In general, interventional clinical trials can have a wide variety of study design approaches all within good clinical practice parameters, but pharmaceutical interventions for hearing loss (PIHL) studies have largely employed heterogenic design elements. The PIHL literature shows a broad range of design elements for clinical studies, including the intensity, duration, and type of noise exposure, dosing schedules, data collection time points, statistical analysis plans, as well as the primary audiometric endpoints. Currently, no Food and Drug Administration (FDA) guidance exists specifically for pharmacological intervention studies for hearing loss or tinnitus (1). However, reproducible, validated human research methodologies of the highest rigor must be employed. At a minimum, all PIHL or otoprotective clinical trials should meet International Conference on Harmonization guidelines for clinical trials (2). Additionally, test equipment, test environment, clinical procedures, and personnel must meet all relevant American Speech-Language Hearing association (ASHA), American Academy of Audiology (AAA), American National Standards Institute (ANSI) standards and guidelines, and military standards where applicable. The following sections are intended to provide researchers in the PIHL field with guidance for the clinical study design elements which may eventually lead to more homogeneic and, therefore, comparable study data and outcomes.
GENERAL TEST ENVIRONMENT
Audiologic testing should be conducted in a quiet testing environment meeting appropriate current standards. If testing is conducted inside a health care facility, additional requirements exist that must be met to maintain Joint Commission or other accreditation related to patient safety. This document focuses on recommended technical requirements.
SOUND FIELD REQUIREMENTS
Audiologic testing should be conducted inside a sound booth. Single-walled or double-walled sound booths may be appropriate depending on the ambient noise levels outside the booth, with double-walled booths providing increased attenuation relative to single-walled booths. Sound levels inside the sound room should be verified annually, or whenever any new noise source is introduced within the vicinity of the audiometric room. It is important that hearing measurements be conducted in an audiometric test room meeting specific standards because, if the ambient noise level exceeds allowable standards, the measured thresholds will be inaccurate because of the masking phenomenon. Excluding all noise from a sound room environment is generally not feasible nor required. A variety of standards specify the maximum allowable noise levels for accurate threshold assessment. In many cases, the appropriate standard will be ANSI S3.1, most recently revised in 2013 (3). The current version of all ANSI standards can be obtained from the Acoustical Society of America (ASA) at http://acousticalsociety.org/standards(4).
ANSI S3.1 specifies the maximum permissible ambient noise levels (MPANLs) allowed during audiometric testing to prevent excessive ambient noise masking (3). This standard is particularly relevant for clinical trials employing normal hearing listeners who may have thresholds at, or below, 0 dB HL (i.e., 0 to −10 dB HL). Importantly, ANSI S3.1 specifies different MPANLs for different test frequencies as some frequencies are more vulnerable to masking by ambient noise (3). Different MPANLs are also specified for different earphone types because some earphones attenuate ambient noise more effectively than others. Within this standard, MPANLs are specified for octave and one-third octave band intervals from 125 to 8000 Hz for the audiometric conditions of testing with ears covered using supra-aural headphones or insert earphones, as well as in the free field (ears not covered). In some settings, other standards may be more appropriate. During audiometric testing conducted as part of a hearing conservation program under the requirements of 29 CFR 1910.95, the permitted noise levels in the audiometric room are increased relative to the ANSI S3.1 (3,5). If testing is being conducted using a mobile van service, a sound-level meter will be required to verify sound levels meet the relevant standards once the van has arrived at the test location. Sound level meters (SLMs) must also meet the relevant specifications (6) (American National Standards Institute, ANSI 2014). Specific performance categories define Type 1 and Type 2 SLMs, with greater measurement precision required for Type 1 (±1 dB) than for Type 2 (±2 dB) devices. A Type 2 meter is the minimum requirement for noise measurements for Occupational Safety and Health Administration (OSHA) purposes (5). We recommend that annual professional calibrations be completed using Type 1 meters.
EQUIPMENT STANDARDS, INCLUDING CALIBRATION
Test equipment, test environment, and test procedures must meet all relevant national standards and guidelines in the United States or the national equivalent in other countries. A variety of sources of information are available as either standards or guidelines. ANSI has established standards for audiometer performance (7) as well as tympanometer performance (8).
ANSI S3.6 provides specifications and tolerances for audiometers (7). An important component of ANSI S3.6 is standard reference threshold levels for different earphones, as well as bone vibrators and loudspeakers. Reference equivalent threshold sound pressure level values specify the values used to convert sound levels from dB SPL to dB HL. Sound levels produced by earphones coupled to an audiometer are measured in dB SPL using microphones during the calibration process. These physical measurements are converted to dB HL, the normative reference population threshold, using reference equivalent threshold sound pressure levels, such that thresholds are consistently expressed in dB relative to the expected population threshold of 0 dB HL. Audiometers must be calibrated according to ANSI S3.6 at least annually, as well as anytime the equipment has been moved, to assure that threshold assessments performed on the same individual with different audiometers provide equivalent results across test sessions, and accurately represent differences between the individual's thresholds and the normative reference thresholds (7). For clinical trial purposes, it is critically important that measurements across time accurately reflect changes in function. If clinical testing includes acoustic immittance or otoacoustic emission measurements, these devices must conform to the appropriate ANSI S3.6 specifications (7).
In addition to annual professional calibration, device users should perform a daily listening check to verify equipment function. Listening checks should be used to confirm that tones are not distorted, that they are produced by both earphones and from the correct earphone, and that no background noise exceeds limits. Wires should be manipulated to ensure no intermittency that would suggest a broken wire with sporadic connection. The results of the listening check should be documented on a daily basis. If the listening check is to conform with 29 CFR 1910.95, a listener with normal, stable thresholds should be tested daily to verify that thresholds are within ±5 dB of the expected thresholds (5). Deviations of 10 dB or greater indicate additional calibration is needed before use of the equipment. It is also appropriate to use a “bio-acoustic simulator” in place of a human listener as it will produce stable thresholds with no human variability. A helpful overview of calibration requirements is provided in the newly released Hearing Conservation Manual (9).
Tympanometric testing is used to assess the functional status of the middle ear, in particular, the tympanic membrane and middle ear ossicles, by measuring the sound level changes of a 226 Hz acoustic stimulus as a function of dynamic pressure changes in the ear canal. Data generated include middle ear pressure (MEP), peak compensated static acoustic admittance (Ytm), and equivalent ear canal volume (Vea) measurements. ANSI S3.39 provides specifications and tolerances for devices that are intended to be used for measurement of acoustic impedance, acoustic admittance, or both, based on a probe-tone frequency of 226 Hz (8). Calibration of devices according to ANSI S3.39 assures that acoustic-impedance or acoustic-admittance measurements will be equivalent when tests are repeated across time, or repeated using different devices (8). Although the standard is based 226 Hz probe tones, this limitation is explicitly not intended to inhibit or restrict the development or incorporation of new features including the use of probe tone frequencies other than 226 Hz. The use of higher frequency probe tones for clinical diagnosis is clearly of increasing interest (see, for example, Carazo and Sun (10)). However, this ANSI standard does not establish normative values for human ears. Jerger and Keith (11) advocate the use of large populations for establishing reference ranges, and normative data are available in other sources.
CLINICAL TRIALS: PERSONNEL, INCLUDING LICENSURE AND CERTIFICATION
When designing and conducting a clinical trial, a critical consideration is for the investigator to carefully select and monitor the personnel required to perform the work. Personnel are often the most expensive component of a clinical study, yet personnel costs can potentially be modified more easily than other fixed study costs, such as space and equipment. Unfortunately, investigators often underestimate the total number of personnel, the time commitment of the personnel to execute the clinical study, and the human resources issues such as training, vacation time, possible turnover during the study, and sick leave. This serious mistake is common, despite the fact that study success is highly dependent upon on the motivation, effectiveness, expertise, and constant availability of the personnel workforce. These considerations apply even when the support required is minimal in nature. Over-reliance on individuals who have other full time or even part-time jobs with no official protected time commitment to the study is another common mistake of principal investigators. In military environments, military personnel generally cannot receive additional pay for their assistance in a clinical trial because they are already fully funded federal employees. Thus, a clinical trial can constitute an extra uncompensated workload for them in addition to their regular duties. It is critical to have committed and funded individuals to successfully navigate the “trials” of any PIHL clinical trial as documented in this article.
To conduct this type of a trial, certain skill sets among study personnel are required as detailed below. While it may seem economically feasible to budget and plan for individuals to fill multiple roles, it is highly recommended that PIHL studies are planned with a minimum of one individual with distinct experience and expertise in each of these areas although some may be part time (e.g., the PI or statistician). With each personnel type listed, we will define the jobs they can perform and the necessary certification and/or licensure.
The Principal Investigator (PI) has overall responsibility for all elements of the clinical trial including study design, regulatory reviews and approvals, study implementation, recruitment, enrollment, study procedures (from consent to experimental elements), scientific analysis, and scholarly presentation as well as any reporting requirements associated with institutional, regulatory, or funding agencies. Given this range of responsibilities, the PI must have the proper background, experience, time, and credentials to carry out these many different roles. For clinical trials in hearing loss, this individual is usually an Otolaryngologist (with potentially a further sub-specialization in Otology/Neurotology) or a PhD (in Audiology or Neuroscience).
Lead Study Coordinator or Project Director
Most investigators experienced in clinical trials will agree that the individual most important to the success of the clinical trial is a good Clinical Research Coordinator or Project Director. This lead individual is responsible for all the administrative elements of the trial including compiling study submission documents such as the Investigator's Brochure, Clinical Protocol, and Case Report Forms, monitoring study document submissions and Institutional Review Board (IRB) approval, implementing all monitoring systems for the study, recruiting and tracking study participants, and, in many cases, staff management. If the clinical trial is a multicenter study, the responsibilities of the lead Clinical Research Coordinator or Project Director or study director would involve coordinating these and a myriad of other elements at each separate site. Preferably, these individuals will have a Masters or higher degree in Public Health (MPH), nursing, or other related/relevant field with additional training specifically in research and study coordination, design, and monitoring and certification from a nationally recognized organization such as the Society of Research Administrators International, Association of Clinical Research Professionals, or The Society of Clinical Research Associates.
The audiologist is the backbone of any clinical study involving hearing and is responsible for working with the PI and study coordinator on study design and implementation, and is the individual with day-to-day oversight of all hearing measures or auditory assessments in the study. The audiologists will be responsible for all testing and for recording the audiologic data files for all patients enrolled in the study. Given that study outcomes are dependent upon data collected by the audiologists, it is critical that these individuals have a significant amount of experience with clinical and research audiology and have appropriate time to dedicate to the project. In general, individuals in this role will have a doctorate in audiology (AuD); however, if they were licensed before the adoption of the AuD as the terminal degree in audiology, they may have a master's degree in audiology combined with completion of a fellowship from a graduate program accredited regionally and by ASHA. In many cases, a PhD audiologist may be desirable for at least the lead audiology role to ensure that the rigorous standards of research are appropriately focused upon and prioritized. However, the PhD audiologist should also have a solid clinical background including not only clinical training and licensure but also extensive clinical experience. In the US, every state has an audiology licensure, so employing licensed audiologists is required for clinical practice and should be standard for clinical trials. For other countries, testing should be conducted by the equivalent personnel in the host country recognizing that in other countries, trained audiometric technicians may be the most skilled and appropriate personnel delivering the audiologic standard of care in that country. Further, it is incumbent on the study team to assure that all staff responsible for audiological testing in foreign-speaking nations both fully understand the protocol and ensure that language barriers do not introduce study variables.
In addition to a team of audiologists that collect the audiometric data, PIs may choose to rely on audiology-trained technicians to carry out certain day-to-day elements of the study under the supervision of the study audiologist and PI. These individuals should be dedicated to the study, and if the individual is part-time, the portion of their time devoted to the study must be carefully specified and monitored. Certification for this individual can be difficult to accurately define. Local IRB certification requirements may vary by institution and by study type (i.e., single versus multicenter), local and national audiology requirements, and device specific certification. State licensure laws for audiologists and audiology technicians may also be a factor. The number and qualifications of the individuals required on the study team will depend on the study size and the timing of interventions/observations and other human resources factors such as vacation time and overtime calculations.
Certain study design elements, such as subject screening for further audiology follow-up testing or longitudinal study follow-up, may warrant testing performed by technicians who are certified by the Council for Accreditation in Occupational Hearing Conservation and under the supervision of an audiologist. For studies including military personnel as subjects specifically for noise-induced hearing loss monitoring, the overarching regulation for testing is the Department of Defense Instruction (DODI) 6055.12 Hearing Conservation Program (HCP) dated December 3, 2010 (12). Section 9(b) (1) states that all hearing conservation audiometric surveillance testing shall, “Be performed by a licensed audiologist, otolaryngologist, or other qualified physician; or by a technician who has attended training approved by the Council for Accreditation in Occupational Hearing Conservation or equivalent military training. A technician who performs audiometric tests shall be responsible to an audiologist, an otolaryngologist, or other qualified physician.”
Depending on the size of the study and the elements of audiology testing/equipment being used, an additional study technician or technicians may be necessary to help with study procedures other than audiologic testing. While using technicians is not often necessary, a good technician with a variety of skill sets can keep equipment up and running and the study on target in a variety of situations. However, for studies involving more challenging audiologic test populations such as pediatrics, geriatrics, disabled or ill patients, testing should be performed by an audiologist for US studies or the most qualified individual appropriate in other countries.
A study statistician is a critical element of any study team and is involved in study design, determination of statistical methods for analyzing the data, power analysis, and the scholarly presentation and publication of the study data. If not a biostatistician by trade and degree, this individual should at least have extensive research experience in hearing as well as an advanced degree with a focus or emphasis in biostatistics, which may be represented across a variety of scientific disciplines.
CLINICAL TRIALS: PROCEDURES FOR PURE-TONE THRESHOLD TESTING
Regardless of study design, audiological exams ought to include a detailed history including any preceding noise exposure and the use of hearing protective devices (HPDs), and whether additional complaints of aural pain, fullness, pressure, or tinnitus are present. A physical examination including otoscopy needs to be conducted with the appearance of the ear drum and any defects or abnormalities of the external ear canal noted. Evidence of middle ear disease including tympanic membrane perforation, retraction, or other deformity requires further medical evaluation by an otolaryngologist. An audiometric examination should include tympanometry, a non-invasive test of middle ear function, and conventional pure-tone air and bone conduction audiometry: behavioral tests of hearing threshold that can identify and characterize conductive, mixed, and/or sensorineural hearing loss, should be completed.
Pure-tone air-conduction threshold testing should be conducted both at baseline and after the noise exposure or within 2 to 4 weeks post noise, to determine whether a temporary (TTS) or permanent threshold shift (PTS) exists. Pure-tone air-conduction testing should be conducted at 0.5, 1, 2, 3, 4, 6, 8 kHz. Several studies have included threshold tests in the extended high frequency (EHF) range (>8 kHz); the utility of EHF testing in PIHL trials has not been confirmed and is not recommended as an essential part of PIHL clinical trials. However, if time and budget allow, EHF testing could become an interesting element for meta-analyses in the future. Bone conduction testing should be conducted at 0.5, 1, 2, 3, and 4 kHz if the pure-tone air-conduction threshold at that frequency is greater than or equal to 15 dB HL.
Pure-tone threshold testing should be conducted using the modified Hughson Westlake procedure (13,14) as follows: initial descent towards threshold is accomplished in 10-dB steps. Beginning with the first non-response, level is increased by 5-dB for each non-response, and decreased by 10-dB after each correct detection response. Threshold is defined as the lowest level at which two responses are obtained out of three presentations on an ascending run.
At the baseline visit, pure-tone air-conduction testing should be immediately repeated at 1 and 2 kHz to determine that the subject provides reliable responses. Responses are considered reliable if retest thresholds at both frequencies do not exceed ±5 dB of the previously obtained threshold response. This method of verifying threshold reliability in clinical populations is based on ototoxicity monitoring protocols described by Fausti et al. (15) and Campbell et al. (16).
The timing of the baseline and follow-up tests may vary by study. Clinically, for acute acoustic trauma inclusion studies, an individual presenting clinically with a complaint of hearing loss, aural pain, or tinnitus needs to be evaluated as soon as possible to determine the extent of the injury and to provide a diagnosis and prognosis.
Although not measures of auditory threshold, otoscopy, and tympanometry, just before each hearing assessment, are advisable to rule out possible outer or middle ear abnormality. Unidentified outer or middle ear disorders may cause air-conduction threshold abnormalities or fluctuant hearing thresholds unrelated to noise exposure. Thus, subjects presenting with potential conductive disorders should be excluded from participation in PIHL clinical trials to avoid incorrect data interpretation later.
Tympanometry should be measured with a standard 226 Hz probe tone generally with a pressure sweep from +200 to −400 daPa. Standardized criterion for pass/fail should be set for the clinical trial.
For clinical trials, or to standardize procedures across sites, tympanometry screeners can provide fast, easy, and standardized measures with a simple “pass” or “fail” determination. These screeners are particularly useful, if audiometric technicians are employed in the data collection process.
DETERMINATION OF TTS, PTS, AND OTOPROTECTION
Pure-tone audiometry is the criterion standard by which sensorineural hearing loss especially NIHL is determined. The conventional audiogram illustrates hearing thresholds at several different test frequencies from 0.5 to 8 kHz in each ear, independently. A typical audiogram from a patient with NIHL will show a greater loss of hearing sensitivity at 3, 4, or 6 kHz versus 0.5, 1, 2, and 8 kHz; this configuration is commonly referred to as a “notched” audiogram. A temporary change in hearing sensitivity or threshold shift (TTS) occurs and recovers within hours to days of the noise exposure, typically 16 to 48 hours (17). During this period, one's ability to discriminate speech might be impaired and the perception of tinnitus is sometimes reported. A permanent change or loss in hearing sensitivity or threshold shift (PTS) is confirmed with an audiometric re-test at least 3 weeks, preferably 30 days, after the noise exposure. The majority of temporary loss resolves within a few days after noise exposure, but hearing can continue to gradually improve. The risk of PTS usually begins with hearing loss persisting at 14 days after noise exposure with the upper limit of recovery being 30 days. However, TTS and PTS outcomes may vary by noise insult and patient factors. Before each audiologic assessment for PTS, the patient should have a quiet period or lack of noise exposure in the 16 to 24 hours immediately preceding the audiometric examination but that quiet period before testing is not required to determine TTS. TTS (H93.24) and PTS or NIHL (H83.3) have specific ICD-10 classifications and should be coded appropriately. Additional comorbidities such as tinnitus (H93.1) or hyperacusis (H93.23 a sensitivity to sound) may also be reported by the subject and should be documented as well.
A significant and clinically relevant loss of hearing occurs when a ≥10 dB loss of hearing sensitivity or threshold shift has been demonstrated, at one or multiple test frequencies. In the work place, OSHA has established the standard threshold shift (STS) criteria, which means that a ≥10 dB average threshold shift occurred at 2, 3, and 4 kHz in the same ear (5). When this shift is confirmed by a re-test, the pure-tone average threshold exceeds 25 dB HL, and the hearing loss is not secondary to recreational noise exposure or other non-occupational factors, then an STS has occurred. STS is considered a significant work related injury (WRI) and must be reported. STS or WRIs are often used to monitor hazardous work areas or conditions and should be used to further improve hearing conservation programs. Unfortunately, many initial STS or PTS that are ≥10 dB are not reported because they do not satisfy this WRI rule. For example, a threshold shift of 0 dB at 2 kHz, 10 dB at 3 kHz, and 15 dB at 4 kHz would average 8.3 dB across 2, 3 and 4 kHz in the same ear; retest would not be required and hearing loss would not be reported. OSHA only requires the reporting of an STS. Changes in 6 kHz hearing are not averaged at all in the determination of an STS, even though TTS and PTS at 6 kHz hearing are frequently observed in noise-exposed individuals (5). The Department of Defense guidelines: Defense Occupational and Environmental Health Readiness System-Hearing Conservation (DOEHRS-HC) guidelines use the same standard for determination of STS as does OSHA. In addition the guidelines provide a guidance for early warning shift STS or early warning flag, which is defined as a 15 dB or greater change at 1, 2, 3, or 4 kHz in either ear (12).
HPDs and other environmental engineering controls are two important and critical instruments or processes of an effective hearing conservation program that reduces an individual's noise exposure and risk of developing either a noise induced TTS or PTS. Any research study aiming to prevent or rescue a noise injury must be able to demonstrate that investigators have taken all precautionary measures, including but possibly above and beyond OSHA mandates, to protect the study participants. OSHA mandates a hearing conservation program including annual audiometric testing be put in place in work environments where the noise exposure reaches or exceeds 85 dBA based on a time weighted average (TWA) over an 8-hour work shift (5). Adequate HPDs must be provided to workers in hazardous or noisy work environments where their individual noise exposure exceeds 90 dBA TWA. Noise exposures or TWAs are determined using sound level meters in the immediate work environment or by personal dosimeters mounted at the shoulder.
For the purposes of this discussion, we will limit the definition of otoprotection to pharmacologic strategies or drugs that aim to reduce, mitigate, prevent or treat a NIHL. Recent research has demonstrated several changes in both sensory and neural processes underlying a noise-induced TTS or PTS. The sensorineural changes observed in the noise-induced TTS and PTS often involve the same cell types such as auditory hair cells and neurons. One difference is that auditory hair cell and neuronal cell death or apoptosis clearly result in a loss of hearing sensitivity, while lesser auditory hair cell and neuronal injury may not result in a loss of hearing sensitivity. At least 50 potential otoprotectants have been tested in several different preclinical models of TTS and PTS (for reviews, see Lynch and Kil (18); Campbell and Le Prell (19); Le Prell and Bao (20)). Typically, these investigational drugs have been injected into laboratory animals such as mice, rats, guinea pigs, or chinchillas, before and after the noise exposure. Only a few of these animal studies have delivered the otoprotective drugs orally and/or after the induction of an NIHL. The drugs that have demonstrated the most effective preclinical otoprotection display either anti-oxidant, anti-inflammatory, or anti-apoptotic properties. While no otoprotective drugs have been FDA approved, several drugs have been tested in man or are in active clinical testing for this indication in an affected population.
Effective otoprotection should be defined as including a significant reduction in the incidence and/or severity of either the TTS and/or the PTS. Several clinically validated measures would be considered important determinants of either a TTS or PTS, such as pure-tone audiometry at 3, 4, and 6 kHz and/or word recognition tests. Although not direct determinants of TTS or PTS, patient reports of tinnitus or hyperacusis may also be a consideration. In the case of an acute or single noise exposure, a reduction in the incidence, severity or duration of the TTS would be considered significant and potentially clinically relevant. For example, a reduction in the incidence of a ≥10 dB threshold shift would be considered significant and clinically relevant, since a 10 dB loss in hearing sensitivity requires a 10-fold increase in sound intensity to evoke an accurate behavioral response using pure-tone audiometry. A 10 dB improvement in hearing sensitivity would potentially allow for better word recognition in the acute period especially in a noisy environment where the signal to noise ratio is low (≤6 dB). A reduction in the duration of the TTS (e.g., from days to hours) would also be considered a clinically relevant improvement in hearing. In the case of chronic or repeated noise exposure that results in a PTS, a reduction in the incidence or severity of the PTS would be considered highly significant and clinically relevant. In the case of chronic NIHL, a reduction in the progression of a PTS would be considered significant and clinically relevant. Acute NIHL (TTS) has been linked to increased or accelerated age-related hearing loss (21), although this outcome is not observed after all exposures resulting in TTS (22). A clinically relevant improvement in hearing or effective otoprotection would reduce the incidence, severity of duration of the noise-induced TTS or PTS, and may have the potential to prevent accelerations in age-related hearing loss.
STUDY DESIGN AND DATA CAPTURE
Experimental design consists of within-subjects serial testing in which baseline standard frequency audiograms (as outlined above) are initially acquired. By comparing similar measures obtained at the end of the study period to the relevant pre-noise exposure or baseline audiogram, reliable noise-induced changes in pure-tone hearing threshold can be identified. Thus, subjects will serve as their own control for identifying a specific change in hearing based on frequency and ear.
The generally accepted criterion standard design to prove safety and efficacy of a pharmaceutical agent is the prospective, double-blind, placebo-controlled randomized controlled trial, reviewed and approved by an IRB for human subjects protections considerations. To accomplish this in a PIHL investigation, several study elements must be considered.
First, the manufacturing of the investigational agent must be carried out with the forethought to fully blind both the study subjects and the investigators. Active and placebo batches need to appear as identical as possible (i.e., size, shape, color, taste). Differences between active and placebo batches perceived by either study subjects or the clinical investigators may introduce a bias, which if detected during the conduct of the study, should be reported. PIs should familiarize themselves with the Current Good Manufacturing Practice regulations (http://www.fda.gov/Drugs/DevelopmentApprovalProcess/Manufacturing/ucm090016.htm) at the forefront of study design (1).
Second, the experimental design itself must be considered. Previous hearing loss clinical trials have employed within-subjects serial testing in which baseline standard frequency audiograms (as outlined above) are initially acquired and then compared with measurements obtained at the end of the study period at a single site. For example, in Kopke et al. (23) investigators determined significant noise-induced changes in pure-tone hearing threshold by collecting a baseline audiogram (per guidance above), followed by 16 days of weapons training, followed by another audiogram in the same booth, with the same equipment, 10 days after the last noise exposure. The goal of this interventional study was to determine if n-acetylcysteine (NAC) could prevent NIHL among US Marine Corps recruits stationed at the Marine Corps Recruit Depot (MCRD) in San Diego, CA. The investigators found that even in monitored HPD-protected subjects, at least 27% of placebo subjects experienced a threshold shift; thus, demonstrating that subjects can adequately serve as their own control for identifying a significant hearing change because of active military training.
In the prevention indication studies to date, the investigators follow cohorts through similar pre-planned noise exposures, using actual or estimated rates of hearing loss based on those specific noise exposures to calculate an appropriately powered clinical trial. However, in a treatment indication study, the investigator will be faced with fluctuating rates of hearing loss, changes in the duration or severity of the hearing loss, and the timing and/or the reporting of the injury by the subject. Therefore, in a trial designed to test a treatment indication, the prospective plan will likely need to expand beyond a single site.
NAC was also tested in a treatment indication study, which involved four US military base installations. This multicenter, prospective, randomized, double-blind, placebo-controlled study began the dosing of NAC or placebo for 7 days within 4 hours of an acute acoustic trauma as documented by audiologic inclusion criteria. The self-report feature of the study design ultimately led to its failure to enroll subjects and the study closed before any significant data was collected. It is imperative that investigators consider such design limitations and employ both reasonable expectations as well as, whenever possible, pilot studies to test the logistics of any study design plan. The success of any prevention or treatment indication trial design will hinge on the rates of injury, the duration of the hearing loss, and timely access to that affected population.
Two additional crucial elements of the study design itself are the noise exposure or otherwise injury-imposing exposure(s) anticipated from which to protect or treat, as well as the timely access to such identified study populations. The primary challenge of the former element is that PIHL research, like other areas of research such as infectious diseases and chemical and biological warfare injury, aims to test a population with a potentially permanent disabling injury (24). Like researchers in anthrax or smallpox, PIHL researchers must consider ethical and feasible limitations of human safety and efficacy trials. If a population is identified which will predictably sustain a permanent or chronic NIHL, there is a moral imperative to introduce protective efforts to prevent it. Moreover, an investigator should not intentionally impose a PTS or chronic NIHL on any human population. Thus, researchers are left searching for plausible-yet-unknown exposures within which to test an otoprotectant. And if permanent injuries are identified through the auditory measurements captured, follow-up procedures must be in place both to treat the studied cohort and to better protect future cohorts within that population. Ideally, to accomplish the former, a computerized study database should be configured so that when the post-noise audiogram is entered in the database, criteria for hearing change will automatically be calculated and flagged, so that the examiner can recheck the frequencies in question and refer for follow-up standard of care for those affected subjects. Appropriate hearing conservation programs ought to be consulted to accomplish the latter.
While identification of a viable study population has its unique challenges in PIHL research, gaining access to that affected population within study design parameters (i.e., hours) may prove to be among the most challenging aspects of the clinical trial. Timing and space requirements for testing as outlined above may not coincide with ideal trial design elements. For example, location of audiometric testing booths may require clinic space far from the location of noise exposure, which may impede ability to test within a certain window pre- or post-exposure. There may also be issues with securing appropriate study testing space in traditional medical treatment facilities, which lack dedicated research space. Further obstacles to population access need to be accounted for during study planning, a particularly important aspect in military population studies where population access may include physical access issues like base installation passes; vulnerable population regulations requiring absent senior ranking officials, and present ombudsman for subject protections; or waivers from subjects’ commanding officers to participate at all.
True to all randomized controlled trials, investigators ought to take care in capturing any confounding elements of noise exposure including concomitant medications, smoking and drinking habits, environmental exposures to solvents or other noise sources, and intermittent use of HPDs. Finally, data capture and quality assurance must be considered. Electronic records may not be available in all cases and anticipating the task of comprehensively combing medical records for pertinent study data elements is critical.
* Test Environment: Maximum permissible ambient noise levels for testing must meet ANSI and OSHA standards. Whether a single or double walled boot is required depends on ambient noise levels.
* Equipment and Calibration: Audiometers, immittance meters, and calibration procedures used must meet ANSI specifications and tolerances. Equipment should be calibrated at least annually in addition to daily listening checks.
* Personnel: Personnel are usually the most expensive component of a clinical trial and most frequently underestimated. The primary personnel for the study includes the principal investigator, the lead study coordinator or project director, the research audiologists, and sometimes study technicians or audiology technicians as appropriate. The specific personnel, including their required licenses and certifications, may vary according to the specific clinical trial site.
* Procedures: The audiometric examination for threshold determination should include otoscopy, tympanometry, and pure-tone air- and bone-conduction audiometry. Procedures and standards are specified by ASHA, ANSI, and depending on the clinical trials site, by OSHA and DOEHRS-HC.
* TTS, PTS and Otoprotection: Specific recommendations for determination and classification of TTS and PTS can vary by agency. Most standards are for PTS only. In clinical trials for otoprotection, no specific guidelines for determination of significant protection exist, and may be tailored to the type of protection under study (e.g., reduction in the duration of TTS or in the incidence of STS for PTS.
* Study Design and Data Capture: The gold standard is the double-blind, placebo controlled clinical trial with each subject serving as their own control relative to baseline hearing thresholds. Pharmaceuticals need to meet FDA good manufacturing practices standards. All elements of the study must meet all human subjects protections including prohibition of deliberately imposing NIHL on any subject where NIHL could have been avoided.
Authors would like to acknowledge the DOD Hearing Center of Excellence for the support of Pharmaceutical Interventions for Hearing Loss (PIHL) working group and publication of this article.