Introduction
Because of the widespread use of imaging modalities, unruptured intracranial aneurysms (UIAs) are detected with increasing frequency. In Germany, the number of patients who were admitted to hospital for an UIA increased by a factor of 2.3 from 2005 to 2017, with a large proportion of patients being older than 69 years [
9].
The rupture of an intracranial aneurysm leads to subarachnoid hemorrhage (SAH), which is associated with high morbidity and mortality [
17]. The overall incidence of aneurysmal SAH is approximately 9/100,000 people per year [
17]. The appropriate management of UIAs depends on the individual risk factors for aneurysm rupture and the anticipated individual risk of treatment-related complications. Over the past two decades, endovascular treatment has evolved as the first-line therapy for UIAs and various devices and endovascular techniques have been developed since then.
Conventional coiling represents the long-term standard technique for endovascular aneurysm treatment. Stent-assisted coiling (SAC) and flow-diverters allow the treatment of morphologically complex aneurysms and can provide higher aneurysm occlusion rates than conventional coiling. However, these techniques are associated with an increased risk for ischemic stroke [
19]. Since the introduction of the Woven Endobridge (WEB) in 2011, intrasaccular flow-disruption has evolved as a proven concept for endovascular aneurysm therapy, in particular for the treatment of wide-necked bifurcation aneurysms. A further benefit of WEB over stent-assisted procedures is that it does not require long-term anti-platelet medication.
Besides clinical and angiographic outcomes, the economic impact of the respective endovascular technique may be relevant for health care decision makers and health care providers. Previous cost analyses revealed that the net material costs (e.g., stent or WEB) highly impact the overall hospital costs. However, little is known about the long-term economic impact of the WEB device for treatment of UIAs, in particular considering the overall treatment costs and long-term health benefits. The objective of this modeling analysis was to compare the cost-effectiveness of WEB treatment with SAC for the treatment of wide-necked unruptured aneurysms from the perspective of the German Statutory Health Insurance (SHI). The cost-effectiveness of conventional coiling was also calculated for reference, although wide-necked aneurysms are usually not treated with this method.
Discussion
In our model calculations, overall treatment costs (including follow-up examinations and potential rehabilitation costs) of the WEB (20,440 €) were slightly lower than that of SAC (23,167 €), while conventional coiling (8200 €) was the least expensive treatment option. Regarding health outcomes, WEB (13.24 years) and SAC (12.92 years) provided slightly more QALYs than coiling (12.68 years).
Notably, net material costs of WEB and SAC were comparable (11,470 € vs. 11,897 €) for the analysed subset of aneurysms (3–11 mm). Likewise, Kashkoush reported similar treatment costs for WEB ($ 18,530) and SAC ($ 18,950) [
23], while Rai et al. reported lower implant costs for the WEB ($ 17,028) than for SAC ($ 23,813) [
32]. The slightly increased lifetime costs of SAC over WEB may be largely ascribed to higher rehabilitation costs in the SAC group resulting from a slightly higher initial morbidity (mRS ≥ 2) for SAC compared to WEB (4.1% versus 1.3%). In contrast, the SAC material cost is not a driver of the lack of cost-effectiveness of the SAC strategy. In our analysis, a reduction in SAC material costs of 83% would be required to attain the same cost-effectiveness than WEB treatment. The initial net material costs of coiling (680 €) are distinctly lower than that of WEB or SAC, however, the comparatively high recanalization and retreatment rates increase the overall costs of this procedure. Nevertheless, overall treatment costs of coiling were the lowest.
Although SAC is associated with a higher complication rate and a higher procedure-related morbidity rate than coiling [
30], the gains of QALYs were similar for both modalities (12.92 vs. 12.68 years). We attribute this effect by an increased recurrence rate and hence a higher risk of rupture of recanalized aneurysms after coiling compared to SAC.
Results from PSA showed that WEB is potentially cost-effective at WTP thresholds of 30,000 €/QALY gained or higher, while at lower WTP coiling would be the preferred alternative. The results were presented for a range of WTP values, because in Germany there is no commonly accepted threshold [
15]. Previous cost-effectiveness analyses assessing interventions aimed at patients with UIA in other settings have applied country-specific WTP thresholds, namely $ 100,000/QALY for the United States, 20,000 to 30,000 £/QALY for the United Kingdom, and 80,000 €/QALY for the Netherlands [
6].
Although the hospital perspective was not the focus of our analyses, it should be noted that the same DRG codes are used for WEB and SAC, and therefore, the reimbursed lump-sums are in principle the same [
37]. However, material costs for WEB and SAC are covered separately though innovation payments. This financing mechanism was implemented in Germany to promote faster adoption of potentially beneficial innovative medical devices. It is used for DRG-tariffs which are currently not cost-covering for the hospital [
10], thereby preventing the provision of innovative devices (e.g., WEB and SAC).
When interpreting the results, it has to be kept in mind that the findings of the current study are only valid for saccular aneurysms between 3 and 11 mm in diameter. Indeed, the indication for the three evaluated modalities differ. In clinical practice, stand-alone coiling is predominantly used for small and large aneurysms with a favorable dome-to-neck ratio. The WEB is suitable for both wide- and narrow-necked aneurysms, however, it is restricted to a saccular shape and an aneurysm diameter ≤ 11 mm [
16]. SAC is a well-established treatment option for complex aneurysms including very large, fusiform, lobulated and partially thrombosed aneurysms and vessel branches arising from the aneurysm sac [
40]. All these aneurysm types are difficult to treat with the other two modalities and justify SAC as primary treatment for these aneurysms.
The costs of coiling and stent-assisted coiling vary depending on the size of the aneurysm. Smaller sized aneurysms may have lower net material costs, mainly due to a reduced number of coils, while larger aneurysms may be also treated by two overlapping stents. These specifics were considered in our analysis, as we retrospectively recorded the number of implanted devices in aneurysms ranging between 3 and 11 mm. For the WEB, the net material costs would remain stable assuming the implantation of a single device with the appropriate dimensions.
Although newer multi-center studies were available for the WEB, such as the WorldWideWEB Consortium [
7], we selected the prospective benchmark studies by Pierot et al. since most WEB studies report on ruptured and unruptured aneurysms and do not differentiate outcome parameters [
44]. In the studies by Pierot et al. the portion of ruptured aneurysms was only 8 and 51% were MCA aneurysms. Among high-quality WEB studies, these features fitted to our model best. Nevertheless, the angiographic results of the WorldWideWEB Consortium were also within the range that was considered in our deterministic sensitivity analysis, which led basically to the same results.
Although the WEB is associated with a reasonable safety profile in numerous studies and in our own experience, the complication and morbidity rates may be underestimated in studies of novel interventions or therapies (optimism/reporting bias) [
27]. In this context, the reported procedural morbidity of WEB treatment in several studies is lower than that of coiling in the employed meta-analysis by Phan et al. [
30] To account for this potential bias, we performed a structural sensitivity analysis assuming similar morbidity rates for coiling and WEB, which showed comparable results to our base case scenario.
Limitations
Although we performed this study with utmost methodological care, model-based cost-effectiveness studies have several inherent limitations. As the base case scenario was established from the perspective of the German SHI, generalization of the results can be difficult, in particular for different countries and health care systems with diverging reimbursement policy. In this context, material and overall treatment costs can also differ markedly between hospitals in the same country, mainly depending on the purchasing pools and discount contracts between individual hospitals, insurances and manufacturers.
The primary outcomes of this study are based on the results of prior studies that included aneurysms with different location, size and morphology and diverging patient characteristics. Therefore, the applied data may not completely apply to our base case scenario as there might be differences in QALYs, life expectancies, treatment risks and angiographic outcomes. The outcome data for SAC and coiling were mainly derived from a large meta-analysis of predominantly retrospective single-center studies, while the data for the WEB were recorded from three prospective multi-center studies. Admittedly, there are several reviews and meta-analyses on the WEB device, however, these did not report clinical and angiographic outcome for ruptured and unruptured aneurysms separately, which would distort outcome results for elective WEB treatment. In this context, the long-term efficacy of the treatment modalities are difficult to determine, in particular for WEB, for which long-term studies are still rare. At least, the employed study on the WEB report 2‑ and 3‑year angiographic results, which were included into this analysis.
The recruitment periods of the employed WEB studies were between 2011 and 2015. Hence, they did not fully cover recent advances in WEB technology and improvement of WEB handling (e.g. learning curve, +1/−1 rule), which might contribute to slightly better cost-effectiveness results. In this context, the employed meta-analysis on SAC included largely studies on conventional intracranial stent. More advanced stents can have a more flexible structure and a surface finishing designed to reduce device thrombogenicity, which might also increase the attractiveness of SAC from a cost-effectiveness perspective.
Another important limitation was that we adopted coiling for retreatment for all modalities due to the lack of relevant systematic studies. However, in clinical practice, retreatment after WEB implantation is more often performed with an additional stent, whereas retreatment after coiling and SAC is mainly performed by simple recoiling [
22,
38]. Retreatment with stents is naturally associated with increased retreatment costs and thromboembolic complications. However, as we assumed a rather low retreatment rate for WEB (3.4%) and SAC (6.2%) in our base case scenario, the main conclusions of this study are probably not significantly distorted by this simplified assumption.
Finally, diverging management strategies (i.e., aneurysm treatment and follow up regimen) and different health care systems might result in different cost-effectiveness of the individual endovascular treatment options. Although the generalizability of the base case result might be limited, we report enough methodological detail to allow adapting this model to other settings.
Despite these limitations, the findings of this study indicate that WEB treatment is a cost-effective treatment alternative to SAC for wide necked aneurysms. The results from our study can be considered in the setting of competing endovascular options for a patient with a standard saccular-shaped unruptured aneurysm to guide an informed decision that optimizes quality of life at affordable costs of intervention and rehabilitation.