Cost-effectiveness Analysis Comparing Osia System to Percutaneous Bone Conduction Devices in Sweden

Abstract

Objectives:

Bone conduction devices (BCD) have been commercially available since the 1980s for patients with conductive hearing loss (CHL), mixed hearing loss (MHL), or single-sided deafness (SSD). BCDs significantly improve the hearing capacity of patients regardless of age, which has a positive impact on their quality of life, mental health, education, and work performance. BCDs are categorized as percutaneous or transcutaneous solutions based on the attachment method. Selecting the optimal solution for patients requires consideration of audiological outcomes, guidelines, safety, patient preferences, and costs. This study evaluated the cost-effectiveness of an active transcutaneous solution (Osia System, Cochlear Ltd.) compared with percutaneous solutions to treat adult patients in Sweden.

Materials and methods:

A Markov model was developed to simulate a cohort of adult patients aged 59 years or older receiving Osia or percutaneous BCDs. The model considered 3 health states: patients who have received a BCD, patients without a device due to explantation (without reimplantation) or elective device non-use, and an absorbing state (death). The final state was based on age-dependent mortality, assuming hearing loss does not impact mortality. Probabilities were determined for transition between the health states with 6-month cycles and a lifetime horizon. Utility scores for each health state were derived from published literature using the Health Utilities Index Mark III (HUI3). Complication rates were also sourced from the literature, and the cost of complications and implantation from the National Board of Health and Welfare in Sweden, Socialstyrelsen. The perspective of the model was the Swedish health care system, and the main outcome was incremental cost per quality-adjusted life year (QALY) gained. Deterministic and probabilistic sensitivity analyses were conducted to assess the robustness of the results.

Results:

Osia was associated with an increase in cost of 79,293 SEK and an increase in QALYs of 0.73 compared with percutaneous devices, resulting in an incremental cost-effectiveness ratio of 108,318 SEK per QALY gained. The main drivers of cost-effectiveness were baseline age, the cost of surgery, and the difference in utility scores between the two interventions. The differences in HUI3 scores may be linked to the reduced risk of complications, improved cosmetic appearance, and improved hearing with Osia.

Conclusions:

Osia is cost-effective compared with percutaneous devices over a lifetime horizon, when applying low to moderate cost per QALY willingness-to-pay thresholds indicated by the Swedish Agency for Health Technology Assessment and Assessment of Social Services. This analysis provides information on the economic impact of different hearing solutions to support decisions on which BCD best meets the needs of patients and health care budgets.

Introduction

Hearing loss has far-reaching consequences, negatively affecting speech and language development, relationships and social interactions, education, employment, quality of life, mental health, and independence at various stages of life.^1–4 Globally, more than 1.5 billion people experience a deterioration of hearing during their lifetime.⁴ Due to the aging population in Sweden and other developed countries, hearing loss is likely to become an increasingly prevalent functional impairment.¹ Thus, there has been a call for multidisciplinary action on hearing health care, including comprehensive screening programs, noise-reduction strategies, and increased access to hearing technologies.⁵

Sweden’s health care system is decentralized, where the Ministry of Health and Social Affairs, supported by national agencies, oversees health care policy and regulations, and the 21 regions handle financing, organizing, and delivering health care. Health care services are publicly funded with minimal direct cost to the patient. Although regions and municipalities must follow national regulations, they have significant autonomy within these directives.^6,7 Priorities in Swedish health care are guided by 3 principles: individual value, need and solidarity, and cost effectiveness.^8,9 Hence, cost-effectiveness analyses serve as a basis to assist in making treatment decisions. Since each region is responsible for managing and prioritizing its own health care resources, guided, but not restricted, by central recommendations, individual priorities are made not only by regions but also by institutions and individual practitioners.¹⁰

BCDs have been commercially available since the 1980s, providing a benefit to eligible patients with conductive hearing loss (CHL), mixed hearing loss (MHL), and single-sided deafness (SSD).¹¹ When conventional hearing aids cannot be worn, for example, in cases of recurrent external otitis, ear canal eczema, ear canal atresia, or after extensive ear surgery, BCDs are the best option.^12,13 Hearing rehabilitation using BCDs is shown to improve functional hearing and health-related quality of life, work performance, and satisfaction.¹⁴ Hearing performance, in terms of aided hearing thresholds and speech understanding in noisy environments, has also been demonstrated to improve when using BCDs.¹⁵

BCDs can be categorized as percutaneous or transcutaneous.¹¹ There is greater potential for adverse skin reactions with percutaneous devices due to protruding, skin-penetrating abutments compared with transcutaneous devices.¹¹ A significantly higher proportion of patients who received transcutaneous BCDs had healthy, normal skin with minimal adverse events (86.5%) compared with patients who received percutaneous devices (45.5%, P = 0.002) in a prospective study of 70 patients who received BCDs.¹⁶ Transcutaneous devices are designed to avoid the cosmetic and skin complications associated with percutaneous devices, while still improving hearing outcomes.¹¹ The Osia System (Osia) (Cochlear Ltd.) is an active transcutaneous solution and is the first osseointegrated steady-state implant approved for people with CHL, MHL, and SSD and bone conduction hearing loss up to 55 dB. Osia uses digital piezoelectric stimulation to bypass nonfunctional areas of the natural hearing system and conducts the sound through bone to the cochlea.¹⁷ Osia is extremely responsive to high frequencies compared with other transcutaneous osseointegrated devices.¹⁸

Selecting the right BCD involves considering patient preferences, audiological outcomes, safety, and potential adverse events, as well as economic implications from budget constraints and pressure on health care systems. The objective of this cost-effectiveness analysis (CEA) was to assess the cost-effectiveness of treating patients with Osia compared with percutaneous devices, from a Swedish health care system perspective, to inform decision-making.

Materials and methods

Model design

The model was developed in Microsoft Excel following the principles of good practice for CEA in economic modeling.¹⁹ The analysis used a Markov model as this best represents evolving chronic diseases and can be used to estimate long-term costs and consequences. The cohort included adult patients with hearing loss aged 59 and older (the average age of adult patients receiving hearing implants in Sweden between 2013 and 2022 according to Socialstyrelsens statistikdatabas), who required a hearing implant and were treated with Osia or percutaneous devices.²⁰ The model analyzed patients from the age of 59 to 100 years (41-year period) to reflect a lifetime horizon. Three health states in total were considered in the model (Fig. 1); 2 health states (“with implant” and “without implant”) and 1 absorbing state (death). In the initial “with implant” state, patients were implanted with either Osia or a percutaneous device. In this state, they could experience no or mild complications that required no further intervention, or moderate or severe complications or explantation (removal of the device, potentially followed by reimplantation), which results in additional costs and patient burden. Patients could transition from “with implant” to a “without implant” state based on a probability of non-use (eg, patients who discontinue their device), and a probability of explantation without reimplantation due to a severe complication. The probability of transitioning from “with implant” to “without implant” was calculated from the rates of explantation and elective non-use from published literature for both Osia and percutaneous implants (Table 1, Supplemental Digital Content 1 and 2, http://links.lww.com/MAO/C318, http://links.lww.com/MAO/C319). The following formula was used to transform rates into cycle-adjusted probabilities:

Figure 1.

Cost-effectiveness model structure.

Table 1.

Model parameters and inputs for osia and percutaneous devices

Model Parameters	Device	Input (Sample Size)a	Source
Time horizon	Both	41 y	Based on baseline age of 59 and a 100-year life expectancy accounting for mortality rate
Cycle length	Both	6 mo	Assumption
Annual discount rate	Both	3%	Recommended by Swedish guidelines21
Age at baseline	Both	59 y	Based on average age of adults requiring hearing implants
Utility values for each health state
Baseline utility value (HUI3)	Without implant	0.67 (N= 75)	Brunner et al, 202422
Mean change in utility from baseline (HUI3)	Osia	0.09 (N= 69)	Brunner et al, 202422
Mean change in utility from baseline (HUI3)	Percutaneous devices	0.048(N= 43)	van Hoof et al, 202023
6 mo complications probabilities
Moderate complications	Osia	1.35% (N= 222)	Key et al, 202424
	Percutaneous devices	2.72% (N= 701)	Teunissen et al, 202425
Severe complications	Osia	1.80% (N= 222)	Key et al, 202424
	Percutaneous devices	1.93% (N= 701)	Teunissen et al, 202425
Explanationc	Osia	6.10%b (N= 222)	Key et al, 202424
	Percutaneous devices	10.12% (N= 701)	Teunissen et al, 202425
Reimplantationd	Osia	78.04% (N= 1,500)	Crowder et al, 202126
	Percutaneous devices	83.24% (N= 701)	Teunissen et al, 202425
Elective non-use	Osia	2.60% (N= 20)	Cowan et al, 202327
	Percutaneous devices	0.58% (N= 701)	Teunissen et al, 202425

^a As reported in the publication used as source.

^b 0.11% is reported as the predicted incidence for explantation in Key et al, 2024, but this was converted to a 6-month probability for the purposes of the Markov model 6-month cycles.

^c As a proportion of all severe complications.

^d As a proportion of all explantations

HUI3 indicates Health Utilities Index III.

$$\mathrm{P}=1-e^{(-r*t)}$$

where, P is the probability, r is rate and t is time.

No direct costs or risks of complications associated with implants were assigned to patients in the “without implant” state. The model did not account for partial non-use of devices, as it was assumed that people either use their devices or choose not to use them. The probability of experiencing a complication at every cycle was applied until the end of the model. However, for non-use, the probability of experiencing a complication was limited to 24 months, as this was considered more realistic and reflected the minimum follow-up period from the clinical studies.^25,27 As no patient-level data were included, no specific ethics committee approval was required. To estimate mortality, 2019 life expectancy data for Sweden were used from the World Health Organization (WHO).²⁸ The probability of death due to age-specific mortality is taken from the WHO Life tables for Sweden and varies by age. All rates and probabilities for complications and mortality were transformed and adjusted to the cycle length of the model, following principles of good practice in economic modeling.²⁹

Utilities

Utilities from the Health Utilities Index Mark III (HUI3) questionnaire were sourced from the literature and were used to calculate the number of quality-adjusted life years (QALY).^22,23 When individuals received an implant, there was an improvement in health-related quality of life that was assumed not to deteriorate over time, provided they continued to use their device. The mean change in utility score for both Osia and percutaneous devices was added to baseline values and was assigned to the patients in the “with implant” health state. As no aggregate health utility data were identified in the literature specifically for percutaneous devices, the utility for the percutaneous hydroxyapatite-coated Cochlear Baha BA400 abutment was used for the broader category of percutaneous devices. Disutilities associated with complications for BCD were not reported in the literature, therefore, in the base case, no disutilities were applied. However, this was explored in a scenario analysis where complication-related disutilities were applied from a Health Quality Ontario health technology assessment of bilateral cochlear implants.³⁰

Costs

As a health care system perspective was used, only direct health care costs incurred during treatment were considered. These costs included surgical implantation procedural costs (which include both the device and initial sound processor costs), the costs of sound processor upgrades, and the cost of complications for moderate and severe complications, explantation, and reimplantation (Table 2). The cost of each surgical procedure and the costs of management required to treat various complications were sourced from the Swedish National Board of Health and Welfare.³¹ Complications were split into moderate complications, which were assumed to require a physician visit and conservative management, and severe complications, which required hospitalization or revisional surgery. For severe complications, a percentage of patients could also have explantation of the device, with or without a reimplantation. Complication and explantation probabilities were sourced from the published literature (Supplemental Digital Content 1, http://links.lww.com/MAO/C318). There are no data specifically from the Swedish population, therefore, these were taken from available published evidence (a systematic literature review of 14 studies from various countries, a US study, and a retrospective cohort study conducted in the Netherlands, Table 1). For percutaneous devices, the Holgers classification was used with grade 2 complications categorized as moderate complications, and grade 3 and grade 4 as severe complications.^24–27 Every 5 years, all patients with an implant receive a replacement sound processor as part of routine processor upgrades. The model also assumes that patients undergo a number of scheduled health care visits, and these are the same in both arms.

Table 2.

Cost inputs for Osia and percutaneous devices

Cost Inputs	Device	Cost	Reference
Procedure cost	Osia	121,334 SEK	NordDRG31
	Percutaneous devices	28,587 SEK	NordDRG31
Moderate complications	Osia	4467 SEK	NordDRG31
	Percutaneous devices	4467 SEK	NordDRG31
Severe complications	Osia	20,060 SEK	NordDRG31
	Percutaneous devices	20,060 SEK	NordDRG31
Explanation	Osia	12,408 SEK	NordDRG31
	Percutaneous devices	5029 SEK	NordDRG31
Reimplantation	Osia	121,334 SEK	NordDRG31
	Percutaneous devices	28,587 SEK	NordDRG31
Sound processor replacement	Osia	58,474 SEK	NordDRG31
	Percutaneous devices	58,474 SEK	NordDRG31

As a proportion of all severe complications.

As a proportion of all explantations.

DRG indicates diagnosis-related group; SEK, Swedish krona.

Outcomes and analysis

The model used a 6-month cycle length, in line with other economic evaluations for bone-conduction implants and to reflect the occurrence of complications in real-world clinical practice.³⁰ A half-cycle correction was used to prevent over or underestimation of the costs and benefits of treatment. A default annual discount rate of 3.0% was assumed for both costs and health outcomes, as recommended by Swedish guidelines.³² The incremental cost-effectiveness ratio (ICER) was calculated by dividing the difference in total costs (incremental cost) by the difference in QALYs (incremental QALYs) over the patient’s lifetime. In Sweden, no public willingness to pay (WTP) thresholds are available, therefore, various thresholds were used to determine cost-effectiveness following the methodology identified by the Swedish Agency for Health Technology Assessment and Assessment of Social Services.²¹

Sensitivity analysis

One-way sensitivity analysis was conducted for all parameters to explore the main drivers of results. All parameters were independently varied by 20% above and below the base-case values, and a tornado diagram was used to evaluate the potential impact of the different input variables. A probabilistic sensitivity analysis (PSA) was also performed to account for uncertainty in the model. Each parameter was simultaneously varied over 1000 Monte Carlo simulations. Results of the PSA were plotted in a cost-effectiveness plane, and cost-effectiveness acceptability curves were used to show the probability of cost-effectiveness of the 2 devices over different WTP thresholds.

Results

Overall, Osia demonstrated cost-effectiveness over percutaneous devices over a lifetime horizon (Table 3). Osia was associated with an increase in cost and QALYs of 79,293 SEK and 0.73, respectively, resulting in an ICER equal to 108,318 SEK per QALY gained. In the scenario analysis where complication-related disutilities were applied, incremental QALYs increased to 0.74 and lowered the ICER to 107,346 SEK per QALY.

Table 3.

Cost-effectiveness results

	Osia	Percutaneous Devices
Total costs per patient	336,642 SEK	257,349 SEK
Total QALYs per patient	15.28	14.55
Incremental cost per patient	79,293 SEK
Incremental QALYs per patient	0.73
ICER	108,318 SEK/QALY

ICER indicates incremental cost-effectiveness ratio; QALY, quality-adjusted life year; SEK, Swedish krona.

Sensitivity analysis

One-way deterministic sensitivity analysis results showed that the main drivers of cost-effectiveness were baseline age, the cost of surgery, and the difference in health utilities between the different hearing loss devices. The tornado diagram shows from top to bottom the variables with the highest impact on the analysis and the ICERs associated with an increase or decrease of the baseline values by 20% (Fig. 2). The PSA yielded distributions of incremental costs and QALYs, which were plotted on the cost-effectiveness plane (Fig. 3A). The majority of simulations lie in the northeast quadrant of the plane, meaning that Osia is more costly but also more effective than percutaneous BCDs. Cost-effectiveness acceptability curves were also used to plot a range of cost-effectiveness thresholds (from 0 to 1 million SEK/QALY) on the horizontal axis against the probability that the intervention will be cost-effective at that threshold on the vertical axis (Fig. 3B). The probability of Osia being cost-effective ranged from 43% (at a WTP of 100,000 SEK per QALY) to 100% (at a WTP of 500,000 SEK per QALY). This can also be seen on the cost-effectiveness plane, where the 100,000 and 500,000 SEK/QALY WTP were drawn as diagonal lines (Fig. 3A). Any estimated ICER that falls below and to the right of the WTP line is to be considered cost-effective.

Figure 2.

Tornado diagram of deterministic sensitivity analysis results.

Figure 3.

A and B, Probabilistic sensitivity analysis results. A, Distribution of probabilistic sensitivity analysis results. B, Cost-effectiveness acceptability curves.

Discussion

This model was developed to support decisions on the use of the Osia active transcutaneous solution compared with percutaneous implants for the treatment of hearing loss in adult patients in Sweden. The results show that Osia is cost-effective over a lifetime horizon across multiple WTP thresholds identified in Sweden. The Swedish Dental and Pharmaceutical Benefits Agency, the Swedish Association of Local Authorities and Regions, and the National Board for Health and Welfare do not provide a specific cost-effectiveness threshold for medical devices and pharmaceuticals in Sweden.³³ However, in an analysis of approved reimbursements between 2005 and 2011, the median approved cost per QALY used by the Swedish Dental and Pharmaceutical Benefits Agency (Tandvårds- och läkemedelsförmånsverket, TLV) was 350,000 SEK per QALY.^21,34 Furthermore, cost per QALY was categorized as low (<100,000 SEK), moderate (100,000 to 499,999 SEK), high (500,000 to 1,000,000 SEK), or very high (>1,000,000 SEK) in the National Guidelines for cardiac care by the Swedish National Board for Health and Welfare, indicating acceptable WTP thresholds.^21,33 Thus, the cost per QALY for Osia is relatively low when compared with previously accepted thresholds. This analysis demonstrates the cost-effectiveness of Osia over percutaneous devices and supports decision-making on the availability of Osia for adult patients in Sweden.

The ICER for Osia is also lower than other commonly implanted medical devices.³³, such as total knee replacement (between approximately 162,384 and 267,752 SEK per QALY gained).^35,36 and unilateral hip replacement surgery (∼369,943 SEK per QALY gained).³⁷

The baseline age in the model was 59 years, as this was the average age of BCD implantation in Sweden between 2013 and 2022 according to the national database, Socialstyrelsens statistikdatabas.²⁰ However, this is likely influenced by the data provided on percutaneous implants, as these have been available for longer. Transcutaneous implants are often implanted in younger patients, and consequently, the baseline age of ≥59 limits the generalizability of these results to younger patients and is likely to be lower in the future. The baseline age was tested in the sensitivity analysis and was found to be one of the inputs with the most impact on the results, underscoring that the results for younger populations could differ meaningfully, which must be taken into account.

The variation in HUI3 scores may be linked to the different complication rates (reflected in the “pain” attribute) and patient preferences (“emotion” attribute). Unfortunately, not all studies measuring HUI3 report scores for all attributes, and comparisons were, therefore, not possible. However, Osia scored higher than a percutaneous device in the “hearing” attribute of HUI3 at 12 months after surgery (0.149 vs 0.080, respectively),^13,23 thus demonstrating improvements in hearing outcomes for patients. Percutaneous devices are still an appropriate option for certain patients, for example, those for whom general anesthesia is contraindicated or who need frequent magnetic resonance imaging (MRI) scans in the head and neck area.³⁸ The newer Osia System with OSI300 enables patients to undergo MRI at 3.0 Tesla without the need for surgical magnet removal.¹⁷ However, this is still contraindicated in MRI scans of the head and neck.

The findings from this study on the cost-effectiveness of Osia are consistent with several recent economic analyses. In a 2024 CEA comparing Osia to the passive transcutaneous Baha Attract System in Australia.²², Osia was found to be cost-effective with an ICER of AUS $29,301 per QALY gained over a 10-year time horizon due to its audiological outcomes, which resulted in a clinically meaningful utility benefit.²² A 2024 cost-utility analysis of cochlear implants for SSD in adults and children in the Netherlands reported the ICER for adults with a cochlear implant as €3,494 per QALY gained over a lifetime horizon compared with BCDs.³⁹ In a cost-effectiveness analysis (2021) of unilateral cochlear implants in Swedish adults with severe hearing loss, the ICER was SEK 140,474 per QALY gained over a lifetime horizon compared with hearing aids.³³ Furthermore, a 2022 cost-utility analysis compared unilateral cochlear implants with hearing aids or no hearing aids for people with severe hearing loss in the UK. Unilateral cochlear implants were deemed cost-effective when compared with hearing aids (£11,946 per QALY) or no hearing aid (£10,499 per QALY).⁴⁰ In addition, a 2024 narrative review of 9 European studies that focused on the cost-effectiveness of cochlear implants for adults with severe to profound hearing loss found that cochlear implants yield positive societal benefits for working-age individuals, despite substantial variability among countries in factors such as time horizons, cost-effectiveness thresholds, methods of cost collection, eligibility criteria, and country-specific health systems.⁴¹ This review recommended further country-specific cost analyses that include detailed information on health systems and all associated costs and benefits. The current model provides data to help fill this gap by comparing Osia with percutaneous implants for the treatment of adults with hearing loss in Sweden.

Limitations

No head-to-head study was identified that compared Osia and percutaneous devices. Thus, the inputs for model parameters had to be sourced from different clinical studies, adding heterogeneity and uncertainty to the analysis. However, this uncertainty was addressed by conducting sensitivity analyses and highlighting the parameters that were most sensitive to change. For complications and explantation rates, there was no published data specifically from a Swedish population, therefore, multinational data were used. This was predominantly from the United States and Europe and is expected to be generalizable to the Swedish population. For unit costs, Swedish data were sourced. An additional limitation is that the published clinical evidence for percutaneous devices generally had longer follow-up and more patients compared with Osia, due to their longer time on the market. Consequently, there is less long-term, real-world data on Osia’s effectiveness and rate of complications. This limitation was addressed by avoiding extrapolation beyond 2 years for non-use, as this would result in an unrealistic number of patients without implants. In addition, this limitation will be addressed when longer-term, real-world data for Osia are available in the future.

The model was designed to support decision-making on which hearing implant to use for the treatment of CHL, MHL, or SSD by understanding the cost-effectiveness of Osia compared with a percutaneous device. Patients who were eligible for a BCD but did not receive an implant for personal or financial reasons were not included, as there is no evidence to suggest that the proportion of these patients would be different between arms. In addition, this group of patients would not be part of the decision on which implant to use. However, it is acknowledged that this could lead to a potential upward bias in utility outcomes. Lastly, no differentiation was considered for patients with CHL, MHL, or SSD, due to the overall lack of disaggregated evidence for the individual conditions. Different types of hearing loss are likely to be associated with different quality of life and, therefore, different health utilities. This is shown in the Abbreviated Profile of Hearing Aid Benefit and HUI3 scores in the work from van Hoof et al, 2020.²³ At baseline, patients with MHL had lower utility scores (as measured by the HUI3) compared with CHL and SSD (0.531 MHL, 0.777 CHL, and 0.707 SSD). However, changes from baseline following treatment with percutaneous devices did not differ significantly by subgroup.²³

Conclusion

In this economic evaluation, Osia was found to be cost-effective compared with percutaneous devices from the perspective of the Swedish health care system over a lifetime horizon for adult patients with hearing loss, when applying low to moderate cost per QALY WTP thresholds. This analysis provides valuable evidence on the economic impact of different hearing solutions to support decisions on which BCD best meets the needs of patients and health care budgets.

Acknowledgments

The authors thank Valid Insight (part of the Bioscript Group) for developing the model.