Techniques for Remotely Programming Children With Cochlear Implants Using Pediatric Audiological Methods via Telepractice

Abstract

Purpose

The purpose of this report is to describe the equipment setup and techniques for successfully testing behavioral thresholds in young children with cochlear implants (CIs) using telepractice. We will also discuss challenges associated with pediatric CI programming that are unique to the use of distance technology, and we will describe ways to overcome those challenges. Last, we will review the results from 2 recent studies specifically targeted toward testing behavioral thresholds in young children with CIs.

Method

Conditioned play audiometry or visual reinforcement audiometry was used to measure behavioral thresholds (T levels) for 35 young children with CIs (n = 19 for conditioned play audiometry and n = 16 for visual reinforcement audiometry). Participants were tested in the traditional in-person condition and in the remote condition using an AB-BA study design over 2 visits.

Results

There was no significant difference in T levels between the in-person and remote conditions, indicating that it is feasible to test young children using conventional pediatric testing procedures via remote technology. The primary challenges encountered were in regard to proper camera and video monitor placement at the remote site and the timing of communication between the audiologist and test assistant.

Conclusions

The results from studies to date suggest that distance technology can be used successfully to program CI sound processors for young children using standard, age-appropriate testing techniques. The alternative of remote testing has substantial implications for reducing time and travel burdens for families, potentially leading to the construction of appropriate maps for young children with CIs in a timelier manner.

Cochlear implantation typically involves a multidisciplinary approach and is a less common procedure compared with other audiological interventions. As a result, cochlear implant (CI) centers are not as geographically widespread as general audiology, hearing aid, or otology practices. This logistical challenge can pose problems for people with CIs who live far away from their center, or who otherwise have travel barriers. From an audiological standpoint, recipients are typically seen at the CI center approximately six to nine times within the first year, with biannual or annual visits thereafter. At our center, children are seen at the initial stimulation (consisting of appointments 2 days in a row), 2 weeks, 1 month, 3 months, 6 months, 9 months, and 1 year post initial stimulation. The audiology visits typically involve creating and fine-tuning sound processor programs, or “maps.” Once the initial map parameters are chosen by the audiologist (e.g., pulse duration, stimulation rate, and strategy), behavioral thresholds (T levels) and either the most comfortable levels (M levels) or the maximum comfortable levels (called C or M levels, depending on the manufacturer) must be measured for all or a subset of electrodes. This process can be time-consuming. An additional challenge for programming CIs for young children is that it can be difficult to obtain enough reliable behavioral information to create a map due to the child's limited attention span and/or level of cooperation. As a result, additional visits beyond the standard six to nine visits could be necessary within the first year to construct a map that would allow the child appropriate access to sound. Missed visits could result in suboptimal programs that could delay speech and language development. Telepractice could therefore be an attractive alternative to in-person visits, particularly for young children.

The majority of studies that have evaluated the feasibility of remote programming for CI recipients have focused on adults (Eikelboom, Jayakody, Swanepol, Chang, & Atlas, 2014; Hughes et al., 2012; McElveen et al., 2010; Ramos et al., 2009; Wesarg et al., 2010). These studies have shown that map levels obtained remotely are not clinically different from those obtained in person in the clinic. Furthermore, the majority of CI recipients and clinicians in these studies rated the remote procedures as an acceptable alternative to in-person visits. Programming CIs for adults who are postlingually deaf is typically straightforward because these recipients have prior experience with sound, so they understand the concepts of soft, comfortable, and loud. This population also has spoken language skills to describe what they hear. As a result, adapting the clinical programming procedure for remote implementation poses few problems.

Young children, on the other hand, require testing techniques that are different from those used with adults. Conditioned play audiometry (CPA) is typically used with children between the ages of approximately 2–5 years and consists of conditioning the child to respond to a sound by engaging in a play task (e.g., drop a block in a bucket, place a puzzle piece in a puzzle). For even younger children, starting at the age at which CIs are approved (12 months in the United States; Working Group on Cochlear Implants, 2003) to approximately 2–3 years of age, visual reinforcement audiometry (VRA) is used. With VRA, the child is conditioned to turn and look at an animated or lighted object when he or she hears a sound. With both CPA and VRA, two testers are typically required. The audiologist uses the clinical CI software to present sounds through the child's implant while a test assistant engages the child in the task, judges the validity of the child's responses to sound, and communicates with the audiologist about those judgments. Because the timing and communication between the CI audiologist and the test assistant is critical, programming young children via telepractice is likely to be more challenging than for the typical adult CI user (Franck, Pengelly, & Zerfoss, 2006; Hughes, Goehring, Miller, & Robinson, 2016).

The purpose of this report is threefold. The first goal is to describe the recommended equipment setup and test procedures on the basis of our experience from two studies evaluating the use of telepractice for programming young children via CPA (Goehring & Hughes, 2017) and VRA (Hughes, Goehring, Sevier, & Choi, in press), respectively. Second, we will identify some of the challenges specific to remote testing with young children and discuss potential solutions to those problems. Last, we will review our recent study results comparing T levels obtained remotely versus in person for each of the two pediatric testing methods (VRA, CPA). The information presented here should provide a framework for CI audiologists to begin implementing remote programming with young children who have CIs.

Test Setup

Figure 1 shows a schematic illustration of our recommended setup for CI programming with young children using distance technology. The left and right panels show the setup at the clinic/audiologist site and the remote/recipient site, respectively. For the clinic site, the CI audiologist should be seated in front of a computer used to access and control the computer at the recipient site. Many remote-access software applications are available to securely control another computer from a remote location (e.g., GoToMeeting, TeamViewer, Remote Utilities, UltraVNC; see Fisher, 2018, for a list of free remote-access software applications). Our studies used Windows Remote Desktop Connection (for Windows 7) because the audiologist and recipient sites were in different rooms within the same building, with both computers connected to the same network. (This setup was used to minimize logistical obstacles for the study.) The videoconferencing system should be located at a 45° to 90° angle to the CI audiologist so he or she can easily see both the videoconferencing screen and the computer screen. The web camera should be zoomed in and positioned so that the test assistant can clearly see the audiologist. The microphone (not shown) should be located on or near the computer desk, close to the audiologist, for optimal sound transfer.

Figure 1.

Schematic illustration of equipment and participant setup for remote testing. Left: cochlear implant (CI) audiologist site. Right: remote site with the child, caregiver, and test assistant. VR = visual reinforcer. Adapted from Hughes et al. (in press).

At the remote/recipient site, a child participating in CPA should sit at a table opposite the test assistant. A younger child participating in VRA should sit on the caregiver's lap at the table with the test assistant sitting opposite them (see Figure 1, right panel). Because CI programming is typically done in a regular clinic room (i.e., a sound booth is not needed), an elaborate VRA setup is not necessary. Visual reinforcers can be battery-powered toys that are operated by the test assistant via a remote control or wired to a pancake switch or foot pedal located on the floor underneath the testing table (Hughes et al., in press). Toys (for CPA), the visual reinforcer (for VRA), and/or distractor objects (for VRA) should be placed on the table. As with typical clinical practice, only the items needed at the time should be kept on the table to avoid distracting the child. Additional toys are kept on the floor near the test assistant, out of the child's view, as backups if the child loses engagement in the task. For VRA, the lighted or animated object used for visual reinforcement (indicated by “VR” in Figure 1) should be placed at the end of the table nearest the web camera so that the child's looking behavior can be seen clearly by the CI audiologist and the reinforcer is not the center of the child's attention. The web camera should be located toward the side of the play table so that both the child and the test assistant can be seen clearly by the CI audiologist. For testing children with bilateral CIs, we recommend placing the visual reinforcer on the same side of the table as the test ear, ensuring that the web camera is placed toward the same side as well. The videoconferencing screen should be located behind and slightly above or to the side of the child and caregiver so that the test assistant has a clear view of the monitor but the child does not. The programming computer at the recipient site is equipped with the clinical programming software, CI interface hardware, and programming cable that will connect to the child's CI sound processor. The computer-to-processor link is starting to go wireless (e.g., Cochlear's N7 platform; Cochlear Ltd., 2017), which will eliminate the need for an interface and programming cable at the remote site. Care should be taken to ensure that the computer monitor or laptop screen is positioned so that it is not visible to the child, thereby avoiding distractions or visual cues to the stimulus presentation. The test assistant is responsible for turning on the videoconferencing unit and the programming computer, receiving the remote access request from the CI audiologist, ensuring all peripheral CI equipment is attached, and connecting the programming cable to the child's sound processor (or activating the wireless link, if applicable). We recommend establishing the videoconferencing link first so that the CI audiologist can assist with troubleshooting any hardware or software problems at the remote site.

Procedure

Once the CI audiologist has gained access to the computer at the remote/recipient site, he or she will open the child's file within the programming software and pull up the currently used map. We do not recommend using remote programming for initial stimulation visits because the external equipment must be fit to the child, which includes ensuring proper magnet strength. Because remote programming is only recommended for follow-up visits, the clinician should have some framework of existing map levels to start with. The CI audiologist and the test assistant will proceed with standard CI programming methods to condition the child using either CPA or VRA and then proceed to obtain T levels for as many electrodes as possible.

It should be noted that the procedures for setting upper comfort (C or M) levels have not yet been validated using remote programming because young children lack the concepts and language to convey loudness percepts (Goehring & Hughes, 2017; Hughes et al., in press). As a result, upper comfort levels are typically set using clinic- or audiologist-specific protocols. These might include setting C or M levels at a specific number of programming units above the T level, or estimating upper comfort levels with objective measures, such as the electrically evoked compound action potential (eCAP) or the electrically evoked stapedial reflex threshold (eSRT; e.g., Gordon, Papsin, & Harrison, 2004). The eCAP can be measured easily with the commercial CI programming software and requires no additional equipment beyond that used for programming. For eSRTs, an immittance bridge is needed in addition to the CI programming software (which provides the stimulus). Although several previous studies have reported on the feasibility of measuring the eCAP remotely (Hughes et al., 2012; Ramos et al., 2009; Shapiro, Huang, Shaw, Roland, & Lalwani, 2008), no studies have empirically evaluated eCAP testing in young children postoperatively using telepractice. We are also unaware of any studies that have evaluated eSRT testing via telepractice. These topics remain an area for further investigation.

When the CI audiologist presents stimuli to the child, it is critical to communicate to the test assistant when stimulus presentation begins and ends so the test assistant can provide accurate reinforcement to the child. In addition, the audiologist should also communicate when the stimulus level will be increased or decreased. For example, once the child is conditioned, the audiologist might say, “Here we go at a lower level; ready, one, two, three, four,” while counting aloud with the presentation of the pulse trains in the software. It should be noted that the microphone on the CI sound processor is disabled during programming, so the child should not be able to hear the discourse between the test assistant and the audiologist. For children who use bilateral CIs or a hearing aid on the opposite ear, practitioners should turn off the device from the nontest ear during programming. In addition, caregivers (particularly for VRA, where they are holding the child) should be instructed not to cue the child to the presence of sound.

During conditioning and testing, especially for VRA, both the CI audiologist and test assistant should be looking for a response from the child. Although the camera placement should allow for the CI audiologist to see the child's responses, sometimes these behaviors can be subtle, requiring confirmation by the test assistant. The test assistant should also closely monitor the child's attention and cooperation level. Being right in front of the child, the test assistant might have a better feel than the audiologist for whether the toys need to be substituted, a break needs to be taken, or the child is ready for the session to be finished. In these cases, it is important for the test assistant and CI audiologist to establish a very clear, continuous, and explicit line of communication. Clear communication is also important for efficient use of time, which will maximize the amount of behavioral information obtained from the child.

In our experience, the test assistant does not need to be a skilled pediatric audiologist or therapist (Hughes et al., in press). For individuals without a pediatric audiology background, training for the role of test assistant in our studies consisted of watching video recordings of in-person and remote test sessions, followed by detailed coaching by the programming audiologist during the first few programming sessions. Minimal training will be needed for an individual to become proficient as a test assistant if the individual is already comfortable interacting with young children.

Challenges and Potential Solutions

One challenge that we encountered in the beginning of our studies was a slight delay between the audiologist's verbal counting of pulse trains and the timing of stimulation from the programming computer at the recipient site. This delay resulted in the test assistant's reinforcement occurring at the wrong time. To alleviate this problem, we enabled the audible notification feature within the programming software on the computer at the recipient site so that it would play time-locked beeps that the test assistant could hear when the stimuli were presented to the child's CI. Given that our studies were conducted with the audiologist and recipient in different rooms within the same building, the time delays that we experienced are likely to be minimal compared with a more realistic situation in which the clinician and recipient are in different cities with potentially more variable Internet speeds. As a result, we recommend enabling the audible notification in the programming software for optimal timing of reinforcement.

VRA testing in the normal clinical setting with acoustic stimulation is typically conducted in a sound-treated booth with a more elaborate setup of visual reinforcement objects. For our remote setting, we used a foot pedal or pancake switch on the floor to activate the lighted or animated toys that were used for visual reinforcement. The pancake switches were obtained through a website for enabling devices; these and the more durable foot pedal switch were connected via a 3.5-mm mini plug to toys that were specifically manufactured with a mini plug for activation. Another option is to use remote-controlled toys, but this requires the test assistant to keep one hand free to operate the remote (preferably under the table, out of the child's sight).

As noted above, the angle and zoom of the web camera is extremely important, as is the physical separation between the camera and videoconferencing screen. It is important to have the camera zoomed in enough to see the child and test assistant clearly, but zoomed out enough to accommodate some movement at the table so that everyone stays in the frame of view. A web camera that is integrated within the videoconferencing equipment cannot be moved to a different location than the video screen. It is important to keep the video screen within view of the test assistant, but out of view of the child to avoid distractions; however, the web camera needs to be placed so that the child's face and the test assistant can both be seen.

Last, it is important to note that the experiments described here (Goehring & Hughes, 2017; Hughes et al., in press) were conducted within the same building, which does not fully represent actual clinical implementation. In practice, additional barriers could include inconsistent quality of Internet service, poorer quality videoconferencing hardware or software platforms, and so forth. As a result, more work is needed to validate CI programming for children in actual clinical implementation.

Current Findings

To date, only one study has empirically compared T levels obtained in person versus remotely for young children with CIs using CPA (Goehring & Hughes, 2017). We have recently completed a similar study with a younger cohort using VRA (Hughes et al., 2016, in press). Both studies utilized an AB-BA design over two closely spaced visits, where the in-person condition (A) was tested first, followed by the remote condition (B) at the first visit. Participants were then tested in the opposite order (BA) at the second visit. Data from these two studies are summarized in Figure 2. For the VRA group, thresholds were obtained for three electrodes (basal, middle, apical) for 17 children between the ages of 1.1 and 3.4 years (M = 1.9 years). A complete data set could not be obtained for one child, so data for the remaining 16 children are shown in Figure 2. For the CPA group, thresholds were also obtained for three electrodes for 19 children between the ages of 2.4 and 7.1 years (M = 5.25 years). The data in Figure 2 were averaged across all test electrodes. The in-person and remote results represent the average across the two A and the two B conditions, respectively. Results from the VRA group (Hughes et al., in press) showed no significant difference in T levels between the in-person (M = 4.9 nanocoulombs [nC]) and the remote (M = 4.9 nC) test conditions. Similarly, data from the CPA group (Goehring & Hughes, 2017) showed no significant difference in T levels between the in-person (M = 3.1 nC) and remote (2.9 nC) conditions. (Readers are referred to Goehring & Hughes, 2017, and Hughes et al., in press, for more detailed methods and statistical results.)

Figure 2.

Bar graphs depicting mean T levels (bar height) and standard errors (whiskers) for measures made in the traditional in-person condition (white bars) versus the remote condition (gray bars). Data are averaged across all participants and tested electrodes. Left panel: results for the younger age group, obtained using visual reinforcement audiometry (VRA; data from Hughes et al., in press). Right panel: results for the older age group, obtained using conditioned play audiometry (CPA; data from Goehring & Hughes, 2017). The number of participants is noted in each panel. Means are denoted on each bar. nC = nanocoulomb.

For both studies, test time was recorded for each session to determine whether the remote procedure resulted in prolonged testing time compared with the in-person session. Results from both studies showed that there was no significant difference between the average duration of the remote and in-person sessions. The mean test times for the remote and in-person sessions for the VRA group were 13.0 and 12.4 min, respectively (Hughes et al., in press). Mean test times for the remote and in-person sessions for the CPA group were 15.4 and 16.4 min, respectively (Goehring & Hughes, 2017). However, it should be noted that these test times were for measuring thresholds for only three electrodes per child. In clinical practice, the overall test time will be longer because more time would be dedicated to obtaining behavioral measures for as many electrodes as possible within the timeframe available for a clinical appointment. Regardless, the study results suggest that testing time is not prolonged due to the remote procedure. It should also be noted that the test time did not include the time it took to establish the videoconferencing connection and to remote into the computer at the recipient site, because the goal was to determine whether the remote procedure itself (e.g., potential signal delays, disrupted video link, or increased need for clinician–assistant communication) prolonged the testing time. For both studies, the average time to establish the videoconferencing link and remote in to the laptop at the recipient site was less than 1 min (Goehring & Hughes, 2017). In summary, these studies showed that similar thresholds and test times were obtained in both conditions.

Conclusions

Empirical results show no significant difference in T levels between remote and in-person conditions, which demonstrates that telepractice is feasible for obtaining behavioral thresholds in young children with CIs (Goehring & Hughes, 2017; Hughes et al., 2016). The main challenges with remote programming were related to the timing of communication between the programming audiologist and the test assistant and the optimal placement of the web camera and videoconferencing monitor at the remote site. These challenges were easily overcome with slight modifications of our process. The option of using remote testing for programming CIs with young children is therefore possible and has substantial implications for reducing time and travel burdens for families. Furthermore, remote programming provides an alternative to additional visits to the center if a child refuses to cooperate at a scheduled visit. More immediate access to CI programming could potentially lead to the construction of appropriate maps for young children with CIs in a timelier manner, potentially fostering optimal listening, speech, and language development.

Funding Statement

This study was supported by the National Institute on Deafness and Other Communication Disorders Grants R01 DC013281 (Primary Investigator: Michelle Hughes) and P30 DC04662 (Primary Investigator: Michael Gorga) and the National Institute of General Medical Sciences Grant P20 GM109023 (Primary Investigator: Walt Jesteadt), each awarded to Boys Town National Research Hospital.