To calculate the real costs of TAVR in the Portuguese healthcare system we selected, as a sample, the information collected from a high-volume Portuguese TAVR center, the cardiology department of Centro Hospitalar de Vila Nova Gaia. Using official hospital information from 2017, we prospectively collected data about all the material used (quantity consumed and price per unit), costs of human resources and costs of hospital facilities. Costs were calculated separately for each type of valve.23

As detailed in Table 1, costs were subdivided into preprocedural, procedural and postprocedural costs.

The costs of preprocedural exams (transesophageal echocardiography, computed tomography and blood tests) were obtained from the official table of costs of complementary diagnostic and therapeutic procedures of the Portuguese Ministry of Health for 2018.24 Before the procedure, the patient usually undergoes assessment by a multidisciplinary heart team that includes an interventional cardiologist, a cardiac imaging expert, an anesthesiologist and a cardiac surgeon, to assess their suitability for the procedure. The attributed cost was 71.5 euros of this assessment, which is the sum paid to the hospital, according to the Ministry of Health's 2018 official table of costs.

Regarding material costs, the price per unit was obtained individually for each item using 2017 official hospital consumption prices, after reductions or discounts. The medications used included heparin, propofol, droperidol, fentanyl, paracetamol, midazolam and cefazolin.

To calculate the costs of human resources, we considered the average salary of a Portuguese doctor (2746 euros/month), which including subsidies, bonuses and other extras, represents an annual cost of 48719 euros. Considering that the mean number of working hours is 151.7 per month, the cost per hour was estimated at 27 euros. The costs are the same for cardiologists or anesthesiologists. Two interventional cardiologists and one anesthesiologist are needed for a standard TAVR procedure. The procedure takes on average two hours, resulting in a total of six hours of medical work. The same method was used to calculate the price per hour of a nurse, which was 10.50 euros. At least four nurses are needed during the procedure, which corresponds to eight hours of nursing work.

Regarding the costs of hospital facilities, no detailed information was previously available about the cost per day of stay in the cardiac intensive care unit or on the cardiology ward. Therefore, this price was determined prospectively using hospital information from 2017. Information on the median time of hospitalization from January 2017 to January 2018 was used to calculate the median time of hospitalization after the procedure.

To estimate the costs per hour of use of the angiography room, we estimated that the total investment for an angiography room (including angiographic and other equipment and construction work) was 900000 euros, which was the total cost of such a room in this hospital acquired in 2017. Maintenance costs over the room's 10-year lifetime, 450000 euros in total, were added to this figure. Considering the number of hours of use (11hours per day in our center) on 250 days per year (to account for weekends, holidays, maintenance days and strikes, according to 2017 data), the estimated cost per hour of use was 55.0 euros.

The costs were subdivided into two types: fixed and variable costs (the latter depending on the risk and type of complications). To more accurately reflect the real impact of variable costs, we retrospectively collected information on the risk and types of complication in this hospital in the previous year (from January 2017 to January 2018), to take into account the TAVR learning curve.25 Because the risks and types of complication can differ according to the type of valve, a separate analysis was carried out for each valve type.

To assess the economic impact at national level, we used public information from the European Association of Percutaneous Cardiovascular Interventions (EAPCI) Valve for Life initiative, which recorded a TAVR penetration rate in Portugal of 54 procedures per million population in 2017.26

Table 1 details the costs of the TAVR procedure. The total cost of TAVR with an SEV was 22134.5 euros, while with a BEV it was 23321.5 euros. Most of this figure was due to the cost of the valve itself (74.5% and 81.5% of the total cost, respectively). Although the cost of the BEV is higher (19000 euros vs. 16500 euros), the valve kit includes some additional material which reduces this price difference. Moreover, the risk of complications was slightly lower with the BEV (296.2 euros vs. 337.4 euros), although this was insufficient to reduce the total cost, because the overall risk of complications was low. It is noteworthy that although the procedure is technically demanding, the human resources costs were very low, representing only 1.1% of the total cost of the procedure.

In 2017, Centro Hospitalar de Vila Nova Gaia performed 120 TAVR procedures: 62 with SEV and 58 with BEV. Therefore, TAVR had a local economic impact of 2 724 986 euros in 2017. At a national level, according to the EAPCI Valve for Life Initiative, 550 TAVR procedures were performed in the country in 2017,26 at a total estimated cost of 12500000 euros.

To estimate the future demand for the technology, three scenarios were constructed. In scenario 1 the estimated number of TAVR candidates was 1948 procedures per year, corresponding to 189 procedures/million population (Figure 2). If TAVR is expanded to patients with intermediate risk (scenario 2, Figure 3) the number of TAVR candidates will increase by 28% to 2488 candidates (241 procedures per million). Finally, if TAVR expands to scenario 3 (Figure 4), the increase will be 107%, corresponding to 4039 TAVR candidates (391 procedures/million).

Figure 2.

Estimated number of candidates eligible for transcatheter aortic valve replacement in Portugal in scenario 1, which represents the use of the procedure according to current guidelines. AS: aortic stenosis; TAVR: transcatheter aortic valve replacement.

(0.27MB).

Figure 3.

Estimated number of candidates eligible for transcatheter aortic valve replacement in Portugal according to scenario 2, which represents an expansion in the indications for TAVR to include intermediate-risk patients. AS: aortic stenosis; TAVR: transcatheter aortic valve replacement.

(1.13MB).

Figure 4.

Estimated number of candidates eligible for transcatheter aortic valve replacement in Portugal according to scenario 3, in which the clinical indication of TAVR is expanded to include low-risk patients at low risk aged over 75 years. AS: aortic stenosis; TAVR: transcatheter aortic valve replacement.

(0.28MB).

If these three scenarios were applied with current procedure costs, the economic impact would increase to 43770586 euros in scenario 1, 55904116 euros in scenario 2 and 90754310 euros in scenario 3.

It is also important to mention the need to improve the limited installed capacity to respond to scenarios 2 and 3, due not only to the predicted number of facilities needed, but also to the future demand for interventional cardiologists specializing in structural valve disease.

The total cost of TAVR hospitalization in Portugal (∼23000 euros) is significantly lower than in the US or even in central European countries.19,20,33,34 In Portugal, most of the cost of the procedure (75-80%) is driven by the cost of the material, a proportion that is higher than other countries, such as in France (65%).35 In absolute terms, the prices of the valves are not very different from the Netherlands19 (17590 euros) or France35 (19500 euros), but are significantly lower than in the US36 (26390 euros). Given the high costs of the procedure and the prospect of increasing numbers of candidates in the near future, it is important to discuss the reimbursement model for public hospitals and to implement strategies to improve it and simultaneously reduce costs, to enable more patients to be treated. It should be borne in mind that the aim of the present study was not to compare data on TAVR with surgical valve replacement or to determine the incremental cost of TAVR to the national healthcare system relative to surgery.

First, considering that the price of the valve itself is the most important driver of the total cost, saving strategies should concentrate on decreasing their price per unit. This can be accomplished by central negotiation of the price to be paid for the valves, by acquiring them at national level, instead of the hospital level.37

Although in our study the price of the SEV was slightly lower, both options should be kept available, so that the main driver of this choice can be the patient's clinical and anatomical characteristics,23 to reduce the risk of complications.

Secondly, TAVR should be performed only in high-volume centers that can further streamline the procedure with dedicated teams. Several studies have shown that this approach can reduce procedure time, decrease costs and reduce complications.38,39

Third, in this study we observed that the use of hospital facilities (especially stay in the cardiac intensive care unit) is the second most important contributor to the overall cost. To lower this cost, it is now possible to design and implement fast-track TAVR protocols in selected patients that allow very short ICU stays (<24hours), decrease total length of hospital stay (<4 days) and decrease costs.40,41 All of these ways to reduce costs should be discussed for application in the Portuguese healthcare system.

Several studies have analyzed the cost-effectiveness of TAVR, with differing results according to the clinical setting.42 In inoperable patients, almost all analyses show that TAVR has an acceptable cost to society,33 depending on the comorbidities of the treated population.43–45 In high- or intermediate-risk patients, clinical trial results show short-term advantages in quality of life and length of hospital stay, with similar survival rates for TAVR and surgical valve replacement.13,14,19 In the latter setting, cost-effectiveness analyses are less favorable, more variable and strongly influenced by methodology, patient characteristics and the jurisdictions and countries where they were applied.42

As shown in Figure 1, the current TAVR penetration rate in Portugal is below the European average,21 which highlights the considerable inequities in access to this technology.

Additionally, the penetration rate in this country is below that expected according to the current TAVR guidelines11 (53 vs. 189/million), which is valuable information for healthcare resource planning, especially to develop policy on investment in new equipment, to predict future needs in human resources, and to plan budgetary requirements.

This number may increase significantly in the near future if TAVR indications are expanded to patients at intermediate or even low risk, as estimated in scenarios 2 and 3. This change has already occurred in some countries, including Germany and Switzerland.17,18

Nevertheless, at current valve prices, it is highly unlikely that TAVR will be considered a cost-effective treatment in these patients.36 It is expected that with more competition and more patients treated, prices will decrease significantly in the near future, as has been seen with other new cardiovascular technologies.46 However, until rigorous cost-effectiveness analysis is available in these lower-risk patients, it is not recommended that current TAVR indications in Portugal should be expanded.

In Portugal there is no specific diagnosis-related group (DRG) for the TAVR procedure, and hospitals are only reimbursed with the DRG associated with traditional surgery (2316.48 euros per patient), which covers only 10.3% of the total cost of TAVR. This means that the remainder of the procedural cost is assumed by the local hospital, using its overall budget or by increasing the hospital's deficit. Some hospitals have accordingly imposed a cap on the number of procedures per year, but this process has been arbitrary and variable, and causes significant regional inequalities in access to the technology.47

Different countries have implemented various models of TAVR reimbursement. For example, in Germany and Switzerland, TAVR is reimbursed using a specific DRG. It is known that hospital financing using DRGs can improve efficiency and transparency and expand hospital activity,48–50 but can also increase total healthcare costs, especially when the DRG funding follows a linear model. In fact, there is a direct association between TAVR penetration rates and reimbursement by DRG.22 However, this system can also have unintended effects. For example, in Germany, the TAVR DRG is 2.3 times higher than the DRG of surgical replacement (∼35000 euros vs. ∼15000 euros),51 which has created an incentive to perform TAVR instead of surgery in lower-risk patients, as recently reported.17,18

In this study, data was collected retrospectively, in contrast to most economic analyses in TAVR, in which costs are usually calculated based on data collected from randomized clinical trials on TAVR.20,34 However, it is known that economic studies based on well-performed observational studies are equally valuable and can more accurately reflect real costs, adjusted for each country.52

To calculate procedural costs, the authors used data from a single TAVR center. Although we tried to include other centers, hospital administrations did not readily deliver the unit costs needed to perform this study. To obtain a clear picture of health expenditures in Portugal on different medical technologies and procedures, there should be more transparency regarding this type of data.

In Portugal, all centers performing TAVR are tertiary hospitals with similar characteristics,53 even though there may be slight differences in the personnel involved in the procedure, with variable ratios of radiology technicians to nurses. Salary costs are similar and therefore we do not expect there to be major differences between hospitals in the cost of the procedure.