June 2020 The Hearing Journal
The COVID-19 pandemic has, by necessity, resulted in all medical practitioners rushing to incorporate telehealth into their practice repertoire. Audiologists, of course, are no different. But we must remember that telehealth is not new, nor is it a separate medical specialty; it is simply an adaptation of how health care is delivered. Health outcomes from care provided via telehealth should meet or exceed those achieved with face-to-face care. Still, telehealth has the potential to yield better results on other metrics such as patient and provider satisfaction, patient access, clinical efficiency, travel and clinical costs, and wait times. So, to what extent do data support positive outcomes for teleaudiology?
This two-part literature review discusses teleaudiology procedures that can be carried out either by the patient alone (e.g., online testing, apps, etc.) or via synchronous care with a clinician in one location and the patient in another. Procedures that require an intermediary to run a test or take a measurement (e.g., remote otoscopy or use of certain threshold measurement apps) and those that require special equipment to be sent to a patient were excluded.
Part 1 focuses on evidence supporting remote assessment processes and procedures, while part two will examine the evidence supporting auditory rehabilitation via teleaudiology. However, it is also essential to first consider the literature pertaining to practitioner and patient attitudes toward teleaudiology, as these will presumably dictate the extent to which telaudiology procedures will be incorporated into audiology practices.
ATTITUDES TOWARDS TELEAUDIOLOGY
The research literature on the attitudes of hearing care providers towards teleaudiology is fairly sparse, but it indicates that, to some extent, attitudes have become more positive over time. Specifically, early work showed that just 30% of surveyed practitioners were willing to try teleaudiology for hearing aid-related appointments. More recent work indicates that the vast majority of audiologists are now ready to use teleaudiology for particular hearing aid-related activities, such as for providing advice and counselling. However, when it comes to cochlear implant mapping, fitting of hearing aids, and allowing patients to make permanent programming changes autonomously, considerable reluctance remains, with fewer than 20 percent being willing to do these activities remotely. Audiologists are also unwilling to use teleaudiology with children younger than 12 years old and adults over 80 years old. The rationale for these opinions is presumably associated with the perceived pros and cons of teleaudiology. In general, audiologists think that it will positively impact wait times, access, patient travel, and views of the profession of audiology but also that it will negatively impact relationships with new patients, the ability to discuss private topics, and the quality of their interactions with patients.
Conversely, however, one study found that audiologists who had programmed smartphone-connected hearing aids felt they got to know their patients better because they asked more in-depth questions about the patients' lifestyles and listening preferences. Only one study examined patient acceptance of teleaudiology, and it reported high levels of patient satisfaction with a hybrid model of audiological care (online hearing screening and audiological rehabilitation, face-to-face audiological assessment, and hearing aid fittings). This finding indicates that whether by necessity or preference, the use of hybrid services is feasible and acceptable to patients.
The question is: Will the COVID-19 crisis result in long-term changes in teleaudiology practices? Preliminary data from our laboratory at Manchester Centre for Audiology and Deafness (ManCAD) at the University of Manchester suggest this might be the case. Initial analysis of data from an ongoing survey of NHS audiologists indicated that 33 percent of respondents used teleaudiology before COVID-19, and having now used it, 83 percent said they would continue doing so.
AUDIOMETRIC EVALUATION FOR ADULTS
Remote assessment of pure-tone hearing thresholds has been limited by the need for calibrated headphones and a sound-attenuated environment. Currently, no method has been validated for remote pure-tone assessment without the required special equipment. Should the need for calibrated signals and transducers be overcome, data show that the differences in pure-tone hearing thresholds methodically obtained via conventional manual and automated air conduction audiometry are small and do not differ from typical test-retest differences obtained via the conventional method. However, a method for overcoming the need for calibrated signals and transducers remains, and until that time, remote audiometric assessment is limited to suprathreshold testing or in-situ audiometry using signals delivered via a hearing instrument.
Suprathreshold testing has been widely used for screening using digits presented in background noise via digits-in-noise (DIN) testing. Rather than measuring a hearing threshold, DIN tests measure a signal-to-noise ratio for a fixed level of performance from which hearing thresholds can be estimated. DIN data are available for versions used in the Netherlands, Belgium, the United States, South Africa, and Australia, and the test has been conducted via telephone, a computer, and a smartphone. Studies have shown that the sensitivity and specificity of the DIN in identifying hearing loss are high at >0.85 for both. The test, however, is not entirely independent of familiarity with the language used in testing. Importantly, scores on the DIN have been shown to correlate significantly with pure-tone thresholds measured using conventional audiometry.
The DIN was recently used in home-based and simulated home-based16 assessments of adult cochlear implant users. The results showed no difference in performance between home-based (or simulated home-based) and clinic-based performance; thus, it is feasible for experienced CI users to perform self-administered speech recognition tests at home. The DIN is, therefore, a well-validated measure than can be used for screening and self-assessment of hearing loss.
In-situ audiometry allows testing of pure tone thresholds using signals generated by a hearing instrument. One advantage of this test is that the signals are presented through the same coupling and transducer as will be used for amplification. There are few studies on the validation of in-situ audiometry. One study reported that hearing thresholds obtained via in-situ and conventional audiometry with one model hearing aid were comparable. Meanwhile, another study showed more variation in the findings. In the latter study, the thresholds for 30 individuals with hearing loss were measured via conventional audiometry and in-situ audiometry using hearing aids from four major manufactures (Resound, Phonak, Starkey, and Oticon). It found that the in-situ thresholds overestimated the severity of hearing loss at low frequencies and underestimated it at higher frequencies. There were also significant variations across each manufacturer's hearing aid, and there was an interaction between the error and degree of hearing loss. It should be noted, however, that the hearing aids evaluated in this study were manufactured over five years ago. It will be interesting to see whether in-situ audiometry with newer models of hearing aid would yield less variable and more comparable findings.
Some data have raised awareness that individual differences in education level, perceived locus of control, and cognitive function impact the ability of participants to independently conduct in-situ audiometry. This is important because these findings likely apply to some degree to all self-administered testing and self-managed rehabilitation (see part two of this article series for more on this topic).
AUDIOMETRIC EVALUATION FOR CHILDREN
The use of teleaudiology for newborn hearing screening and for assessing very young children (under 3 years of age) is currently not possible because of the limitations of using objective tests, such as tympanometry, otoacoustic emissions, and auditory brainstem tests in a remote care setting. Several studies, however, have compared results of pure tone hearing screening using a tablet or computer with results obtained using in-person screening audiometry for children aged three and older. In general, the result showed acceptable sensitivity and specificity relative to those conducted in-person, although unsurprisingly, age effects were evident, especially among very young children. The DIN test has also been used to successfully assess the hearing of children, although data indicate that adult norms would have to be adjusted for a paediatric population when interpreting the outcome.
REMOTE RISK ASSESSMENT
A major concern about remote-only audiology practice, such as in the case of direct-to-consumer hearing aids, is that in the absence of a medical evaluation, ear diseases will go unnoticed, resulting in serious adverse health consequences. In the general population, the prevalence of ear conditions that require medical intervention is low. For example, the prevalence of chronic otitis media is 4.5%, cholesteatoma is 0.01%, and vestibular schwannoma or other retrocochlear 0.002%. Data from Medicare beneficiaries show that about 11% of patients complaining of hearing loss would be expected to have an active otologic disease or medically-treatable conditions that affect their hearing. Nonetheless, if remote-only audiology is to be safe, there is a need to distinguish symptoms that are red flags for serious conditions from those that are not. To that end, a literature review was conducted to identify conditions likely to occur in adults complaining of hearing loss. The key symptoms associated with these conditions were documented, and a questionnaire was developed with questions that aimed to identify these symptoms. Since many presenting symptoms of these conditions are similar (e.g., asymmetric hearing loss and sudden-onset hearing loss), the resulting questionnaire, known as the Consumer Ear Disease Risk Assessment (CEDRA), only has 15 items. This online questionnaire includes yes/no and multiple-choice questions and has a reading grade level of 5.7, which falls within the recommended reading difficulty level for health materials and is better than most other hearing and health questionnaires. Studies have shown that the sensitivity of the CEDRA varies from 76 percent to 91 percent, and the specificity varies from 72 percent to 80 percent, depending on the test sample and computations used. It seems that its validity has not been assessed with at-risk populations; however, further research is ongoing to test and refine the tool. Nonetheless, it is a highly promising tool for triaging patients to remote care versus in-person medical assessment.
This review of teleaudiology processes and procedures associated with hearing assessment indicates that considerable work has gone into validating remote assessment approaches. Two major limitations of remote hearing assessment include (1) the inability to conduct objective tests relevant to the evaluation of newborns and very young children—which also applies to people with special needs who cannot manage behavioural testing or provide reliable self-report—and (2) the need for calibrated signals and transducers to measure hearing thresholds. It would seem, however, that when threshold testing is to be used for hearing aid fitting, in-situ testing using signals generated by a hearing aid can provide adequate information for at least an initial fit (see part 2 of this article for further discussion of remote hearing aid fitting). These limitations notwithstanding, audiologists should remain optimistic about remote hearing assessments. With continued efforts to examine the validity of new approaches, clinicians will be able to provide evidence-based care to their patients.