Advertisement

Cost-Effectiveness Analysis of SAPIEN 3 Transcatheter Aortic Valve Implantation Procedure Compared With Surgery in Patients With Severe Aortic Stenosis at Low Risk of Surgical Mortality in France

Open AccessPublished:November 17, 2021DOI:https://doi.org/10.1016/j.jval.2021.10.003

      Abstract

      Objectives

      The clinical and cost-saving benefits of transcatheter aortic valve implantation (TAVI) over surgical aortic valve replacement (SAVR) in patients with severe aortic stenosis who are at high or intermediate risk of surgical mortality are supported by a growing evidence base. The PARTNER 3 trial (Placement of AoRTic TraNscathetER Valve Trial) demonstrated clinical benefits with SAPIEN 3 TAVI compared with SAVR in selected patients at low risk of surgical mortality. This study uses PARTNER 3 outcomes in combination with a French national hospital claim database to inform a cost-utility model and examine the cost implications of TAVI over SAVR in a low-risk population.

      Methods

      A 2-stage cost-utility analysis was developed to estimate changes in both direct healthcare costs and health-related quality of life using TAVI with SAPIEN 3 compared with SAVR. Early adverse events associated with TAVI were captured using the PARTNER 3 data set. These data fed into a Markov model that captured longer-term outcomes of patients, after TAVI or SAVR intervention.

      Results

      TAVI with SAPIEN 3 offers meaningful benefits over SAVR in providing both cost saving (€12 742 per patient) and generating greater quality-adjusted life-years (0.89 per patient). These results are robust with TAVI with SAPIEN 3 remaining dominant across several scenarios and deterministic and probabilistic sensitivity analyses.

      Conclusions

      This model demonstrated that TAVI with SAPIEN 3 was dominant compared with SAVR in the treatment of patients with severe symptomatic aortic stenosis who are at low risk of surgical mortality. These findings should help policy makers in developing informed approaches to intervention selection for this patient population.

      Keywords

      Introduction

      Since the first intervention was performed in Rouen, France, by Alain Cribier in 2002, transcatheter aortic valve implantation (TAVI) has emerged as the first option therapy in inoperable or patients at high risk of surgical mortality with severe aortic stenosis. Currently, the number of TAVIs exceeds the number of surgical aortic valve replacement (SAVR) in many countries, including France.
      MCO par diagnostic ou acte. ScanSanté.
      There exists real-world evidence supporting the beneficial performance of SAVR in patients at low risk of surgical mortality.
      • Thourani V.H.
      • Suri R.M.
      • Gunter R.L.
      • et al.
      Contemporary real-world outcomes of surgical aortic valve replacement in 141,905 low-risk, intermediate-risk, and high-risk patients.
      The PARTNER 3 study was a pivotal multicenter randomized controlled trial that compared treatment outcomes of the SAPIEN 3 transcatheter heart valve (Edwards Lifesciences, Irvine, CA) delivered through a transfemoral approach with SAVR in selected patients with symptomatic severe aortic stenosis, who were evaluated as at low risk of death with surgery (Society of Thoracic Surgeons risk of mortality calculation < 4%).
      • Mack M.J.
      • Leon M.B.
      • Thourani V.H.
      • et al.
      Transcatheter aortic-valve replacement with a balloon-expandable valve in low-risk patients.
      The study demonstrated that TAVI with SAPIEN 3 reduced death, stroke, or rehospitalization after 1 year in this low-risk population compared with SAVR.
      • Mack M.J.
      • Leon M.B.
      • Thourani V.H.
      • et al.
      Transcatheter aortic-valve replacement with a balloon-expandable valve in low-risk patients.
      An improvement in efficacy over previous versions of the device used in the original PARTNER trial (SAPIEN valve, in patients ineligible for surgery or at high risk of mortality) and in PARTNER 2 (SAPIEN-XT valve, in patients at intermediate risk) was demonstrated.
      • Leon M.B.
      • Smith C.R.
      • Mack M.
      • et al.
      Transcatheter aortic-valve implantation for aortic stenosis in patients who cannot undergo surgery.
      • Smith C.R.
      • Leon M.B.
      • Mack M.J.
      • et al.
      Transcatheter versus surgical aortic-valve replacement in high-risk patients.
      • Leon M.B.
      • Smith C.R.
      • Mack M.J.
      • et al.
      Transcatheter or surgical aortic-valve replacement in intermediate-risk patients.
      There is a trend toward using TAVI in patients at lower risk of surgical mortality, and support for this approach was presented in two 5-year data sets after TAVI procedures in patients at high or intermediate surgical risk.
      • Mack M.J.
      • Leon M.B.
      • Smith C.R.
      • et al.
      5-year outcomes of transcatheter aortic valve replacement or surgical aortic valve replacement for high surgical risk patients with aortic stenosis (PARTNER 1): a randomised controlled trial.
      ,
      • Makkar R.R.
      • Thourani V.H.
      • Mack M.J.
      • et al.
      Five-year outcomes of transcatheter or surgical aortic-valve replacement.
      Nevertheless, before PARTNER 3, there was a scarcity of evidence in patients at low risk.
      • ElGuindy A.
      PARTNER 2A & SAPIEN 3: TAVI for intermediate risk patients.
      The recent American College of Cardiology/American Heart Association guidelines on valvular heart disease recommend the use of TAVI for all severe aortic stenosis patients, whatever the risk score and the stratification as made by age, life expectancy, and patient’s anatomy (transfemoral approach feasibility).
      • Otto C.M.
      • Nishimura R.A.
      • et al.
      Writing Committee Members
      2020 ACC/AHA guideline for the management of patients with valvular heart disease: executive summary: a report of the American College of Cardiology/American Heart Association joint Committee on Clinical Practice guidelines [published correction appears in J Am Coll Cardiol. 2021;77(9):1276].
      As the clinical characteristics of TAVI become more clearly defined, demonstration of the cost-effectiveness of the procedure becomes essential in supporting policy decisions and healthcare resource allocation. Previous studies have examined the cost-effectiveness case using older iterations of valve replacement technology in relatively high-risk patients, with mixed findings.
      • Orlando R.
      • Pennant M.
      • Rooney S.
      • et al.
      Cost-effectiveness of transcatheter aortic valve implantation (TAVI) for aortic stenosis in patients who are high risk or contraindicated for surgery: a model-based economic evaluation.
      • Tarride J.E.
      • Luong T.
      • Goodall G.
      • Burke N.
      • Blackhouse G.
      A Canadian cost-effectiveness analysis of SAPIEN 3 transcatheter aortic valve implantation compared with surgery, in intermediate and high-risk severe aortic stenosis patients.
      • Baron S.J.
      • Wang K.
      • House J.A.
      • et al.
      Cost-effectiveness of transcatheter versus surgical aortic valve replacement in patients with severe aortic stenosis at intermediate risk.
      A recent 10-year cost study using the latest-generation TAVI devices found notable cost saving with TAVI over SAVR in patients at intermediate risk of surgical mortality.
      • Zhou J.
      • Liew D.
      • Duffy S.J.
      • Walton A.
      • Htun N.
      • Stub D.
      Cost-effectiveness of transcatheter aortic valve implantation compared to surgical aortic valve replacement in the intermediate surgical risk population.
      In patients at low risk, TAVI, particularly with balloon-expandable prostheses, was suggested to be a cost-effective option for patients with severe aortic stenosis.
      • Tam D.Y.
      • Azizi P.M.
      • Fremes S.E.
      • Chikwe J.
      • Gaudino M.
      • Wijeysundera H.C.
      The cost-effectiveness of transcatheter aortic valve replacement in low surgical risk patients with severe aortic stenosis.
      Using data from the PARTNER 3 study and cost perspectives based on a French national hospital claim database, the cost-effectiveness of TAVI and SAVR in patients with severe aortic stenosis and at low risk of surgical mortality was assessed in this study.

      Methods

       Model Structure

      A cost-utility analysis was developed to estimate changes in both direct healthcare costs and health-related quality of life (HRQOL) with the use of TAVI with SAPIEN 3 compared with SAVR in people with aortic stenosis and at low risk of surgical mortality (Society of Thoracic Surgeons < 4%). The model was developed following a targeted literature review of economic models developed on TAVI. At the time of development, no published models on the use of TAVI in those at low risk of surgical mortality were identified. Models in patients at moderate or high risk of surgical mortality comprised Markov models based around complications or New York Heart Association class.
      • Zhou J.Y.
      • Liew D.
      • Duffy S.J.
      • Walton A.
      • Htun N.
      • Stub D.
      Cost-effectiveness of transcatheter versus surgical aortic valve replacement in low-risk patients with severe aortic stenosis.
      ,
      • Goodall G.
      • Lamotte M.
      • Ramos M.
      • Maunoury F.
      • Pejchalova B.
      • de Pouvourville G.
      Cost-effectiveness analysis of the SAPIEN 3 TAVI valve compared with surgery in intermediate-risk patients.
      The New York Heart Association class was not judged granular enough to capture the differences between treatments in patients at low risk of surgical mortality, and thus, a model based around complications was used. The health states and events included in our model were similar to those modeled previously
      • Zhou J.Y.
      • Liew D.
      • Duffy S.J.
      • Walton A.
      • Htun N.
      • Stub D.
      Cost-effectiveness of transcatheter versus surgical aortic valve replacement in low-risk patients with severe aortic stenosis.
      and were aligned with events from the PARTNER 3 trial experienced by a meaningful proportion of patients or with significant differences between treatment arms.
      A lifetime time horizon (30 years) was chosen to reflect all potential consequences to people with aortic stenosis over their lifetime. The model had 2 stages. Early adverse events (AEs) linked to the TAVI procedure were first captured using the 30-day AE data set from the PARTNER 3 trial in a decision tree (Fig. 1A).
      • Mack M.J.
      • Leon M.B.
      • Thourani V.H.
      • et al.
      Transcatheter aortic-valve replacement with a balloon-expandable valve in low-risk patients.
      These data were fed into a subsequent Markov model that captures longer-term outcomes of patients, after TAVI or SAVR intervention (Fig. 1B).
      • Mack M.J.
      • Leon M.B.
      • Thourani V.H.
      • et al.
      Transcatheter aortic-valve replacement with a balloon-expandable valve in low-risk patients.
      Figure thumbnail gr1
      Figure 1The cost-effectiveness model had 2 stages: (A) early AEs from the PARTNER 3 trial were captured in a decision tree, which were fed into a (B) Markov model that captures longer-term outcomes of patients. ∗Short-term AEs include new permanent pacemaker, hospitalization, nondisabling stroke, transient ischemic attacks, myocardial infarction, bleeding, acute kidney injury with renal replacement therapy, and aortic reinterventions.
      AE indicates adverse event; AF, atrial fibrillation; SAVR, surgical aortic valve replacement; TAVI, transcatheter aortic valve implantation.
      The model is constructed with 4 distinctive health states:
      • 1.
        “Alive and well”: patients have undergone the procedure and survived with only short-term or no AEs. Patients in this health state can transition to disabling stroke, treated atrial fibrillation (AF), or death at any point during the model time horizon.
      • 2.
        “Treated AF”: patients have undergone the procedure and survived but now have AF and a specific treatment was required. This can either occur within the first 30 days or during the rest of the time horizon of the model. Patients in this health state can transition to disabling stroke or dead at any point during the model time horizon.
      • 3.
        “Disabling stroke”: patients have undergone the procedure and survived but had a disabling stroke. This can either occur within the first 30 days or during the rest of the time horizon of the model. Patients in this health state can only transition into the dead state at any point during the model time horizon.
      • 4.
        “Dead”: this is the absorbing state in the model. All patients in the model are at risk of dying because of general all-cause mortality. Patients with treated AF and stroke are at an increased risk of dying.
      HRQOL was included in the analysis using quality-adjusted life-years (QALYs) measured using EQ-5D for the different health states in the model, using decrements taken from a published study and age-adjusted population norms.
      This cost-effectiveness model used monthly cycles, with half-cycle corrections applied when appropriate for costs and QALYs. The only costs that were not half-cycle corrected were the procedure costs, because these occur at the point of entry to the model. All patients will incur these costs assuming they do not die during the procedure, which is unlikely given they are at low risk of surgical mortality. Similarly, QALYs were half-cycle corrected except for disutility associated with the procedure, which is assumed to be 0 in the base case. Discounting of future costs and benefits was applied at a rate of 2.5% per French guidelines.
      Choices in methods for economic evaluation – HAS. Haute Autorité de Santé.

       Model Inputs

       Trial overview

      The first stage of this model was informed by the PARTNER 3 trial. Patients with clinical frailty, bicuspid aortic valves, or other anatomical features that increased the risk of complications associated with either TAVI or surgery were excluded. Among the 1000 enrolled patients, 503 were randomized to undergo SAPIEN 3 and 497 to SAVR. The “as treated” population was defined by 496 patients in the SAPIEN 3 group versus 454 patients in the surgical group. The primary endpoint was a composite of death from any cause, stroke, or rehospitalization at 1 year after the procedure. The trial comprised patients with an average age of 73 years, of whom 69% were male.

       Clinical events in the cost-effectiveness model

      Predictions of clinical developments in the model (see Appendix Table 1 in Supplemental Materials found at https://doi.org/10.1016/j.jval.2021.10.003) were based on a decision tree that captured all AEs experienced up to 30 days after the procedure, as reported in PARTNER 3 (see Appendix Fig. 1 in Supplemental Materials found at https://doi.org/10.1016/j.jval.2021.10.003), where more than one patient had experienced the event. The rate of new onset AF was based on only those patients that received an action or treatment and excluded those with “no action.” This was assumed to capture those with more permanent chronic AF rather than those with AF occurring just in the acute phase. All AE rates at 30 days and death within the first 30 days after the procedure were based on the PARTNER 3 trial.
      • Mack M.J.
      • Leon M.B.
      • Thourani V.H.
      • et al.
      Transcatheter aortic-valve replacement with a balloon-expandable valve in low-risk patients.
      Costs and disutilities were only applied to events that would not fall within complications included in the initial procedure cost or disutility, to prevent double counting.
      Transition probabilities for the Markov model were informed through several sources. “Alive and well” to “treated AF” was captured through new-onset AF between 30 days and 1 year, as reported in Table S9 of PARTNER 3.
      • Mack M.J.
      • Leon M.B.
      • Thourani V.H.
      • et al.
      Transcatheter aortic-valve replacement with a balloon-expandable valve in low-risk patients.
      The probability from “alive and well” to “treated AF” from year 2 onward was also based on the PARTNER 3 trial 2-year outcomes.
      • Leon M.B.
      • Michael J.M.
      • Rebecca T.H.
      • et al.
      outcomes 2 years after transcatheter aortic valve replacement in patients at low surgical risk.
      “Alive and well” to “disabling stroke” was captured using Stroke Alliance for Europe data from 2017,
      The burden of Stroke in France. SAFE.
      with specific burden of illness relating to France, and “treated AF” to “disabling stroke” was captured from a large study conducted in 61 582 patients hospitalized for AF in France.
      • Cotté F.E.
      • Chaize G.
      • Kachaner I.
      • Gaudin A.F.
      • Vainchtock A.
      • Durand-Zaleski I.
      Incidence and cost of stroke and hemorrhage in patients diagnosed with atrial fibrillation in France.
      The PARTNER 3 data were not used to inform these “treated AF” to “disabling stroke” estimates because of the very low number of stroke events in the trial.
      The survival estimates using parametric survival analysis from the 2-year outcomes of the PARTNER 3 trial produced clinically implausible estimates in the model because of a very low rate of death in the study.
      • Mack M.J.
      • Leon M.B.
      • Thourani V.H.
      • et al.
      Transcatheter aortic-valve replacement with a balloon-expandable valve in low-risk patients.
      Nevertheless, they are included as options in the model and are explored in scenario analyses in combination with general mortality data for France to ensure that patients in the model cannot survive for longer than would be expected for the general population. All-cause mortality in the base case was determined from general population normal mortality risk, with relative risks applied from published literature corresponding to each health state (see Appendix Table 2 in Supplemental Materials found at https://doi.org/10.1016/j.jval.2021.10.003).
      Transitions to “treated AF” and “disabling stroke” were assumed to remain constant over the time horizon from year 2 onward, regardless of time in the model or patient age. Hospitalization data used in the model were based on 1- and 2-year data from the PARTNER 3 trial and were assumed to remain constant over the time horizon of the model after 2 years. Similarly, the rate of reinterventions was assumed to remain constant after 22 years,
      • Bourguignon T.
      • Bouquiaux-Stablo A.L.
      • Candolfi P.
      • et al.
      Very long-term outcomes of the Carpentier-Edwards Perimount valve in aortic position.
      because of a paucity of longer-term data. In the base case, the same reintervention rate was applied to all treatment options. Nevertheless, the impact of increased risk of reintervention with SAPIEN 3 was explored in scenario analyses, and reinterventions were associated with an increased risk of procedure-related mortality.
      Myocardial infarctions were captured as intercurrent events between 30 days and 1 year from the PARTNER 3 trial.
      • Mack M.J.
      • Leon M.B.
      • Thourani V.H.
      • et al.
      Transcatheter aortic-valve replacement with a balloon-expandable valve in low-risk patients.
      The risk of these events was applied to all patients who were alive in the model during this period, regardless of health state, with an associated monthly cost. Permanent pacemaker was integrated as a postoperative complication during the index hospitalization and was associated with additional cost (device, procedure, and monthly follow-up cost). Rehospitalization and reintervention events were also captured for the remainder of the lifetime horizon and applied to all patients alive in each cycle regardless of health state beyond the 30-day period; these events had a one-off cost associated with them. A time-varying rate was used for rehospitalization and reintervention modeling, based on how long a patient has been in the model.

       Utilities

      Utility values used age-adjusted population utility norms. A 3-level EQ-5D index value (time trade-off value set) was used for the age-adjusted population utility norms, specific to the French population.
      A pragmatic search of the literature was conducted to identify disutilities that could be most appropriately combined with each health state. The best source identified by the search was a study by Lanitis et al,
      • Lanitis T.
      • Cotté F.E.
      • Gaudin A.F.
      • Kachaner I.
      • Kongnakorn T.
      • Durand-Zaleski I.
      Stroke prevention in patients with atrial fibrillation in France: comparative cost-effectiveness of new oral anticoagulants (apixaban, dabigatran, and rivaroxaban), warfarin, and aspirin.
      which estimated the cost-effectiveness of oral anticoagulants in patients with AF in France and reported utility values of 0.727 for AF and 0.514 for severe stroke. Although used in previous French cost-effectiveness analyses, the decrements themselves are from a UK population. The reported scores were subtracted from the reported population norm utility value for age groups 55 to 64 and 65 to 74 years, respectively (utility score for age range 55-64 years 0.853; utility score for age range 65-74 years 0.81), which resulted in disutilities of 0.13 for “treated AF” and 0.30 for “severe/disabling stroke.” The disutility for severe stroke, rather than moderate or mild stroke, was used because it was judged to be more reflective of the health state “disabling stroke.” This is tested in sensitivity analyses. No procedure-related disutility was included to avoid double counting with health state disutilities. Disutilities were not included for any intercurrent events because it was considered that this may induce a double counting error where health state disutilities were applied to patients with treated AF and disabling stroke.

       Cost inputs

      The cost perspective was based on the French Programme de Médicalisation du Système d’Information (PMSI) national hospital claim database. Using PMSI, both procedure (“hospitalisation en médecine, chirurgie, obstétrique et odontologie [MCO]”) and rehabilitation (“soins de suite et de réadaptation [SSR]”) costs were estimated from an exhaustive data collection from 2017 to 2018 for TAVI and SAVR (isolated), valued using the 2017 French national frame of reference for hospital costs (Echelle Nationale des Coûts) and actualized to 2020. TAVI and pacemaker device costs were determined from the liste des produits et prestations remboursables (LPPR) French reimbursement prices. Follow-up cost, cost linked to postoperative complications, and acute and chronic rehospitalization costs have been estimated from the PMSI/Echelle Nationale des Coûts or relevant publications and are summarized in Table 1.
      Table 1Costs associated to TAVI and SAVR (procedure, complications, long-term).
      Unit cost componentsSAPIEN 3SAVRSource
      Procedure
       Intervention (without pacemaker)€8139€22 603PMSI 2017-2018

      ENC 2017
       Intervention (with pacemaker)€8858€25 432PMSI 2017-2018

      ENC 2017
       TAVI bioprosthesis (including VAT)€15 419LPPR 2019
       Pacemaker (including VAT)
      Weighted average price based on PMSI analysis.
      €2990€2990PMSI 2017-2018 LPPR 2019
       Transfer to other hospital MCO€3077€3441PMSI 2017-2018

      ENC 2017
       Rehabilitation SSR€6594€5333PMSI 2017-2018

      ENC 2017
      Acute postoperative complications
       Reintervention€25 445€25 935PMSI 2017-2018

      ENC 2017
      Associated to health states
       Treated AF, month 1€2569€2569PMSI 2017-2018

      ENC 2017
       Treated AF, ≥ month 2€156€156Le Heuzey et al 2014
       Disabling stroke, month 1€6774€6774PMSI 2017-2018

      ENC 2017
       Disabling stroke, ≥ month 2€1261€1261Schmitz et al 2015
       Caregiver for disabling stroke, ≥ month 1€2100€2100Barrel et al 2019
       Alive and well, year 1€26€30SNIR 2017
       Alive and well, year 2+€13€13SNIR 2017
      Other costs considered
       Pacemaker procedure€8884€8884PMSI 2017-2018

      ENC 2017
       Pacemaker complications (monthly)€144€144Clémenty et al 2019
       Rehospitalization€2226€2226PMSI 2017-2018

      ENC 2017
       Reintervention€25 445€25 935PMSI 2017-2018

      ENC 2017
       One-off cost of death€733€733PMSI 2017-2018

      ENC 2017
      AF indicates atrial fibrillation; ENC, Echelle Nationale des Coûts; LPPR, liste des produits et prestations remboursables; MCO, médecine, chirurgie, obstétrique et odontologie; PMSI, Programme de Médicalisation du Système d’Information; SAVR, surgical aortic valve replacement; SNIR, système national inter-régimes; SSR, soins de suite et de réadaptation; TAVI, transcatheter aortic valve implantation; VAT, value added tax.
      Weighted average price based on PMSI analysis.

       Model Outputs

      The cost-effectiveness model generated total per patient costs and QALYs in each arm over the patients’ lifetime and an incremental cost per QALY gained for SAPIEN 3 compared with SAVR in a population of French patients at low risk of surgical mortality.

       Sensitivity Analyses

      To evaluate uncertainty in the results, one-way deterministic sensitivity analyses were performed by varying inputs using confidence intervals and ranges from the literature where available and plausible ranges where data were unavailable as specified in Appendix Table 3 in Supplemental Materials found at https://doi.org/10.1016/j.jval.2021.10.003. The tornado diagram ranks and displays the 20 input parameters to which the model’s results are most sensitive and displays results in terms of net monetary benefit using a cost-effectiveness threshold of €50 000/QALY. No official cost-effectiveness threshold is reported for France; hence, this is used as a proxy.
      • Toumi M.
      • Motrunich A.
      • Millier A.
      • et al.
      Analysis of health economics assessment reports for pharmaceuticals in France - understanding the underlying philosophy of CEESP assessment.
      Overall parameter uncertainty was addressed by a probabilistic sensitivity analysis (PSA). Probability distributions for all input parameters were specified and 1000 Monte Carlo simulations were run using random draws of all parameters from within their assigned distributions (see Appendix Table 4 in Supplemental Materials found at https://doi.org/10.1016/j.jval.2021.10.003). Several scenario analyses were performed to examine the impact of major structural assumptions. All analyses were performed using Microsoft Excel (Microsoft Corporation, Redmond, WA).

      Results

       Base Case

      TAVI with SAPIEN 3 offers meaningful benefits compared with SAVR in that it is both cost saving (by €12 742 per patient) and generates greater QALYs (by 0.89 per patient). Base case results over a 30-year time horizon are presented in Table 2. Cost savings are driven by both a lower initial procedure cost with SAPIEN 3 compared with SAVR, when rehabilitation is considered, and lower long-term costs relating to the treatment of AF and disabling stroke (see Appendix Fig. 2 in Supplemental Materials found at https://doi.org/10.1016/j.jval.2021.10.003). Additional QALYs are gained with SAPIEN 3 because of lower risk of mortality and better HRQOL.
      Table 2Base case results with acute and lifetime costs.
      Summary resultsTAVI with SAPIEN 3SAVRIncremental
      Cost per patient€38 992€51 734−€12 742
      Life-year gained (undiscounted)14.0713.180.89
      Median survival (years)17.2514.672.58
      QALYs per patient8.447.550.89
      ICERDominant
      Incremental NMB€57 486
      Incremental NHB1.15
      Acute phase cost (first hospitalization and rehabilitation)
       Index hospitalization (without pacemaker)€8139€22 603−€14 464
       Valve LPPR (including VAT)€15 419€15 419
       Transfer MCO
      Incidence of event multiplied by average cost.
      €270€529−€259
       Rehabilitation SSR
      Incidence of event multiplied by average cost.
      €1378€3999−€2621
       Pacemaker
      Incidence of event multiplied by average cost.
      €239€231€9
       Acute phase costs€25 445€27 361−€1916
      Additional costs at 1 year
       MI€182€105€78
       Costs of pacemaker complications€99€60€39
       Costs of hospitalizations€152€228−€75
       Reintervention costs€118€119−€1
       Alive and well health state costs€288€226€62
       Treated AF health state costs€206€1518−€1312
       Disabling stroke costs€16€247−€231
       Death costs€17€24−€7
       Total costs at 1 year€26 523€29 888−€3365
      Additional lifetime costs
       Costs of pacemaker complications€1168€675€492
       Costs of hospitalizations€280€264€17
       Reintervention costs€4845€4293€551
       Alive and well health state costs€1430€898€532
       Treated AF health state costs€1888€7107−€5219
       Disabling stroke costs€2355€8099−€5744
       Death costs€503€509−€6
       Additional lifetime costs€12 468€21 846−€9378
       Total lifetime costs€38 992€51 734−€12 742
      AF indicates atrial fibrillation; ICER, incremental cost-effectiveness ratio; LPPR, liste des produits et prestations remboursables; MCO, médecine, chirurgie, obstétriqueet odontolgie; MI, myocardial infarction; NHB, Net health benefit; NMB, net monetary benefit; QALY, quality-adjusted life-year; SAVR, surgical aortic valve replacement; SSR, soins de suite et de réadaptation; TAVI, transcatheter aortic valve implantation.
      Incidence of event multiplied by average cost.

       Deterministic Sensitivity Analyses

      Univariate sensitivity analysis is displayed in a tornado diagram (see Appendix Fig. 3 in Supplemental Materials found at https://doi.org/10.1016/j.jval.2021.10.003), which shows that SAPIEN 3 TAVI remains cost saving regardless of changes in individual model parameters (note that only those 20 parameters with the greatest influence on the model’s results are displayed). The model is most sensitive to the starting age of patients entering the model and the various transition probabilities toward treated AF and disabling stroke for both SAPIEN 3 and SAVR.
      Additional sensitivity analyses were conducted around the utility decrements applied for stroke because these were not from a French population and related to disabling stroke. Even where this decrement is set to 0 (ie, disabling stroke has no impact on HRQOL), SAPIEN 3 TAVI remains dominant (see Appendix Table 5 in Supplemental Materials found at https://doi.org/10.1016/j.jval.2021.10.003).

       PSA

      The results of the PSA are presented on the cost-effectiveness plane (Fig. 2A) and in a cost-effectiveness acceptability curve (Fig. 2B). At a threshold of €50 000/QALY or above, SAPIEN 3 TAVI is estimated to be cost-effective compared with SAVR in 100% of iterations. SAPIEN 3 TAVI is markedly less costly and more effective (dominant) than SAVR in approximately 82% of iterations.
      Figure thumbnail gr2
      Figure 2(A) Probabilistic sensitivity analysis and (B) cost-effectiveness acceptability curve for TAVI with SAPIEN 3 compared with SAVR.
      PSA indicates probabilistic sensitivity analysis; QALY, quality-adjusted life-year; SAVR, surgical aortic valve replacement; TAVI, transcatheter aortic valve implantation.

       Scenarios

      A series of scenario analyses were conducted to assess the impact of changing various assumptions on the results of the model. TAVI with SAPIEN 3 remained dominant regardless of the time horizon considered because of the cost savings generated in the acute phase once rehabilitation was considered (Table 2). In model scenarios, SAPIEN 3 TAVI was also more cost-effective (dominant) than SAVR, despite:
      • 1.
        Changes to the discount rates used
      • 2.
        Increasing the risk of reintervention for device replacement with SAPIEN 3
      • 3.
        Using HRQOL data by treatment from PARTNER 3
      • 4.
        Artificially increasing the risk of stroke with SAPIEN 3 to produce results aligned with outcomes in PARTNER 3
      • 5.
        Including AE costs for those occurring within the first 30 days after surgery (considering events occurring after discharge from index hospitalization)
      • 6.
        Using survival data from the PARTNER 3 study and extrapolating this over the model time horizon (after adjusting for background mortality)
      • 7.
        Alternative severity level 1 used to cost TAVI and SAVR procedures
      Results of each of the scenario analyses conducted are presented in Appendix Table 5 in Supplemental Materials found at https://doi.org/10.1016/j.jval.2021.10.003.

      Discussion

      Driven by lower long-term management costs associated with treated AF and disabling stroke and considering a lower intervention cost (when taking into account postsurgery rehabilitation), SAPIEN 3 TAVI presents a compelling value-based case for patients in France with aortic stenosis and a low risk of surgical mortality. The modeled cost-effectiveness case remained robust at a threshold of €50 000/QALY, despite input modifications and challenging scenarios.
      In this model, transitions to “treated AF” and “disabling stroke” were assumed to remain constant over the time horizon from year 2 onward, regardless of time in the model or patient age. This is a simplified assumption and is likely to understate transitions between the disease states in later years because it is likely that stroke risk will increase with age.
      • Kelly-Hayes M.
      Influence of age and health behaviors on stroke risk: lessons from longitudinal studies.
      For the SAPIEN 3 TAVI outcomes, this is likely to represent a conservative assumption because it is anticipated that the risk of stroke would be higher in the SAVR treatment arm because of increased incidence of treated AF; therefore, this would likely increase the incremental difference in stroke incidence between treatment arms.
      Hospitalization data used in the model are based on 1- and 2-year data from the PARTNER 3 trial. A simplifying assumption was used whereby this rate remained constant over the time horizon of the model after 2 years; the impact of this assumption is unknown because people in both treatment arms in the model are at risk of hospitalization.
      The rate of reinterventions is assumed to remain constant after 22 years, because of a paucity of longer-term data. It is expected that at this point this would have very little impact on modeled outcomes because it is expected that only approximately 11% of the patients would still be alive in the model, with a limited need for reintervention. Nevertheless, there remains some uncertainty regarding reintervention rates and durability of the TAVI device particularly in younger patients. Data at 5 years on the latest TAVI devices show no early signs of concern with the durability of TAVI.
      • Pibarot P.
      • Hahn R.T.
      • Weissman N.J.
      • Monaghan M.J.
      Assessment of paravalvular regurgitation following TAVR: a proposal of unifying grading scheme.
      Nevertheless, scenarios were conducted in the model to show the impact of increasing the rate of reintervention with TAVI compared with SAVR as discussed under “scenario analyses,” and this did not change the direction of the results. This analysis was undertaken in a subgroup of the population eligible for TAVI (those at low risk of surgical mortality); thus, aside from the age of the patient population, it is expected to be relatively homogeneous. The impact of age on reintervention and on the results of the model is discussed earlier.
      Disutilities were not included for any intercurrent events because it was judged that this may risk double counting them with the health state disutilities being applied to patients in the “treated AF” and “disabling stroke” states. This is a conservative assumption because the rates of intercurrent events are lower with SAPIEN 3 than SAVR, except for pacemaker complications which may be higher in the SAPIEN 3 treatment arm.
      With regard to the cost inputs specific to the French healthcare system, the cost of the TAVI procedure may have been overestimated because PMSI data contain mostly inoperable or high-risk patients that could be considered as more “costly” overall than patients at low risk of surgical mortality. The costs and impact on HRQOL of a disabling stroke were similar to those reported in other France-based analyses.
      • de Pouvourville G.
      • Karam P.
      Le coût hospitalier des événements cérébro-vasculaires associés à une fibrillation atriale en France.
      ,
      • de Pouvourville G.
      • Blin P.
      • Karam P.
      The contribution of real-world evidence to cost-effectiveness analysis: case study of dabigatran etexilate in France.
      The disutility inputs were derived largely from the cost-effectiveness study performed by Lanitis et al
      • Lanitis T.
      • Cotté F.E.
      • Gaudin A.F.
      • Kachaner I.
      • Kongnakorn T.
      • Durand-Zaleski I.
      Stroke prevention in patients with atrial fibrillation in France: comparative cost-effectiveness of new oral anticoagulants (apixaban, dabigatran, and rivaroxaban), warfarin, and aspirin.
      ; this study based its data on an EQ-5D study conducted in a UK population.
      • Sullivan P.W.
      • Slejko J.F.
      • Sculpher M.J.
      • Ghushchyan V.
      Catalogue of EQ-5D scores for the United Kingdom.
      Consideration should be given to how well these data truly generalize to a French population.
      There are a few publications demonstrating the dominance of TAVI versus SAVR in low-risk patients. First, this model was accepted by Haute Autorité de Santé as part of a health technology assessment (HTA) submission, with slightly different assumptions (shorter time horizon set at 15 years and higher price of the valve).
      SAPIEN 3 Modèle 9600 TFX bioprothese valvulaire aortique avec systeme de mise en place Edwards CommanderTM. Haute Autorité de Santé.
      ,
      SAPIEN 3 Titement de la sténose aortique sévère symptomatique en france chez les patients à faible risque chirurgical. Haute Autorité de Santé.
      SAPIEN 3 was found to be dominant versus SAVR with €7737 savings and 0.33 LYG. Second, both the Irish HTA (HIQA) published in late 2019
      Health Technology Assessment of transcatheter aortic valve implantation (TAVI) in patients with severe symptomatic aortic stenosis at low and intermediate risk of surgical complications. Health Information and Quality Authority.
      and the Norwegian HTA also found TAVI to be dominant versus SAVR.
      Transcatheter aortic valve implantation (TAVI) versus surgical aortic valve replacement (SAVR) for patients with severe aortic stenosis and low surgical risk and across surgical risk groups: a health technology assessment.
      A recent analysis conducted from a Canadian third-party payer perspective estimated the cost-effectiveness of balloon-expandable TAVR and self-expandable TAVR related to SAVR using data from the PARTNER 3 trial up to 1-year follow-up.
      • Tam D.Y.
      • Azizi P.M.
      • Fremes S.E.
      • Chikwe J.
      • Gaudino M.
      • Wijeysundera H.C.
      The cost-effectiveness of transcatheter aortic valve replacement in low surgical risk patients with severe aortic stenosis.
      The analysis reported a similar model structure with health states reflecting “alive and well,” “procedure,” “disabling stroke,” and “> moderate paravalvular leak (PVL).” The key difference in our model structures is the choice of “treated AF” as a health state rather than PVL. In our analysis, treated AF was chosen because there was a large difference in incidence between treatment arms (4.1% TAVR vs 35.8% SAVR at 30 days), and it was judged that this could have a long-term impact on survival, costs, and quality of life (QOL) and also in terms of additional risk of stroke. PVL was not explicitly included in this analysis because the first-choice treatment is reintervention; it was judged that this would already be captured within the rate of reintervention that was included throughout the patient’s lifetime in the model. Furthermore, it was judged that this would unlikely have a meaningful impact on results because of the low incidence reported in the PARTNER 3 study (0.8% SAPIEN 3 vs 0% SAVR at 30 days, 0.6% SAPIEN 3 vs 0.5% SAVR at 1 year, and 0.8% SAPIEN 3 vs 0.9% SAVR at 2 years; these differences were not statistically significant).
      The Canadian analysis estimated lifetime QALYs of 9.16 and 9.09 in balloon-expandable TAVR and SAVR, respectively (an incremental difference of 0.1).
      • Tam D.Y.
      • Azizi P.M.
      • Fremes S.E.
      • Chikwe J.
      • Gaudino M.
      • Wijeysundera H.C.
      The cost-effectiveness of transcatheter aortic valve replacement in low surgical risk patients with severe aortic stenosis.
      This compares with estimates of 9.11 and 8.12 in our analysis when using the Canadian discount rate of 1.5% (an incremental difference of 0.99).
      • Tam D.Y.
      • Azizi P.M.
      • Fremes S.E.
      • Chikwe J.
      • Gaudino M.
      • Wijeysundera H.C.
      The cost-effectiveness of transcatheter aortic valve replacement in low surgical risk patients with severe aortic stenosis.
      The difference in estimation of incremental QALYs is largely explained by the inclusion of treated AF beyond the initial 30-day period in this model. In our model, the removal of the QOL decrement and increased RR of mortality for the treated AF health state results in slightly higher QALY estimates than the Canadian model of 9.39 and 9.21 but reduces the incremental difference (0.18) between the 2 treatments, which could be explained by not considering PVL or using a more conservative risk of death value associated with stroke (2.3 instead of 5.4 in the Canadian analysis). The difference in costs is highly dependent on the healthcare systems so it is difficult to comment on and compare differences in individual cost input parameters used in each of the analyses. Costs linked to AF and rehabilitation were not factored in the Canadian analysis and could partially explain why SAPIEN 3 is not dominant but instead cost-effective with an incremental cost-effectiveness ratio of CAN$27 196/QALY.
      This cost-effectiveness model provides compelling insight for the application of TAVI with SAPIEN 3 in patients with intermediate or high risk of surgical mortality where, to date, there is a paucity of data from randomized controlled trials (previous trials were conducted with earlier versions of the device). By demonstrating dominance over SAVR in the population that is most suited to favorable outcomes with surgery, it is plausible that cost savings with TAVI would be even greater in patient populations at intermediate or high risk of surgical mortality. As surgical complications become more likely, the safety and cost proposition for selecting TAVI become stronger.
      The findings of this study are meaningful in 3 key areas. From the patient perspective, an intervention associated with a lower risk of complication and rehospitalization, and with improved recovery rate and QOL gains would be preferable. For the appropriate population of patient with severe symptomatic aortic stenosis (in particular, for those without complex coronary artery disease or severe heart valve disease and for whom transfemoral access is suitable), TAVI may become the reference treatment.
      From the perspective of the healthcare provider, TAVI with SAPIEN 3 can offer a reduced organizational and resource impact (ie, lower general anesthesia) compared with SAVR, with shorter hospital stays, lower risk of infection, and fewer complications.
      Policy makers may consider the benefits of TAVI over SAVR in an overall societal–medical perspective. In this example using the French healthcare system, a holistic view would not consider procedure costs (MCO), rehabilitation costs (SSR), and budget (LPPR) in silos, but as a “whole” when making policy decisions for the management of severe symptomatic aortic stenosis.

       Study Limitations

      A specific limitation of this initiative is that high-risk anatomy (eg, annular calcification, unfavorable coronary anatomy) and bicuspid aortic valves were excluded from PARTNER 3,
      • Mack M.J.
      • Leon M.B.
      • Thourani V.H.
      • et al.
      Transcatheter aortic-valve replacement with a balloon-expandable valve in low-risk patients.
      making the conclusions difficult to generalize to the overall aortic stenosis population. As described, there are many inherent limitations to a cost-effectiveness model, including assumptions made in the presence of “best fit” data or a paucity of data, extrapolations into time horizons that are modeled beyond the scope of existing input data, and the underestimations and overestimations potentially caused by differences in healthcare systems or by the criteria for intervention and treatment selection within that specific system. The disutility scores were based on values obtained from Lanitis et al,
      • Lanitis T.
      • Cotté F.E.
      • Gaudin A.F.
      • Kachaner I.
      • Kongnakorn T.
      • Durand-Zaleski I.
      Stroke prevention in patients with atrial fibrillation in France: comparative cost-effectiveness of new oral anticoagulants (apixaban, dabigatran, and rivaroxaban), warfarin, and aspirin.
      which were originally informed by a study
      MCO par diagnostic ou acte. ScanSanté.
      using EQ-5D scores from the United Kingdom for a range of different conditions. Therefore, the generalizability of these scores to a French population should be considered. This model used available inputs to enable a cost-effectiveness assessment in patients with aortic stenosis in France, and the findings cannot be assumed to be accurate in other healthcare systems.

      Conclusions

      This model demonstrated that TAVI with SAPIEN 3 was dominant compared with SAVR in the treatment of patients with severe symptomatic aortic stenosis at low risk of surgical mortality. The model seems to be robust and handled uncertainty introduced via a range of scenarios and sensitivity variables. The model was based primarily on data available for TAVI with SAPIEN 3 and as such cannot necessarily be generalized to all transcatheter heart valves.
      When assessing the healthcare budget impact of TAVI and SAVR, policy makers are encouraged to take a broad, unsiloed view. The authors suggest that these findings support the prioritization of TAVI with SAPIEN 3 over SAVR in patients with severe symptomatic aortic stenosis for whom TAVI is suitable, across the range of risks for surgical mortality.

      Article and Author Information

      Author Contributions: Concept and design: Gilard, Eltchaninoff, Iung, Lefèvre, Spaulding, Mutuon, Roussel, Candolfi, de Pouvourville, Green, Shore
      Acquisition of data: Gilard, Mutuon, Roussel, Candolfi
      Analysis and interpretation of data: Gilard, Eltchaninoff, Iung, Lefèvre, Spaulding, Dumonteil, Mutuon, Roussel, Candolfi, de Pouvourville, Green, Shore
      Drafting of the manuscript: Gilard, Spaulding, Dumonteil, Candolfi, de Pouvourville, Green, Shore
      Critical revision of the paper for important intellectual content: Gilard, Eltchaninoff, Iung, Lefèvre, Spaulding, Dumonteil, Roussel, Candolfi, de Pouvourville, Green, Shore
      Statistical Analysis: Gilard, Candolfi
      Provision of study materials or patients: Gilard
      Administrative, technical, or logistic support: Gilard, Candolfi
      Supervision: Gilard, Candolfi
      Other: Gilard
      Conflict of Interest Disclosures: Dr Iung reported receiving speaker’s fees from Edwards Lifesciences outside the submitted work. Dr Lefèvre reported receiving personal fees for proctoring from Edwards Lifesciences and Boston Scientific during the conduct of the study and proctoring fees from Abbott outside the submitted work. Dr Spaulding reported being a consultant for Techwald, Edwards Lifesciences, Medtronic, and Stentys outside the submitted work and reported receiving speaker fees from AstraZeneca, Boehringer Ingelheim, Abbott, and Bristol Myers Squibb outside the submitted work. Dr Dumonteil reported receiving consultancy and proctoring fees from Abbott Vascular, Boston Scientific, Edwards Lifesciences, and Medtronic during the conduct of the study. Drs Roussel and Candolfi are employed by Edwards Lifesciences. Dr de Pouvourville reported receiving personal fees from Edwards Life Sciences for manuscript preparation and being a former coauthor of a previous publication and personal fees from Edwards Life Sciences for support for submission to French market access authorities and cowriting of a previous article outside the submitted work. Drs Green and Shore reported receiving consultancy fees from Edwards Lifesciences during the conduct of the study. No other disclosures were reported.
      Funding/Support: The article was sponsored by Edwards Lifesciences .
      Role of the Funder/Sponsor: Drs Roussel and Candolfi were involved in the study design and conduct of the analysis and interpretation of the data; preparation, review, and approval of the manuscript; and the decision to submit the manuscript for publication.

      Acknowledgment

      The authors acknowledge Sandrine Baffert, Marc-Anthony Chechoi Mienje, Julien Robert, and Cécile Fabron from CEMKA for their valuable input regarding the costing and the literature review specific to France. Writing support was supplied by Zenith Healthcare Communications Ltd (Chester, United Kingdom) and Griffin Scientific Ltd (Gloucestershire, United Kingdom), and this was funded by Edwards Lifesciences.

      Supplemental Material

      Figure thumbnail figs1
      Figure thumbnail figs2
      Figure thumbnail figs3

      References

      1. MCO par diagnostic ou acte. ScanSanté.
        • Thourani V.H.
        • Suri R.M.
        • Gunter R.L.
        • et al.
        Contemporary real-world outcomes of surgical aortic valve replacement in 141,905 low-risk, intermediate-risk, and high-risk patients.
        Ann Thorac Surg. 2015; 99: 55-61
        • Mack M.J.
        • Leon M.B.
        • Thourani V.H.
        • et al.
        Transcatheter aortic-valve replacement with a balloon-expandable valve in low-risk patients.
        N Engl J Med. 2019; 380: 1695-1705
        • Leon M.B.
        • Smith C.R.
        • Mack M.
        • et al.
        Transcatheter aortic-valve implantation for aortic stenosis in patients who cannot undergo surgery.
        N Engl J Med. 2010; 363: 1597-1607
        • Smith C.R.
        • Leon M.B.
        • Mack M.J.
        • et al.
        Transcatheter versus surgical aortic-valve replacement in high-risk patients.
        N Engl J Med. 2011; 364: 2187-2198
        • Leon M.B.
        • Smith C.R.
        • Mack M.J.
        • et al.
        Transcatheter or surgical aortic-valve replacement in intermediate-risk patients.
        N Engl J Med. 2016; 374: 1609-1620
        • Mack M.J.
        • Leon M.B.
        • Smith C.R.
        • et al.
        5-year outcomes of transcatheter aortic valve replacement or surgical aortic valve replacement for high surgical risk patients with aortic stenosis (PARTNER 1): a randomised controlled trial.
        Lancet. 2015; 385: 2477-2484
        • Makkar R.R.
        • Thourani V.H.
        • Mack M.J.
        • et al.
        Five-year outcomes of transcatheter or surgical aortic-valve replacement.
        N Engl J Med. 2020; 382: 799-809
        • ElGuindy A.
        PARTNER 2A & SAPIEN 3: TAVI for intermediate risk patients.
        Glob Cardiol Sci Pract. 2016; 2016e201633
        • Otto C.M.
        • Nishimura R.A.
        • et al.
        • Writing Committee Members
        2020 ACC/AHA guideline for the management of patients with valvular heart disease: executive summary: a report of the American College of Cardiology/American Heart Association joint Committee on Clinical Practice guidelines [published correction appears in J Am Coll Cardiol. 2021;77(9):1276].
        J Am Coll Cardiol. 2021; 77: 450-500
        • Orlando R.
        • Pennant M.
        • Rooney S.
        • et al.
        Cost-effectiveness of transcatheter aortic valve implantation (TAVI) for aortic stenosis in patients who are high risk or contraindicated for surgery: a model-based economic evaluation.
        Health Technol Assess. 2013; 17: 1-86
        • Tarride J.E.
        • Luong T.
        • Goodall G.
        • Burke N.
        • Blackhouse G.
        A Canadian cost-effectiveness analysis of SAPIEN 3 transcatheter aortic valve implantation compared with surgery, in intermediate and high-risk severe aortic stenosis patients.
        Clinicoecon Outcomes Res. 2019; 11: 477-486
        • Baron S.J.
        • Wang K.
        • House J.A.
        • et al.
        Cost-effectiveness of transcatheter versus surgical aortic valve replacement in patients with severe aortic stenosis at intermediate risk.
        Circulation. 2019; 139: 877-888
        • Zhou J.
        • Liew D.
        • Duffy S.J.
        • Walton A.
        • Htun N.
        • Stub D.
        Cost-effectiveness of transcatheter aortic valve implantation compared to surgical aortic valve replacement in the intermediate surgical risk population.
        Int J Cardiol. 2019; 294: 17-22
        • Tam D.Y.
        • Azizi P.M.
        • Fremes S.E.
        • Chikwe J.
        • Gaudino M.
        • Wijeysundera H.C.
        The cost-effectiveness of transcatheter aortic valve replacement in low surgical risk patients with severe aortic stenosis.
        Eur Heart J Qual Care Clin Outcomes. 2021; 7: 556-563
        • Zhou J.Y.
        • Liew D.
        • Duffy S.J.
        • Walton A.
        • Htun N.
        • Stub D.
        Cost-effectiveness of transcatheter versus surgical aortic valve replacement in low-risk patients with severe aortic stenosis.
        Heart Lung Circ. 2021; 30: 547-554
        • Goodall G.
        • Lamotte M.
        • Ramos M.
        • Maunoury F.
        • Pejchalova B.
        • de Pouvourville G.
        Cost-effectiveness analysis of the SAPIEN 3 TAVI valve compared with surgery in intermediate-risk patients.
        J Med Econ. 2019; 22: 289-296
      2. Szende A. Janssen B. Cabases J. Self-Reported Population Health: an International Perspective Based on EQ-5D. Springer, Dordrecht, NL2014
      3. Choices in methods for economic evaluation – HAS. Haute Autorité de Santé.
        • Leon M.B.
        • Michael J.M.
        • Rebecca T.H.
        • et al.
        outcomes 2 years after transcatheter aortic valve replacement in patients at low surgical risk.
        J Am Coll Cardiol. 2021; 77: 1150-1161
      4. The burden of Stroke in France. SAFE.
        • Cotté F.E.
        • Chaize G.
        • Kachaner I.
        • Gaudin A.F.
        • Vainchtock A.
        • Durand-Zaleski I.
        Incidence and cost of stroke and hemorrhage in patients diagnosed with atrial fibrillation in France.
        J Stroke Cerebrovasc Dis. 2014; 23: e73-e83
        • Bourguignon T.
        • Bouquiaux-Stablo A.L.
        • Candolfi P.
        • et al.
        Very long-term outcomes of the Carpentier-Edwards Perimount valve in aortic position.
        Ann Thorac Surg. 2015; 99: 831-837
        • Lanitis T.
        • Cotté F.E.
        • Gaudin A.F.
        • Kachaner I.
        • Kongnakorn T.
        • Durand-Zaleski I.
        Stroke prevention in patients with atrial fibrillation in France: comparative cost-effectiveness of new oral anticoagulants (apixaban, dabigatran, and rivaroxaban), warfarin, and aspirin.
        J Med Econ. 2014; 17: 587-598
        • Toumi M.
        • Motrunich A.
        • Millier A.
        • et al.
        Analysis of health economics assessment reports for pharmaceuticals in France - understanding the underlying philosophy of CEESP assessment.
        J Mark Access Health Policy. 2017; 51344088
        • Kelly-Hayes M.
        Influence of age and health behaviors on stroke risk: lessons from longitudinal studies.
        J Am Geriatr Soc. 2010; 58: S325-S328
        • Pibarot P.
        • Hahn R.T.
        • Weissman N.J.
        • Monaghan M.J.
        Assessment of paravalvular regurgitation following TAVR: a proposal of unifying grading scheme.
        JACC Cardiovasc Imaging. 2015; 8: 340-360
        • de Pouvourville G.
        • Karam P.
        Le coût hospitalier des événements cérébro-vasculaires associés à une fibrillation atriale en France.
        J Gestion Econ S. 2019; 5: 399-409
        • de Pouvourville G.
        • Blin P.
        • Karam P.
        The contribution of real-world evidence to cost-effectiveness analysis: case study of dabigatran etexilate in France.
        Eur J Health Econ. 2020; 21: 235-249
        • Sullivan P.W.
        • Slejko J.F.
        • Sculpher M.J.
        • Ghushchyan V.
        Catalogue of EQ-5D scores for the United Kingdom.
        Med Decis Making. 2011; 31: 800-804
      5. SAPIEN 3 Modèle 9600 TFX bioprothese valvulaire aortique avec systeme de mise en place Edwards CommanderTM. Haute Autorité de Santé.
      6. SAPIEN 3 Titement de la sténose aortique sévère symptomatique en france chez les patients à faible risque chirurgical. Haute Autorité de Santé.
      7. Health Technology Assessment of transcatheter aortic valve implantation (TAVI) in patients with severe symptomatic aortic stenosis at low and intermediate risk of surgical complications. Health Information and Quality Authority.
      8. Transcatheter aortic valve implantation (TAVI) versus surgical aortic valve replacement (SAVR) for patients with severe aortic stenosis and low surgical risk and across surgical risk groups: a health technology assessment.