A systematic review of economic evaluations on stent-retriever thrombectomy for acute ischemic stroke

Our objective was to review economic evaluations on stent-retriever thrombectomy (SRT) added/not added to intravenous (IV) tissue plasminogen activator (t-PA) in acute ischemic stroke (AIS) due to large-vessel occlusion (LVO). We conducted a systematic review using several electronic databases and searching for studies published from January 2009 to September 2017. Inclusion criteria: any publication type reporting the incremental cost-effectiveness ratio of SRT in people with AIS secondary to LVO. Quality assessment was undertaken with the CHEERS and the Philips' checklists. Eight original articles (four from North America/four from Europe) were included; of these, seven were model-based cost-effectiveness studies and one was a study conducted alongside a clinical trial. The perspective was the healthcare system in seven studies, and societal in one. The time horizon was lifetime (minimum 20 years) in all but two studies where it was 1 and 5 years. Overall, studies were rated of good quality (mean score 79%; range 70-90). Data sources, effectiveness outcomes and other input parameters were heterogeneous across studies. In three studies, SRT was dominant (less expensive and more effective). In five studies, SRT was more expensive and generated more quality-adjusted life years but had a high probability (79-100%) to be cost-effective at conventional thresholds. This review shows that SRT added/not added to IV t-PA is likely to be cost-effective or even dominant, which is consistent with the opinion from several Health Technology Assessment bodies recommending SRT. However, our findings are supported by primary studies with substantial methodological heterogeneity.