Cost-effectiveness of endobronchial valve treatment in patients with severe emphysema compared to standard medical care

INTRODUCTION

The randomized-controlled STELVIO trial performed in our hospital showed that endobronchial valve (EBV) treatment significantly improved pulmonary function, exercise capacity, physical activity and quality of life after 6 months in COPD patients characterized by emphysema and the proven absence of interlobar collateral ventilation compared with standard of care.1, 2 Furthermore, these improvements were still present after 1-year follow-up.3 Afterwards, two larger randomized-controlled multicentre trials using the same technique in both homogeneous and heterogeneous emphysema confirmed our findings.4, 5 Therefore, the EBV treatment is an important asset to the treatment options for this patient group who have hardly any treatment options left. Only for a few of these patients, invasive surgical procedures such as lung volume reduction surgery (LVRS) and lung transplantation remain. Recently, the EBV treatment was included for the first time in the Global Initiative for Chronic Obstructive Lung Disease (GOLD) guidelines.6

Although showing significant efficacy, and being important for the quality of life of our patients, it is also important to perform an economic evaluation of this new treatment. Such an evaluation can provide justification for the costs of the treatment by relating it to the clinical benefit and allows comparison with other treatment modalities in support of policy decisions regarding reimbursement. Until now, only one study investigated the cost-effectiveness of the EBV treatment in a study population from Germany and concluded that the EBV treatment is cost-effective in the German healthcare system.7

The aim of our study was to perform a cost-effectiveness analysis in the short term (6 months) and long term (using a simulation model up to 10 years) of EBV treatment compared with standard medical care (SoC) in patients with severe emphysema in the Dutch healthcare system.

METHODS

Long-term economic evaluation

For the long-term cost-effectiveness evaluation, a Markov simulation model was constructed using Excel2010 (Microsoft Corp., Redmond, WA, USA). Transition probabilities derived from the STELVIO trial and subsequent follow-up and mortality data from the literature14 were used to model the monthly transition of patients (Fig. 1). Because of the detailed information for the first 6 months of the study, the time scale in the simulation model was in months. Table 2 shows the transition probabilities, costs and utilities (all fixed parameters). As prognosis in our patient group is closely linked to their GOLD stage, we distinguished three GOLD stages in our model, that is, GOLD II, III and IV. Transitions between the GOLD stages were modelled according to the short-term data from the randomized controlled trial (RCT) up to 6 months. After that, no transitions between GOLD stages were modelled and patients could only leave their GOLD stage by dying. Mortality estimates for these transitions were taken from the literature.14 In addition, background mortality due to other causes was included based on age-specific data from Statistics Netherlands.15 For each month that a simulated patient was in a certain state, costs corresponding to this state were calculated using half-cycle correction. Costs of dying and costs of transition between GOLD stages were set to zero. The total costs per month were summed up to present total costs of EBV treatment and usual care for the duration of the time horizon of the model. Summation was done with and without discounting, that is, with and without reflecting depreciated value of future costs and effects. The net effect of discounting is that early costs and effects receive more weight in the summation than late costs and effects. Common annual discount rates in economic evaluations are 3–5%, and we applied annual discount rates of 4%.

Table 2. Long-term economic evaluation of transition probabilities, costs and utility scores

Parameter	Untreated	EBV	Source
Mortality
GOLD II	0.06/year	0.06/year	Afonso et al.14
GOLD III	0.11/year	0.11/year	Afonso et al.14
GOLD IV	0.25/year	0.25/year	Afonso et al.14
Costs
0–6 months
Treatment	€0	€13.197	STELVIO trial
Events	€86/month	€506/month	STELVIO trial
Re-bronchoscopy	€0	€263/month	STELVIO trial
6–12 months
Re-bronchoscopy	€0	€90/month	STELVIO trial
>12 months
Re-bronchoscopy	€0	€67/month	STELVIO trial
Utility scores
SGRQ 0 month	0.52	0.53	STELVIO trial
SGRQ 1 month	0.55	0.71	STELVIO trial
SGRQ 6 months	0.55	0.74	STELVIO trial
SGRQ 12 months	0.58	0.68	STELVIO trial + extrapolated
EQ5D 0 month	0.66	0.63	STELVIO trial
EQ5D 1 month	0.72	0.75	STELVIO trial
EQ5D 6 months	0.67	0.75	STELVIO trial
EQ5D 12 months	0.66	0.77	STELVIO trial + extrapolated

• EBV, endobronchial valve; EQ5D, EuroQol5-Dimensions; GOLD, Global Initiative for Chronic Obstructive Lung Disease; SGRQ, St George’s Respiratory Questionnaire; €, Euro.

The model was populated with 100 patients in each group and distributed across the GOLD stages in accordance with the RCT population. The average patient age at the start of the Markov model simulation was set at 50 years. In view of the limited availability of long-term estimates of patient survival and the known poor overall survival of this patient group,14 the time horizon of the long-term evaluation was restricted to 10 years.

RESULTS

Short-term evaluation

The resource use of both groups is shown in Table 1. In the EBV group, per patient on average 1 Chartis catheter, 1.4 EBV catheters and 4.5 EBV were used. The mean average cost of the EBV treatment in the STELVIO trial was €13 197. At 6 months follow-up, the mean cost difference between the EBV group and SoC group per patient was €16 721 (95% CI: €16 675–16 766 after bootstrapping). In the randomized 68 patients, the mean difference in change in EQ5D utility score between groups at 6 months was 0.12 (95% CI: 0.01–0.24, P = 0.04) and for SGRQ utility score was 0.16 (95% CI: 0.07–0.24, P < 0.01) (Table 2). Unfortunately, there was a significant difference between groups at baseline in EQ5D utility score (control group: 0.66 and EBV group: 0.63). As a consequence, the ICER was not reliable for this outcome (−€1 941 250). After bootstrapping and adjusting for the 1-year time period, the ICER per additional quality-adjusted life-year (QALY) was €205 129 (95% CI: €203 547–206 709) using the SGRQ.

We also calculated the ICER for the change in 6MWD and for the change in SGRQ total score (Fig. 2). The ICER for the 6MWD at 6 months follow-up was €160.46 (95% CI: €159.66–161.27), meaning that 1-m improvement in 6MWD costs €160 extra in comparison with SoC. The ICER for SGRQ was €1240.71 (95% CI: €1253.99–1227.42), meaning that 1-point improvement in SGRQ total score costs €1241 extra.

DISCUSSION

We investigated the cost-effectiveness for the EBV treatment in severe emphysema patients with proven absence of collateral ventilation using data from a randomized-controlled trial.1 For the first 6 months, the incremental costs per QALY gained was €205 129 and the ICER for the 6MWD was €160 and for the SGRQ total score was €1241. Consequently, reaching the minimal important difference (MID) in 6MWD of 26 m would cost €4160 extra in comparison with SoC and reaching the MID of 7 points in SGRQ total score would cost €8687 extra. In the long term, the cost-effectiveness ratios were much more favourable, as higher costs of EBV treatment were partially compensated by lower mortality and better quality of life. The additional costs per life-year gained were around €71 500 and per QALY gained were around €39 000 with a 5-year time horizon. With a 10-year horizon, the incremental costs were around €29 000 per life-year gained and around €21 500 per QALY gained.

The other study investigating the cost-effectiveness of the EBV treatment in Germany found an ICER of €46 322 per QALY gained over 5 years and an ICER of €25 142 per QALY over 10 years.7 We found comparable values despite the differences in modelling structure and input variables (5 years: €39 638, 10 years: €22 375 using the same SGRQ). For example, there is a large difference in the price of the EBV treatment. In the German population, the treatment price per patient was €9851 compared with €13 197 in our population. This could be explained by the higher average number of valves used in our trial, the inclusions of the (Chartis) catheters in the price evaluation and the difference in costs of hospital admissions between countries. Unfortunately, no short-term economic evaluation is reported in the German study. Regarding the generalizability of our results in a Dutch population from an international perspective, it is positive that similar results were observed in a German population. However, even though Germany and The Netherlands are similar countries with respect to population, unit prices differed between these countries which will influence the actual cost-effectiveness ratio. Therefore, the generalizability of our cost-effectiveness ratios should be evaluated taking the country’s healthcare system and differences in unit prices into account.

To put the cost of the EBV treatment in perspective, we compared the costs-effectiveness analyses of other treatments for advanced emphysema patients such as lung volume reduction coil treatment, LVRS and lung transplantation. Another bronchoscopic lung volume reduction treatment option is the treatment with coils.16, 17 The REVOLENS trial investigated the lung volume reduction coil treatment and reported an ICER of $782 598 (approximately €738 400) per additional QALY, calculated over a 1-year period.18 This is much higher than our calculated ICER of €205 129 using the same SGRQ at 1 year. This could be explained by the higher cost difference of the coil treatment versus the EBV treatment ($47 908/€45 071 vs €16 721) and the less distinct efficacy improvements of the coil treatment. Therefore, the relative cost-effectiveness is in favour of EBV treatment. However, there is a large group of patients who cannot be treated with EBV but could benefit from the coil treatment. For LVRS, one study in Canada found a cost-effectiveness ratio of €90 611 per QALY gained ($133 900, Canadian Dollars) compared with SoC over a 2-year time horizon. Furthermore, the NETT trial investigating LVRS reported a cost-effectiveness of LVRS versus medical therapy of €125 521 per QALY gained ($140 000, American Dollars) at 5 years and €48 415 per QALY gained ($48 000, American Dollars) at 10 years.19 Therefore, the cost-effectiveness of LVRS compared with EBV treatment also seems in favour of EBV treatment. For lung transplantation, a study in the UK reported costs-effectiveness of €43 413 ($48 241) for single lung, €29 410 ($32 803) for double lung per QALY gained over a 15-year time period.20 Therefore, the cost-effectiveness of EBV treatment also seems to be superior to lung transplantation. However, lung transplantation is only an option for a small group of patients.

There is a large difference between the costs of the treatment in terms of QALY in the short term compared with the long term. Unfortunately, until now, no efficacy and safety results in the long term are published, but we believe that potentially the patient will benefit on average from the treatment at least until 5-year follow-up. The mortality risk of this severe patient group is very high (Table 2 14) and the positive clinical effects support the risk/benefit ratio and additional costs of this treatment. Furthermore, long-term follow-up data on survival of successfully treated patients with EBV confirm this assumption.21 However, this study did not include an SoC group, and therefore, more research is needed to gain insight in the potential improved survival after the EBV treatment. Therefore, the costs related to treatment should be allocated over the total duration of the time that patients benefit. Unfortunately, statistical modelling is necessary for this evaluation as no 5-year follow-up data are available and also it is not ethical to have a control group for this long time period. However, modelling will always be a simplification of the real-world situation. The short-term economic evaluation was based on actual measured data, which is a strength of this analysis.

A limitation of this trial is that the follow-up of the control group was only 6 months before these patients received the EBV treatment. A longer control group follow-up for the cost-effectiveness evaluation would have been better as a QALY is calculated per year and to avoid the use of estimates based on clinical experience.

Another limitation could be that we only included the indirect costs and in the long term only included the EBV-related complication costs like revision bronchoscopies. We did not have information on, for example, the number of exacerbations in the long term in a control group or EBV group, and therefore, we did not include this in the model. Similarly, we did not have data on healthcare use outside of our hospital, such as general practitioner visits. However, we expect that including these costs would probably be in favour of the EBV treatment and that by excluding them we do not overestimate the real difference.

In conclusion, our results suggest that the EBV treatment has a favourable cost-effectiveness profile, also when compared with other treatment modalities for this patient group. We believe that the favourable cost-benefit profile would justify this treatment for this severe patient group with hardly any other treatment options left.

Disclosure statement

D.-J.S. is a physician advisor to PulmonX Inc., USA, an investigator for trials sponsored by PulmonX Inc. and received lecture and travel fees for case support, educational and scientific sessions. PulmonX Inc. had no involvement of any kind in this trial.