Introduction
Adjuvant treatments for breast cancer (BC), including chemotherapy, endocrine therapy, and oophorectomy can lead to treatment-induced menopausal symptoms [
1,
2]. These symptoms, and in particular hot flushes and night sweats (HF/NS), negatively affect health-related quality of life (HRQL) [
3‐
5] and cause some women to discontinue their endocrine treatments [
6,
7]. Although medications such as gabapentin, clonidine, and antidepressants are moderately effective in reducing HF/NS, they are accompanied by bothersome side effects [
8‐
11]. In contrast, cognitive behavioral therapy (CBT) programs are without side effects, are effective, and are favored by BC survivors [
12‐
16].
CBT programs have often been delivered in group format [
14‐
16]. However, BC survivors have reported practical and scheduling barriers to attending such group sessions [
16]. Therefore, these programs have been translated into an online format [
17,
18]. Our recent randomized controlled trial (RCT) comparing Internet-based CBT (iCBT), with and without therapist support, with a waiting list control group demonstrated that women allocated to iCBT experienced a greater reduction in overall levels of menopausal symptoms and perceived impact of HF/NS. Significant reductions in the frequency of HF/NS and improvement in sleep quality were also observed [
19]. When asked about preferences for a specific format, only a minority of women showed a strong preference for guided (16%) or self-managed (21%) iCBT. Although the magnitude of the effects favored the guided over the self-managed iCBT group, the former is associated with higher costs due to the added therapist support.
The observed differences in effectiveness and costs between the iCBT formats and the reality of budget restrictions underscore the need for an economic evaluation to assist policymakers in deciding whether to allocate healthcare resources to this program. Moreover, it may also guide practitioners in choosing which specific format to adopt [
20]. Although a previous study by Mewes et al. [
21] indicated that face-to-face group-based CBT was cost-effective, it is unknown whether online-delivered CBT will lead to favorable cost-effectiveness ratios as well. Moreover, there are no studies reporting the budget impact of iCBT for treatment-induced menopausal symptoms, commonly used to estimate the impact on national, regional, or local health budget plans [
22].
The objective of the current study was to evaluate the cost-utility and cost-effectiveness of guided and self-managed iCBT compared to a waiting list control group in terms of quality-adjusted life years (QALYs) and the primary clinical outcomes of the associated RCT (i.e., overall levels of menopausal symptoms and perceived impact of HF/NS), incorporating a healthcare perspective over a 5-year time period. An additional aim was to establish the estimated annual budget impact of implementing guided and/or self-managed iCBT in the Netherlands.
Discussion
This is the first study to investigate the cost-utility, cost-effectiveness, and budget impact of iCBT to alleviate treatment-induced menopausal symptoms in BC survivors. The results show that both the guided and self-managed formats of iCBT are associated with a small gain in QALYs over a 5-year time horizon, a decrease in menopausal symptoms, and a decrease in perceived impact of HF/NS. These improvements were accompanied with an increase in costs due to additional intervention and healthcare costs. However, analyses showed that ICURs are well below the proposed international WTP threshold of €30,000/QALY for both formats [
39]. The probability that the ICERs are considered acceptable ultimately depends on the willingness to pay for a clinically significant decrease in menopausal symptoms and/or perceived HF/NS. Our results indicate that, to accomplish a significant reduction in overall levels of menopausal symptoms or perceived impact of HF/NS, an investment between €1026 and €1525 for the guided and €193–€753 for the self-managed iCBT format will be necessary (the range reflects the perspective, i.e., only intervention costs, or intervention and healthcare costs). The annual Dutch budget impact (i.e., treating 600 patients) of implementing this program is estimated to be between €74,592 and €192,990 for the guided and between €28,752 and €74.592 for the self-managed iCBT. Additionally, sensitivity analyses showed that self-managed iCBT remains cost-effective (below the threshold of €30,000/QALY) for all variations in input parameters and assumptions, except when utility in the state ‘Reduction in Menopausal Symptoms’ decreases to its lower extreme value. For guided iCBT, shorter duration of intervention effects, increase in costs, decrease in utilities, and decrease in probability of obtaining a reduction in menopausal symptoms may result in unacceptable cost-effectiveness ratios, i.e., around €35,000/QALY or even higher ratios when utilities decrease unfavorably.
Compared to the economic evaluation of the group-based CBT program for alleviating menopausal symptoms in BC survivors [
21], we observed similar costs per QALY outcomes for the guided format, but a reduction of more than €10,000/QALY for self-managed iCBT. We also observed higher incremental costs per clinically significant reduction in overall levels of menopausal symptoms and perceived impact of HF/NS for the guided format (± €500), and lower incremental costs per clinically significant reduction for the self-managed iCBT, when compared with the group-based CBT format [
21]. This indicates that an Internet-delivered CBT program, particularly when self-managed, would be a viable alternative to face-to-face group sessions, with the added advantage of decreasing practical barriers as previously reported that hamper attendance at group sessions [
16,
21]. In addition, the estimated budget impact is low in comparison with the total healthcare costs associated with the treatment of cancer in the Netherlands [
46].
The increase in QALYs observed in our study and that of Mewes et al. [
21] are relatively small. We believe this to be inherent to the aim of the current program, which is not primarily focused on improving overall HRQL, but rather on reducing overall levels of menopausal symptoms and perceived impact of HF/NS. When using a generic indicator of HRQL such as the SF-36, important gains in more specific domains are often missed due to the lack of responsiveness of the instrument [
47], hence explaining the results from the deterministic sensitivity analysis. Therefore, cost-utility analyses should be supplemented by cost-effectiveness analyses in which the cost per condition-specific outcome are measured and taken into account in reimbursement decisions. Moreover, we encourage the development and testing of condition-specific preference-based instruments which can be used within the QALY framework [
47].
Based on our findings, we would recommend implementing the iCBT program according to a stepped care approach [
48] in which the self-managed program serves as the primary treatment option. Dependent on available budgets, patient preferences, and support needs, the iCBT program could be supplemented by therapist support. To keep the related costs of this guided format to a minimum, it is advisable to centralize the program within a limited number of treatment centers and have a relatively limited number of trained therapists. Future research is needed to be able to predict which women will benefit most from which format. Finally, as many BC survivors report a range of (interrelated) psychosocial and physical problems [
49‐
51], we would recommend efforts to combine and integrate various iCBT interventions (e.g., for cancer-related fatigue, sleep problems, etc.) to better serve BC survivors and possibly reduce overall costs of psychosocial care in oncology settings.
Limitations and strengths
This study has some limitations that should be noted. First, due to a lack of data, we did not include costs related to medication uptake. However, based on Mewes and colleagues [
21], we expect these costs to be relatively low and similar across the intervention and control group. Second, we assessed healthcare consumption via generic questions that did not inquire specifically about the reason for utilization. It is likely that the differences in healthcare costs may not so much reflect the costs associated with the different formats of the iCBT program, but rather other factors. Third, there is increasing interest in conducting economic evaluations from a societal perspective, including costs associated with, among other things, productivity loss [
52,
53]. While we had planned to include this perspective, it was evident to us that the productivity losses that were found during the trial could not be attributed to menopausal symptoms, but mainly to comorbid health conditions with which many BC survivors are faced.
This study also had noteworthy strengths. These included the RCT design, multicenter participation, high response rates, including both a healthcare and intervention perspective, evaluating both cost-utility, cost-effectiveness, and budget impact, and incorporating the intervention specific endpoints.
Conclusion
This economic evaluation of guided and self-managed iCBT supports its cost-effectiveness in three respects. First, the cost-utility analysis indicates a cost per QALY well below frequently used thresholds. Second, the cost to obtain a clinically relevant reduction of menopausal symptoms and/or perceived impact of HF/NS is modest for both formats. Third, the budget impact of both programs is negligible when compared to the total healthcare expenditure for treating cancer in the Netherlands. Additionally, while treatment effects were only slightly greater in the guided format, the self-managed format was associated with substantially lower costs and more stable results when testing various assumptions and/or parameters in sensitivity analyses. Taken together, our results tend to favor the self-managed version of the iCBT program over the guided format, and thus we would favor a stepped care approach in which the self-managed version of the program is the default option, with the guided version being reserved for those situations where women have a strong preference for such support and where sufficient funding is available for the additional costs involved.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.