Auditory changes following firearm noise exposure, a review^a)

Firearms produce peak sound pressure levels (dB peak SPL) between ∼130 and 175 dB peak SPL (Murphy and Tubbs, 2007; Beck, 2011; Lobarinas et al., 2016), creating significant risk of noise-induced hearing loss (NIHL) in those exposed to firearm noise (Johnson and Riffle, 1982; Cox and Ford, 1995; Nondahl et al., 2000; Stewart et al., 2001; Moon et al., 2011; Moon, 2007; Heupa et al., 2011; Kirchner et al., 2019; Meinke et al., 2017; Laffoon et al., 2019). NIHL typically presents as an audiometric notch between 2 and 6 kHz (McBride and Williams, 2001). This risk exists not only for individuals using firearms for recreational and/or occupational exposures but also those in bystander locations (Flamme et al., 2011). It is therefore perhaps not that surprising that tinnitus and hearing loss are the top two service-connected disabilities of military personnel following their time in service (U.S. Department of Veterans Affairs, 2020). Service members are exposed to both weapon-fire and to other less impulsive noise sources (Yankaskas, 2013; Jokel et al., 2019). Acute exposure to firearm noise is documented to result in hearing loss in service members and others (Keim, 1970; Segal et al., 1988; Moon, 2007; Moon et al., 2011). Monitoring, prevention, intervention, and treatment for NIHL in U.S. service members generate significant economic costs (Alamgir et al., 2016).

Noise-induced hearing loss is a high-risk occupational hazard for safety officers, similar to service members, exposed to firearm noise among other sources of noise exposure. Police officers were 1.4 times more likely to have NIHL than a control group without reported occupational noise exposure (other municipal or state civil servants from the same city, including administrative staff, research workers, technicians, and janitors; see Lesage et al., 2009). Several previous studies have detected hearing loss rates of 22%–85% in safety officers both within and outside of the United States, with low awareness levels or conservation programs among this population (Singh and Mehta, 1999; Tubbs and Murphy, 2003; Barbosa and Cardoso, 2005; Thomas et al., 2007; Sharif et al., 2009; Lesage et al., 2009; Shrestha et al., 2011; Gupta et al., 2015; Win et al., 2015; Thirugnanam et al., 2017; Sonstrom Malowski and Steiger, 2020). Furthermore, as has been previously documented, the recreational use of firearms, such as that used for hunting or target practice, also presents a similar risk to the shooter (Counter and Klareskov, 1990; Nondahl et al., 2000; Stewart et al., 2001; Stewart, 2002).

The audiogram has been widely used to document significant threshold shift (STS) after noise exposure, although it should be noted that there are a variety of possible operational definitions for STS (for review, see Campbell, 2016; Le Prell, 2021). Regardless of the specific definition, permanent changes in pure-tone detection thresholds are often the result of compromised function of the outer hair cells, although other pathology is also possible, including, for example, damage to the reticular lamina and strial cell density (for review, see Hu, 2012). Changes measured at short post-noise times may reflect temporary threshold shift (TTS) with the potential for recovery as noise injury heals, whereas threshold changes measured at longer post-noise times (typically 2–4 week post-noise) are considered to be permanent threshold shifts (PTS) with little additional recovery expected based on permanent pathology. Both regulatory STS criteria (e.g., OSHA, 1983; NIOSH, 1998) and definitions of STS used by professional organizations (e.g., ASHA, 1994; AAA, 2009) are based on permanent changes in the audiogram.

The high prevalence of PTS/STS in firearm-users, measured using changes in the audiogram, suggests that firearm exposure can cause significant cochlear pathology including outer hair cell loss sufficient to result in permanent threshold shift. However, it is also well known that the effects of noise exposure are cumulative, and some 20%–40% of outer hair cells can be lost before threshold sensitivity is affected (e.g., Hamernik et al., 1989; Hamernik et al., 1996; Chen, 2018; CDC, 2020). The audiogram can therefore miss injury to the ear which might have been detected using other more sensitive assays, particularly when data are collected after a single shooting episode as opposed to measurement of long-term cumulative effects of firearm noise. Distortion product otoacoustic emissions (DPOAEs) and transient evoked otoacoustic emissions (TEOAEs) have been suggested to provide alternative and more sensitive measurement tools for detecting immediate auditory dysfunction following firearm noise exposure (Plinkert et al., 1999; Pawlaczyk-Luszczyńska et al., 2004; Rezaee et al., 2012; Laffoon et al., 2019). Changes in OAEs reflect changes in the outer hair cell active process. Multiple studies have shown OAEs to be affected prior to changes in the behavioral audiogram (e.g., Attias et al., 2001; Lapsley Miller et al., 2006; Nottet et al., 2006; Marshall et al., 2009). Other functional consequences of noise exposure are also possible, including, for example, difficulty hearing-in-noise (HIN), which can occur when hearing is audiometrically within normal limits and has been associated with outer hair cell pathology inferred from OAE deficits (Parker, 2020) as well as neural pathology inferred from evoked potential measurements (Liberman et al., 2016; Grant et al., 2020; Mepani et al., 2020; Mepani et al., 2021). Furthermore, tinnitus is another common clinical consequence of firearm noise exposure (Dancer et al., 1991; Moon, 2007; Wu and Young, 2009; Barreto et al., 2011; Moon et al., 2011; Büchler et al., 2012).

Interest in, and demand for, interventions to prevent and/or treat NIHL in high-risk populations has driven the conduct of several clinical trials investigating prevention of TTS and/or PTS in military personnel exposed to firearm noise (Le Prell et al., 2011; Lindblad et al., 2011; Kopke et al., 2015; Campbell, 2016). The review by Le Prell (2021), which broadly reviewed clinical trial design for NIHL prevention or treatment studies, identified variability in study design across clinical trials evaluating potential NIHL prevention. This review identified variability in how exposure was measured, when exposure was measured, how hearing loss was measured, when hearing loss was measured (relative to exposure), and the population from which hearing loss was measured, in addition to variability in drug-related factors such as drug administered, method of drug administration, onset of drug treatment relative to onset of noise exposure, and duration of drug treatment subsequent to noise exposure.

Interestingly, the majority of PTS prevention studies, and some TTS prevention studies, assessed prevention of firearm-associated NIHL and highlighted the need for better understanding of auditory disorders that may develop prior to changes in the audiogram (see Le Prell, 2021). To better inform the design of future studies investigating firearm-associated NIHL and its prevention, this paper provides a scoping review of the literature on NIHL due to firearm noise exposure, including study protocols, end point definitions, and assessment methodologies. With the identification of overall trends regarding the assessment and interpretation of TTS and/or PTS resulting from firearm-noise, recommendations for an appropriate audiologic test battery to assess NIHL in high-risk populations are proposed, with the goal of improving the sensitivity of both clinical monitoring and clinical trials assessing the prevention of noise injury.

Articles investigating auditory changes following firearm noise exposure were identified using PubMed searches completed with a variety of search terms with searches performed between January 2020 and September 2021. The search included terms such as noise-induced hearing loss and firearms, hearing loss and firearms, auditory change and firearms, auditory dysfunction and firearms, audiometry and firearms, OAEs and firearms, TTS and firearms, STS and firearms, PTS and firearms, and threshold shift and firearms; all of these terms were also searched along with “gun,” and “weapon” in place of firearm. When articles published in the English language were identified, the articles were reviewed to determine if inclusion criteria were met (discussed below) and to identify other relevant references not detected during the PubMed search. All articles meeting inclusion criteria, whether identified in the primary search (PubMed) or during the secondary review of references within the primary literature, were carefully reviewed and specific methodological data extracted. Methodological data extracted and organized in table form included the following variables: publication, threshold shift definition and/or criteria, population and demographic criteria for eligibility, firearm(s) + rounds, hearing protection device(s), environment, primary outcomes (assessments used), secondary outcomes (assessments used), and outcomes (observed auditory change(s)). Criteria for primary and secondary outcomes were established either by being explicitly defined as a primary or secondary outcome within the paper, or was assigned as primary or secondary outcome by the reviewer which depended on the amount of emphasis, which, for most studies, was on the audiogram and OAE data. Data from studies meeting inclusion criteria were entered into the data table (Table I).

The inclusion criteria were as follows: Subjects needed to be enrolled in firearm training, without any type of blast exposure (e.g., improvised explosive devices, flash bangs), as blast exposure can result in mechanical injury to the auditory system (Slepecky, 1986; Roberto et al., 1989; Garth, 1994; Patterson and Hamernik, 1997). The purpose of this inclusion criteria was to limit the search results to sensorineural hearing loss to the greatest extent possible. Second, articles needed to include both a pre-exposure audiogram and a follow-up audiogram that assessed either TTS or PTS. When studies include short-term “TTS” measurements without a later PTS test, it is not clear if the deficits are, in fact, temporary, and in those cases, the authors' TTS nomenclature was used despite uncertainties about recovery. Finally, subjects provided with any type of prophylaxis or therapeutic intervention were excluded. For clinical trial or intervention studies included, only data from placebo groups were reviewed and included in the data table (e.g., Le Prell et al., 2011; Lindblad et al., 2011; Campbell, 2016).

Of 39 articles/reports reviewed, a total of 20 met the criteria for further analysis, and 19 were excluded from the analysis. Table I presents data extracted from studies included, and Table II presents data regarding studies that did not meet inclusion criteria. Of the articles included, heterogeneity of the data prevented a meta-analysis review of hearing loss attributed to firearm noise.

There were 19 articles that did not meet inclusion criteria; some studies were excluded for more than one reason. The most common reason for exclusion was blast exposure in parallel with firearm exposure. Forty-seven percent (9/19) of the excluded articles investigated NIHL associated with blast exposure along with firearms, including that from IEDs, hand grenades, mortars, howitzers, or other weapons (Dancer et al., 1992; Hotz et al., 1993; Munjal and Singh, 1997; Nondahl et al., 2000; Konopka et al., 2005; Muhr et al., 2006; Nottet et al., 2006; Marshall et al., 2009; Kopke et al., 2015). Higher intensity exposures can result in additional mechanical damage (Slepecky, 1986; Roberto et al., 1989; Garth, 1994; Patterson and Hamernik, 1997) beyond that expected with firearm exposure alone.

The second most common reason for exclusion was lack of baseline testing. Forty-two percent (8/19) of the excluded articles did not include audiometric baseline measurements which precludes a true measure of both TTS and PTS (Hotz et al., 1993; Stewart et al., 2001; Stewart et al., 2002; Nottet et al., 2006; Moon, 2007; Bockstael et al., 2008; Moon et al., 2011; Barretto et al., 2011). Some of these studies only included otoacoustic emissions as their primary outcome (Barreto et al., 2011; Bockstael et al., 2008; Hotz et al., 1993), whereas others only compared the test group to a control group (Stewart et al., 2001; Moon et al., 2011). One study included an observation of hearing loss by age group, with no baseline measurements (Stewart et al., 2002), while another only tested participants after target practice if they complained of tinnitus, with no baseline (Nottet et al., 2006).

In addition to the above, one article was excluded based on the statement that subjects did not follow instructions and went shooting prior to baseline measurements, likely compromising the baseline hearing test and skewing the amount of hearing change measured post-shooting (Lindblad and Olofsson, 2003). This study also used Bekesy audiometry for testing, which differs from all other studies that utilized Hughson-Westlake audiometry protocols for estimating thresholds.

Sixteen percent (3/19) of articles were excluded as they reported threshold sensitivity in individuals with long-term firearm noise exposure, rather than threshold shift associated with a discrete firearm exposure (Ylikoski, 1987; Counter and Klareskov, 1990; Ylikoski and Ylikoski, 1994).

Finally, pharmaceutical intervention studies for the treatment and/or prevention of NIHL were part of our exclusion criteria unless a placebo group was included, and data from the placebo group was clearly separated from data from the intervention group. There were no articles excluded on this basis.

Of the 20 articles meeting inclusion criteria, 30% (6/20) explicitly defined the criteria for audiometric TTS or PTS (Dancer et al., 1991; Attias et al., 1994; Plinkert et al., 1999; Dhammadejsakdi et al., 2009; Campbell, 2016; Blioskas et al., 2018), while 70% (14/20) did not provide criteria for what defined a significant shift in hearing. Of those that defined TTS and/or PTS criteria, 20% (4/20) used an audiometric threshold shift of ≥10 dB hearing level (HL) at any frequency (Dancer et al., 1992; Attias et al., 1994; Plinkert et al., 1999; Campbell, 2016). The other two articles defined TTS but not PTS; Blioskas et al. (2018) defined TTS as a shift of 5 dB HL or greater at any frequency, and Dhammadejsakdi et al. (2009) defined TTS as significant if thresholds were greater than 25 dB HL after firearm noise exposure.

Sixty-five percent (13/20) of the articles included OAE measurements. Of the 13 articles that included OAE measurements, only 15% (2/13) provided criteria defining what constituted a significant shift in OAE amplitude following exposure (Plinkert et al., 1999; Balatsouras et al., 2005). Both studies defined significance using a criterion of ≥6 dB SPL shift for DPOAEs and ≥4 dB SPL shift for TEOAEs. All other articles defined any decrease in amplitude as being consistent with potential auditory dysfunction following noise exposure.

Criteria for subject eligibility varied substantially across articles. Sixty percent of the included studies (12/20) included criteria of hearing within normal limits; however, the values used to define normal limits ranged from as low as 5 dB HL (Duvdevany and Furst, 2006) to as high as 30 dB HL (Dancer et al., 1991) on the audiogram. Other criteria commonly reported across studies included a lack of aminoglycoside treatment and/or other ototoxic medications, middle ear pathologies, acoustic trauma, head trauma, tinnitus, neurologic or mental disturbances, and/or a history of excessive noise exposure; some studies required normal kidney function or had other health criteria but this was more common in pharmaceutical studies than non-intervention studies. Surprisingly, there was no eligibility criteria reported for 25% (5/20) of the studies that met our inclusion criteria (Konopka, 2001; Tubbs and Murphy, 2003; Pawalczyk-Łusczcynska et al., 2004; Bapat and Tolley, 2007; Lindblad et al., 2011).

The age range for participants was fairly consistent across studies. Specifically, subjects were within the age range of 17–35 years in 70% (14/20) of studies (Dancer et al., 1991; Attias et al., 1994; Plinkert et al., 1999; Balatsouras et al., 2005; Konopka, 2001; Olszewski et al., 2005; Duvdevany and Furst, 2006; Olszewski et al., 2007; Dhammadejsakdi et al., 2009; Le Prell et al., 2011; Büchler et al., 2012; Rezaee et al., 2012; Blioskas et al., 2018; Saedi et al., 2013). Likewise, it was a consistent finding that most articles included subjects who were enlisted in the military. As shown in Table I, 80% (16/20) of articles included subjects who were in the military. Considerations for research with military personnel are discussed in detail by Hecht et al. (2019).

Firearm type varied across studies. The most common firearms were rifles, used in 70% (14/20) of the included studies (Dancer et al., 1991; Attias et al., 1994; Tubbs and Murphy, 2003; Pawalczyk-Łusczcynska et al., 2004; Olszewski et al., 2005; Duvdevany and Furst, 2006; Bapat and Tolley, 2007; Olszewski et al., 2007; Dhammadejsakdi et al., 2009; Büchler et al., 2012; Rezaee et al., 2012; Saedi et al., 2013; Campbell, 2016; Blioskas et al., 2018). However, the caliber and rifle type (e.g., AK-47 vs M-16 vs H.K. 33) varied across studies. The second most common firearm reported was the automatic machine gun, used in 15% (3/20) of studies (Plinkert et al., 1999; Le Prell et al., 2011; Lindblad et al., 2011). To note, shots fired varied significantly across studies; for example, 420 rifle shots were fired in the Attias et al. (1994) study, ten rifle shots were fired during the Blioskas et al. (2018) study, and five rifle shots were fired in Bapat and Tolley (2007). Sound level measurements were recorded in 80% (16/20) of studies; however, there was variability in terms of measurement weighting and overall sound pressure levels obtained across studies. In general, firearm peak SPL levels ranged from 110 to 172 peak SPL, with the weighting varying between A and C, or weighting not specifically stated. Only three studies reported both external (reference microphone on a stand two meters from weapon) and internal (within the ear canal under the shooters' hearing protection device) sound measurements, at 164–166 and 135–154 dB peak SPL, respectively (Lindblad et al., 2011), 104–127 and 74–107 dBA SPL, respectively (Wu and Young, 2009), and 148 and 172 dB peak SPL, respectively (Tubbs and Murphy, 2003). Interestingly, one of the two studies noted that the variation in the exposure levels was not significantly related to hearing test results (Lindblad et al., 2011).

As shown in Table I, HPDs were used by shooters in 60% (12/20) of the included studies; 5% (1/20) did not report on HPD use. Of the studies that reported participants use of HPDs, earplugs were worn by shooters in 33% (4/12) of the studies, earmuffs were worn by shooters in 42% (5/12) of the studies, earplugs and earmuffs were worn by shooters in 17% (2/12) studies; 5% (1/20) did not report on the type of HPD used. Labeled attenuation varies across HPD, however, measured attenuation values can vary from labeled values. Of the studies that reported attenuation values, levels ranged from 10 to 52 dB (Attias et al., 1994; Tubbs and Murphy, 2003; Bapat and Tolley, 2007; Blioskas et al., 2018); however, it was not clear in all studies if these were measured or labeled. Finally, the environment (e.g., outside, inside, bunker) and physical position (e.g., standing vs laying) of each shooter varied across studies.

Of the 20 articles reviewed, all studies included pure-tone audiometric findings (since this was an inclusion criteria) and 65% (13/20) included both audiometric and OAE results. Of the 13 that included OAE data, 62% (8/13) suggested that OAEs were sensitive to noise-induced outer hair cell damage and/or more sensitive than audiometry. Five authors reported OAE changes following firearm noise exposure (Konopka et al., 2001; Balatsouras et al., 2005; Olszewski et al., 2005; Büchler et al., 2012; Blioskas et al., 2018). Additionally, four studies revealed OAEs to be a more sensitive measure of change as compared to pure tone audiometry (Plinkert et al., 1999; Konopka et al., 2001; Pawlaczyk-Luszczyńska et al., 2004; Rezaee et al., 2012). However, definitions of temporary or permanent threshold shift criteria were often not provided, reducing confidence in conclusions regarding relative sensitivity.

Several authors suggested that OAEs are a more sensitive indicator of auditory change than the audiogram, likely because OAEs are affected initially from less noise exposure than is needed to cause a change to the audiogram. However, similar to the previous observation, few studies provided specific criteria to define a significant OAE shift, whether temporary or permanent (Plinkert et al., 1999; Balatsouras et al., 2005). Decreased OAE amplitudes of any dB SPL were found in 62% of all included articles (8/13) (Konopka et al., 2001; Pawlaczyk-Luszczyńska et al., 2004; Balatsouras et al., 2005; Olszewski et al., 2005; Duvdevany and Furst, 2006; Büchler et al., 2012; Rezaee et al., 2012; Blioskas et al., 2018). A decrease in the average measured OAE amplitude was found in 62% of all studies using OAEs (8/13); however, the average change in OAEs was often less than 3 dB SPL (e.g., Konopka et al., 2001; Pawlaczyk-Luszczyńska et al., 2004; Rezaee et al., 2012).

Pure tone audiometric threshold shifts of 10 dB HL or greater at any frequency following firearm noise exposure were found in at least some participants in 60% (12/20) of studies (Dancer et al., 1991; Attias et al., 1994; Plinkert et al., 1999; Konopka et al., 2001; Balatsouras et al., 2005; Duvdevany and Furst, 2006; Bapat and Tolley, 2007; Dhammadejsakdi et al., 2009; Le Prell et al., 2011; Büchler et al., 2012; Blioskas et al., 2018). There was variability in the percentage of individuals that exhibited threshold shifts of 10 dB HL or greater. For example, Konopka et al. (2001) found shifts greater than 10 dB HL in 100% of participants, Balatsouras et al. (2005) in 50% of participants, Blioskas et al. (2018) in 53% of participants, Bapat and Tolley (2007) in 48% of participants, Dancer et al. (1991) in 38% of participants, Tubbs and Murphy (2003) in 7%–21% of participants based on firearm type, and Dhammadejsakdi et al. (2009) in only 2% of participants.

Only 10% (2/20) of the included studies reported on extended high-frequency audiometry. The largest threshold shifts for extended high-frequency audiometry were reported between 11 and 14 kHz (Plinkert et al., 1999; Büchler et al., 2012). As seen in Table II, several of the articles that were excluded included extended high frequency measures but this measure was not commonly included across studies.

Several regulatory and governmental agencies use defined audiometric shifts on the audiogram as the required method for identifying STS for reporting purposes. As such, there are readily available guidelines that researchers could use when investigating the effects of noise exposure using traditional audiometry. There are challenges to the adoption of such definitions within clinical trials assessing NIHL prevention, however, given the very low rate at which clinical trial participants would, within the time frame of the clinical trial, be expected to accumulate PTS meeting the regulatory STS criteria specified by OSHA (1983) or NIOSH (1998) (for detailed discussion see Le Prell et al., 2019; Le Prell 2021). These challenges have prompted the suggestion that threshold shift of 10 dB HL or greater at one or more frequencies (through 8 kHz) be considered as a primary end point in NIHL prevention trials (see Le Prell, 2021). A major challenge for the replicability of research design across studies and comparisons of drug effects across clinical trials is the variation in operational definitions of STS and/or NIHL.

In addition to inconsistencies in the operational definitions used to document noise injury, reported by Le Prell (2021) for the clinical trial literature, inspection of the broader firearm-induced hearing loss literature also revealed several articles that lacked any specific operational definition for STS and/or NIHL. For example, Büchler et al. (2012) reported that all participants had “normal hearing” prior to firearm noise exposure, defined as less than 25 dB HL, and that 32% of the exposed participants had an audiometric notch consistent with NIHL after exposure. However, it is not clear if the noise notches were present prior to the noise exposure, as the vast majority of participants had thresholds of 25 dB HL or less after the exposure, even at the notch frequencies. If hearing is defined as within normal limits when thresholds are equal to or better than 25 dB HL, a noise notch can still be observed within the 0–25 dB HL range (Sonstrom Malowski and Steiger, 2020). Thus, the level of new NIHL was not clear in Büchler et al. (2012). For some studies, the definition of TTS was provided, however, the specific criteria used to establish the TTS was not provided (Tubbs and Murphy, 2003; Rezaee et al., 2012; Saedi et al., 2013). In total, only 40% (8/20) of studies provided definitions and criteria for TTS.

Differences in the operational definition of STS and strategies for measurement or categorization of NIHL precluded comparisons of hearing loss rate and degree across studies. It would be helpful to have common time points, such as pre-shooting baseline, immediately post-shooting, one-day post-shooting, one-week post-shooting, and four week post-shooting. In addition, it would be helpful if minimal reporting standards included average measured thresholds at baseline, average threshold shift at each tested frequency, the rate at which threshold shifts of 10 dB or greater are observed (through 8 kHz), the rate at which threshold shifts of 15 dB or greater are observed at 0.5, 1, 2, 3, 4, or 6 kHz (NIOSH, 1998), and the rate at which the average threshold shift at 2, 3, and 4 kHz exceeds 10 dB (following OSHA, 1983 and DoD 6055.12). Inclusion of all the above would allow careful comparison of the rate and severity of hearing loss at immediate and longer post-noise times, permitting comparisons of both TTS and PTS.

The current data point to the possibility that OAEs could provide an alternative and perhaps more sensitive measurement tool, relative to the audiogram, for identification of outer hair cell pathology following firearm noise exposure both immediately and at longer post-noise test times (Plinkert et al., 1999; Konopka, 2001; Pawlaczyk-Luszczyńska et al., 2004; Rezaee et al., 2012; Laffoon et al., 2019). Changes in OAEs reflect changes in the outer hair cell active process. A consistent finding from this review was that OAE amplitudes were reduced following noise exposure even when audiometric thresholds were not altered. It is probable that changes in OAEs reflect dysfunction of a subset of the outer hair cells after noise exposure, with the damaged subset not sufficient to alter audiometric thresholds without additional cumulative outer hair cell loss with additional continued exposure to noise. This interpretation is consistent with conclusions from other studies investigating functional consequences of exposure to non-firearm noise exposure (e.g., Attias et al., 2001; Konopka, 2001; Lapsley Miller et al., 2006; Nottet et al., 2006; Marshall et al., 2009).

A limitation of this finding specific to this review is the presence of high variability in what is considered a “significant” OAE amplitude shift. The establishment of recommendations for definitions of significant OAE shifts are needed within hearing conservation research. Guidance on the use of OAEs has been provided for both hearing conservation research (see Konrad-Martin et al., 2012) and also more broadly as a clinical trial outcome measure (Konrad-Martin et al., 2016). However, less than half of the clinical trials on NIHL prevention have included DPOAEs as an outcome measure (for review, see Le Prell, 2021).

One possible barrier to use in NIHL monitoring is the uncertain clinical significance of the OAE shift within the subset of studies investigating effects of firearm noise; it will be important to consider other literature in which this has been more carefully considered. For example, the use of OAEs to monitor significant drug-induced ototoxicity is more advanced (Reavis et al., 2008; Dille et al., 2010; McMillan et al., 2012). At this time, none of the ototoxicity classifications and grading scales (i.e., ASHA STS, CTCAE, TUNE, Brock, Chang, SIOP) include DPOAE measures (for review, see King and Brewer, 2018), but DPOAEs are nonetheless often recommended for ototoxicity monitoring (see Lord, 2019; Paken, 2019) and are routinely included in ototoxicity monitoring programs (see Konrad-Martin et al., 2018). Several investigators have proposed amplitude decreases of 4–6 dB SPL be considered a significant change (for review see Lord, 2019). From a practical perspective, counseling on HPD use when OAE deficits are detected is highly feasible, and could be implemented when OAE deficits first emerge. The broad opportunity for NIHL intervention contrasts with the limited ability to change an ototoxic drug regimen that has life-saving therapeutic purpose, thus, relatively smaller OAE changes may be more relevant for NIHL prevention than ototoxicity monitoring, if test-retest reliability allows smaller changes to be reliably detected.

Based on a review of the NIHL clinical trial literature and factors that affect DPOAE outcomes, Le Prell (2021) concluded that DPOAES were more appropriate as a secondary outcome in clinical trials. In this review, we have focused solely on detection and documentation of the effects of firearm-exposure on hearing drawing on literature outside the context of clinical trials. DPOAEs may have a more substantive role within hearing conservation monitoring programs than in clinical trials as they appear to reveal the earliest injury to the outer hair cells in the inner ear, prior to the development of clinically significant deficits, which are the focus of clinical trials. To be useful in monitoring individual injury for those exposed to firearm noise, it will be critical to account for OAE variability including test/retest reliability, which can be affected by several factors, including probe fit, environmental noise, subject noise, equipment, tester approach, or procedure, for example (Keppler et al., 2010). There is extensive literature pertaining to OAE test/retest reliability (Hoi-Yee and McPherson, 2005; Barboni et al., 2006; Wagner et al., 2008; Keppler et al., 2010). A commonly used clinical “rule of thumb” when measuring OAEs is that test/retest reliability should be replicable within ±5 dB SPL (Hall and Mueller, 1997). Otoacoustic emissions were replicated in 53% (7/13) studies included in this review (Plinkert et al., 1999; Konopka et al., 2001; Olszewski et al., 2005; Olszewski et al., 2007; Büchler et al., 2012; Saedi et al., 2013: Blioskas et al., 2018). Not all studies included a replicability requirement, a few were consistent with an acceptable OAE response if repeatability was 60% or higher (Olszewski et al., 2005; Olszewski et al., 2007; Saedi et al., 2013; Blioskas et al., 2018). Variability across firearm literature presents challenges with trying to define what level of amplitude reduction is considered significant, although the 4–6 dB reductions noted in the ototoxicity literature (for review see Lord, 2019) could be considered. At minimum, the amount of amplitude reduction should fall outside typical test/retest values to reliably define OAE changes as significant.

While research studies show statistically significant shifts in average DPOAE amplitude among groups of subjects, it should be noted that some individuals are more susceptible than others to auditory changes following noise exposure. Marshall et al. (2009) found that subjects with reduced OAEs at baseline were more likely to have an OAE shift following noise exposure. Furthermore, it was found that low-level OAEs indicated an increased risk of future hearing loss by as much as ninefold (Marshall et al., 2009), emphasizing the importance of including OAEs as part of the test battery for studies assessing NIHL. All these factors should be taken into consideration when working with noise-exposed individuals clinically or through research.

There is also a need for further research pertaining to EHFA. In this review, only 10% (2/20) studies included the use of (or a discussion about) EHFA (Plinkert et al., 1999; Büchler et al., 2012). The data supporting the value of EHFA is limited, but expanding, with a small number of additional studies that did not meet the inclusion criteria also reporting EHFA thresholds (Ylikoski, 1987; Konopka et al., 2005; Kopke et al., 2015; Laffoon et al., 2019). EHFA thresholds were considered a useful tool in the long term audiometric evaluation of patients exposed to firearm noise by Büchler et al. (2012). EHFA was more sensitive than traditional audiometry and OAEs in identifying early effects of hazardous noise exposure in a large study investigating the effects of occupational noise exposure (Mehrparvar et al., 2014), leading to the suggestion EHFA provides a means for early diagnosis of NIHL (Mehrparvar et al., 2011). Frequencies affected the most were 14 000 and 16 000 Hz (Mehrparvar et al., 2014). Other studies have reported similar findings (e.g., Ahmed et al., 2001, Somma et al., 2008; Valiente et al., 2015). Thus, it is not surprising that a recent review of 15 published articles concluded populations exposed to workplace noise had significantly higher thresholds compared to a control group at frequencies from 9 to 18 kHz, with the greatest differences at 16 kHz (Škerková et al., 2021).

Although the body of literature is growing regarding the use of EHFA for detection of cochlear pathology in noise-exposed populations, there is a challenge for long-term monitoring in that EHFA thresholds also deteriorate as a function of age, with age as the primary predictor and noise exposure as the secondary predictor of hearing thresholds from 10 to 18 kHz (Ahmed et al., 2001). In addition, EHFA thresholds have significantly greater test-retest variability thresholds in the conventional range (through 8 kHz). Consequently, although significant EHFA threshold differences have been reported for several noise-exposed young adult populations (e.g., Kumar et al., 2017; Le Prell et al., 2013; Liberman et al., 2016; Prendergast et al., 2017; Grose et al., 2017), the small between group differences and large test-retest variation will make it difficult to identify noise-induced EHFA shifts in individual patients. If pre-post measurements are collected after acute firearm noise, shifts that exceed 10 dB HL could be reliably documented. Some studies suggest that while EHFA is useful for early detection of NIHL in younger populations, status of EHFA between exposed and unexposed populations equalizes with age as aging will affect the basal outer hair cells even in the absence of noise exposure (Ahmed et al., 2001; Somma et al., 2008). If a pre-post study design is not possible, the inclusion of age-matched control groups is necessary and these limitations should be taken into consideration.

No studies included any data on speech-in-noise testing with firearm noise exposure from our review. Further research is needed to establish the potential value of including speech-in-noise testing for detecting and tracking auditory changes that may be related to firearm noise exposure. A major question is the extent to which hearing in noise deficits are driven by outer hair cell pathology, neural pathology, or the combination of these two pathologies. Parker (2020) found hearing in noise difficulties to be statistically significantly correlated to DPOAE amplitude, particularly in frequencies affected by noise exposure. This statistically significant correlation was present in both subjects whose hearing was within normal limits and those with minimal SNHL, but it was not observed in subjects with hearing loss that was mild or greater. In contrast, several reports from Liberman and colleagues (Liberman et al., 2016; Grant et al., 2020; Mepani et al., 2020; Mepani et al., 2021) have documented relationships between sound evoked neural potentials (inferred as reflecting neural pathology) and hearing in noise ability, with no statistically significant relationships between DPOAE amplitude and hearing in noise ability.

No studies included any data on sound evoked neural potentials after firearm noise exposure from our review. However, evidence from animal models clearly shows the effects of noise exposure on electrophysiologic measurements even in the absence of PTS (for review, see Liberman and Kujawa, 2017). Changes in auditory evoked potential amplitude have been shown after impulse noise exposure in rodents (Altschuler et al., 2019; Harrison and Bielefeld 2019; Gratias et al., 2021), and seem to be linked with exposure to weapons fire in both Veterans and civilian firearm users (Bramhall et al., 2017). Given the potential for neural injury secondary to firearm exposure, additional research is warranted to understand the relationship between outer hair cell loss, neural damage, and supra-threshold auditory issues (for discussion, see Bramhall et al., 2019). Neural metrics most likely to be associated with auditory dysfunction continue to be a topic of high interest (Liberman et al., 2016; Grant et al., 2020; Mepani et al., 2020; Mepani et al., 2021).

Clinical hearing conservation in occupationally-exposed individuals has largely been based on the principles of OSHA or NIOSH standards. Currently, the primary clinical hearing conservation test battery consists of pure tone audiometry from 500 Hz to 6 kHz and STS is based on detecting and tracking PTS over time at 2, 3, and 4 kHz. In reviewing the literature, 31% (4/13) of the studies that included OAEs suggested that OAEs were more effective at detecting acute noise injury than the traditional audiogram exposure (Plinkert et al., 1999; Konopka, 2001; Pawlaczyk-Luszczyńska et al., 2004; Rezaee et al., 2012), suggesting a possible role for OAEs in monitoring noise injury. In addition, it has been suggested that EHFA thresholds could be monitored to track changes over time, with short-term changes attributable to noise injury and changes over long periods reflecting the combined effects of both aging and noise on EHFA thresholds (Ahmed et al., 2001; Dreisbach et al., 2008; Somma et al., 2008; Mehrparvar et al., 2011; Büchler et al., 2012; Mehrparvar et al., 2014; Laffoon et al., 2019; Jiang, et al., 2021).

Duration and type of noise exposure, hearing protection devices used, subject history of noise exposure, weighting of noise measurement, definition of TTS/PTS, and shift criterion for secondary outcomes should all be considered when developing a NIHL study or hearing conservation plan. As noted above, governmental regulations and standards exist that outline threshold shift criteria for pure tone audiometry. However, research clearly documents that significant damage has already accumulated by the time OSHA's STS criteria are met. While the existing regulations do reliably document STS in the regulated workforce, biomarkers for noise injury such as OAE amplitude and EHF injury are not required to be monitored. A major goal of hearing conservation research is the identification of noise injury metrics that would enable early identification of injury and enable early intervention that protects against further damage and the development of material hearing loss (defined by NIOSH, 1998, as an average hearing test level for both ears that exceeds 25 dB at 1000, 2000, 3000, and 4000 Hz).

In the case of firearm noise exposure, consistent clinical protocols and standardized outcome measures must be developed and implemented so that significant change can be monitored using consistent approaches. In addition, if the goal is early detection and intervention, emphasis must be placed on the detection of any noise-induced shift, mild or severe, with the overarching goal that detection of sub-clinical pathology is used to identify those at risk for permanent damage and in need of hearing protection strategies. Sub-clinical pathology, reflected in DPOAEs, EHFA thresholds, or TTS, should serve as a red flag for intervention and/or modification of currently used conservation protocols.

In regards to the combined audiogram and OAE test battery, correlations between changes in audiometric thresholds and changes in OAE amplitudes have not been found (Lapsley Miller et al., 2006). In their longitudinal study including 338 volunteers exposed to noise over a six month period on an aircraft carrier, average OAE amplitudes reduced significantly; however, there were no changes in average audiometric thresholds following the exposure (Lapsley Miller et al., 2006). However, this is not surprising if OAE deficits precede, rather than parallel, the onset of audiometric threshold shift. This is consistent with several studies identified in this review, with those studies suggesting that outer hair cells are highly vulnerable to noise injury, with audiometric thresholds not shifting until a significant population of outer hair cells have been damaged. Nevertheless, both objective OAE and subjective threshold measures should be included in a standard test battery. The inclusion of routine audiometry through 8 kHz and EFHA, along with OAEs, has been previously recommended to identify cochlear auditory damage that may not be identified using traditional audiometry alone (Dreisbach et al., 2008; Büchler et al., 2012).

Clinical protocols and recommendations should be evidence-based, originating from past and current research and updated when new research results suggest changes are warranted. The current review regarding auditory changes following firearm noise exposure revealed a low level of rigor in that several important factors were not controlled and/or not reported across studies. The lack of rigor decreases confidence in the sensitivity and utility of DPOAEs, EHFA tests, hearing in noise tests, and evoked potential measurements for use as possible routine elements within hearing conservation program measures and clinical trials investigating prevention of firearm noise induced hearing deficits. However, OAEs reflect noise (and drug) induced pathology prior to audiogram changes, and there is also evidence suggesting EHFA is sensitive in revealing noise injury to the basal cochlea even after accounting for age effects. Therefore, incorporating these additional measures into monitoring programs could be valuable, especially in detection of early changes which can allow for early intervention to increase awareness for vulnerable populations. Data from a study enrolling safety officers revealed an overall lack of awareness regarding NIHL (Sonstrom Malowski and Steiger, 2020), paralleling data from other at-risk groups (Chesky, 2011; Feder et al., 2017), and the need for hearing loss prevention programs in high-risk populations has previously been emphasized (Saunders and Griest, 2009). In studies with firearm users, it would be helpful to document knowledge about NIHL and HPDs, access to HPDs, self-efficacy for hearing protective behaviors, and motivation for hearing loss prevention, all of which are necessary for improving health related behaviors (see Fernandez et al., 2009; Portnuff, 2016). Future studies should focus on developing appropriate study design using evidence from prior studies, including well-defined criteria and study outcomes for measuring deficits in auditory function, which should include metrics beyond the audiogram. Long-term, this can help to establish and maintain “best practice” guidelines for patient clinical care and the development of hearing conservation programs for high-risk populations.

This review found significant variability in (1) overall study design, (2) criteria for auditory change(s), and (3) primary and secondary outcomes. Inconsistent test protocols, as well as ambiguous criteria and definitions, created challenges for the planned comparison of data across the literature and compromises planning for future studies. The results of this review suggest current protocols used for the assessment and interpretation of changes in hearing following firearm-induced noise exposure should be reevaluated and consistent standards established. Audiologic protocols should exceed conventional audiometry and include measures that have been previously found to be sensitive to detecting NIHL, including OAEs and EHFA. While TEOAEs and DPOAEs could be useful objective measures for detecting outer hair cell injury as a supplement to TTS and PTS metrics, agreed on definitions for significant OAE shift are needed. Standard definitions for clinically significant shifts generally refer to changes in function that are detectable to the patient and there are no data directly probing the degree of OAE shift that results in patient-reported change in function. Nonetheless, OAE shifts could provide a key biomarker for noise injury, and quantitative, objective biomarkers for NIHL are urgently needed given that significant damage has already occurred by the time the audiogram changes. Early intervention is critical to long-term protection.