (72% vs. 33%, P = <0.001).

Independent predictors for anaesthesiologist-directed sedation

Independent predictors for use of ADS for DT testing included the presence of coronary artery disease (OR 1.60, 95% CI 1.05–2.72; P = 0.03) and hypertrophic cardiomyopathy (OR 2.64, 95% CI 1.19–5.85; P = 0.02, Table 2). A reduced left ventricular ejection fraction, history of ventricular tachycardia, history of cardiac arrest, or NYHA III or IV heart failure symptoms were not significantly associated with use of ADS.

Adverse and safety outcomes

Overall, there were 14 intraoperative adverse events in ADS group (2.2%) compared to 3 (0.5%) with the PDS approach (P = 0.01, Table 3). Intraoperative composite adverse events were significantly higher in the ADS group compared to DT testing performed by PDS (OR 4.70, 95% CI 1.35–16.5; P = 0.02). This association remained significant after multivariable (OR 4.60, 95% CI 1.29–16.5; P = 0.02) and propensity score (OR 4.47, 95% CI 1.25–16.0; P = 0.02) adjustment. This was primarily driven by higher rate of intraoperative hypotension necessitating use of vasoconstrictors for more than 15 min (OR 8.01, 95% CI 1.00–64.2; P = 0.05, Supplementary material online, Table S1). Six patients required non-elective intubation during DT testing in the ADS group, vs. one patient in the PDS group and two patients experienced arterial line complications in the ADS group and none with a PDS approach (Supplementary material online, Table S1). There was no significant association between in in-hospital adverse events and approach to sedation (P = 0.41). No in-hospital deaths occurred for either ADS or PDS but there were higher heart failure events requiring intravenous diuretics or inotropes in the ADS group (1.3% vs. 0.5%, P = 0.22).

At 30-days, there were 51 (8.2%) primary safety outcomes in ADS group compared to 30 primary safety outcomes in PDS (4.9%, P = 0.02). The primary safety composite outcome was higher in ADS compared to PDS (OR 1.74, 95% CI 1.10–2.78; P = 0.02; multivariable-adjusted: OR 1.72, 95% CI 1.06–2.81; P = 0.03; propensity score adjusted: OR 1.72, 95% CI 1.06–2.80; P = 0.03), however, there was no significant difference in the secondary safety composite outcome (P = 0.11). In the ADS group five patients died and no patient died with PDS (Supplementary material online).

Discussion

To the best of our knowledge, this is the first study evaluating adverse and safety outcomes with ADS compared to PDS in a large cohort of patients who had undergone DT testing at the time of ICD implant. Intraoperative and in-hospital adverse events occurred in <1% of patients undergoing PDS for DT testing. In the ADS group, there was a 4.5 times higher odds associated with intraoperative adverse events and 1.7 times higher odds of the primary safety composite outcome at 30-days compared to PDS group. Although, higher rates of in-hospital adverse events and secondary 30-day safety composite events also occurred with ADS, no significant differences were observed between these two approaches.

The safety of PDS for DT testing at ICD implant has been investigated with few single-centre, retrospective studies.^8–10 Only a single study from the late 1990’s has compared ADS and PDS approaches.¹¹ We found an intraoperative adverse event rate of 0.5% in patients undergoing DT testing using PDS with a variety of anaesthetic agents, which is lower than prior studies that reported adverse event rates ranging between 5.9% and 9.2% using midazolam^9,10 and 10% for serious adverse events and 38.7% for non-serious adverse events when using propofol.⁸ However, the definition of adverse events differed among studies. When comparing similar adverse events i.e. hypotension requiring inotropic support, we found a rate of 0.2% compared to 4.6% and 8.1% from two retrospective analyses using midazolam or propofol, respectively.^8,9 Respiratory depression requiring intubation rarely occurred in our study and others with a PDS approach and there were no documented intraoperative deaths.^8–11 When comparing PDS to ADS, we found higher rates of both intraoperative hypotension requiring inotropic support and need for intubation with ADS for DT testing. Our study is novel in assessing in-hospital adverse events and 30-day safety outcomes for both ADS and PDS approaches for DT testing. We found higher event rates with ADS compared to PDS for each of these additional outcomes and there were five deaths at 30-days with ADS.

The higher adverse event rates found in the ADS group compared to the PDS group in our study and other work with a PDS only approach^8,13 may, in part, be explained by the use of propofol. This anaesthetic agent was used in 71.5% of our patients undergoing DT testing using ADS but only 32.8% using a PDS approach. Propofol is known negative inotrope and has been associated with cardiovascular complications in patients with advanced stages of heart failure and diminished cardiac reserve capacity.¹⁴ Although, there was no difference in NYHA class between patients in the ADS and PDS groups, left ventricular ejection fraction, and blood pressure were both lower in the ADS group. In the study using PDS with propofol, patients had more severe heart failure and reduced ejection fraction than our study population regardless of whether patients were in the ADS or PDS group.⁸ Worsening NYHA class and use of propofol infusion were independent predictors of the high non-serious adverse event rate. We found the most common type of anaesthetic agent used in the PDS group was Benzodiazepines (e.g. midazolam). Prior studies have found fewer intraoperative adverse events with use of midazolam for PDS of DT testing compared to propofol and shown to be safe for a cardiologist-only approach to sedation for electrical cardioversion of atrial fibrillation.^15,16 Although we didn’t specifically obtain data regarding amnesia related to sedation, this has been documented with a PDS approach to DT testing using etomidate, propofol, and midazolam.^10,11

Clinical implications

The safety regarding approach to sedation for DT testing is important despite increasing evidence that DT testing does not improve mortality or ICD efficacy and may not be necessary for de novo implants^12,17,18 because there are considerable populations where data for the elimination of DT testing has not been clearly demonstrated such as device replacement, secondary prevention indication, hypertrophic cardiomyopathy, channelopathies, and subcutaneous implants. In addition, DT testing is still being performed at first implant in many centres, ranging anywhere from 20% to 71%^5,19 and likely reflects physician preference to ensure systemic integrity, reliable sensing and to allow for immediate system revision for DT failures. PDS for DT testing is appealing because it eliminates challenges related to scheduling of anaesthesiology, particularly a cardiac anaesthetist, may reduce procedure and staff time, have fewer cardiorespiratory complications and lower costs.^5–7 This approach should require an EP lab that is well equipped to deal with emergencies; nurses appropriately trained with sedation and experienced operators.

Limitations

There are several limitations of our study. Firstly, this is a retrospective post hoc analysis of a randomized control trial that was conducted to answer a different question, namely the utility of DT testing at ICD implant. Secondly participants in this study were not randomized to PDS or ADS, thus potentially resulting in a selection bias. However, we tried to address this through multivariable logistic regression and propensity score adjusted analyses. Sedation strategies were decided based on the site practices and operator preference. Third, because event rates were low, we were unable to evaluate whether the type of sedation used was associated with adverse and safety outcomes. Fourth, we did not capture the type of narcotic used. Finally, the cost-effectiveness of PDS was not studied however, prior work has demonstrated that PDS results in lower costs.^5–7

Conclusions

This large study of complex cardiac patients undergoing ICD implant demonstrates that PDS is safe for DT testing and compared to ADS has significantly fewer intraoperative adverse events and safety outcomes at 30-days. There were no differences in the in-hospital adverse events between the two approaches to sedation although event rates were higher with ADS. Further confirmatory research is needed to help develop practice standards regarding sedation approach for DT testing.

Supplementary material

Supplementary material is available at Europace online.

Acknowledgements

Dr Healey has a personnel award from the Heart and Stroke Foundation, Ontario Provincial office (MC7450).

Funding

The SIMPLE trial was funded by Boston Scientific.

Conflict of interest: none declared.

References