Table 1: Summary of Model Parameters.
Chan et al.30 | Rodgers/McKenna et al.31 | Reynolds et al.21 | Eckard et al.33 | Assasi/Blackhouse et al.34 | Ollendorf et al.36 | Reynolds et al.38 | Aronsson et al.37 | |
Year of Study | 2006 | 2008/ 2009 | 2009 | 2010 | 2010/ 2013 | 2010 | 2014 | 2014 |
Country | USA | UK | USA | Sweden | Canada (Ontario) | USA | UK | Europe |
AF Type | All Types | Mostly paroxysmal | Paroxysmal | Paroxysmal and persistent | Paroxysmal | Proxysmal and persistent | Paroxysmal | Paroxysmal |
Ablation Type | Radiofrequency | Radiofrequency | Radiofrequency | Radiofrequency | Radiofrequency | Radiofrequency | Cryoballoon | Radiofrequency |
First- or Second-Line Therapy | First-line | Second-line | Second-line | Second-line | Second-line | First- and second-line | Second-line | First-line |
Therapeutic Comparator | Rate Control, AAD (amiodarone) | AAD (amiodarone) | AAD (sotalol/flecainide first, amiodarone second), then RC | AAD (amiodarone, flecainaide) | AAD (amiodarone) | AAD (amiodarone), then RC | AAD (propafenone, first sotalol second, amiodarone third), then RC | AAD (amiodarone flecainaide, propafenone, sotalol) |
Model Type | Markov Model | Markov Model | Markov Model | Markov Model | Markov Model | Markov Model | Markov Model | Markov Model |
Model Inputs | Published literature, Medicare data | Published literature, costs from practicing MD | Published literature, Medicare data, FRACTAL registry | Published literature, Swedish clinical and national data | Published literature, FRACTAL registry, Canadian mortality, stroke, cost data | Published literature, Medicare data | Published literature, STOP-AF trial data | MANTRA-PAF trial data |
Time Horizon | Lifetime | 5 years / lifetime | 5 years | Lifetime | 5 years | 5 years / lifetime | 5 years | Lifetime |
AF Ablation Success Rate | 80 % | 74–84 % | 90 % | 78 % | 76 % | 82 % (paroxysmal), 79 % (persistent) | 71 % at 12 months | 85 % at 24 months |
Number of Ablation Attempts | 1.3 | 1.3 | 1.3 | 1.4 | 1.3 | N/A | N/A | 1.5–1.6 based on patient age |
Anticoagulation | Warfarin or ASA based on risk | Warfarin, ASA or none | Assumed same between groups | Warfarin | Warfarin | Warfarin or ASA based on risk | Warfarin, then ASA if bleeding | Warfarin |
Model Patient Assumptions | 3 model groups: 65 y.o. low risk 65 y.o. mod risk 55 y.o. mod risk | Mean age: 52 years 80% male CHADS2 of 1 | 60 years old Male | 3 groups: Age>69 years Age 70–79 years Age>80 years | 65 years old Male CHADS2 of 2 | 3 groups (Male): 60 y.o. low risk 65 y.o. mod risk 75 y.o. high risk | Based on STOP-AF trial population | Based on MANTRA-PAF trial population |
Assumptions with greatest impact on sensitivity analysis | AF ablation success rate, RC success rate, AF stroke rate, Warfarin utility, Warfarin bleed risk | AF Ablation success rate, NSR and AF utility, NSR stroke risk | Time horizon, NSR and AF utility, Ablation cost | AF ablation success rate over lifetime | Time horizon, NSR and AF utility | NSR and AF utility, NSR and AF stroke risk | Time horizon, Follow-up care costs for recurrent AF, Ablation costs | Readiness of offering crossovers, Ablation costs |
AF = atrial fibrillation, AAD = anti-arrhythmic drug, RC = rate control, ASA = aspirin, NSR = normal sinus rhythm, y.o. = year old, mod = moderate, CAD = Canadian dollar, ICER = incremental cost-effectiveness ratio, QALY = quality-adjusted life year.