This decision-tree model cost-effectiveness analysis, the first one that compares fesoterodine and other salient antimuscarinics in Spain, showed that the treatment with fesoterodine resulted in similar overall costs but greater QALY gain than treatment with either tolterodine or solifenacin. In other words, fesoterodine acquisition cost is outweighed by the lower costs related to both direct and indirect resources used considered in this study. This analysis tried to be as much robust as possible, including the relevant comparators (the ones that represent 96% of the Spanish year 2010 OAB treatment market), the adequate efficacy data and the adequate costs. Additionally, several univariate sensitivity analyses were done, to demonstrate the strength of the assumptions. As explained in the results section, fesoterodine continued being the dominant option in most sensitivity analyses done, particularly when the time horizon of analysis is set at 52 weeks. When the time horizon is changed to 12 weeks only (this happening in some particular patients who abandon the OAB treatment rapidly), fesoterodine would still be a cost-effective option when the NHS perspective is considered only: incremental cost-effectiveness ratios per QALY gained of 574€ and 14,568€ compared to tolterodine and solifenacin, respectively. However, if it is considered the societal perspective in the 12-weeks scenario, fesoterodine would be cost-effective relative to tolterodine only. Nevertheless, the plausibility of this scenario is really low even considering than persistence rate with antimuscarinics has been showed to be lower than with other treatments for chronic conditions . At least, in the case of solifenacin persistence, it has been communicated to be above 90% after 12 weeks of starting the therapy .
The model is considered to be comprehensive by including several aspects of OAB affecting the overall economic burden of disease. These include the cost of incontinence pads and the cost of lost productivity due to impairment at work or lost time at work due to OAB symptoms. Also, it includes the cost of associated co morbidities arising with uncontrolled OAB. The main reason of incorporating indirect costs is due to their relevance for OAB. Kobelt et al (2003)  conclude that in an otherwise healthy and professionally active group of individuals, waking at night to void, significantly diminishes their overall well-being, vitality and productivity, leading to a significant level of indirect and intangible costs. Additionally, over 21% of the OAB population worried about interrupting meetings with frequent trips to the toilet and 3% of the population changed jobs or were fired because of their bladder control problems . Considering these figures, lost productivity and lost wages were included in this analysis. Nevertheless, results are consistent with base case scenario even if indirect costs are not included in the analysis and the perspective of National health System is analyzed solely.
As in all scientific works, this analysis includes limitations and strengths. One of the strengths of the model is the use of the clinical trial data for the comparison of all alternatives, which ensures comparable patient populations among the treatments. However, unlike in the actual trial, the model assumes patients whose OAB is not controlled by treatment at each assessment point will discontinue therapy at that point, replicating a treatment pattern more reflective of real clinical practice. On the other hand, solifenacin was not included in the clinical trial and was modelled using published data. In order to diminish the possible impact in the results of solifenacin data, the meta-analysis of phase III clinical trials by Cardozo and colleagues was chosen , which examined data from 2,030 incontinent subjects. Inclusion criteria for these studies were rather similar to the fesoterodine trials. Another possible limitation is that our analysis was not able to incorporate the so-called out-of-pocket expenses, i.e.: incontinence pads not funded by the NHS, particularly because studies used as a sourcing data for the analysis were unable to differentiate it. Finally, this economic model was designed as deterministic, then; only point estimates are showed not allowing for statistical comparison between drugs in both costs and effectiveness. However, these results encourages for development of more sophisticated economic modelling of such drugs using, i.e., probabilistic approach.
As previously mentioned, this is the first cost-effectiveness analysis performed with Spanish data that compares fesoterodine with another antimuscarinics highly used in our health context. However, several economic analyses have been previously published with different antimuscarinic drugs in other settings. Considering the comparators included in this analysis, four economic evaluations have compared solifenacin to ER tolterodine in three different settings [38–41]. These analyses showed that treatment with solifenacin was less costly and more effective than tolterodine. Only one of the four economic evaluations includes fesoterodine as a comparator , with controversial findings in comparison with the results include in our work. The article included some information based in the experience in clinical practice use of all antimuscarinics except for fesoterodine since it was not marketed yet at the time of the analysis included in the paper (year 2008). Due to the lack of available data, fesoterodine inputs were based on extrapolated assumptions from the use of other antimuscarinics instead of data from clinical trials. Another difference is the type of resources and corresponding cost considered in the analysis, since only were considered some direct medical costs, while in the present economic analysis, following Spanish guidelines, indirect costs were also included . Drug-specific treatment persistence data were obtained from the Information Management System database and covered the 12-month period ending in April 2008. As fesoterodine was only on the market since July 2008, treatment persistence data from longitudinal databases were not yet available. To approximate the percentage of patients stopping and switching treatment with fesoterodine in the base-case analysis, Cardozo et al.  used the persistence rates for tolterodine ER, which could potentially represent a bias against fesoterodine. Finally, the percentage of patients who stop or switch treatment due to poor compliance was based on expert opinion.