The cost-utility of catheter ablation of atrial fibrillation : a systematic review and critical appraisal of economic evaluations

9 p.
BACKGROUND: A health technology assessment (HTA) of catheter ablation for atrial fibrillation (CA-AF) was commissioned by the Belgian government and performed by the Belgian Health Care Knowledge Centre (KCE). In this context, a systematic review of the economic literature was performed to assess the procedure's value for money. METHODS: A systematic search for economic literature about the cost-effectiveness of CA-AF was performed by consulting various databases: CRD (Centre for Reviews and Dissemination) HTA and CDSR (Cochrane Database of Systematic Reviews) Technology Assessment, websites of HTA institutes, NHS EED (NHS Economic Evaluation Database), Medline (OVID), EMBASE and EconLit. No time or language restrictions were imposed and pre-defined selection criteria were used. The two-step selection procedure was performed by two persons. References of the selected studies were checked for additional relevant citations. RESULTS: Out of 697 references, seven relevant studies were selected. Based on current evidence and economic considerations, the rationale to support catheter ablation as first-line treatment was lacking.The economic evaluations for second-line catheter ablation included several assumptions that make the results rather optimistic or subject to large uncertainty. First, overall AAD (antiarrhythmic drugs) use after ablation was higher in reality than assumed in the economic evaluations, which had its impact on costs and effects. Second, several models focused on the impact of ablation on preventing stroke. This was questionable because there was no direct hard evidence from RCTs to support this assumption. An indirect impact through stroke on mortality should also be regarded with caution. Furthermore, all models included an impact on quality of life (QoL)/utility and assumed a long-term impact. Unfortunately, none of the RCTs measured QoL with a generic utility instrument and information on the long-term impact on both mortality and QoL was lacking.