gms | German Medical Science

At the end of 2019, the Digital Healthcare Act came into force, which enables doctors and psychologists to prescribe certain apps as digital health applications (German abbreviation DiGAs). Since 2020, around 80 DiGAs have been included in the healthcare directory by the Federal Office for Drugs and Medical Devices (German abbreviation BfArM) of the Federal Ministry of Health, and 13 have so far been finally approved (2023) for the diagnosis and treatment of mainly chronic diseases. A certain degree of proof of efficacy had to be provided for provisional approval.

The number of prescriptions for these apps has risen continuously until 2023, to around 170,000 activation codes for the substitute health insurance funds (by the end of June 2023) and 70,000 for the BKK umbrella organization. Similar figures are reported for the locally organized insurance companies (AOKs). However, in the view of the German Medical Association, the proof of benefit is not yet sufficient and, above all, has not been verified independently from the manufacturer. The health insurance funds also criticize one-sided price fixing by the manufacturers [1].

A general overview of existing DiGAs for ENT medicine was published in 2023 by a working group from Munich and Nuremberg [2]. A total of 6 DiGAs were identified and their evidence – if available – was evaluated in terms of therapeutic success. Two of these DiGAs are intended for the treatment of tinnitus. One app („My Tinnitus App“) contains a mere tinnitus counseling as a basis for further therapeutic interventions.

This app is marketed by a pharmaceutical company (Schwabe) and states „digital counseling“ as the goal of the therapy: The aim is to reduce the burden of tinnitus and to improve quality of life. There are several lessons, each lasting 90 minutes, which essentially provide information about the development of tinnitus, the function of hearing and the structure of the ear. The entire program should be completed over 10 weeks. The app also includes relaxation techniques, videos and mini games. The app costs € 449, but this is covered by the health insurance as DiGA.

The other app (Kalmeda) was approved as a DiGA solely on the basis of manufacturer information without any scientific publications. For the provisional inclusion in the DiGA catalog, a sample data collection of 67 participants was submitted, which demonstrated a reduction in tinnitus stress using the Mini-TF (short form of the tinnitus questionnaire according to Goebel and Hiller [3]).

In the meantime, this DiGA „Kalmeda“ has received final approval as a „mobile tinnitus therapy on prescription“ and has been heavily advertised ever since. The developer of the app, ENT specialist Uso Walter, has since sold his company („MyNoise“) to a pharmaceutical company (Pohl Boskamp) and is now only involved as a consultant. The app costs € 189 for 90 days, should be used for a total of one year, up to 4 activation codes (i.e. € 756) are covered by the statutory health insurance.

According to the distributing pharmaceutical company, the app offers guideline-compliant behavioral therapy for the treatment of chronic tinnitus. The clinical studies have demonstrated the app's convincing effectiveness. The corresponding publication was published in an online journal (PLOS Digit Health) in September 2023 and was only submitted there in November 2022 [4].

The app itself was presented and commercially distributed back in 2018. The study to prove the effectiveness of the app began in August 2020 and the last patient was included in March 2021. According to the authors of the study, it is intended to provide app-supported cognitive behavioral therapy and thus comply with the strong recommendations of the therapy guidelines of the AWMF guideline „Chronic tinnitus“ [5]. However, it is not the current version of the guideline (published in 2021 [6]) that is referenced, but the previous version (published in 2014). The app provides psychoeducation paired with relaxation and sound applications in 5 stages. Over an application period of 3 to 12 months, users of the app should recognize and assess harmful patterns of thought and behaviour, increase their mindfulness and learn to improve their resilience characteristics. In addition, natural sound signals are used as background noise for calming and tinnitus-related information is conveyed.

A total of 187 patients were recruited for the study and divided into an intervention group and a control group, which only started using the app after 3 months (waiting group). The average age of the study group was 48.2 years, 51.9% were male and 48.1% were female. 17.1% (16 out of 94) of the participants in the intervention group and 8.6% (8 out of 93) of the control group discontinued the study after 3 months, mainly for unknown reasons, while 163 patients completed the study over 9 months and completed the corresponding tinnitus questionnaires at least once. A total of 158 patients (76 intervention group, 82 control group) were evaluated, but only at T1, i.e. 3 months after the start of the study. The study protocol did not provide for the collection of audiometric data, such as the tone threshold or tinnitus parameters. In the intervention group, a reduction of more than 6.65 points (minimum clinically significant reference) in the total score determined using the 52-question tinnitus questionnaire [2] was observed in 56 patients (73.7%), while only 22% of patients in the control group showed a comparable reduction in the total tinnitus score. In the intervention group, an increase of tinnitus burden measured by the total tinnitus score was observed in 7.9% of patients.

When evaluating the secondary outcome parameters of depression (PHQ-9 questionnaire), stress (PSQ-20 questionnaire) and self-efficacy-optimism-pessimism (SWOP-K9 questionnaire), there was a significant reduction in the PHQ-9 total score in the intervention group, while no effect was observed in the control group. Similarly, the PSQ-20 score (stress questionnaire) showed a significant reduction in the intervention group, while no effect was recorded in the control group. There were no side effects in the first 3 months of the study. With regard to the tinnitus burden of the cohort studied, it should be noted that the baseline burden in the overall tinnitus score was 39.7 points on average, while the proportion of patients with decompensated tinnitus (grade III or IV, at least 47 points in the overall tinnitus score) was 30.1%. The authors conclude that the app used in the study can be used successfully for treatment in the same way as conventional cognitive behavioral therapy. Commenting on the significance of the study, the authors concede that the study participants only reported moderate tinnitus and that they were already open to psychotherapeutic treatment approaches. Therapeutic support from an audiologist or psychotherapist was not offered. It was astonishing that no placebo effects, i.e. no improvement, were observed in the control group in the questionnaires and that although 7.9% using the app worsened, this was not regarded as a side effect („no adverse effects“; Figure 1 [Fig. 1], [4]).

The health insurance company BIG direkt has presented a very detailed patient survey and evaluation of the use and effectiveness of the Kalmeda app [7]. A total of 409 patients who had been prescribed the app (received an activation code) were contacted. 234 finally responded and a questionnaire completed by these patients was evaluated. Only 15.4% completed the therapy over 12 months, while 60.3% discontinued after just 3 months. Good effectiveness was reported by significantly less than 50% of app users. Only 32.9% of participating app users reported a subjective reduction in tinnitus perception, while 55.1% felt no change and 12% did not provide any information on this. Furthermore, only 35.9% of app users were able to develop a positive attitude towards tinnitus, while 52.1% were unable to do so (Figure 2 [Fig. 2]). The main reasons given for premature termination were difficulties with the handling of the app, problems with data protection (required access to photos), but also too slow progress in accessing the individual modules (Figure 3 [Fig. 3], [8]).

In international tinnitus guidelines as well as in the German S3 guideline “Chronic Tinnitus“ [4], cognitive, tinnitus-specific behavioral psychotherapy (CBT) is proposed with a strong recommendation for treatment. Evidence exists for both direct psychotherapeutic and internet-based interventions [9]. While internet-based tinnitus therapies are usually accompanied and guided by audiologists or psychologists, the Kalmeda smartphone application is a therapy approach that takes place practically without therapeutic support and is based solely on documents and sound applications. What is extremely problematic about this is that deterioration or even danger is not recognized and therefore it is not possible to react, which is possible at any time in CBT, including internet-based CBT. In Germany, such app-supported therapy approaches were introduced on the basis of the DiGA regulation, and the app for tinnitus treatment was one of the first to receive provisional approval. This was despite the fact that no results were available at the time. The study data for the Kalmeda app reported in September 2023 were published with a long delay, although the distribution company had already announced publication of the data long beforehand.

The specific study results were collected two years ago and published by the app manufacturer in advertisements. The publication was not in a renowned journal from the ENT field, but in an online journal (PloS Digital Health), which, by its own admission, places particular emphasis on originality and innovation.

It can be assumed that subject-specific journals with high-quality reviews would have or even did refuse to publish the manuscript, partly because all authors stated that they had significant conflicts of interest, as they received fees from the companies involved. This can certainly be seen as a problematic “bias of funding“ in principle.

The authors of the Kalmeda app study only provide a fragmentary description of how the study participants were recruited. The role of the two ENT practices in conducting the study remains unclear. Recruitment probably took place via the Internet. The data and statistical calculations presented are very confusing and barely comprehensible, while statements regarding possible effects are clearly stated. This makes otherwise incomplete data on the included patients appear complex to calculate. It is particularly problematic that audiological data such as the severity of the respective hearing loss and the quality of the ringing in the ears were not recorded in any way and therefore could not be analyzed. Accompanying or ongoing therapies were not recorded at all; only psychotherapy is said to have been excluded.

So far, the authors of the Kalmeda app study have only been able to present results after a three-month evaluation. In view of the high dropout rate, the question arises as to whether the plan to publish the results of the app usage after 9 or 12 months can be implemented.

It is also astonishing that in the control group, measured by the total tinnitus score over the three-month waiting period, no improvement and significantly more worsening was achieved than in the treatment group, in which a worsening of the tinnitus burden occurred in 8% of cases. This observation is not consistent with the experience from previous studies, in which there was always a certain improvement in the tinnitus burden in the control group, even without treatment.

The thorough survey conducted by the health insurance company BIG-direkt sheds a different light on the success of the Kalmeda app. Both the high drop-out rate (60.3%) and a significantly lower effect size cast doubt on the results mentioned in the publication on “Kalmeda“.

Therefore, a truly well-founded evaluation and, if necessary, a recommendation for such app-mediated tinnitus therapy will only be possible when independent and valid study data are available, which primarily include audiometric data and hearing status.

The authors declare that they have no competing interests.