Ringing in insights: analyzing real-world data from a tinnitus management application and its potential for transforming care

Given the current barriers to treatment accessibility and the increasing digitization of healthcare, mobile applications are becoming key tools for managing highly prevalent conditions such as tinnitus by providing data-driven and personalized care. One digital therapeutic for tinnitus, the SilentCloud application, was designed to incorporate evidence-based treatment modules, including education and counseling, Internet-based cognitive behavioral therapy (iCBT), and sound therapy, to support tinnitus patients in managing their condition. Retrospective observational analyses of routinely collected real-world data showed a significant reduction in tinnitus-related distress, measured using the Tinnitus Handicap Inventory, among users who completed the iCBT module [mean change = −14.7 points, t (62) = 6.67, p < 0.001]. Despite these promising results, high dropout rates of users before starting the therapy modules and long-term adherence remain a challenge. Factors that may affect adherence include motivation, habits, perceived benefits, and the availability of professional support. Digital health applications offer effective tools to overcome barriers to treatment access. However, achieving the best outcomes for patients still requires a collaborative and interdisciplinary approach.

With the rise of digital health technologies, mobile applications are emerging as powerful tools for managing chronic conditions, providing new opportunities for data-driven insights and personalized care. This is especially relevant for highly prevalent conditions such as tinnitus, which represents a major public health issue.

Tinnitus is experienced by approximately 14.4% of the global adult population, affecting more than 740 million people (Jarach et al., 2022). Considering that approximately 1 in 6 of these individuals reports a severe form of tinnitus (Jarach et al., 2022), which is associated with significant medical implications for the patient (Langguth, 2011), substantial healthcare and societal costs can be expected (Trochidis et al., 2021).

To date, tinnitus remains an enigma, primarily due to its extreme heterogeneity in terms of etiology, perceptual characteristics, severity, comorbidities, and patients' response to therapeutic interventions (Cederroth et al., 2019). While several, mostly complementary, pathophysiological models exist, a unified model that adequately explains this heterogeneity remains elusive (Langguth et al., 2024). This poses a considerable challenge for tinnitus management, as a one-size-fits-all solution targeting a specific mechanism does not exist. Consequently, individualized management approaches are paramount.

Currently, numerous treatment options are available, with substantial variation in the quality of their supporting evidence (Kleinjung et al., 2024). Cognitive behavioral therapy (CBT) is considered the gold standard, supported by a large body of research corroborating its beneficial effects (Fuller et al., 2020) and recommended in multiple clinical guidelines (Langguth et al., 2023). CBT consists of different components that provide patients with tools to, for example, restructure intrusive and maladaptive thoughts that catastrophize tinnitus and replace them with functional ones (Fuller and Hoare, 2024). However, since specialized centers providing adequate psychological treatment are often lacking, digital forms of CBT [i.e., Internet-based CBT (iCBT)] have emerged (e.g., Beukes et al., 2019; Weise et al., 2016) and are sometimes recommended to be considered as a first-line intervention (e.g., by the National Institute for Health and Care Excellence (NICE) guidelines; Lewis et al., 2020).

Over the past decade, mobile health (mHealth) applications have seen substantial growth. In 2024 alone, more than 50,000 applications were launched, summing up to more than 337,000 mHealth applications in the top app stores worldwide (IQVIA, 2024). Such digital applications hold the potential to revolutionize healthcare delivery by improving accessibility, addressing care gaps, supplementing therapies delivered by primary caregivers, reducing costs, and empowering patients to manage their own health (Messner et al., 2023; Portz et al., 2024). Recognizing the transformative potential, healthcare systems in many countries are starting to establish specific reimbursement pathways for digital therapies. Despite their tremendous prospects, there is a significant gap in providing a solid scientific foundation for these tools (Mathews et al., 2019). The majority of applications available today have not been rigorously tested for their efficacy or effectiveness, raising concerns about their reliability and potential benefits for users (Messner et al., 2023). In the context of digital tinnitus management, the situation is similar. While several solutions are already available, only a few have been scientifically evaluated (Habeeb et al., 2023; Mehdi et al., 2024; Sattel et al., 2024). However, uncertainty regarding the effectiveness of mHealth applications is a crucial factor influencing both a patient's decision for adoption (Schueller et al., 2018) and a healthcare professional's trust and willingness to recommend them (Wu et al., 2022). Demonstrating clinical benefits with real-world data is therefore vital to mitigate concerns among both patients and healthcare professionals and to support informed decision-making. It is important to distinguish between mHealth applications and digital therapeutics. While mHealth tools broadly support health monitoring and behavior change, digital therapeutics are evidence-based, clinically validated programs subject to regulatory oversight, such as approval by the Food and Drug Administration (FDA) or CE marking.

This study describes the application-based tinnitus management tool and digital therapeutic named SilentCloud and presents the first real-world user insights based on routinely collected application data from five European countries (Belgium, Ireland, Italy, the Netherlands, and Switzerland). The application provides evidence-based management modules, including education and counseling, iCBT, and different types of sound therapy, delivered in a structured format to facilitate the user journey. We report a quantitative, retrospective observational evaluation of the iCBT module. The effectiveness analysis was restricted to iCBT because, in the application version underlying this dataset, iCBT was the only module for which an outcome questionnaire was administered both at baseline and again upon completion, specifically the Tinnitus Handicap Inventory (THI; Newman et al., 1996), enabling pre–post comparisons. The primary outcome was the change in THI score from pre- to post-iCBT. We hypothesized that tinnitus-related distress would decrease among users completing the iCBT module and, therefore, completing the anticipated user journey.

Application and dataset description

The digital therapeutic described here is a mobile tinnitus management application that has received Medical Device Regulation (MDR; EU 2017/745) CE marking and FDA 510k clearance and has been launched commercially in several markets. The application is divided into four main modules, which can be used either consecutively or independently. The following sections provide a more in-depth overview of the application's key components.

The current dataset consists of user data from five European countries (Belgium, Ireland, Italy, the Netherlands, and Switzerland). The data include 17,271 user entries collected over a span of approximately 2 years, with registrations occurring from November 2022 to January 2025. Users are eligible to utilize the application upon reaching the age of 18 years. Additionally, medical clearance from an ear, nose, and throat specialist is required to access the advanced tinnitus management modules, ensuring that contraindications for the advanced therapies are excluded. The application integrates a range of clinically validated questionnaires for tinnitus assessment and three main modules of tinnitus management: education and counseling, iCBT, and sound therapies (Figure 1).

Data collection procedure

We conducted a quantitative, retrospective observational analysis of routinely collected real-world data from the SilentCloud application. Data included user-reported questionnaire responses and in-app usage information. For the present analysis, we analyzed users who completed the iCBT module. In the application version underlying this dataset, iCBT was the only tinnitus management module in which an outcome questionnaire was administered twice, specifically, the THI, once after registration and again upon completion of the iCBT module, allowing within-subject pre–post comparisons. Users may have used the other modules in parallel; however, detailed application or module usage data were not analyzed for this report.

Tinnitus assessment—questionnaires

As assessing the burden of tinnitus is of utmost importance to identify the need for therapy and to monitor treatment success, several questionnaires are administered to the application users (see Table 1). Case history and tinnitus characteristics were assessed using the Tinnitus Sample Case History Questionnaire (TSCHQ; Langguth et al., 2007). The TSCHQ comprises 35 items that have been agreed upon by a consortium of tinnitus experts and is considered a standard in the field. The application uses a shortened version of 13 items to minimize redundancy in the data collected. For instance, personal characteristics are gathered during registration, while some tinnitus characteristics are collected throughout various stages of using the application (e.g., tinnitus pitch in Your Sound Therapy). For the current data analysis, the TSCHQ data were used exploratorily to compare iCBT completers and non-completers.

In the current version of the application, only the Tinnitus Handicap Inventory (THI; Newman et al., 1996) is utilized both before and after the completion of the iCBT module. Consequently, the analyses focused on the THI as the primary outcome measure when evaluating the effectiveness of the iCBT module. The THI is a 25-item questionnaire consisting of different subscales (i.e., functional, emotional, and catastrophic). Questions are answered on a 3-point scale (0 = no; 2 = sometimes; 4 = yes), resulting in a total score ranging from 0 to 100. Since the distinctiveness of the three subscales has repeatedly been questioned (Baguley and Andersson, 2003; Wakabayashi et al., 2020), the total score was used in all further analyses. To evaluate the results from a clinical perspective, a minimal clinically important difference (MCID) of seven was considered (for a discussion, see Langguth and De Ridder, 2023; Zeman et al., 2011).

Tinnitus management modules

SilentCloud is organized as a modular intervention with a structured user journey. After registration and the initial tinnitus assessment, users first complete the education and counseling module. In the application version used for this data analysis, completion of this module was required before advanced management options could be started. Access to the advanced modules (i.e., sound therapy and iCBT) is unlocked by a clinician to exclude contraindications. Once unlocked, users may engage with sound therapy and iCBT according to individual needs and at their own pace, including their concurrent use.

Education and counseling

Your Tinnitus Tutor is a psycho-education module based on leading clinical guidelines. Education and counseling are well-established approaches that are recommended by all clinical guidelines as a first and fundamental step of tinnitus management (Cima et al., 2019; Langguth et al., 2023) and are known to help reduce tinnitus-related burden (Schoisswohl et al., 2025). The module consists of 17 individual lessons, each taking 15–20 min to complete, guiding the user in understanding their tinnitus and how to manage it. It addresses and dispels myths related to tinnitus, encourages users to explore their own tinnitus (e.g., pitch and lateralization), defines its severity, and provides initial concrete action steps, such as guided relaxation. This module is self-paced, and the lessons can be completed consecutively. Completion of this module is required before the advanced tinnitus management modules (i.e., Sound Therapy and iCBT) can be unlocked.

Advanced therapy: sound therapy

Your Sound Therapy comprises three types of sound therapies administered according to the patient's tinnitus characteristics. Foundational Therapy utilizes pink noise, allowing users to personalize the volume to their own comfort level. Custom Sound Therapy consists of five different noise bands that are adjusted by the user based on their individual minimum masking levels. Both Foundational Therapy and Custom Sound Therapy are intended to provide distraction from and/or suppression of the tinnitus via partial masking and therefore to provide relief from the tinnitus (Henry et al., 2006). Furthermore, both therapies are suitable for atonal tinnitus. Advanced Tonal Therapy is the third option, which entails the acoustic coordinated reset approach for tonal tinnitus (Tass et al., 2012). This neuromodulation approach aims to disrupt or desynchronize enhanced neural synchrony that potentially gives rise to tinnitus and thereby reduce the tinnitus-related burden (Theodoroff et al., 2022). Further details about the method can be found in other publications (e.g., Hauptmann et al., 2015; Tass et al., 2012). The usage time for sound therapies is recommended within the application but is limited to 4 h per day.

Advanced therapy: iCBT

Your Counseling Therapy is a core component of the application offering iCBT. This innovative approach is grounded in established and rigorously evaluated CBT protocols, which have been thoughtfully adapted for digital platforms to enhance accessibility and user engagement (Kröner-Herwig et al., 2003). CBT is a well-researched therapeutic method that empowers individuals to develop healthier coping strategies and improve their emotional wellbeing. By integrating these principles into the application, users can benefit from a structured framework that promotes self-awareness and resilience. The application is entirely self-guided, allowing users to progress at their own pace while utilizing a variety of self-help strategies. The application does, however, limit daily usage to one iCBT lesson to prevent users from rushing through the program. Over the course of 40 comprehensive lessons, each taking approximately 20 min to complete, users are introduced to effective techniques aimed at breaking free from the negative thought spiral often triggered by tinnitus. The exercises within the application include clear action steps, such as relaxation techniques, guided imagery exercises, and cognitive restructuring practices. Users learn to challenge and modify unhelpful cognitions related to their tinnitus, fostering a more positive mindset and enhancing their overall quality of life. By combining the principles of CBT with the convenience of digital technology, Your Counseling Therapy not only provides users with valuable tools for managing tinnitus but also encourages a proactive approach to mental health and wellbeing.

Statistical analyses

A one-sided paired samples t -test was used to evaluate the change in tinnitus severity (i.e., THI score) pre- and post-iCBT. Cohen's d av was used as a standardized effect size, as recommended for within-subject designs when paired samples exhibit moderate to strong correlations (see Lakens, 2013). Sample characteristics between iCBT completers and non-completers were compared using Welch's t -test, χ 2 test, Fisher's exact test, and two-proportion z -tests. Welch's t -test was used as it is generally recommended when comparing groups of unequal sizes and/or when the assumption of variance homogeneity is not met (Delacre et al., 2017). The family-wise error rate for post-hoc pairwise comparisons was controlled using the Holm–Bonferroni method (Holm, 1979). All statistical tests were conducted in R (version 4.3.1; R Core Team, 2023).

Effectiveness of iCBT

By January 2025, a total of 17,271 users had registered with the application (M = 51.9 years, SD = 13.8). Users reported an average tinnitus duration of 8.5 years (SD = 10.7), with 79.3% reporting tonal tinnitus and 20.7% reporting atonal tinnitus (Figure 2). Among all users, a total of 63 users successfully completed all 40 sessions of the iCBT module. Statistical comparison of THI scores pre- and post-iCBT indicated a significant reduction in tinnitus-related distress [ t (62) = 6.67, p < 0.001; d av = 0.69, 95% CI (0.34, 1.05); Figure 3A]. Upon completing the iCBT, users reported a lower THI score (M = 33.7 score points, SD = 20.6) compared to baseline (M = 48.3, SD = 21.9). On average, THI scores were reduced by 14.7 points [95% CI (10.3, 19.1)]. Overall, 77.8% of participants showed a reduction in THI scores, with 63.5% experiencing a clinically meaningful improvement (≥7 points) after completing the iCBT (Figure 3B).

Comparing completers and non-completers

In a further step, sample characteristics of individuals who completed all iCBT sessions were compared to those who did not. On average, completers were older (M = 55.8, SD = 11.0) than non-completers [ M = 51.9, SD = 13.8; t (62.72) = −2.85, p = 0.006; Figure 4A]. Additionally, completers had a higher THI score at baseline (M = 48.3, SD = 21.9) than non-completers [ M = 40.5, SD = 25.0; t (62.78) = −2.84, p = 0.006; Figure 4B]. Moreover, a two-sided Fisher's exact test revealed a significant difference in the number of treatments participants had received prior to enrolling (p < 0.001; Figure 4C). Post-hoc pairwise comparisons showed that a notably higher proportion of completers reported having undergone one (p adj < 0.001) or several treatments (p adj < 0.001) before using this application, while a greater proportion of non-completers indicated that they had not received any prior treatments (p adj < 0.001). There was no significant difference in the proportion of participants who reported having undergone many other treatments before using this application (p adj = 1). When participants were asked whether sounds caused them pain or physical discomfort, the responses showed a significant difference between completers and non-completers (χ (2) 2 = 8.95, p = 0.011; Figure 4D). A higher proportion of completers responded “yes” compared to non-completers (z = 2.92, p adj = 0.011).

The analysis of real-world data from people with tinnitus who have been using SilentCloud as a digital therapeutic demonstrates that the application-based cognitive behavioral therapy can effectively reduce perceived tinnitus burden. This conclusion is based on pre–post changes in tinnitus-related distress quantified with the THI. Among users who completed the full iCBT module, THI scores decreased on average by 14.7 points from baseline (i.e., pre-iCBT) to post-treatment [ t (62) = 6.67, p < 0.001]. Nearly two-thirds of completers exceeded the commonly used threshold for clinically meaningful change on the THI (≥7 points), indicating that the observed improvements are not only statistically significant but also likely clinically relevant. However, since SilentCloud is a modular intervention and users engage with education and counseling and/or sound therapy alongside iCBT, the observed pre–post reduction in tinnitus-related distress cannot be attributed to iCBT alone in this retrospective observational dataset. Nonetheless, these first findings support the hypothesis that digital therapeutics for tinnitus can serve as valuable tools for patients in managing their symptoms (Habeeb et al., 2023; Kleinjung and Langguth, 2020; Sattel et al., 2024). Additionally, the application's capability to track treatment success over time allows healthcare professionals to monitor the wellbeing and treatment effectiveness of patients. The therapy modules within the application, previously validated in separate studies using a traditional in-clinic setup, have now been retrospectively analyzed in this self-paced application format, showing effectiveness even with minimal or no expert support.

Publicly available digital therapeutics can provide access to much-needed therapies for a broad range of individuals, especially for those who might face barriers to accessing traditional healthcare services (Snoswell et al., 2023). When adoption is high, large-scale datasets like the one in this study (N = 17,271 users) can yield substantial information that helps identify patterns, trends, and correlations that may not be evident in smaller clinical datasets. This can lead to new insights and discoveries in medical research. The current analysis shows that patient characteristics differ between users who completed the iCBT module (“completers”) and those who did not complete the therapy (“non-completers”). Notably, and perhaps unsurprisingly, users who completed and benefited from iCBT had a higher tinnitus burden before starting the therapy, as indicated by higher THI scores. This suggests that people with higher tinnitus distress are more willing to undergo therapies such as iCBT, which require a heightened level of effort and commitment from the patient. This greater effort likely contributes to therapeutic success (Probst et al., 2019). Another noteworthy observation is that most “completers” had previously tried several treatments. Due to a greater burden (i.e., higher THI), individuals may have made multiple attempts to explore new methods for finding relief. This prior experience may have fostered a greater commitment to adhering to the treatment, as they were likely motivated to find a solution for their tinnitus burden and more aware of the necessity of completing a treatment to fully experience its benefits.

In the current analysis of the user sample, there is a significant disparity between the number of individuals who completed the therapy and those who did not engage with the application for a longer period. One of the biggest challenges of digital health applications is maintaining user adherence over time. Research shows that approximately 80% of the users engage only at a minimum level (Jakob et al., 2022; Meyerowitz-Katz et al., 2020). Several potential reasons have been discussed that hinder users from finishing treatment. Adherence to digital health interventions is influenced by various behavioral factors, such as motivation, habits, and perceived benefits (Baumeister et al., 2014; Jakob et al., 2022). Furthermore, empirical studies have indicated that digital health applications lacking professional support often face challenges in maintaining user engagement and adherence; integrating expert guidance can enhance motivation, provide accountability, and result in improved outcomes (Linardon et al., 2019). This may be particularly relevant for tinnitus management and CBT-based interventions, where progress can be gradual and users may benefit from expectation management, encouragement, and assistance in translating exercises into daily routines. The relevance of accessible professional support, e.g., in the form of brief, regular follow-ups, could help users overcome early barriers, acknowledge improvement, and reduce dropout. Managing health conditions, especially the ones that target mental and/or chronic health challenges, for example, as associated with tinnitus, can be complex and overwhelming. Users may find it difficult to navigate the application or understand how to apply the information to their individual situation, which can result in frustration and eventual discontinuation of the application. Similarly, a rather complex application journey due to the necessity of a medical prescription for e.g., reimbursement or medical clearance, especially for digital therapeutics, might prevent patients from completing the full journey. Consequently, it is important to prioritize a user-centered experience involving multiple health experts and to tailor the content to individual needs, using gamified strategies to enhance motivation and engagement and to foster trust in the solution (Jakob et al., 2022). For the application presented here, potential issues affecting the user experience have been analyzed and will be addressed in future versions.

SilentCloud brings the advantage of a modular approach, allowing for comparison of the effects of single vs. multiple or parallel treatments (i.e., iCBT plus sound therapy). As the number of application users grows, the insights gained from treatment outcomes can be utilized to identify patterns in patient histories and associated results, paving the way for a decision-support tool that recommends personalized treatment pathways for optimal outcomes. Overall, this digital therapeutic takes a holistic approach to tinnitus management suitable for self-paced, remote use. In the long term, integration into existing healthcare systems could help address gaps in access to tinnitus care while demonstrating cost-effectiveness compared to no treatment (Patel et al., 2022).

Limitations

This retrospective data analysis has several limitations that should be considered when interpreting the findings. First, data were not collected in a controlled study environment but were obtained through naturalistic, real-world use of the application. Consequently, the setup of data collection did not allow randomization or the inclusion of a control group. While we compared users who completed the iCBT module with those who did not complete the iCBT module, this comparison does not substitute for random allocation and remains susceptible to systematic differences between the groups. At the same time, real-world data offer the advantage of providing insights into user experience and effectiveness under routine conditions, rather than efficacy under idealized trial conditions. Accordingly, only limited information on users was available beyond the routinely collected variables included in this report. In addition, the substantial attrition observed in the real-world setting implies that effectiveness estimates primarily reflect outcomes among motivated users who completed the iCBT module. Therefore, the results may not generalize to users with partial engagement or early discontinuation. Second, access to outcome measures was limited, resulting in the use of a single patient-reported measure that was administered pre–post within the iCBT pathway only in the application version underlying this dataset. This restricts conclusions about broader, clinically relevant domains such as sleep, mood, or overall quality of life. It also limits the ability to disentangle which element of the modular program drove the observed changes. Third, the retrospective, real-world observational design is prone to selection bias and confounding. Therefore, observed pre–post improvements may partially reflect non-specific effects (e.g., spontaneous improvement, expectancy) rather than causal effects of iCBT. Finally, professional support was not standardized or systematically tracked. Although many users were connected to an audiologist and could request contact via an in-application messaging feature, the frequency and content of follow-up varied and cannot be quantified here, complicating the interpretation of adherence, dropout, and potential guidance effects. A multicenter randomized controlled trial is currently underway and is expected to address key limitations of the present real-world analysis, including the use of a broader outcome battery and the collection of detailed application-usage and module-exposure data to enable component-level analyses. Furthermore, in future iterations of the application, additional validated questionnaires will be integrated longitudinally during application use to enable continuous monitoring of treatment progress, capture changes across multiple outcome domains where benefits may occur, and assess treatment trajectories even among users with partial engagement or early discontinuation.

This analysis demonstrates that the use of a publicly available digital therapeutic, the SilentCloud application, effectively reduced the burden of tinnitus for those users who are motivated to complete the anticipated user journey. Digital health applications provide the benefit of broad, low-threshold access to treatment for patients. Additionally, large-scale datasets are a powerful tool for advancing medical research, improving patient care, and enhancing the efficiency of healthcare systems. However, high dropout rates remain a barrier to a broad deployment of digital therapeutics, highlighting the need for a stronger interdisciplinary approach that places the patient at the center.