In recent years, information technology solutions have been developed that allow professionals to treat patients via teletherapy. With regard to rapidly increasing health care expenses owing to the aging of society and even faster medical and technical advances, cost-effective rehabilitation is both a priority and a challenge for users and therapists [1]. This phenomenon has been stressed by the current COVID-19 pandemic crisis, which is transforming our society and has implications for health care [2-4]. Telemedicine has been shown to be an option in previous outbreaks, such as severe acute respiratory syndrome–associated coronavirus or Middle East respiratory syndrome coronavirus [4,5]. An additional benefit is that these digital solutions have the potential to reduce health care costs associated with supervision and high-frequency training [6-8].

So far a teletherapeutic approach is often used in psychotherapeutic sessions with a high level of satisfaction and compliance [9-11]. A positive outcome after home-based therapy has also been reported in patients with chronic pain [12] and those who received knee or hip replacements [13].

Auditory training is an important part of rehabilitation after cochlear implantation. Several consensus papers have reported that it is necessary for the brain to adapt to the new auditory stimulus transmitted by the implant [14-16]. However, rehabilitation after cochlear implantation differs among countries. In some countries, postoperative rehabilitation programs are not routinely offered because of a lack of reimbursement by health insurance companies and a shortage of specialized therapists [17], whereas in others, cochlear implant (CI) recipients follow an intensive rehabilitation regime that is regularly covered by the general health insurance for at least 2 years after surgery [16]. Auditory training usually takes place in a face-to-face setting in specialized centers; computer-based applications are used only as an additive to standard (face-to-face) therapy. In a previous study, we developed Train2hear, which is a highly individualized digital training platform that combines different components of adaptivity, feedback, and motivation to allow CI users to receive computer-based auditory training (CBAT) that is tailored to their specific therapeutic needs. The first evaluation, within the setting of an applicant’s workshop, clearly demonstrated that CI users enjoyed using Train2hear [18]. A challenge faced by teletherapy is to achieve the same efficiency as standard face-to-face therapy.

Few studies have assessed the effectiveness of digital auditory rehabilitation in adult CI users, and these studies also have only analyzed some aspects in a small number of participants (ie, less than 20) [19-21]. In Schumann et al [22] 15 CI users received 3 weeks of training on phoneme discrimination. A control group and follow-up assessments were not included. Fu et al [23] used a similar approach and studied phoneme discrimination in 10 participants over 4 weeks. In addition to improved performance in trained skills, a transfer effect on sentence comprehension was observed. This observation contrasts with that of Stacey et al [24] who found a significant improvement in consonant discrimination but not in sentence comprehension. Self-perceived improvement was reported in only 2 of the 11 participants. The only publication so far that has compared standard face-to-face therapy with a computer-based approach was by Bernstein et al [6] who analyzed speech tracking ability in 9 patients after a 4-week period. In their study, the tracking rate was improved, but no difference was observed between the two methods.

Furthermore, only a few studies have investigated the ability to listen in noise after CBAT [19,21,25]. However, there were conflicting results, with small number of participants. In Ingvalson et al [21], 5 CI users with postlingual deafness and at least one year of hearing experience showed improved speech perception only in quiet conditions. In contrast, Oba et al [19] reported a significant improvement in babble and steady noise after a 4-week digit training in 10 participants with CI. Even Green et al [25] observed in 9 participants with postlingual deafness that the thresholds to understand 50% of the sentences presented in noise significantly improved after 4 weeks of training in noise, but transfer effects on phoneme discrimination and memorization could not be demonstrated. In short, a systematic evaluation of a complete teletherapeutic rehabilitation program is lacking.

Therefore, the aims of this study are (1) to assess the usability and feasibility of the CBAT platform Train2hear in adult CI users; (2) study the objective and subjective auditory development as well as the economic benefit after a 3-week tablet-based rehabilitation as compared with a 3-week conventional face-to-face setting; and (3) analyze the impact of sociodemographic variables on outcomes.

In total, 20 adult CI users were included in this study (Table 1). To be included in the study, potential participants had to be adults (≥18 years); CI users with postlingual bilateral hearing loss and a CI experience of at least 3 months; have no significant motor, visual, or cognitive impairment; be willing and able to complete the tasks inherent in the study; and to give their informed consent. All subjects attended weekly face-to-face therapy at the implant center before the study (range: 7-48 sessions; SD 10.3).

During the intervention period, costs were measured for both CBAT and face-to-face therapy according to the international guidelines for conduction cost analysis [26,27]. Cost-related data covering costs relevant to the health center and costs for the patients were assessed on a standardized cost sheet for each patient. Subsequently, the costs of the two treatment modalities were compared.

All participants performed at least seven therapeutic face-to-face sessions in the rehabilitation center before the start of the study (mean 26.3, SD 10.3). Internet access and an audio loop were required to use the telerehabilitation system at home. The tablets were provided by the clinic.

Participants completed 3 weeks of conventional face-to-face rehabilitation followed by 3 weeks of self-training with the home-based digital auditory training program, Train2hear. All participants were assessed at baseline, after the 3-week face-to-face rehabilitation, and after the 3-week digital training program, as shown in Figure 1.

First, a descriptive analysis of the data using the mean value and SD was performed.

Thereafter, rank-analysis of variance (ANOVA; Friedman test) was performed to prove that there were significant changes between the three measurements (T1, T2, and T3). Afterward, the Wilcoxon signed-rank test for paired samples was applied to evaluate participants’ results after the two types of therapy. For rank binding, a sign test was applied. The exact U test was used to determine the correlation between the results and sex.

Further correlations between outcome and continuous sociodemographic factors, such as age and years of education, were calculated. If there was no normal distribution, then the Kendall rank correlation in case of rank binding was applied. The significance level was set at P=.05. Statistical analyses were performed using Medas (Grund).

Ethics committee approval (19-6618-BR) was received from the Ethics Committee of the Medical Faculty of Ruhr University Bochum.

All participants completed face-to-face and CBAT training sessions. For each participant, therapists spent 360 minutes (3×120 minutes) on face-to-face procedures and 60 minutes (30 minutes of videoconferencing and 30 minutes of introducing the digital program) for CBAT. To participate in the 3-week face-to-face training, participants needed approximately 6 hours of travel time (range: 1.5-12 hours).

Previous experience of using digital media differed among the participants: 13 had regularly used a computer (daily or several times a week), whereas 4 had never worked with a computer. In total, 11 participants regularly used a tablet several times a week, whereas three did not. All other participants had previous experience of using digital devices (tablets, smartphones, and computers). Before the study, 16 participants had no experience with videoconferencing, whereas 4 had used videoconferencing to communicate with family members or friends.

Digital experience did not correlate with speech understanding assessed by the test battery at T2 or T3. In contrast, affinity to digital media had a significant impact on the assessment of the usability of the program. Participants who frequently used a computer stated significantly more often that the videoconference was easy to use (P=.03; Bochum Usability Questionnaire Q16). A significant positive correlation among questions 3, 5, 6, and 7 of the SUS and digital experience could be detected. Participants with more experience judged the program to be easier to use (P=.03; SUS Q3) and more often stated that the different functions were well integrated (P=.02; SUS Q5). Experienced users also stated, significantly more often, that the handling of the program could be learned quickly (P=.007; SUS Q7). Nevertheless, regular tablet users still found the program cumbersome to use (P=.02; SUS Q8).

The results of the test battery at baseline (T1), after face-to-face therapy (T2), and after CBAT (T3) are shown in Multimedia Appendix 4. Tests that did not significantly differ in rank-ANOVA (Friedman test) between the three test times were not further investigated (Multimedia Appendices 4 and 5).

To attend the face-to-face session, patients had to travel 237 km (SD 80.7), which entailed spending on an average of 234 minutes (SD 58.6) on the road. Therapists devoted 450 minutes for a standard face-to-face therapy and 90 minutes per patient for CBAT (including the time of preparation and documentation). Therefore, costs could be reduced from €262.50 (US $320.25) to €52.50 (US $64.05) for the study period (Multimedia Appendix 9).

If standard face-to-face therapy, which regularly included 20 sessions of speech therapy (each of which lasted 120 minutes) was completely replaced by CBAT, then the costs would decrease from €1750.00 (US $2134.00) to €350.00 (US $427.00) based on the data obtained in this pilot study.

This study is, to the best of our knowledge, one of the first to demonstrate that a digital auditory rehabilitation program might reduce adult CI users’ dependence on human resources while ensuring that they receive a clinical outcome similar to that of standard therapy, that is, conventional face-to-face rehabilitation at a specialized rehabilitation center.

A comparison of the two auditory training methods (face-to-face and CBAT) revealed a greater benefit in sentence comprehension in background noise after CBAT. This may be explained by the application of the training schedule. Teletherapeutic tasks were performed five to seven times a week, whereas outpatient therapy was performed only once a week. This is in line with Vu et al [36] who found significant differences in log-in frequency and learning activities between successful and unsuccessful learners in web-based training for teachers. The most remarkable improvements were detected in phoneme discrimination and speech tracking, which are closely related to interactive communication [6,37].

Overall, the Train2hear program was rated as highly usable by the participants. The fact that older participants rated usability worse than younger participants may be related to the lower level of technical experience among older people. This result was also mentioned by Ferguson and Henshaw [38], who stated that access to hardware and lack of skills in using hardware hinder access to computer-based training. Regardless of this age-related difference, all participants agreed that Train2hear is easy to use. Nonetheless, external support may be helpful for older users. This could be done either by the user’s partner or family or friends or by therapists via videoconferencing.

In contrast, older participants had significantly higher scores on motivation and ease of adherence to the training schedule than did younger participants. In general, age did not significantly influence performance; the Freiburg monosyllabic test was the only speech understanding test in which older participants scored worse than their younger counterparts as compared with face-to-face therapy and CBAT (P=.004). Prolonged training intervals might have a positive effect because of a slower learning curve in older adults [6].

Nonusage has been known to be an important barrier in the field of web-based training [39], especially in interventions using automatic functions with minimal human involvement. In this study, 100% (20/20) of the participants completed the 3-week digital training program. As compared with other studies, this adherence rate can be interpreted as extraordinarily high [39].

A possible explanation for the high adherence rate could be that the Train2hear software is highly individualized, which includes a basic assessment of the user’s demands and needs and automatically adapts the training schedule to their performance.

Furthermore, the Train2hear platform contains various motivational elements that might lead to better user adherence, for example, a close feedback system and reminders [38-40]. However, it remains to be seen if such levels of adherence would continue at a long-term follow-up. In a study on patients with stroke, Jurkiewicz et al [41] found that adherence in the initial period was significantly higher than that in the long-term follow-up. Previous works have shown that incorporating an avatar can increase motivation and engagement with a training application and the time spent in training [24,42,43]. With this observation in mind, we added a train conductor as an avatar to the new training platform.

Educational level had a significant impact on the handling of the software. This result is in line with Kriwy and Glöckner, who reported that the higher an individual’s level of education, the better they could take part in computerized health programs [44].

Furthermore, significant correlations were observed between the total duration of hearing loss and improvements after T2. Generally, the shorter the duration of hearing loss, the greater the improvement in speech understanding. This result was also assumed by Ihler et al [45] in their study on home-based auditory training of speech recognition on the telephone in 20 CI users with postlingual hearing loss. Whether auditory training over a longer period can lead to greater improvements in speech comprehension, even in people with a long duration of hearing loss, has yet to be proven.

Participants’ self-evaluated hearing abilities remained nearly unchanged after both face-to-face training and CBAT. Previous studies have reported this result. No, or only minor, self-reported improvements of listening abilities after auditory training periods have also been reported by Stacey et al [24] (after 5 days a week for 3 weeks) and Bernstein et al [6] (once a week for 8 weeks). The question is, if despite objectively shown improvements in speech understanding, a training period of 3 weeks is too short to have an impact on self-perceived hearing status.

There is currently an acute need to study the effectiveness of therapeutic interventions in speech language pathology and audiology. Studies designed and conducted in accordance with evidence-based criteria provide a rational basis for therapeutic approaches that are missing in large parts of auditory therapy [46]. CBAT might be an appropriate tool for future multicenter studies because the protocol is well defined (although highly individualized) and therefore comparable. In addition, CBAT enables a large amount of data to be obtained during the entire training procedure. This process can help speech and language pathologists to more precisely investigate the progress of CI users and to evaluate and refine the therapeutic approach.

Regarding the time- and resource-saving potential of CBAT, each therapist saved more than 5 hours per participant during the 3-week training period, including the time they would have needed to prepare the lessons. The participants saved a mean of almost 4 hours of traveling. Regarding the intense rehabilitation program that is regularly offered to CI recipients in Germany for 2 years and reimbursed by the general health insurance, home-based training might save an enormous amount of economic and human resources even if it might be suitable only for selected CI users and limited to only some parts of the rehabilitation process.

Furthermore, at the time of writing, much of the world is under lockdown or some form of restriction due to the COVID-19 pandemic. In such situations, people are not able to access office-based therapy; therefore, CBAT might be an appropriate and crucial tool for successful hearing rehabilitation, especially for older CI users and those with weakened immune systems.

The few studies that exist on CBAT have a limited scope. They evaluated only a few aspects of auditory training [6,22,45] and generally had short or no follow-up assessments [19,23,25]. Previous studies have described improvements in speech comprehension and communication skills after several weeks of CBAT [20,47,48].

Overall, studies usually included only a small sample size [47]. Only Bernstein et al [6] compared the standard face-to-face regimen with digital auditory training. They conducted their study on speech tracking performance in CI users. Similar to our results, they found that CI users had an improved tracking rate (P<.001) and sentence recognition (P<.001) using both therapeutic approaches.

Although this study is one of the largest on CBAT in terms of the number of participants, 20 participants were still a limited study group. In our study, we chose only a period of 3 weeks because of the regulations of the research project. It must be kept in mind that this period is short compared with the long rehabilitation period of 2 years, which is regularly performed in Germany after cochlear implantation. Furthermore, future studies would benefit from increasing the duration of the training period and analyzing the long-term effects to better evaluate how effective CBAT is and how well users adhere to the training program.

Due to the study design, it cannot be completely ruled out that the positive outcome after CBAT is partially due to the long-term effects of the conventional face-to-face training sessions. However, all participants had experience with face-to-face therapy before the study. This is a bias that all therapeutic studies are faced with. A complete stop of the training over a longer period would be necessary to rule out long-term effects, and this is not ethically justifiable.

Even the inclusion of a control group could not have solved this problem because CI recipients widely differ in terms of age, duration of hearing loss, socioeconomic status, etc. Therefore, we cannot completely rule out the effects of age, sex, duration of hearing loss, technical experience, and hearing experience on treatment outcomes. However, these correlations did not show a significant association. Large multicenter studies should be conducted in the near future to confirm the presented data.

Due to global demographic changes and the pressure under the current COVID-19 pandemic, there is an enormous and increased need for computerized therapeutic interventions in speech language pathology and audiology. Computer-based auditory therapy is an evidence-based and standardized yet highly individualized approach that has the potential to save human and economic resources. Outcomes seem to be quite similar to face-to-face therapy although due to the small number of participants, the results have to be confirmed. However, the promising results of this pilot study justify further investigation and evaluation of the Train2hear program in a large multicenter study over a longer period.