This MTA sought to assess the available evidence to support the use of dual-chamber pacemakers to treat symptomatic bradycardia due to SSS without AV block in comparison with single-chamber atrial pacemakers. It is a partial update of NICE’s TA88,18 which had a wider remit investigating dual-chamber pacemakers for the treatment of symptomatic bradycardia due to SSS and/or AV block. With regard to the subset of patients of interest to this research, NICE’s TA88 recommends single-chamber atrial pacemakers for patients with SSS in whom, after full evaluation, there is no evidence of impaired AV conduction.
The TAG’s systematic review of the clinical effectiveness identified six RCTs in the population of interest. Three RCTs19,37,41 had a parallel-group design and three were crossover studies.34,35,43 The crossover trials were generally small (12–21 patients) with limited follow-up (up to 3 months), which limited their opportunity to inform the outcomes of interest for this research. The parallel-group RCTs were relatively large (50–1415 patients), had longer follow-up than the crossover studies (1–5.4 years) and measured outcomes that were of direct interest to this research.
There was limited opportunity to combine the results using meta-analysis from the six RCTs identified from the published literature. When this was possible, the results were predominantly influenced by the largest trial, DANPACE,19 which accounted for over 80% of the weight in change in pacing mode, all-cause mortality and stroke. In no instance did the level of significance of an outcome from the DANPACE trial19 change as a result of its combination in a meta-analysis of that outcome [e.g. the OR for change in pacing mode was 0.52 (95% CI 0.38 to 0.71) from the DANPACE trial19 and 0.50 (95% CI 0.37 to 0.67) from the meta-analysis including the DANPACE trial,19 Albertsen et al.37 and Nielsen et al.41].
In this review, dual-chamber pacing was associated with a lower risk of AF and fewer reoperations than single-chamber atrial pacing. No statistically significant difference between the pacing modes was found for mortality, HF, stroke or QoL, and there were limited data on adverse effects of pacemaker implantation. However, in the case of patients younger than 75 years, the risk of HF seems to be higher with a dual-chamber pacemaker than with a single-chamber atrial pacemaker, and for patients older than 75 years the risk seems to be lower with dual-chamber pacing compared with single-chamber atrial pacing.
The DANPACE study19 is a relatively large trial of good quality with long follow-up, which gives a reasonable evidence base for dual-chamber pacing compared with single-chamber atrial pacing for people with SSS without evidence of impaired AV conductance. Although the time horizon in the DANPACE trial19 was reasonable, the increase in AV block increases the number of patients requiring a change in pacing mode which requires an operation for patients with a single-chamber pacemaker. Additionally, the DANPACE trial19 did not allow pacemaker algorithms designed to minimise VP in patients with intact AV conduction, which have become more common since the start of this trial. Although the DDDR pacemakers in the DANPACE trial19 were programmed in a way intended to reduce unnecessary VP, the incidence of VP was still 65% (SD 33%), which may offset some of the benefit of implanting a dual-chamber pacemaker.
Patients with single-chamber atrial pacing who do not go on to develop AV block will be paced appropriately and avoid any unnecessary VP, which may have adverse consequences for cardiac function. Implanting a single-chamber atrial pacemaker may also have additional benefits in terms of reducted time for the implantation procedure, a lower risk of complications associated with the implantation of a second lead and shorter (i.e. less time) follow-up appointments. However, patients who have a dual-chamber pacemaker implanted and who go on to develop AV block will be protected by the presence of a ventricular lead and will not need a further operation to upgrade the pacemaker and insert a second lead, which is likely to be associated with higher risk of complications than first-time implantation. Additionally, the DANPACE trial19 has shown that the risk of developing paroxysmal AF is lower with dual-chamber pacing than with single-chamber atrial pacing. In addition, subgroup analysis identified that, in patients younger than 75 years, the risk of HF may be higher with a dual-chamber pacemaker than with a single-chamber atrial pacemaker, whereas, in patients older than 75 years, the risk may be lower with dual-chamber pacing than with single-chamber atrial pacing.
The systematic review49 of existing cost-effectiveness analyses identified only one study that compared dual-chamber pacemakers and single-chamber atrial pacemakers for the treatment of symptomatic bradycardia due to SSS without AV block in a UK setting and which was based on the research carried out to inform NICE’s TA88.18 In addition, in the update of the cost-effectiveness systematic review, a study by Oddershede et al.54 was identified, which was based on the perspective of the Danish health-care system. This study is of particular interest, as it includes the DANPACE trial19 as well as two other Danish RCTs: the pilot study for the DANPACE trial (Nielsen et al.41) and Andersen et al.45
One of the strengths of the Oddershede et al.54 approach is that it was based on IPD that allowed the researchers to account for baseline characteristics such as age, sex, previous myocardial infarction and history of AF. The researchers were also able to categorise patients as low risk or high risk (and, by deduction, the remainder of patients as at moderate risk) of a subsequent event. For each of the risk categories, and for an evaluation based on all patients, Oddershede et al.54 found that the probability of dual-chamber pacemakers being cost-effective compared with single-chamber atrial pacemakers was > 50% at a WTP threshold of £20,000. This fell to > 40% at a WTP threshold of £30,000. This is likely to be a result of the incremental QALY decrement associated with dual-chamber pacemakers in their analysis. However, the model developed by Oddershede et al.54 focused primarily on the occurrence of stroke and death, which may have restricted the comprehensiveness of the analysis to assess fully costs and benefits.
As no pre-existing economic evaluation adequately presents the cost-effectiveness of dual-chamber pacemakers in comparison with single-chamber atrial pacemakers for the treatment of symptomatic bradycardia due to SSS without AV block in a UK setting, the TAG developed a de novo economic model to help inform this important question.
As there were concerns around potential clinical heterogeneity as a result of different patient populations (e.g. prior history of AF) and different device programming used (e.g. different per cent VP) in the RCTs identified, the decision was made to base the model on the DANPACE trial. The base-case results of the TAG’s economic model demonstrate that dual-chamber pacemakers are more expensive but also more effective than single-chamber atrial pacemakers, resulting in an ICER of £6506. Probabilistic sensitivity analysis reduced this figure to £5989, principally because the incremental cost was lowered. This reduction in the difference in cost is likely to be because of the non-linearity of the minimum–maximum cost of implant/implantation of a single pacemaker compared with the minimum–maximum cost of implant/implantation of a dual pacemaker. The likelihood that dual-chamber pacemakers are cost-effective was found to be over 70% at a threshold of either £20,000 or £30,000.
As the deterministic results and the probabilistic results were so similar all subsequent analyses were based on the deterministic model.
A structural sensitivity analysis looking at a more granular approach to incorporating risk of reoperation using the available Kaplan–Meier data from the DANPACE trial19 reduced the ICER from £6506 to £3425. A second structural sensitivity analysis reducing the time horizon to 5 years more than doubled the base-case ICER to £14,261. In essence, halving the time horizon halved the incremental benefit.
One-way sensitivity analysis highlighted the key drivers of cost-effectiveness in the economic model. Those likely to increase the deterministic ICER to over £20,000 were:
lowest risk of HF (dual-chamber pacemakers dominated by single-chamber atrial pacemakers)
highest cost of implant/procedure for dual-chamber pacemakesr (ICER of £23,010)
lowest cost of implant/procedure for single-chamber atrial pacemakers (ICER of £27,409).
The result for the lowest risk of HF is being driven by an increase in cost of £710 and a modest reduction in benefit (–0.01 QALYs) compared with dual-chamber pacemakers.
A series of scenario analyses were undertaken to test the impact on the results when using alternative sources for parameter estimates or challenge assumptions in the model. The scenario analyses that raised the ICER above the base case were:
assuming no difference in HF (ICER of £22,213)
using the risk of stroke from the TAG’s meta-analysis (ICER of £6438)
using spell-level costs of pacemaker implantation (ICER of £7605)
using monthly cost of HF from NICE’s TA88
18 (ICER of £7140)
using reprogramming/device replacement for AF of 0% (ICER of £10,897)
using a discount rate of 6% (ICER of £6938).
Only when we assume that the risk of developing HF is the same regardless of implanted device does the ICER increase beyond £20,000, albeit by a modest amount, to £22,213.
A cumulative worst-case scenario was also conducted that combined the monthly cost of HF from NICE’s TA88,18 the risk of stroke from the meta-analysis conducted by the TAG, the spell-level costs of implantation, reprogramming/device replacement for AF of 0% and the assumption of no difference in risk of developing HF between the two types of implant. This resulted in an ICER of £48,738.
The results of the scenario analysis and the OWSA highlight how sensitive the results are to risk of HF, with dual-chamber pacemakers being considered cost-effective or dominated by single-chamber atrial pacemakers depending on the data used. Subgroup analysis from the DANPACE trial19 identified a significant difference in HF owing to age (p = 0.05), but no significant differences were found in each subgroup (p > 0.31).40 When the risk of HF is assessed by age, the ICER is reduced compared with the base case in patients aged > 75 years (£4918 vs. £6506, respectively), whereas dual-chamber pacemakers are dominated by single-chamber atrial pacemakers in patients aged ≤ 75 years (i.e. they are more costly and less effective).
Statement of principal findings
This MTA uses the best available evidence to explore the clinical and cost-effectiveness implications for using dual-chamber pacemakers rather than single-chamber atrial pacemakers to treat symptomatic bradycardia due to SSS without AV block. The DANPACE trial19 found that dual-chamber pacemakers significantly reduced the risk of reoperation to change mode of pacing compared with single-chamber atrial pacemakers (9.3% vs. 0.6%, p < 0.001). The difference is primarily because of the development of AV block requiring upgrade to a dual-chamber device. The DANPACE trial19 also demonstrated a reduced risk of paroxysmal AF with dual-chamber pacing compared with single-chamber atrial pacing (OR 0.75, 95% CI 0.59 to 0.96). No statistically significant difference between the pacing modes was found for mortality, HF, stroke or QoL. However, the risk of developing HF may vary with age and device.
The de novo economic model developed by the TAG shows that dual-chamber pacemakers are more expensive and more effective than single-chamber atrial devices resulting in a base case ICER of £6506. The ICER remains below £20,000 in probabilistic sensitivity analysis, structural sensitivity analysis and most scenario analyses and OWSA.
A potentially important finding of this MTA is the impact that HF may have on the decision to use dual-chamber pacemakers or single-chamber atrial pacemakers to treat symptomatic bradycardia due to SSS without AV block. The results of an analysis based on age (> 75 years or ≤ 75 years) and risk of HF indicate that dual-chamber pacemakers are the dominant treatment option in older patients (i.e. are less costly and more effective), whereas, in younger patients, dual-chamber pacemakers are dominated by single-chamber atrial pacemakers (i.e. the former are more expensive and less effective than the latter). However, these results are based on a subgroup analysis and should be treated with caution.
Strengths and limitations of the assessment
Weaknesses
The number of RCTs available to inform this decision question was limited, and those trials that were identified failed to report the results in a consistent manner.
The technologies under investigation are developing rapidly, such that the single-chamber atrial pacemakers or dual-chamber pacemakers used in current trials are likely to be superseded by newer implants (and/or pacing algorithms) prior to their completion.
A cohort approach using the adjusted trial level data from the DANPACE trial
19 was used to populate the efficacy parameters within the economic model rather than a microsimulation informed by IPD.
The costs of the individual pacemakers under consideration were unavailable for use within the economic model and so the average costs reported within the appropriate HRG codes were used.
Uncertainties
The DANPACE study19 is the single largest RCT that has been conducted to compare single-chamber atrial pacemakers and dual-chamber pacemakers in patients with symptomatic bradycardia due to SSS and no evidence of AV block. However, it does not conclusively answer the clinically relevant questions concerning a difference in risk of HF, stroke and all-cause mortality. It seems unlikely that larger studies will be conducted to investigate these outcomes, and will not use the same pacemakers used in the DANPACE trial,19 as pacemaker design and pacing modes have rapidly changed over time and look likely to continue to change in the future.
Typically in a cost-effectiveness analysis, the acquisition costs of the interventions are known and the uncertainty in costs lies elsewhere. However, as the manufacturers declined the opportunity to make a submission and were unable to supply costs for devices in the time allowed, the costs for the individual pacemakers under consideration in this MTA were unavailable. We had to use the average costs reported within the appropriate HRG codes, which incorporate the cost of device plus the cost of implantation. There was considerable uncertainty in the economic evaluation as a result of implementing these costs. It was not possible to disentangle the uncertainty relating to cost of devices and cost of implantation.
Other relevant factors
Based on criteria outlined by NICE, the TAG considers that neither dual-chamber pacemakers nor single-chamber atrial pacemakers are eligible for consideration as end-of-life treatments.
