Using tablet-based technology to deliver time-efficient ototoxicity monitoring

Abstract

Objective: The goal of this article is to highlight mobile technology that is not yet standard of care but could be considered for use in an ototoxicity monitoring programme (OMP) as an adjunct to traditional audiometric testing. Current guidelines for ototoxicity monitoring include extensive test protocols performed by an audiologist in an audiometric booth. This approach is comprehensive, but it may be taxing for patients suffering from life-threatening illnesses and cost prohibitive if it requires serial clinical appointments. With the use of mobile technology, testing outside of the confines of the audiometric booth may be possible, which could create more efficient and less burdensome OMPs. Design: A non-systematic review of new OMP technology was performed. Experts were canvassed regarding the impact of new technology on OMPs. Study sample: OMP devices and technologies that are commercially available and discussed in the literature. Results: The benefits and limitations of portable, tablet-based technology that can be deployed for efficient ototoxicity monitoring are discussed. Conclusions: New mobile technology has the potential to influence the development and implementation of OMPs and lower barriers to patient access by providing time efficient, portable and self-administered testing options for use in the clinic and in the patient’s home.

Key Words:

• Hearing conservation/hearing loss prevention
• instrumentation
• otoacoustic emissions
• tele-audiology/tele-health

Introduction

Prospective audiological monitoring for ototoxicity is considered to be a best practice (ASHA 1994; AAA 2009) for patients undergoing therapy with ototoxic medications, including platinum-based chemotherapy drugs and aminoglycoside antibiotics. The audiologist designs testing protocols, provides or supervises patient testing, interprets test results and manages ototoxic symptoms when detected (AAA 2009). The optimal clinical protocol for ototoxic monitoring should incorporate a full audiological-vestibular diagnostic baseline evaluation, with regular follow-up appointments in the audiology clinic during and after treatment. However, the actual implementation of an ideal ototoxicity monitoring programme (OMP) that fully complies with these requirements can be challenging (Konrad-Martin et al. This issue). First, it places a major time burden on the patient, who may already be managing a complicated schedule of clinical appointments while suffering from a life-threatening illness. It may also require patients to spend time-performing tests that do not significantly contribute to the OMP’s primary goal of detecting changes in threshold. Further, some audiology clinics may not have the personnel resources or clinic space to provide the level of scheduling flexibility and timeliness required to meet the needs of all patients taking ototoxic medications. Finally, sustainable OMPs require providers to establish ototoxicity monitoring as a high priority in the treatment plan of these very ill patients.

Because of these challenges, both the American Speech-Language-Hearing Association (ASHA) and American Academy of Audiology (AAA) have guidelines that provide leeway for shortened screening protocols to be used for ototoxicity monitoring (ASHA 1994; AAA 2009). Although audiologic evaluation is ideally conducted in a sound-treated booth, the ASHA guidelines recognise that, even with shortened protocols, full booth-based audiometric monitoring may not be feasible in all clinical environments. Boothless audiometry in which threshold results are the goal requires technology that utilises standards for environmental noise tolerances by frequency. American National Standards Institute (ANSI) S3.1 specifies Maximum Permissible Ambient Noise Levels (MPANLs) in octave and one third octave bands by frequency (up to 8000 Hz) permitted in any audiometric test room. The key elements to achieve reliable mobile boothless testing for OMP are described in this paper.

The goal of this paper is to review mobile technology options that, while not yet standard of care, could be considered for future use with an OMP as an alternative to traditional audiometric booth testing. These portable audiometers and tablet- and application-based systems have the potential to test patients in the hospital or clinic, thereby alleviating some of the appointment-based burdens placed on both patients and audiology staff. Our aim is not to provide an exhaustive review of portable audiometric options available for routine hearing screening, but rather aims to provide a brief overview of the possible options for boothless audiometry and a more in-depth review highlighting two unique tablet-based portable audiometers designed specifically to address the emerging new technological needs for boothless ototoxicity monitoring during treatment.

Unique requirements for ototoxic monitoring

OMPs have considerations that differentiate them from other aspects of audiology practice. Hearing loss following treatment with ototoxic medications typically occur first in the high-frequency region (>8 kHz). Frequencies most important for speech understanding (<8 kHz) can generally be affected with additional dosing (Fausti et al. 1994). AAA (2009) recommends early detection of extended high-frequency audiometric thresholds (9–20 kHz). Further, as discussed in more detail by Brewer and King (This issue) in this Special Issue, various criteria are used to define or grade an ototoxic hearing change.

ASHA defines a “significant” ototoxic hearing shift as a worsening of results as determined by a comparison of the baseline test to the follow-up test for any one frequency (of 20 dB or greater) or two adjacent test frequencies (of 10 dB or greater) that occurs in either the conventional (0.5 kHz through 8 kHz) and/or extended high frequencies (typically 9–20 kHz). An ototoxic shift can also be noted when responses are no longer obtained at three consecutive test frequencies where responses were previously obtained. This criterion is limited to the highest frequencies tested. Although ASHA criteria state that changes should be measured relative to a baseline, some patients taking ototoxic medications do not receive a baseline hearing test before the start of treatment, making it difficult to define a shift in hearing.

Other criteria, such as the National Cancer Institute (NCI) Common Terminology Criteria for Adverse Events (NCI CTCAE, v4.03, 2010), have provisions for assessing ototoxicity when the patient is not enrolled in a monitoring programme. In this latter instance, criteria may be based on subjective hearing change or the subjective need for hearing aids or other interventions that were not indicated by the patient prior to treatment. The CTCAE is primarily concerned with changes that occur in the conventional audiological frequency range up to 8 kHz. The audiologist interprets when these changes are likely to diminish hearing for speech communication.

Additionally, both ASHA and AAA guidelines for the audiological monitoring of ototoxicity refer to the possible use of abbreviated test protocols (e.g., tympanometry and air conduction testing only) and screening protocols to conduct monitoring (although these abbreviated methods are not required for standard of care). One screening protocol specifically described in these guidelines identifies a patient-specific “Sensitive Range for Ototoxicity (SRO)” that can be used as a time-efficient early detection of ototoxicity screening test (Vaughan et al. 2002; Fausti et al. 2003). The SRO method determines an operationally defined highest audible frequency, at which the patient can reliably detect a tone of 100 dB SPL or less. Thresholds tested for this highest audible frequency and the next six frequencies measured at one-sixth octave intervals constitute the SRO. To administer this protocol, audiometers need to be capable of generating high-level, high-frequency stimuli in one-sixth octave intervals. Generally, most clinical audiometers have this capability as do some portable systems.

Advances in portable technology for OMP

Portable audiometry is not new, but interest in testing hearing in less than ideal boothless environments has grown recently and vendors have responded with increased product lines. The need for boothless audiometry has arisen from improvements to humanitarian efforts, school-based screenings, hearing conservation programmes (Meinke et al. 2017), teleaudiology and OMPs. In this section, we describe multiple methods of portable audiometry.

Basic and advanced portable audiometers

Basic portable audiometers are primarily used for hearing screening. Air-conduction audiometry is often the only test included. However, more advanced portable audiometers can be used for diagnostic purposes. The portability of audiometers varies widely and tends to correspond to the quantity of frequencies that can be tested and the complexity of software functionality. Considerations when choosing an audiometer for an OMP should include, size, availability of transducers, types of testing desired (e.g. air conduction, bone conduction or speech capability), operator (manual) versus patient-controlled testing (automated), room noise monitoring with capability to stop testing should noise exceed MPANLs, means to assess reliability of the test results, and telehealth capability to send test results to the audiology clinic for analysis.

Most portable audiometers are designed to be operated by an audiologist in a manual mode where the tester places the earphones on the patient, adjusts the level and frequency of the tone presentation and records the listener’s responses. Other systems (several of which are described in detail below), include automated capabilities, meaning that testing can be performed via software so that a self-administered hearing test can be completed by the listener. As an example, software may use widely accepted guidelines for hearing testing methodology (ANSI S3.21). The listener will be given instructions to follow regarding earphone placement and how to take the hearing test. Software options should allow the audiologist to define the number of test frequencies to be tested or the option to use abbreviated protocols. Inclusion of teleaudiology capability might mean that the audiologist need not travel to the chemotherapy treatment unit to view and evaluate the test results but could instead have the results arrive to their desktop computer or phone application for review during a typical clinical day. Due to the variety of basic and advanced portable audiometers currently available, the administrator of an OMP must choose a product based on the needs of the programme.

Application-based audiometric test systems

Several vendors have developed applications (apps) that can be downloaded on portable consumer electronic devices, such as mobile phones or tablets, and used to conduct audiometric screening or diagnostic tests, which can either be self-administered by the patient or be manually administered by a clinician. These apps rely on the internal sound card of the commercial device to generate the test signals and a standardised electroacoustic transducer, such as a pair of headphones, to present the sound to the listener. Some apps allow for the use of personal ear/headphones while other apps require the purchase of a specific set of headphones (i.e. Sennheiser HDA 280s).

Multiple application-based audiometer systems have been empirically researched, including uHear^TM, EarTrumpet, Hearing Test™, hearScreen^TM and the ShoeBOX Audiometer. uHear^TM and EarTrumpet are applications that rely on the user’s personal mobile device, including personal headphones (Foulad et al. 2013; Abu-Ghanem et al. 2016). Neither the uHear™ nor the EarTrumpet are recommended for diagnostic audiometry. However, EarTrumpet and Hearing Test™ (Renda et al. 2016) integrate filter coefficients to correct for the different frequency responses of the specified headphones. hearScreen^TM (Mahomed-Asmail et al. 2016) requires the use of either supra-aural or circumaural headphones and is calibrated to ANSI standards. hearScreen^TM has the capability to test up to 16 kHz using Sennheiser HD 202 headphones and up to 25 kHz with Sennheiser HS 280 Pro headphones. The ShoeBOX Audiometer (Thompson et al. 2015; Rourke et al. 2016) is calibrated to ANSI S3.6 and meets ANSI S3.1 for testing outside of a sound booth. The ShoeBOX Pro version can function as either a diagnostic or a screening audiometer with options for testing with puretones or speech stimuli. Sound delivery options include the following recommended transducers: Sennheiser HDA 280 headphones, TDH-50 (60 ohm) headphones, E-A-RTONE 3A insert earphones, or B81 RadioEar bone vibrator. Frequencies that can be tested range from 0.25 to 16 kHz. Lastly, ShoeBOX Audiometer includes both automated and manual testing. Good sensitivity and specificity outcomes have been shown for automated testing using ShoeBOX Audiometer in a quiet room compared to manual testing in a sound booth performed by an audiologist (Thompson et al. 2015). It should be noted that the ShoeBOX Audiometer is the first portable iPad-based audiometer to be listed as a Class II device by both the U.S. Food and Drug Administration (FDA) and Health Canada. As an FDA listed product that is in compliance with ANSI standards the ShoeBOX audiometer can be used clinically, which reduces reimbursement barriers as it is not considered an experimental device. While many of the apps discussed above have been designed to expand accessibility of hearing healthcare to patients outside of the clinic, the ShoeBOX Audiometer was designed to improve the flexibility of services offered within the audiology clinic. As innovations in technology continue, the service delivery model will be an important aspect to consider in the usability of the product.

A final note should be made about the importance of calibration in the use of apps for audiometric measurements. The most sophisticated apps, like the one from ShoeBOX Audiometer, have been demonstrated to be in compliance with the electroacoustic requirements for audiometric test equipment laid out in ANSI S3.6. These standards confirm that the system is capable of reliably producing audio test signals over the range of levels and frequencies required to perform audiometric threshold measurements. However, compliance with ANSI S3.6 does not address the issue of accurately measuring absolute threshold levels. If the apps are intended for measuring clinical audiometric thresholds, then the specific combination of headphone and portable electronic device would need to be calibrated annually in accordance with the same ANSI standards used to calibrate clinical audiometers. Without this calibration, clinicians interpreting the thresholds measured by these systems would have to assume that there could be some variation between these results and those that would be obtained in a standard clinical audiogram (or even in an audiogram measured with a different example of the same portable audiometric system).

Advances in portable technology for OMP

Portable computer-based audiometric systems

Portable automated audiometric test systems are computer-based devices that combine a clinical-grade audiometer (ANSI S3.6) with a tablet or another human interface device that allows the user to respond to auditory stimuli. The devices discussed below allow users to conduct self-administered audiometric threshold testing using a modified Hughson–Westlake method. Some devices also incorporate other features, like speech audiometry, otoacoustic emissions (OAE) measurement and real-time ambient noise monitoring. Here, we discuss two devices that have been evaluated in peer-reviewed articles and were specifically designed for use in support of an OMP: the Ototoxicity Identification Device (the OtoID) and the Creare Wireless Audiometer. Both portable audiometric test systems were designed to provide a flexible way to administer a wide variety of audiometric tests in boothless environments.

OtoID developed by the Department of Veterans Affairs (VA), National Centre for Auditory Rehabilitative Research (NCRAR) located in Portland Oregon, USA

The original Ototoxicity Identification Device (OtoID), described in detail elsewhere (Jacobs et al. 2012), is a portable analogue audiometer with a microchip-controller that has options to enable automated testing of the sensitive range of seven frequencies chosen from 0.5 to 20 kHz in one-sixth up to octave step sizes. The SRO screening protocol can be implemented in the automated or manual mode. Unlike other portable audiometer options, this system is capable of monitoring ambient room noise for reliable out of booth testing and can transmit test results to a provider’s email inbox. The accuracy of threshold measurements relative to pure tone air conduction testing in a sound booth, and threshold test–retest difference reliability metrics have been assessed for the original OtoID. On 40 participants (n = 80 ears), both young and older, with and without hearing loss, the repeatability of thresholds within 5 dB of the initial test generally exceeded 95% except at the very highest frequencies 16–20 kHz where it decreased to 85%. False-positive rates for an ASHA-significant hearing shift were <6% (n = 3 tests) in the automated test mode and <4% (n = 2 tests) in the manual mode (Jacobs et al. 2012).

A more flexible tablet-based format for the OtoID was recently developed that utilises a digital signal processor (DSP) sound generation and stimulus control so that more complex sounds (e.g. speech, chirps, clicks) can be presented to the patient during testing (Figure 1). This new system, called the OtoID-Tablet, has the following additional features: (i) digital generation of stock or provider-determined audio signals; (ii) integration of a cisplatin-induced hearing loss prediction tool for each patient based on their projected dosing and measured pre-exposure hearing (Dille et al. 2012); (iii) capability to allow distortion-product otoacoustic emissions (DPOAEs) testing for identification of outer hair cell damage; and (iv) administration of questionnaires for subjective patient information such as, the presence of new or aggravated tinnitus, quality of life changes during and following treatment and impacts of hearing loss and tinnitus on daily living. The OtoID-Tablet retains the original wide range of audiometric frequencies from 0.5 to 20 kHz with levels spanning 105 dB and includes both manual and automated evaluation formats using the modified Hughson–Westlake hearing threshold procedure. The provider can select the frequency range, step size and test order. It also incorporates software designed to collect and analyse the threshold data needed to implement the SRO screening test and ambient noise monitoring and uses Wi-Fi for transmission of results to a designated provider. When used with hard-wired circumaural headphones, the original and new tablet versions of OtoID are in full compliance with ANSI S3.6-2010 specification for a Type-4 high-frequency (HF) audiometer.

Figure 1. The OtoID-tablet is shown below. In this picture, a Veteran is testing his own hearing on the Chemotherapy Treatment Unit during treatment. Using the self-testing modality, a screen directs the veteran to listen for a tone. After the listening interval, he is asked to respond (yes or no) if a tone was heard. A tone may or may not be presented. The percentage of intervals without tones (“catch trials”) being played can be varied in the software. Noise monitoring microphones (not shown) monitor the room for excessive noise. If noise exceeds the MPANLs for a test room, a tone will not be presented and the screen will instruct the veteran to seek help in quieting the room. Bilateral automated SRO pure tone testing takes approximately 15 min.

Creare wireless audiometer developed by Creare LLC

The Creare Wireless Audiometer is a tablet-based, stand-alone audiometric test device. It incorporates a design where a commercial tablet computer running a custom app is used as the user interface. This tablet connects wirelessly via Bluetooth to a headset that houses the electronics, transducers and battery required for the audiometer. The headset is fully circumaural such that the transducers are not in direct contact with the pinna but slightly recessed inside the cup (Figure 2). With the addition of high-quality ear seals, the headphones provide passive attenuation that is on par with or better than a single-wall sound booth (30–40 dB of attenuation from 0.25 kHz and up measured according to ANSI S12.6; Meinke et al. 2017).

Figure 2. The Creare Wireless Audiometer is shown. The contents of the device include the noise attenuating earcups, wireless audiometer circuit, speaker and microphone, faceplate and protective fabric, and an ear seal. The battery for the device is located within the left ear cup.

In addition to the Bluetooth chip used to communicate with the tablet computer, the headset includes a DSP to control the sound pressure level of the speakers and to apply logic to automated hearing tests. The Creare Audiometer can be controlled through an open source tablet application called TabSINT (Shapiro and Galloza 2016), designed to administer customised hearing tests and questionnaires on mobile devices located across multiple sites. The application provides a user interface to control the audiometer and select specific tests. The Creare Audiometer and TabSINT, together, provide access to a number of hearing tests, including: automated audiometry with a modified Hughson–Westlake algorithm; automated audiometry with a Békésy-like tracking algorithm (fixed frequency); manual audiometry; automated screening; manual screening; several speech in noise tests including the Modified Rhyme Test (MRT), the triple-digit test and the Hearing In Noise Test (HINT) (licensed from Hearing Test Systems, LLC). Specific to OMPs, TabSINT has the capability to measure hearing thresholds from 0.25 to 20 kHz with 1/24th octave step sizes Maximum output levels are 80 dB HL at .125 and .25 kHz, 100 dB HL at .5–4 kHz, 90 dB HL at 5–8 kHz and 70 dB HL above 8 kHz.

Practical considerations in the use of portable systems for ototoxic monitoring

In this section, we discuss some practical considerations that should be considered when using a portable system, such as a tablet, to implement an OMP.

Self-administered versus provider-administered testing

An important consideration in the design and implementation of a portable audiometric test device is whether the tests are intended to be manually administered by an audiologist or self-administered by the patient. Both manual and self-administered audiometric threshold testing typically use standard methodologies such as, the Hughson–Westlake procedure for selecting the level of each stimulus presentation and deciding when a threshold measurement is reliably obtained. In the self-administered test, the algorithm is programmed by the provider to test selected individualised (by patient) frequencies, sequence the presentations of the tones including catch trials (intervals with no stimulus) to ensure reliable responses. In manual mode, the audiologist selects the frequency and intensity of each tone presentation and its presentation timing. The time required to measure a threshold varies from system to system, but manual audiometry performed by an experienced audiologist is generally faster than self-administered testing. This is due primarily to the necessity of “catch” trials (intervals with no tone presentation) to insure the most reliable test result.

Strategies for background noise in boothless environments

Portable technology used in OMPs should provide ways to monitor and/or mitigate ambient noise levels to ensure reliable thresholds are obtained. The use of headphones or insert earphones that provide high levels of sound attenuation can decrease the amount of ambient noise reaching the eardrum as could active noise reduction in environments with increased ambient noise (Bromwich et al. 2008). In addition, recent developments in portable audiometric systems have made it possible to use electronic monitoring to determine when the ambient noise levels may compromise the accuracy of a threshold measurement. Figure 3 displays the maximum permissible ambient noise levels (MPANLs) for the measuring audiometric thresholds with ears uncovered. Shown are the Creare Wireless Audiometer headphones, the Sennheiser HDA200 and HDA300 and ER3A insert earphones (deeply inserted into the ear canal). Also shown in the figure are ambient noise measurements for two hospital environments: Johns Hopkins University Hospital (JHU) and the Portland VA Medical Center chemotherapy treatment unit (PVAMC) (Gordon et al. 2005). Note that the MPANLs increase with the attenuation values of the transducer, so the transducers that have the greatest attenuation (higher on the chart) allow testing in noisier environments than those that do not provide as much attenuation. For comparison, the figure plots ambient noise measurements obtained from the two hospital environments, JHU and the PVAMC chemotherapy treatment unit. Both the JHU and PVAMC display the average noise by frequency obtained during ambient room noise measurement over time. Although the JHU noise was generally louder than the PVAMC environmental noise, both had similar spectral shapes, the noise levels were highest at low frequencies, and dramatically decreased at frequencies above 2 kHz. Comparing the MPANLs to the noise measurements, it is apparent that all the transducers provided enough attenuation to allow reliable threshold measurements at frequencies above 3 kHz in both hospital environments, and that some of the transducers provided enough attenuation for testing all frequencies. However, noise levels can vary over time, and ambient noise level monitoring is highly recommended to ensure that the background noise for each audiometric threshold measurement (including thresholds down to 0 dBHL) does not exceed the MPANLs for the transducer used and the frequency being tested.

Figure 3. Headphone Ambient Noise Levels. This figure shows the calculated Maximum Permissible Ambient Noise Levels (MPANLs) for octave band frequencies in dB SPL re: 20 µPa for four different circumaural and insert transducers. For reference purposes, this figure also shows the average measured ambient noise obtained from two different hospital environments and the ANSI Maximum Permissible Ambient Noise Levels (MPANLs) for ears uncovered, that is, testing in the sound field (ANSI S3.1-1999R2008). The MPANLs for each transducer was calculated by adding the measured or published passive attenuation of each transducer to the “ears uncovered” MPANL standard. Finally, the average noise levels on the Portland VA Medical Center Chemotherapy Treatment Unit—CTU (open squares) and at Johns Hopkins University (JHU) Hospital (open circles) are shown for comparison (Busch-Vishniac et al. 2005; Gordon et al. 2005). Some transducers, such as the Creare (filled diamond) or ER3A Deep insertion (filled triangle), may provide better protection during testing from interfering room noise in the lower frequencies (<1000 Hz) compared to the Sennheiser HDA200 (filled circle) and HDA300 (filled square) earphones, especially true for JHU hospital noise though both types of Sennheiser earphones provide sufficient protection fron interfering noise at 4000 Hz and above which is consistent with their selection for use on the OtoID by providing capability for high-frequency testing (>8000 Hz).

Management of and distribution of patient data

All health providers in the United States must meet the privacy protection requirements of the Health Insurance Portability and Accountability Act (HIPAA), which places strict requirements on how identifiable patient information can be stored and transmitted on electronic devices and across electronic networks. Portable test devices must be certified as HIPAA compliant if they store identifiable patient information. One strategy for addressing this issue is to only store non-identifiable data on the tablet and then connect this information with the patient at the time it is uploaded to the electronic health record. Walter Reed National Military Medical Center is experimenting with an optical scanning technique where test results from the Creare Wireless Audiometer are presented visually on the tablet screen in the form of a QR code that can be optically scanned into the patient’s record. This can be very helpful in environments where security requirements make it impossible or impractical to have the audiometric test system on the same network as the patient record database.

If the device is HIPAA compliant and it can be given access to patient data, the flexible user interface of the tablet can provide great advantages for managing patient care in an OMP. This could, for example, give the provider access to previous baseline and monitoring test results when testing the patient. Wireless capabilities can also be very useful for transmitting the test results to the audiologist. For example, the OtoID has a feature that allows wireless transmission of de-identifiable results (subject is given a number) directly to the secure inbox of a provider who has the key that links to subject number to his/her identifiers) for interpretation (Dille et al. 2015). This allows the audiologist to review the test results and notify the patient’s care team in order for the physician to make timely recommendations about possible therapeutic changes. This can be especially helpful in cases where someone other than the audiologist conducts the hearing test in a remote location.

Future considerations for ototoxic monitoring assessments

Thus far, we have focussed primarily on OMPs based on the traditional fixed-frequency audiometric threshold. However, the technological capabilities of portable audiometric test systems go far beyond traditional fixed-frequency threshold testing, and research is now underway to explore additional test measures that could augment traditional audiogram-based protocols for OMPs.

Békésy tracking audiometry

High-frequency audiometry using a Békésy-style approach is typically administered at individual fixed frequencies and the level is adjusted to find the patient’s threshold. An alternative approach to find the highest audible frequency is to fix the intensity of the tones while varying the acoustic frequency, increasing the acoustic frequency while the patient presses the button until the patient can no longer hear the tones. Then, the patient releases the button and the acoustic frequency decreases until the patient hears it again and presses the button. The highest audible frequency is calculated as the average threshold over a set number of trials. This fixed-level Békésy approach, or fixed-level frequency thresholds (FLFT), can cover a large frequency range in a short time period. Because this approach can focus on the highest audible frequencies at suprathreshold levels, it is inherently noise tolerant, possibly allowing patients to be tested outside a sound-treated room. The fixed level used can be adjusted for different populations or can be set adaptively based on the users’ initial responses. The FLFT is easily automated and allows providers not specifically trained in audiology techniques to administer the test, which can reduce the difficulties associated with ototoxicity monitoring in a cancer care unit or office.

The FLFT approach has been compared to SRO monitoring in a group of 29 normal subjects not being monitored for ototoxicity (Rieke et al. 2017). On average, the FLFT took approximately 30 s to complete, while the SRO using modified Hughson–Westlake audiometry took approximately 4.5 min. Over four testing sessions at least a week apart, the Békésy-style FLFT was repeatable within 1/12 octave, or one step-size in the testing procedure used. The FLFT agreed well with the highest-audible frequency determined using fixed frequency audiometry at 80 dB SPL. Overall, the FLFT appears to be a suprathreshold, noise-tolerant test amenable to automated and self-administration, which may provide a short, accurate, noise-tolerant ototoxicity screening test that could improve the monitoring of extended high-frequency thresholds.

Comprehensive OAE testing

Field studies show that OAE testing is sensitive to early detection of physiologic changes to the outer hair cells in the cochlea, the structures believed to be the site of initial damage from noise and most ototoxic medications. Changes in these structures can lead to threshold shifts. (Miller et al. 2006; Dille et al. 2010; Konrad-Martin et al. 2012). For ototoxicity monitoring, Reavis et al. (2011) also found that DPOAEs provide a good measure of ototoxicity from cisplatin. Although the standard DP-gram is often used to assess ototoxicity, DPOAE level/phase (L/P) mapping may provide a more comprehensive assessment of cochlear function (Martin et al. 2005; Meinke et al. 2013). This technique measures DPOAEs at multiple f2 frequencies and f2/f1 frequency ratios, providing an overall picture or map of outer hair cell responses throughout the cochlea. Preliminary data using DPOAE L/P mapping before and after loud music exposures showed cochlear changes that were not detected from the DP-gram obtained at a single frequency ratio (Buckey et al. 2015). The use of DPOAE mapping for ototoxicity monitoring is currently being assessed by Geisel School of Medicine at Dartmouth and Creare LLC with support from the Office of Naval Research. One major limitation of OAE monitoring for ototoxicity, however, is that many patients receiving ototoxic drugs for head and neck and other cancers may not have robust or even detectable OAEs at baseline (Newlin et al. 2010). But, OAEs have the advantage of being quick and portable, and they do not require a behavioural response. Additional papers in this issue discuss the use of OAEs in OMPs (Brooks and Knight This issue; Garinis et al. This issue; Konrad-Martin et al. This issue).

Speech-based testing

The ultimate goal of ototoxicity monitoring is to prevent functional hearing deficits that might impair the patient’s long-term quality of life. This goal is closely tied to the preservation of speech intelligibility, particularly when listening to speech in background noise. Thus, one might argue that speech perception would be a reasonable metric in an OMP, even if it is only for the purpose of having a baseline for future comparison for patients who develop hearing deficits related to exposure to ototoxic medications. Portable audiometric test systems are ideally suited for measuring speech perception. In cases where closed-set test material is used, patients can respond using a multiple-choice selection allowing automation of the speech perception test on a platform such as a tablet. Speech testing may also be advantageous as it is less sensitive to background noise, and small differences in presentation level due to present background noise will not cause a large difference in audibility during speech testing compared to pure tone threshold measurement. Speech-in-noise testing is typically done with a broadband noise masker that will mask out environmental sounds that might leak in through the headset, which should lead to consistent results in a broader range of noise environments.

One example of a speech test that is sensitive to hearing loss and well-suited for boothless administration is the triple-digit test. Results from this test has been shown to be correlated with hearing loss when administered over a commercial phone line, which is band-limited and subject to uncontrolled variation in overall level (Watson et al. 2012). Its sensitivity to changes in hearing caused by ototoxic exposure is unknown for a broader range of noise environments and testing outside of a sound booth, but it may be a useful area for future research.

Vestibular and balance testing

Vestibular issues, though common during the course of treatment with some ototoxic medications, are not typically directly addressed in OMPs (Black and Pesznecker 1993). Recent papers suggest that the accelerometers and gyroscopes in commercial smart phones and tablet computers can be sensitive enough to allow diagnostic measurements comparable to the Sensory Organisation Test in clinical environments (Alberts et al. 2015a) and that they can enhance the sensitivity of field measures of postural stability like the Balance Error Scoring System (BESS) (Alberts et al. 2015b). Care would need to be taken to develop vestibular tests appropriate for the patients in ototoxic treatment regimens, who may generally be in poor health. However, the possibility of an objective screening measure for postural stability, balance and/or vestibular function that could be administered to ototoxic patients is attractive. See additional details on vestibulotoxicity discussed in Handelsman (This issue).

Surveys/Questionnaires

The interactive tablet interface on most automated audiologic test systems makes them ideal for administering surveys to patients, which can help monitor their audiological symptoms and could provide valuable supplemental information to audiometric testing for tracking changes in ability over time. Surveys could be useful for identifying vestibular symptoms (Dizziness Handicap Inventory), monitoring tinnitus related to ototoxicity (Tinnitus Handicap Inventory, Tinnitus Ototoxicity Monitoring Index) and changes in quality of life and hearing handicap (Hearing Handicap Inventory) not generally captured well by hearing testing alone. Portable systems, particularly tablet-based systems, offer the convenience of automated scoring, and they may make it easier to get survey information uploaded into the medical record than traditional survey methods based on optical scanning of survey instruments filled out on paper by the patient.

Conclusions

Ototoxicity is an extremely unfortunate side effect of many life-saving drug therapies and can significantly impact quality of life of individuals during and following treatment. Communication with family and friends becomes of primary importance among those who survive as well as for those transitioned from treatment to hospice or palliative care. Ideally, the effects of ototoxicity could be alleviated either through the use of otoprotective agents or through the development of new, less ototoxic treatments or treatment regimens. However, until that happens, OMPs will play a key role in the management of patients. In this paper, we have discussed new technological advances that may make OMPs accessible to more patients. These devices have the potential to enable efficiencies by broadening where testing could be completed and expanding who could administer the tests. Automated, portable systems allow for the possibility of individuals, other than audiologists, to be trained to administer a limited monitoring protocol under the supervision of an audiologist. The devices described here also allow for the patient to test him- or herself with results sent to the OMP audiologist for interpretation. This self-testing approach can be performed in patient care facilities or in the patient’s home using automated protocols. The audiologist can be alerted when testing has been completed, at which time results can be reviewed, analysed and recommendations can be made. To the extent that compliance with OMPs is reduced by logistics related to overburdened patient and clinic schedules, these new technologies may allow OMPs to be more feasible for facilities that have difficulty overcoming these barriers. While many systems discussed here already have been validated for use with patients (e.g. OtoID), as outlined by Swanepoel de et al. (2010), it will remain essential that any portable and automatic procedures used in boothless technology be validated before use in patient care; we are hopeful that the right combination of these technological advances will lead to more effective, lower cost OMPs resulting in better patient outcomes.

Declaration of interest

Dr. Odile Clavier is an employee of Creare LLC. Edare Inc. is an affiliate of Creare LLC, with common ownership. Edare manufactures the “Creare wireless audiometers” currently commercially available. Dawn Konrad-Martin, Marilyn Dille, Samuel Gordon were employees of the U.S. Government during the period of time when this manuscript was written. The work was prepared as part of their official government duties and no revenue is being received for this intellectual property.

The opinions and assertions presented are the private views of the authors and are not to be construed as official or as necessarily reflecting the views of the Department of the Army, Department of the Navy, Department of the Air Force, the Department of Defense, Department of Veterans Affairs, the National Institutes of Health or the U.S. Government. Title 17 U.S.C. §105 provides that ‘Copyright protection under this title is not available for any work of the United States Government.’ Title 17 U.S.C. §101 defines a U.S. Government work as a work prepared by a military service member or employee of the U.S. Government as part of that person’s official duties. The contents of this publication do not represent the views of the U.S. Department of Veterans Affairs, Department of Defence or the United States Government.

Abbreviations
AAA	=	American Academy of Audiology
ANSI	=	American National Standards Institute
ASHA	=	American Speech-Language-Hearing Association
BESS	=	Balance Error Scoring System
CTCAE	=	common terminology criteria for adverse events
DPOAEs	=	distortion product otoacoustic emissions
DSP	=	digital signal processing
FLFT	=	fixed-level frequency threshold
HIPAA	=	Health Insurance Portability and Accountability Act
HINT	=	hearing in noise test
JHU	=	Johns Hopkins University Hospital
MPANLs	=	maximum permissible ambient noise levels
MRT	=	modified rhyme test
NCI	=	National Cancer Institute
OAE	=	otoacoustic emissions
OMP	=	ototoxicity monitoring program
OtoID	=	ototoxicity identification device
PVA	=	VA Portland healthcare system
SRO	=	sensitive range for ototoxicity

Acknowledgements

Funding for this work was partially provided by U.S. Army Public Health Command in support of the Army Hearing Program, and by the U.S. Department of Veterans Affairs Office of Rehabilitation Research & Development (RR&D) Service (Grant #C0239R). This research was also supported in part by an appointment to the Postgraduate Research Participation Program at the Army Public Health Center (Provisional) administered by the Oak Ridge Institute for Science and Education through an interagency agreement between the U.S. Department of Energy and the APHC (Prov). Development of the Creare wireless audiometer was supported by the National Institute on Deafness and Other Communication Disorders (NIDCD) of the National Institutes of Health under Award Number R44DC012861 to Creare LLC.