
==== Front
Cochrane Database Syst Rev
Cochrane Database Syst Rev
14651858
10.1002/14651858
The Cochrane Database of Systematic Reviews
1469-493X
1361-6137
John Wiley & Sons, Ltd Chichester, UK

38757544
10.1002/14651858.CD015066.pub2
CD015066.pub2
CD015066
Medicine General & Introductory Medical Sciences
Hearing protection field attenuation estimation systems and associated training for reducing workers’ exposure to noise
Cochrane Work Group
Morata Thais C
Gong Wei
Tikka Christina
Samelli Alessandra G
Verbeek Jos H
Division of Field Studies and Engineering National Institute for Occupational Safety and Health (NIOSH) Cincinnati OH USA
School of Medicine, Institute of Public Health and Clinical Nutrition University of Eastern Finland Kuopio Finland
Department of Physical Therapy, Speech-Language-Hearing Sciences, and Occupational Therapy Universidade de São Paulo São Paulo Brazil
Cochrane Work, Department of Public and Occupational Health Amsterdam UMC, University of Amsterdam Amsterdam Netherlands
17 5 2024
2024
17 5 2024
2024 5 CD015066Copyright © 2024 The Authors. Cochrane Database of Systematic Reviews published by John Wiley & Sons, Ltd. on behalf of The Cochrane Collaboration.
https://creativecommons.org/licenses/by-nc/4.0/ This is an open access article under the terms of the Creative Commons Attribution-Non-Commercial Licence , which permits use, distribution and reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes.
Abstract

Background

Global Burden of Disease studies identify hearing loss as the third leading cause of years lived with a disability. Their estimates point to large societal and individual costs from unaddressed hearing difficulties. Workplace noise is an important modifiable risk factor; if addressed, it could significantly reduce the global burden of disease.

In practice, providing hearing protection devices (HPDs) is the most common intervention to reduce noise exposure at work. However, lack of fit of HPDs, especially earplugs, can greatly limit their effectiveness. This may be the case for 40% of users. Testing the fit and providing instructions to improve noise attenuation might be effective. In the past two decades, hearing protection fit‐test systems have been developed and evaluated in the field. They are called field attenuation estimation systems. They measure the noise attenuation obtained by individual workers using HPDs. If there is a lack of fit, instruction for better fit is provided, and may lead to better noise attenuation obtained by HPDs.

Objectives

To assess: (1) the effects of field attenuation estimation systems and associated training on the noise attenuation obtained by HPDs compared to no instruction or to less instruction in workers exposed to noise; and (2) whether these interventions promote adherence to HPD use.

Search methods

We used CENTRAL, MEDLINE, five other databases, and two trial registers, together with reference checking, citation searching, and contact with study authors to identify studies. We imposed no language or date restrictions. The latest search date was February 2024.

Selection criteria

We included randomised controlled trials (RCTs), cluster‐RCTs, controlled before‐after studies (CBAs), and interrupted time‐series studies (ITSs) exploring HPD fit testing in workers exposed to noise levels of more than 80 A‐weighted decibels (or dBA) who use hearing protection devices. The unit 'dBA' reports on the use of a frequency‐weighting filter to adjust sound measurement results to better reflect how human ears process sound. The outcome noise attenuation had to be measured either as a personal attenuation rating (PAR), PAR pass rate, or both. PAR pass rate is the percentage of workers who passed a pre‐established level of sufficient attenuation from their HPDs, identified on the basis of their individual noise exposure.

Data collection and analysis

Two review authors independently assessed study eligibility, risk of bias, and extracted data. We categorised interventions as fit testing of HPDs with instructions at different levels (no instructions, simple instructions, and extensive instructions).

Main results

We included three RCTs (756 participants). We did not find any studies that examined whether fit testing and training contributed to hearing protector use, nor any studies that examined whether age, gender, or HPD experience influenced attenuation. We would have included any adverse effects if mentioned by the trial authors, but none reported them. None of the included studies blinded participants; two studies blinded those who delivered the intervention.

Effects of fit testing of HPDs with instructions (simple or extensive) versus fit testing of HPDs without instructions

Testing the fit of foam and premoulded earplugs accompanied by simple instructions probably does not improve their noise attenuation in the short term after the test (1‐month follow‐up: mean difference (MD) 1.62 decibels (dB), 95% confidence interval (CI) ‐0.93 to 4.17; 1 study, 209 participants; 4‐month follow‐up: MD 0.40 dB, 95% CI ‐2.28 to 3.08; 1 study, 197 participants; both moderate‐certainty evidence). The intervention probably does not improve noise attenuation in the long term (MD 0.15 dB, 95% CI ‐3.44 to 3.74; 1 study, 103 participants; moderate‐certainty evidence).

Fit testing of premoulded earplugs with extensive instructions on the fit of the earplugs may improve their noise attenuation at the immediate retest when compared to fit testing without instructions (MD 8.34 dB, 95% CI 7.32 to 9.36; 1 study, 100 participants; low‐certainty evidence).

Effects of fit testing of HPDs with extensive instructions versus fit testing of HPDs with simple instructions

Fit testing of foam earplugs with extensive instructions probably improves their attenuation (MD 8.62 dB, 95% CI 6.31 to 10.93; 1 study, 321 participants; moderate‐certainty evidence) and also the pass rate of sufficient attenuation (risk ratio (RR) 1.75, 95% CI 1.44 to 2.11; 1 study, 321 participants; moderate‐certainty evidence) when compared to fit testing with simple instructions immediately after the test. This is significant because every 3 dB decrease in noise exposure level halves the sound energy entering the ear.

No RCTs reported on the long‐term effectiveness of the HPD fit testing with extensive instructions.

Authors' conclusions

HPD fit testing accompanied by simple instructions probably does not improve noise attenuation from foam and premoulded earplugs. Testing the fit of foam and premoulded earplugs with extensive instructions probably improves attenuation and PAR pass rate immediately after the test. The effects of fit testing associated with training to improve attenuation may vary with types of HPDs and training methods. Better‐designed trials with larger sample sizes are required to increase the certainty of the evidence.

Plain language summary

Do earplug fit tests plus training on how to fit earplugs help people who work in noisy environments?

Key messages

• Earplug fit testing paired with simple instructions for properly fitting earplugs probably does not improve noise reduction.

• Earplug fit testing paired with extensive instructions probably improves protection against noise compared to receiving no instructions or simple instructions.

Why are noisy work environments a problem?

Noise exposure at work is associated with serious health conditions, from hearing loss and tinnitus (ringing or buzzing sounds in the ears without an external source), to injuries and heart/blood vessel problems. Hearing protection devices (HPDs) – such as earplugs and earmuffs – are commonly used to reduce people’s noise exposure at work. Especially for earplugs, it is well‐known that many workers struggle to insert them properly into the ear canal. Ill‐fitting earplugs will not protect one's hearing sufficiently.

What are hearing protection fit‐test systems?

Hearing protection fit‐test systems are technologies that measure the extent to which noise is reduced (attenuated) at the ear for a person wearing earplugs or another hearing protection device.

What did we want to find out?

We wanted to find out if hearing protection fit testing plus training to wear earplugs (or other protection devices) properly is effective at reducing noise experienced by workers, and whether they continued to wear earplugs consistently. We also wanted to find out if people’s age, gender, earplug type, and experience of using hearing protection devices made any differences to the effectiveness of fit testing and training.

What did we find?

We found 3 studies involving 756 participants. They examined the effects of giving people simple or extensive instructions for fitting their hearing protection devices (all the studies used foam or premoulded earplugs) together with fit tests that measured how much noise was reduced. We did not find any studies that examined whether fit testing and training promoted consistent hearing protection use, nor any studies that examined whether age, gender, or hearing protection experience influenced noise reduction.

Main results

• Earplug fit testing accompanied by simple instructions probably does not improve personal attenuation ratings (PAR) from foam and premoulded earplugs compared to no instructions.

• Fit testing with extensive instructions probably improves workers’ protection against noise, measured immediately after the test and training, compared to having no or simple instructions. An increase in protection of about 3 dB will halve the sound's power at the ear, increasing the protective effect of earplugs.

• The effects of fit testing may depend on the type of hearing protection device and training methods.

• None of the included studies measured or reported harmful effects from fit testing and training, and we did not think it was likely we would find any.

What are the limitations of the evidence?

We have moderate to low confidence in the evidence. Our confidence was reduced because the studies were small, and we had concerns about how participants were allocated to groups.

How up‐to‐date is the evidence?

The evidence is current to February 2024.

These authors contributed equally to this work.

New
==== Body
pmcSummary of findings

Summary of findings 1 Effects of instructions versus no instructions for fit testing HPDs

Population: workers exposed to noise using earplugs
Setting: industry
Intervention: fit testing of earplugs with extensive or simple instructions
Comparison: fit testing of earplugs without instructions	
Outcomes	Anticipated absolute effects* (95% CI)	№ of participants
(studies)	Certainty of the evidence
(GRADE)	Comments	
PAR with fit testing of earplugs without instructions	Effect on PAR with fit testing of earplugs with instructions	
PAR (dB) at short‐term follow‐up: range 1 month to 4 months after simple instructions	The mean PAR (dB) at short‐term follow‐up was 25.92 dB at 1‐month and 24.39 dB at 4‐month follow‐up.	1 month: MD 1.62 dB higher (0.93 lower to 4.17 higher)
4 months: MD 0.40 higher (2.28 lower to 3.08 higher)	285
(1 RCT)	⊕⊕⊕⊝
Moderatea	Fit testing of earplugs with simple instructions probably has no or little effect on PAR (dB) in the short term compared to no instructions.	
PAR (dB) at long‐term follow‐up: mean 12 months after simple instructions	The mean PAR (dB) at long‐term follow‐up was 26.19 dB	MD 0.15 dB higher
(3.44 lower to 3.74 higher)	103
(1 RCT)	⊕⊕⊕⊝
Moderatea	Fit testing of earplugs with simple instructions probably has no or little effect on PAR (dB) in the long term compared to no instructions.	
PAR (dB) at immediate follow‐up after extensive instructions	The mean PAR (dB) at immediate follow‐up was 5.54 dB	MD 8.34 dB higher
(7.32 higher to 9.36 higher)	100
(1 RCT)	⊕⊕⊝⊝
Lowb	Fit testing of premoulded earplugs with extensive instructions may improve PAR (dB) at immediate follow‐up compared to no instructions.	
PAR pass rate (%)	No studies measured or reported PAR pass rate.	
Adverse effects	No studies measured or reported adverse effects.	
Proportion of workers who wear HPDs	No studies measured or reported the proportion of workers who wear hearing protectors.	
*The effect on PAR in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group.
CI: confidence interval; MD: mean difference; PAR: personal attenuation rating	
GRADE Working Group grades of evidence
High certainty: we are very confident that the true effect lies close to that of the estimate of the effect.
Moderate certainty: we are moderately confident in the effect estimate; the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.
Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.
Very low certainty: we have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect.	
aDowngraded by one level due to imprecision: fewer than 400 participants.
bDowngraded by two levels due to study limitations: lack of information on randomisation and imprecision due to fewer than 400 participants.

Summary of findings 2 Effects of extensive instructions versus simple instructions for fit testing HPDs

Population: workers exposed to noise using HPDs
Setting: military
Intervention: fit testing of foam earplugs with extensive instructions 
Comparison: fit testing of foam earplugs with simple instructions	
Outcomes	Anticipated absolute effects* (95% CI)	Relative effect
(95% CI)	№ of participants
(studies)	Certainty of the evidence
(GRADE)	Comments	
PAR with fit testing of foam earplugs with simple instructions	Effect on PAR with fit testing of foam earplugs with extensive instructions	
PAR (dB) ‐ Fit testing of foam earplugs (immediate)	The mean PAR (dB) ‐ Fit testing of foam earplugs was 25.2 dB	MD 8.62 dB higher
(6.31 higher to 10.93 higher)	‐	321
(1 RCT)	⨁⨁⨁◯
Moderatea	Fit testing of earplugs with extensive instructions probably improves PAR (dB) (immediate effect) compared to simple instructions.	
PAR pass rate (%) ‐ Fit testing of foam earplugs (immediate)	50.5%	88.1%
(72.6 to 100.0)	RR 1.75
(1.44 to 2.11)	321
(1 RCT)	⨁⨁⨁◯
Moderatea	Fit testing of foam earplugs with extensive instructions probably increases PAR pass rate (%) (immediate effect) compared to simple instructions.	
Adverse effects	No studies measured or reported adverse effects.	 	
Proportion of workers who wear HPDs	No studies measured or reported the proportion of workers who wear hearing protectors.	 	
*The effect on PAR in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: confidence interval; MD: mean difference; PAR: personal noise attenuation; RR: risk ratio.	
GRADE Working Group grades of evidence
High certainty: we are very confident that the true effect lies close to that of the estimate of the effect.
Moderate certainty: we are moderately confident in the effect estimate; the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.
Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.
Very low certainty: we have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect.	
aDowngraded by one level due to imprecision: fewer than 400 participants.

Background

Noise is an environmental agent that impacts several aspects of health and well‐being. High noise levels can cause irreversible damage to hearing and trigger other auditory symptoms, including tinnitus and hyperacusis (noise sensitivity or reduced tolerance to sound) (Nelson 2005). Less intense noise levels can cause annoyance, sleep disturbance, negative cardiovascular and immunological outcomes, and other stress‐mediated effects in both children and adults (Basner 2014; Morata 2005). Noise has long been recognised as an occupational hazard, even before industrialisation, and evidence of the burden on society it represents has continued to accumulate over time (ACGIH 2018; Teixeira 2021). Noise‐induced hearing loss, increased injury risk (ACGIH 2018), and cardiovascular disease are three of the noise‐related adverse work conditions receiving increased attention (Teixeira 2021).

A 2021 systematic review conducted as part of the Global Burden of Disease (GBD) initiative examined population‐representative surveys on hearing loss prevalence from 1990 to 2019 (GBD 2019 Hearing Loss Collaborators). It estimated that 1.57 billion (1570 million) people (95% confidence interval (CI) 1.51 to 1.64) globally had hearing loss in 2019 alone, which corresponds to one in five individuals (20.3%, 95% CI 19.5 to 21.1). Differences in prevalence between countries were attributed to disparities in healthcare access and quality, but also to the prevalence of occupational noise exposure. By 2019, 7 million years lived with disability (YLDs) (95% CI 4.76 to 10.1) were attributable to occupational noise exposure. The GBD study group has identified noise‐reduction strategies as one of the urgently needed multidisciplinary actions to improve hearing health care (GBD 2019 Hearing Loss Collaborators).

The relevance of hazardous noise exposure in the workplace to the prevalence of acquired hearing loss has been recognised for decades (ISO 2013; Nelson 2005; WHO 1997). This recognition prompted not only research but also the development and implementation of public health policy, and early intervention and prevention programmes (Suter 2017; Themann 2019). Numerous countries require workplaces from different economic sectors (including manufacturing/industry, utilities, transportation, the military, construction, agriculture, and mining) to comply with governmental regulatory requirements to control hazardous noise exposures and implement hearing conservation programmes (Suter 2022). The evidence for reducing risks from various workplace hazards favours controlling the source of exposures (primary prevention). In practice, however, the most common attempt to reduce noise exposure consists of distributing hearing protection devices (HPDs) (Themann 2013a). A 2017 Cochrane review on the effectiveness of interventions to prevent work‐related noise‐induced hearing loss identified strategies that effectively reduced occupational noise exposure and hearing loss: the implementation of stricter legislation, better use of hearing protection devices as part of hearing loss prevention programmes (HLPPs), and training in the proper use of earplugs. However, the overall quality of evidence was very low to moderate, and there was limited follow‐up of participants receiving training for insertion of earplugs (Tikka 2017). The authors concluded that further research is likely to have an important impact and could modify the conclusions reached (Tikka 2017). Our review is an elaboration and update of the fit testing of earplugs with accompanying instructions.

Since it is clear that the fit of HPDs determines to a large extent their noise attenuation, it is necessary to accurately measure the individual noise attenuation (ANSI 2018). That information reveals who needs instructions to improve the fit and, thus, the protection against the adverse effects of noise. Currently, the most accurate technology to measure individual fit and attenuation is the field attenuation estimation system, commonly referred to as HPD fit testing (ANSI 2018; Berger 2022). Since the mid‐1970s, laboratory studies have used many forms of hearing protection fit‐test technology to investigate the actual acoustic attenuation hearing protectors provide, often to inexperienced users. Technology to measure the noise attenuation obtained by individual workers wearing HPDs has become commercially available in the past decade, and research has also looked into the implementation of this technology (Byrne 2018; Voix 2022). A few countries have published standards recommending (Argentina, Australia, Brazil, Italy, Russia, Uruguay, and Venezuela) or regulating (Canada, Germany, Malaysia, USA) the use of fit testing as part of hearing conservation interventions (Wells 2024).

Description of the condition

Long‐term exposure to noise levels greater than 80 A‐weighted decibels (dBA) carries an excess risk of hearing loss (ISO 2018; Prince 1997). (The unit 'dBA' reports on the use of a frequency‐weighting filter to adjust sound measurement results to better reflect how human ears process sound.) The risk increases with noise level and can ultimately lead to hearing impairment. The impairment is characterised by a limitation in situational awareness, as well as in the capacity to engage in conversation in meetings or social activities. Hearing difficulties can therefore become a significant barrier to establishing or maintaining emotional relationships. Such outcomes have prompted studies which reported the association between hearing loss and depression, cognitive decline, increased hospitalisations and healthcare costs, and mortality (Basner 2014; Themann 2019). Workers with hearing loss have higher rates of absenteeism; may be at increased risk of work‐related injuries; and are more likely to be underemployed or unemployed (Dzhambov 2017; Kerr 2017; Themann 2013b). The condition is permanent, and there is no effective treatment for it. However, the risk of noise‐induced hearing loss can be greatly minimised if noise is reduced to below 80 dBA (NIOSH 1998).

Despite decades of study, workplace interventions, and regulations, noise‐induced hearing loss is consistently amongst the most common self‐reported occupational illnesses or injuries (Nelson 2005). Information is also available for self‐reported hearing difficulty and tinnitus amongst workers and non‐workers (Masterson 2016a), as well as disability‐adjusted life years (Masterson 2016b). A 4‐kHz audiometric 'notch', a decrease in hearing acuity at the frequency of 4 kHz, is understood to be one of the first signs of noise‐induced hearing loss (Nelson 2005). Worldwide, 16% of disabling hearing loss in adults is attributed to occupational noise. Leigh and colleagues calculated a global annual incidence of noise‐induced hearing loss of 1,628,000 cases, which means an annual incidence rate of almost two new cases per 1000 older workers (Leigh 1999). The risk for work‐related hearing loss is present in more economic sectors than originally anticipated. As discussed by Themann 2019, industry‐ and occupation‐specific estimates indicate risk in unexpected sectors, such as transportation, warehousing, and real estate. Workers outside manufacturing are rarely offered preventive interventions, such as risk assessment, audiometric testing, or access to hearing protection (Themann 2019). The prevalence of HPD non‐use was 53% amongst all workers in the USA who were exposed to hazardous workplace noise (Green 2021).

When properly selected and fitted, hearing protectors can be an effective means to prevent hearing loss (Tikka 2017). However, many workers are not fully protected when they do not wear their hearing protectors correctly or consistently (Groenewold 2014). Gong 2021a found that, of 503 workers who were exposed to noise levels between 80 and 100 dBA, 28% had insufficient protection because their earplugs did not fit well. After receiving proper earplug fitting instructions, noise attenuation was improved by an average of 15 dB (Gong 2021a).

Description of the intervention

Technically, most hearing protectors can attenuate noise exposure enough to prevent hearing disorders (Voix 2022). A Cochrane review on the effectiveness of interventions to prevent work‐related hearing loss concluded that the use of HPDs in a well‐implemented hearing loss prevention programme was associated with less hearing loss, and that instruction on how to insert plugs into the ears was key to their effectiveness in noise reduction (Tikka 2017).

The noise attenuation provided by HPDs is described by a numerical rating calculated from measurements in a laboratory with human subjects (Berger 2008). Different rating systems are in use around the world. In the USA, the noise reduction rating (NRR) is required to be measured in the laboratory and published on the primary label of every HPD by the U.S. Environmental Protection Agency (EPA 1979). The NRR is then used to assess the adequacy of the attenuation of a hearing protector for a given noise exposure. For example, an HPD with an NRR of 22 dB would reduce an environmental noise level of 100 C‐weighted decibels to 80 dBA (a 2‐dB C‐ to A‐weighted correction is applied, Berger 2008) in the ear, if it worked as properly as in the laboratory. Consequently, in the industrial setting, HDPs are chosen only based on the published NRR value to ensure that efficient noise reduction is possible. In practice, this theoretical reduction is rarely achieved because of improper application of the devices (Berger 1983; Murphy 2011). Earplugs can be especially difficult to insert due to the small dimensions of the ear canal or lack of skills to do it properly (Berger 2008).

In contrast to the NRR reported by the manufacturer of the HPDs, the personal attenuation rating (PAR) is the attenuation of the given HPD achieved by an individual at the moment of fit testing, in the real world (Voix 2022). Different technologies estimate PAR, and results are presented in similar ways. Current methods used in the field to estimate a personal attenuation rating (PAR) are: real‐ear attenuation at threshold (REAT, ISO 2018), microphone in the real‐ear (MIRE, ANSI 2018), and the alternating binaural loudness balance under headphones (Voix 2022). The REAT measures the difference in the minimum sound level one can detect when wearing and when not wearing a hearing protector. The MIRE measures the difference between the sound pressure levels in the ear canal with and without an earplug in place (result expressed as insertion loss) and the difference in sound pressure levels outside or underneath the HPD, referred to as noise reduction. For the alternating binaural loudness balance, the individual is asked to indicate with and without earplugs when the loudness between ears is balanced (Rimmer 1997). Each of the different fit‐test methods combines the attenuation measurements at different frequencies to determine an overall PAR for the individual being tested. Some fit‐test systems may also present visual cues such as pass/fail indicator lights (Voix 2022). The HPD's fit testing per se is aimed at measuring an individual worker's attenuation level with a particular protector. Achieving the desired attenuation requires proper selection and use of HPDs; this knowledge is likely to have motivated the incorporation of instructions/training components to the fit‐test intervention (Voix 2022).

HPD fit testing is neither an invasive technology nor a clinical medical treatment, and the test is usually completed quickly (Federman 2021; Murphy 2007). These factors suggest that there would be no adverse effects from it, but it is not always safe to assume this. There might be a minor risk of fatigue or discomfort completing HPD fit tests. Workers may experience discomfort from notifications about insufficient noise attenuation received from the HPDs they wear. The adverse effects of HPD fit testing cannot be prespecified as they are not currently known or suspected to be associated with the intervention. Therefore, potential adverse effects associated with HPD fit testing still need to be examined.

How the intervention might work

Fit‐test interventions aim to help an individual worker achieve his/her needed personal attenuation level with a particular protector. Instructions will result in a better fit which can be objectively measured, but workers will also gain a sense of what proper insertion feels like and what that level of attenuation sounds like. It has been suggested that these effects may also prompt greater adherence to hearing protection use, as a consequence of the objective results and an increased sense of self‐efficacy (Schulz 2011; Stephenson 2011).

Better noise attenuation whilst wearing hearing protection and better adherence to hearing protection use will eventually lead to less hearing loss amongst workers (Sayler 2018). In a study of 14 metal manufacturing plants, Sayler and colleagues reported that the four plants which spent the most on training and conducted fit testing of hearing protectors for at least five years had the lowest hearing loss amongst workers, measured as percentages of age‐corrected standard hearing threshold shifts. These plants also had the lowest 10‐year average high‐frequency hearing loss rates (Sayler 2018). In a study conducted in the mining industry, an association between PAR values and hearing loss was reported, with greater hearing loss associated with smaller PARs, as measured by fit testing (Ullman 2021).

Why it is important to do this review

The global burden of disease attributed to noise‐induced hearing loss, together with the widespread provision of HPDs to reduce this burden, warrant a systematic review of the effects of HPDs on noise levels at the ear. There is uncertainty regarding the effects of this developing fit‐testing technology and associated training on the attenuation of noise levels at the ear and workers' adherence to hearing protection use. A review of these initial practices can assess their effectiveness, and has the potential to guide future research and the implementation of new standards.

Objectives

To assess: (1) the effects of field attenuation estimation systems and associated training on the noise attenuation provided by HPDs compared to no instruction or to less instruction in workers exposed to noise; and (2) whether these interventions promote adherence to HPD use.

Methods

Criteria for considering studies for this review

Types of studies

Evaluations of hearing loss prevention interventions can be biassed by factors that affect noise exposure (Tikka 2017). Randomisation is the best protection against such bias. Studies in which individual workers are randomised to an intervention and a control group are difficult to perform in the work setting because of the presence of many stakeholders and high levels of social interaction between participants (Tikka 2017). Cluster‐randomisation, in which whole companies or departments are randomly assigned to the intervention and control groups, is a way to replace randomisation at the individual level whilst still leading to bias reduction. Therefore, we chose to include RCTs and cluster‐RCTs.

Given the difficulty of randomisation in the workplace, we also included controlled before‐after (CBA) studies in which the outcome is measured at least once before and once after the intervention in both the intervention and control groups.

With data that are often routinely gathered, interrupted time‐series (ITS) studies are also a feasible non‐randomised study design. Hearing loss data are often routinely collected. Therefore, we also included ITS studies in which outcomes were measured at least three times before and three times after the intervention in one group of study participants (EPOC 2012; Ramsay 2003).

We also considered including uncontrolled before‐and‐after (UBA) studies which compare individual results before and after the intervention (Reeves 2021). However, we felt that the lack of a control group in UBA studies makes it difficult to determine if the effects can be attributed to the intervention. Thus, we decided to consider findings from UBA studies in the Discussion section of the review, but we did not use these data in the review's analyses or conclusions.

We excluded studies whose participants were not routinely exposed to occupational noise in the workplace because we considered they could differ from our target population in multiple ways, including demographics, educational level, awareness of risk, knowledge of hearing protection devices, and motivation to protect hearing.

Types of participants

We included studies conducted with adult workers at workplaces with noise levels of more than 80 dBA as a time‐weighted average (TWA) over an entire work shift, working day, or part of the work shift. The unit 'dBA' reports on the use of a frequency‐weighting filter to adjust sound measurement results to better reflect how human ears process sound. The selected exposure metric identifies populations that are likely to be at risk of hearing loss and likely to be offered the studied intervention in the workplace.

Types of interventions

We included studies that evaluated field attenuation estimation systems with the objective of examining if a needed personal attenuation rating (PAR) was provided by various types of HPDs that workers use. The fit‐testing system should also have included instructions for workers on how to properly select and fit HPDs. We considered any type of instruction sufficient for a study to be included in the review.

We included interventions based on any type of technology (REAT, MIRE, and the alternating binaural loudness balance) to perform fit tests with any type of HPDs (earmuffs, canal caps, foam earplugs, premoulded earplugs, push‐in earplugs, etc).

We considered studies eligible regardless of whether the intervention was part of a work‐related hearing loss prevention programme. We included interventions provided as part of corporate safety and health programmes, by either in‐house personnel or third parties. We also included studies reporting measurements for either a specific facility or a specific type of worker, and regardless of the frequencies measured (Hz).

We compared: (1) fit testing systems that measured noise attenuation at the ear (in dB) with training to fit testing systems without training; and (2) fit testing systems with different levels of training (simple or extensive). We defined extensive training as any training format with an individualised instruction component, and simple training as any training format without an individualised instruction component.

We excluded laboratory studies, as it is difficult to assess if their results can be applied to real‐world situations.

We excluded studies on other types of interventions, such as engineering controls to reduce or eliminate the noise source, changing materials, processes or workplace layout (Cohen 1997), and administrative control measures that involve changing work practices.

Types of outcome measures

We included studies measuring workers' personal noise attenuation obtained through their personal hearing protection using either MIRE, REAT, or loudness balance technologies.

Primary outcomes

Noise attenuation at the ear, measured as an overall PAR. The PAR is calculated as the average noise attenuation, as measured by the fit testing system, for frequencies measured at both ears, weighted according to the A‐filter (a type of frequency weighting filter used to adjust sound measurements to better reflect how human ears perceive noise).

Noise attenuation at the ear, measured as a PAR pass rate. The PAR pass rate indicates the percentage of participants whose HPDs sufficiently attenuated their environmental noise levels.

Any adverse effects of HPD fit testing.

We included studies evaluating immediate (same day), short‐term (up to six months of follow‐up), and long‐term effects (longer than six months of follow‐up).

Secondary outcomes

We planned to include the proportion of workers who wear hearing protectors as a secondary outcome.

Search methods for identification of studies

We conducted systematic searches for RCTs, CBA, ITS, and UBA studies of field attenuation estimation systems. We developed a search strategy based on the concepts of occupational noise exposure, hearing protection, and fit testing. The search strategy is shown in Appendix 1.

We imposed no restrictions on language, publication year, or publication status.

Two authors (TM, WG) searched the reference lists of all the included studies. The databases were last searched on 29 February 2024.

Electronic searches

We searched the following databases:

Cochrane Central Register of Controlled Trials (CENTRAL) via the Cochrane Library (including Cochrane Ear, Nose and Throat Disorders Group's Trials Register and Cochrane Work's Trials Register), from 1800 to 29 February 2024;

MEDLINE (Ovid), from 1946 to 29 February 2024;

Embase (Ovid), from 1988 to 29 February 2024;

PsycINFO (Ovid), from 1806 to 29 February 2024;

CINAHL (EBSCOhost) (Cumulative Index to Nursing and Allied Health Literature), from 1981 to 29 February 2024;

Scopus, from 1788 to 29 February 2024;

NIOSHTIC-2 (National Institute for Occupational Safety and Health), from 1930 to 29 February 2024;

ClinicalTrials.gov, from 2000 (inception) to 29 of February 2024;

World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP), from 2005 (inception) to 29 February 2024.

Searching other resources

An information specialist from the US Centers for Disease Control and Prevention, with experience in systematic reviews, advised on the strategy, search terms used and tailored, and ran all electronic searches.

To identify additional studies (published, unpublished, conference presentations, dissertations, theses, and ongoing trials), we scanned reference lists of identified studies for further papers. Using the same electronic search strategy, we also searched the Trip medical database, National Health Service (NHS) Evidence ‐ Ear, Nose, Throat and Audiology (formerly NLH ENT & Audiology Specialist Library) for systematic reviews, and two authors (TM, WG) scanned their reference lists for additional studies until 29 February 2024. We also searched the Open Researcher and Contributor IDentifier (ORCID), Google Scholar, and ResearchGate pages of principal investigators identified from relevant publications using the same search terms as used for the databases until February 2024. The latest search was on 29 February 2024. We have reached out to Mr Elliott Berger, M.S., who retired from the 3M Personal Safety Division in 2018. He currently serves as the National Hearing Conservation Association historian and maintains a comprehensive list of references on hearing protector effectiveness as of 23 November 2023.

Data collection and analysis

We used Covidence for abstract and full‐text screening, and for collecting data on study characteristics and outcomes extracted from the included studies (Covidence).

Selection of studies

We selected eligible studies in two stages.

First, three review authors (TM, WG, AS) independently screened the titles and abstracts of all studies identified by our systematic search. We excluded studies that either did not fulfil the inclusion criteria or that fulfilled the exclusion criteria. Second, we retrieved the full‐text study reports or publications for all included references. Two review authors (of TM, WG, AS, CT) independently and in duplicate assessed full texts for eligibility. We resolved disagreements through discussion or by consulting a third review author (JV), if required, to make a final decision.

We recorded reasons for exclusion of studies at the full‐text stage. We identified duplicates and collated multiple reports of the same study under one study ID. We recorded the selection process in a PRISMA study flow diagram (Figure 1). When our systematic search identified studies conducted by the authors of this review (AS, WG, TM), these authors did not participate in making decisions on inclusion or risk of bias assessment to avoid conflict of interest.

1 PRISMA study flow diagram

Data extraction and management

Two review authors (WG, TM) extracted study characteristics using the data collection form that was used to produce the Cochrane review 'Interventions to prevent occupational noise‐induced hearing loss' (Tikka 2017).

We collected information on methods, study design and setting (including industrial sector, occupation, and country), participant characteristics (including age, gender, type of workplace), study eligibility criteria, details of the interventions given, outcomes assessed, source of study funding, and any conflicts of interest reported by the investigators. For RCTs, we were interested in the effect of the assignment to the intervention.

Two of three review authors (WG, TM, AS) independently extracted outcome data from each of the included studies. We resolved any disagreements by consensus or by involving a third review author (JV). One review author (WG) contacted the study authors where key information was missing, and transferred the data to Review Manager (RevMan 2024). We double‐checked that data had been entered correctly by comparing the data presented in the systematic review with the study reports. A second review author (TM) spot‐checked study characteristics for accuracy against the trial report.

For studies published in languages other than English, German, Dutch, Chinese, or Portuguese, we planned to arrange for a native speaker or sufficiently qualified person outside the review team to help with data extraction, but there were no such studies.

Assessment of risk of bias in included studies

For randomised studies, we used the updated Cochrane risk of bias tool (RoB 2) (RoB 2 Excel template; Risk of bias tools), following the guidance in the Cochrane Handbook for Systematic Reviews of interventions on assessing the risk of bias for a specific outcome (Higgins 2023). Two authors (WG, JV) assessed the risk of bias independently and in duplicate for intention‐to‐treat effects for the primary 'PAR' and 'PAR pass rate' outcomes, measured using fit‐test technology on immediate (same day), short‐term (until six‐month follow‐up), and long‐term (longer than six‐month follow‐up) time points after the start of the intervention (Higgins 2023). We attempted to resolve any disagreements first through discussion; failing that, we deferred to a third review author (CT).

We graded the risk of bias as high, some concerns, or low for each domain and for an overall risk‐of‐bias judgement for a specific result, using the signalling questions built into the tool. We supported our rating with a study quote where possible and with a justification for our judgement. We summarised the risk of bias judgements for each outcome across all the studies for each of the following risk of bias domains:

bias arising from the randomisation process;

bias due to deviations from intended interventions;

bias due to missing outcome data;

bias in measurement of the outcome;

bias in selection of the reported result.

If information on risk of bias related to unpublished data or correspondence with a trial, we planned to note this in the risk of bias table, but there was no need for this step.

We judged each result to have a high risk of bias overall if at least one domain was judged to have a high risk of bias. If no domain was at high risk but one or more domains were judged to have some concerns, the overall result was judged at some risk of bias. We judged the result to have a low risk of bias overall if all key domains were judged to have a low risk of bias.

We planned to use the Risk Of Bias In Non‐randomised Studies of Interventions (ROBINS‐I) tool to assess the risk of bias in CBA studies (Sterne 2016). We also planned to use the risk of bias criteria developed by the Cochrane Effective Practice and Organisation of Care (EPOC) group to assess the risk of bias in ITS studies (EPOC 2012). However, we did not include any CBA or ITS studies in the review.

We did not plan to assess the risk of bias of the included uncontrolled before‐and‐after studies as we included these in the Discussion section only to compare them with RCTs, given the limitations of their study design and the high risk of bias.

Assessment of bias in conducting the systematic review

We conducted the review according to a published protocol (Morata 2021), and have reported differences between the protocol and review in the appropriate section (see Differences between protocol and review). We included tables that profile the evidence in more detail, reporting study characteristics and the consensus decisions for the signalling questions of risk of bias.

Measures of treatment effect

We entered the outcome data for each study into the data tables in RevMan 2024 to calculate the treatment effects. We used risk ratios (RRs) for dichotomous outcomes (PAR pass rate), and mean differences (MDs) for continuous outcomes (PAR). We would have used standardised mean differences if MDs had been reported on different scales.

If only effect estimates and their 95% confidence intervals (CIs) or standard errors were reported in studies, we would have entered these data into Review Manager using the generic inverse‐variance method (RevMan 2024). We ensured that higher scores for continuous outcomes have the same meaning for a given outcome, and we explained the direction to the reader and reported where the directions were reversed, if that was necessary. If the results could not have been entered in either way, we planned to describe them in the 'Characteristics of included studies' table or enter the data into 'Additional tables', but there was no need to take these steps.

We would have recalculated an SMD into an MD from the pooled effect size using the median standard deviation of the included studies in the formula: (pooled MD = pooled SMD × median standard deviation).

For ITS studies, we would have extracted data from the original papers and re‐analysed them according to the recommended methods for analysis of ITS designs for inclusion in systematic reviews (Ramsay 2003). We would have used the standardised change in level and change in slope as effect measures. However, no ITS studies were included in the review. For CBAs, we would have reported odds ratios (ORs) and MDs or SMDs, using baseline data for comparison.

We provided 95% confidence intervals for all effect sizes.

Unit of analysis issues

For studies that employed a cluster‐randomised design and reported sufficient data to be included in the meta‐analysis but did not make an allowance for the design effect, we would have calculated the design effect based on a fairly large assumed intracluster correlation coefficient of 0.10. We base this assumption of 0.10 as being a realistic estimate by analogy to studies about implementation research (Campbell 2001). We would have followed the methods described in the Cochrane Handbook for Systematic Reviews of Interventions for the calculations (Deeks 2021). However, in this version of the review, there were no such studies.

If cross‐over trials reported continuous outcomes for which the authors did not report a paired analysis, we would have performed a paired analysis based on a reported or imputed correlation between the outcomes of the intervention and the control condition (Higgins 2021). For dichotomous outcomes, we would have adjusted the CIs for the paired analysis according to Elbourne 2002. However, there were no such studies in this version of the review.

Where multiple trial arms were reported in a single trial, we included only the relevant arms. If two or three comparisons were combined in the same meta‐analysis with the same control group, we would have halved or divided the control group into two or three to avoid double‐counting (Analysis 2.1; Analysis 2.2).

Dealing with missing data

If data were missing that we considered important for assessing the results or the risk of bias, we contacted the investigators or study sponsors to verify the key study characteristics and to obtain missing numerical outcome data where possible. We did so for two of the included studies in this version of the review (Federman 2021; Murphy 2007). In cases where this was not possible, we would have calculated the missing data from available information, such as missing standard deviations (SDs) from P values, as described in the Cochrane Handbook (Higgins 2023), but there were no studies where this step was needed.

Assessment of heterogeneity

We assessed the homogeneity of the results of the included studies based on similarity of the study design, population, intervention, comparator, outcome, and follow‐up time.

We considered each study design, as stated in the Types of studies section, as different and did not combine their results.

We considered all workers exposed to noise above 80 dBA as similar enough to combine studies.

We considered all fit‐test systems described in the Types of interventions section as similar enough to be combined. For control interventions, we considered the results separately for studies with no instruction, simple instruction, and extensive instruction.

We considered the follow‐up times of immediate, short term, and long term as different and did not combine study results across these categories.

We would have tested for statistical heterogeneity by means of the I2 statistic, as presented in the meta‐analysis graphs generated by Review Manager (Deeks 2021; RevMan 2024), but we did not do any meta‐analysis. If this test statistic was greater than 50%, we would have considered there to be substantial heterogeneity between studies (Deeks 2021).

Assessment of reporting biases

We planned to assess publication bias through funnel plots (and the tests used for examining funnel plot asymmetry) but there were too few studies to do so.

Data synthesis

We would have pooled data from studies deemed sufficiently homogeneous regarding interventions, participants, and settings, but none of the comparisons had multiple studies. If more than one study provided useable data in any single comparison, we would have performed a meta‐analysis. We would have used a random‐effects model because we believe that the type of intervention and study designs included always lead to heterogeneity.

The primary analysis included all eligible studies. We would have performed sensitivity analyses to show how conclusions might be affected if studies at high risk of bias were excluded, but there were too few studies to conduct this analysis.

Subgroup analysis and investigation of heterogeneity

We planned to carry out the following subgroup analyses: by age, gender, earplug type, and HPD use experience. We planned to use the following outcomes in the subgroup analyses: PAR and PAR pass rate. We planned to use the Chi2 test to test for subgroup interactions in Review Manager (RevMan 2024). However, we did not carry out subgroup analyses because none of the comparisons included multiple studies.

Sensitivity analysis

We would have performed sensitivity analysis defined a priori to assess the robustness of our conclusions. This would have involved studies with low risk of bias versus studies with high risk of bias. However, none of the comparisons included multiple studies.

Summary of findings and assessment of the certainty of the evidence

We created summary of findings tables for the following comparisons:

effects of fit testing HPDs with instructions versus fit testing HPDs without instructions;

effects of fit testing HPDs with extensive instructions versus fit testing HPDs with simple instructions.

We reported all outcomes for these comparisons. We used the five GRADE considerations (risk of bias, inconsistency, imprecision, indirectness, and publication bias) to assess the certainty of a body of evidence as it related to the studies that contributed data for the prespecified outcomes. We used the overall RoB 2 judgements to feed into the GRADE assessment. Two review authors (WG, TM) independently performed the GRADE assessments and compared results. We used the methods and recommendations described in Chapter 12 of the Cochrane Handbook for Systematic Reviews of Interventions (McKenzie 2021), employing GRADEpro GDT software (GRADEpro GDT). We justified all decisions to downgrade or upgrade the certainty of the evidence using footnotes and comments in each table.

We based our conclusions on studies with the lowest risk of bias designs, such as RCTs. If there was no difference in risk of bias across designs, we would have used all the available studies for our conclusions, but there were no non‐randomised studies that fulfilled our inclusion criteria.

Results

Description of studies

Results of the search

Our searches, which were run up to 29 February 2024, yielded 3015 references in total (see Figure 1). We assessed 54 articles in full text.

Three RCTs fulfilled our inclusion criteria (Federman 2021; Murphy 2007; Salmani 2014). These studies were reported in five articles and included a total of 756 participants. One study reported results in three articles (Murphy 2007).

We identified 17 UBA studies with complete datasets (Assunção 2019; Casali 1991; Chiu 2020; Copelli 2021; Federman 2016; Gong 2019; Gong 2021a; Hayes 2022; Kim 2019; Liu 2018; Liu 2020; Martin 2019; Murphy 2016; Park 1991; Takada 2020; Tsukada 2008; Zhong 2023). The results from one UBA study were reported in two articles (Casali 1991; Park 1991). One RCT article reported a UBA component on PAR for the same participants before and after training (Federman 2021). The 17 UBA studies included a total of 2932 participants. We reviewed these studies to support the discussion, but not the conclusions, of this review.

Of the remaining 32 articles, we discarded four articles reporting on UBA trials because their datasets were incomplete (Cassano 2015; Johnson 2011; Kunz 2014; Smith 2014). We listed the remaining 28 studies (28 articles) as 'excluded studies' with reasons.

We did not identify any cluster‐RCTs, CBA or ITS studies in our search results. All included articles were published in English. We did not list any studies as 'awaiting classification'. To the best of our knowledge, there are no relevant ongoing studies.

We did not find any studies that examined whether fit testing and training contributed to hearing protector use, nor any studies that examined whether age, gender, or HPD experience influenced attenuation.

Included studies

Design

Three studies used a randomised design (RCT) in which participants were randomised to different types of instructions (Federman 2021; Murphy 2007; Salmani 2014).

Sample size

Of 807 participants who were recruited in the three RCTs, a total of 756 were eligible to participate. Federman 2021 included 321 participants and Salmani 2014 included 150. Murphy 2007 included 285 workers, of which 209, 197, and 103 participants participated in the one, four, and 12 months of follow‐up, respectively.

Setting

Murphy 2007 was carried out in American manufacturing companies, and Federman 2021 in the US Marine Corps. Salmani 2014 was conducted in Iran, but did not specify the type of industry nor the type of jobs included in the study. Two RCTs reported immediate effects of the intervention (Federman 2021; Salmani 2014), and one RCT reported short‐term and long‐term effects (Murphy 2007).

Participants

All participants were described as being exposed to noise at work.

The participants in Federman 2021 were Marine Corps training recruits. Recruits were exposed during training to rifle noise with an average peak sound pressure level of approximately 165 dB (measured at 15 cm from the shooter's ear) (Federman 2021). The United Auto Workers in Murphy 2007 were exposed to noise levels above 85 dBA. As noted above, Salmani 2014 did not describe the type of industry nor the type of job studied.

Of the three RCTs, only Salmani 2014 reported participants' socio‐demographic information. Participants were between 19 and 39 years old and mostly male.

Interventions

Fit testing HPDs included five independent elements: fit‐test technology, type of HPDs, instruction intensity, the time of follow‐up, and HPDs' noise reduction rating (NRR) label. For the three included RCTs, we found the following combinations of intervention elements.

Fit‐test technology

All studies used the REAT method to measure the effect on personal attenuation ratings (PAR).

Type of hearing protection devices (HPDs)

All studies used earplugs. Studies evaluated the effect of fit testing of foam earplugs (Federman 2021), premoulded earplugs (Salmani 2014), and foam and premoulded earplugs (Murphy 2007).

Instruction intensity

One study compared simple instructions to no instructions (Murphy 2007). One study compared extensive instructions to no instructions (Salmani 2014). One study compared extensive to simple instructions (Federman 2021).

Simple instructions were: receiving a written brochure accompanied by minimal instructions (Murphy 2007), or receiving a small group video instruction (Federman 2021). In the self‐training session, participants practised automatic threshold testing for familiarisation with the testing procedure.

Extensive instructions were: receiving individualised training instructions (Federman 2021), receiving small group video training followed by individualised training (Federman 2021), or 15 minutes of training in the correct methods of wearing the earplugs and self‐fit under the supervision of a trainer (Salmani 2014).

Follow‐up

Two studies measured the outcome immediately after the intervention (Federman 2021; Salmani 2014). One study measured the effect at short‐term (1‐ and 4‐month follow‐up) and long‐term follow‐up (12 months) (Murphy 2007).

Noise reduction rating (NRR)

Studies used HPDs with various NRRs. Values ranged from 22 dB to 37 dB (Table 3).

1 RCT intervention descriptions

Study	Fit‐test technology	HPD	NRR (dB)	Instruction (Intervention group)	Instruction (Control group)	Effect	Outcome	
Federman 2021	REAT	Foam earplug	33	Extensive instructions:
Group 1: One‐on‐one instructions;
Group 2: integrated short video and one‐on‐one instructions	Simple instructions:
Short video instruction	Immediate	PAR, PAR pass rate	
Murphy 2007	REAT	Multi‐earplugs	26 ‐ 33	Simple instructions: Counselling group (minimal instructions)	No instructions	Short term, long term	PAR	
Salmani 2014	REAT	Two types of premoulded earplugs	25 ‐ 30	Extensive instructions:
15‐minute instruction on correct methods of wearing earplugs and wearing earplugs under supervision of a trainer (NRR = 25 dB)	No instructions:
Wearing earplugs without any training, group 1: NRR = 25 dB; Group 2: NRR = 30 dB.	Immediate	PAR	
HPD: hearing protection device; NRR: noise reduction rating; PAR: personal attenuation rating; REAT: real‐ear attenuation at threshold

Outcomes and measures

Primary outcomes

PAR (dB)

All included studies measured and reported PAR (dB).

Murphy 2007 reported the mean PAR levels achieved from each tested earplug. We contacted the author who provided us with the individual PAR values of the participants. We then calculated the mean and SDs of PARs from all tested earplugs for each group at different visits (see Table 4).

In Salmani 2014, the SDs reported were unrealistically small, and did not match with the box plot in the published figure. The study authors did not reply when we asked for clarification. Therefore, we extracted the interquartile ranges from the box plot and divided them by 1.35 to obtain a more realistic estimate of the SDs (Table 4), according to guidance in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2023). We summed the square SD at each frequency by adding them all up, divided the sum by the number of SDs, and then calculated the square root of the result, resulting in an overall SD over all frequencies.

PAR pass rate (%)

Only one RCT carried out by the US Marine Corps reported PAR pass rates determined by the target 25 dBA pass criterion (Federman 2021). This means that the PAR had to be at least 25 dBA to attain sufficient attenuation for the environmental noise levels to which the workers were exposed.

Adverse effects

We did not locate any studies which evaluated or mentioned potential adverse effects of the studied interventions.

2 Recalculation of study data for review results and meta‐analysis

Murphy 2007#	
Test	Group	n	Recalculation PAR (dB)	
Mean	SD	
Initial test	Group B	153	27.57	10.5	
 	Group C	132	26.63	8.9	
1‐month follow‐up test	Group B	100	25.92	9.0	
 	Group C	109	24.30	9.8	
4‐month follow‐up test	Group B	103	24.39	9.7	
 	Group C	94	23.99	9.5	
12‐month follow‐up test	Group B	49	26.19	8.7	
 	Group C	54	26.04	9.9	
Salmani 2014	
Study data	Published PAR (dB)	Recalculation PAR (dB)	
Group	Mean	SD	Mean	SD	
Group 1	5.5	1.5	5.5	2.4	
Group 2	13.9	2	13.9	2.8	
Group 3	11.3	2.45	11.3	3.2	
#Description of the recalculation methods can be found at Included studies (Results > Description of included studies > Outcome measures > Primary Outcomes).

PAR: personal attenuation rating; SD: standard deviation

Secondary outcome

Proportion of workers who wear hearing protectors

We did not find any studies that examined whether fit testing and training had an effect on the proportion of workers who wear hearing protectors.

Excluded studies

We excluded 28 studies because they did not meet all the inclusion criteria (Biabani 2017a; Biabani 2017b; Byrne 2011; Cassano 2013; Davis 2011; De Vito 2017a; De Vito 2017b; Edwards 1978; Edwards 1979; Gaudreau 2008; Giardino 1996; Gibbs 2017; Gong 2020; Green 1989; Kabe 2012; Lee 2020; Lempert 2010; Mlynski 2014; Murphy 2013; Neitzel 2006; Pfeiffer 1992; Rocha 2016; Samelli 2015; Samelli 2018; Smoorenburg 1982; Ullman 2021; Upegui‐Rincon 2019; Wu 2016). Most of them compared the PARs or noise attenuation, as reported by the participants, with the HPD manufacturers' noise reduction ratings.

We excluded one study because the participants were not routinely exposed to occupational noise in the workplace (Murphy 2013). We excluded two cross‐sectional studies which were carried out at a public university. The experimental group was composed of 40 workers while the control group consisted of students and workers (Samelli 2015; Samelli 2018). We excluded the remaining 25 studies because they did not provide measurements from workers for PAR and PAR pass rate both before the intervention and after the intervention.

Risk of bias in included studies

The results of the risk of bias assessment per domain are presented in Figure 2 and Appendix 2.

2 Risk of bias graph: review authors’ judgements about each risk of bias domain presented as percentages across all outcomes from the included RCT studies

Randomisation process

All included RCTs reported that the participants were randomised into intervention and control groups. One study used a random number table (Salmani 2014); however, the study authors did not explain how this was used. Murphy 2007 reported using a random card draw. When contacted, the authors explained that after volunteers signed an informed consent agreement, they drew a card from a deck which was shuffled by the researchers. Federman 2021, the Marine Corps study, did not report which randomised method was used. When contacted, the study authors explained that they screened 821 participants in order to identify people who failed to achieve the minimum target PAR of 25 dB (Federman 2021). Participants who failed were randomly assigned to three test groups based on the penultimate digit in the participants' social security number (Social Security Bulletin 1982).

None of the included studies reported allocation concealment. However, the authors of two studies confirmed that the allocation sequence was concealed until the participants were enroled and assigned to intervention groups (Federman 2021; Murphy 2007).

Only one study reported no or little difference in age or gender between the intervention and control groups (Salmani 2014). We identified no or little difference in the baseline PAR levels between the intervention and control groups in all three included RCTs.

Deviations from intended interventions

Given the nature of field study with different instruction methods or different types of tested HPDs, it is not possible to blind participants or people delivering the intervention. One study did not report if participants in the intervention group read or followed the written instructions they received (Murphy 2007), which led to some concerns about participants adhering to the intervention. However, it was unlikely that the deviation from the intended interventions arose because of the trial context.

Missing outcome data

All studies reported that the outcomes were available for all participants (Salmani 2014), or nearly all (Federman 2021; Murphy 2007), which led to a low risk of bias in this domain.

Measurement of the outcome

Two studies reported how the blinding of outcome assessors was done (Federman 2021; Salmani 2014). Furthermore, all studies applied the same measurement technology to all randomised participants and followed the same measurement procedure. Since this was an objective measurement and did not include a subjective element, we judged all studies to have a low risk of bias.

Selection of the reported results

The HPD fit‐test studies evaluated the effectiveness of HPDs in reducing workers' exposure to noise. Protocols for Federman 2021 and Salmani 2014 were not published, which led to some concerns over the reported results possibly deviating from the protocol.

Overall bias

We judged that the results for the comparison of fit testing plus simple instructions versus fit testing without instructions at different follow‐up times had low risk of bias overall. The two other comparisons (extensive instruction versus none; extensive versus simple instruction) were judged as having some concerns about risk of bias overall.

Effects of interventions

See: Table 1; Table 2

Comparison 1. Fit testing HPDs with instructions versus no instructions

See Table 1.

Primary outcomes

Personal attenuation rating (PAR) (dB)

Simple instructions versus no instructions (earplugs, short‐term follow‐up)

HPDs: multiple types of earplugs (26 to 33 dB NRR)

Participants: n = 285

Results: fit testing earplugs with simple instructions probably has little or no effect on PAR in the short term compared to no instructions. PAR values were 1.62 dB higher after one month (95% CI ‐0.93 to 4.17; Z = 0.21, P = 0.21; 1 study, 209 participants) and 0.40 dB higher after four months (95% CI ‐2.28 to 3.08; Z = 0.29, P = 0.77; 1 study, 197 participants) for workers receiving simple instructions compared to workers receiving no instructions (moderate‐certainty evidence; Analysis 1.1). We downgraded the certainty of evidence by one level because of imprecise results from a small sample size (fewer than 400 participants) (Murphy 2007).

Simple instructions versus no instructions (earplugs, long‐term follow‐up)

HPDs: multiple types of earplugs (26 to 33 dB NRR)

Participants: n = 103

Results: fit testing earplugs with simple instructions probably has little or no effect on PAR values in the long term compared to no instructions. After 12 months, PAR values for workers receiving simple instructions for fit testing earplugs were 0.15 dB higher compared to those receiving no instructions for fit testing earplugs (95% CI ‐3.44 to 3.74; Z = 0.08, P = 0.93; 1 study, 103 participants; moderate‐certainty evidence; Analysis 1.1). We downgraded the evidence by one level to moderate certainty because of imprecise results from a small sample size (fewer than 400 participants) (Murphy 2007).

Extensive instructions versus no instructions (earplugs, immediate follow‐up)

HPDs: premoulded earplugs (25 dB NRR)

Participants: n = 100

Results: fit testing premoulded earplug with extensive instructions for better fitting (Table 3) compared to no instructions may increase the mean PAR by 8.34 dB (95% CI 7.32 to 9.36; Z = 15.99, P < 0.001; 1 study, 100 participants; low‐certainty evidence; Analysis 1.1). We rated the certainty of evidence as low because of lack of information about randomisation and imprecise results from a small sample size (fewer than 400 participants) (Salmani 2014).

Personal attenuation rating (PAR) pass rate (%)

We did not locate any studies which evaluated this outcome.

Adverse effects

We did not locate any studies which evaluated or mentioned potential adverse effects of the studied interventions.

Secondary outcome: proportion of workers who wear hearing protectors

We did not find any studies that examined whether fit testing and training affected the proportion of workers who wear hearing protectors.

Comparison 2. Fit testing HPDs with extensive instructions versus fit testing with simple instructions

See Table 2.

Primary outcomes

Personal attenuation rating (PAR) (dB)

Extensive instructions versus simple instructions (earplugs, immediate follow‐up)

HPDs: foam earplugs (33 dB NRR)

Participants: n = 321

Results: one study with two experimental arms evaluated the immediate effect of extensive instructions, and found an increase in the mean PAR of 8.62 dB compared to simple instructions (95% CI 6.31 to 10.93; Z = 7.31, P < 0.001; 321 participants; moderate‐certainty evidence; Analysis 2.1). We downgraded the evidence by one level due to imprecision from having fewer than 400 participants (Federman 2021).

Personal attenuation rating (PAR) pass rate (%)

Extensive instructions versus simple instructions (earplugs, immediate follow‐up)

HPDs: foam earplugs (33 dB NRR)

Participants: n = 321

Results: one study with two experimental arms found that the PAR pass rate probably increases for fit testing foam earplugs with extensive instructions compared to simple instructions (risk ratio (RR) 1.75, 95% CI 1.44 to 2.11; Z = 5.72, P < 0.001; 321 participants; moderate‐certainty evidence; Analysis 2.2). We downgraded the evidence by one level due to imprecision from having fewer than 400 participants.

Adverse effects

We did not locate any studies which evaluated or mentioned potential adverse effects of the studied interventions.

Secondary outcome: proportion of workers who wear hearing protectors

We did not find any studies that examined whether fit testing and training affected the proportion of workers who wear hearing protectors.

Discussion

Summary of main results

Effects of instructions versus no instructions for fit testing HPDs

Fit testing with simple instructions probably has no or little effect on PAR at short‐term follow‐up (1 month: MD 1.62 dB, 95% CI ‐0.93 to 4.17; 4 months: MD 0.40 dB, 95% CI ‐2.28 to 3.08; 1 study; moderate‐certainty evidence; Table 1) or at long‐term follow‐up (MD 0.15 dB, 95% CI ‐3.44 to 3.74; 1 study, moderate‐certainty evidence; Table 1). It is possible that this finding resulted from there being little room for improvement in PAR amongst the study participants. They already had very high PAR results (of 24 dB) at baseline.

We found that fit testing with extensive instructions compared to no instructions may immediately increase the mean PAR from premoulded earplugs in workers by 8.34 dB (95% CI 7.32 to 9.36; low‐certainty evidence; Table 1). This is significant because every 3 dB decrease in noise exposure level halves the energy entering the ear.

RCTs on the short‐ and long‐term effects of fit testing of HPDs with extensive instructions versus fit testing of HPDs without instructions are missing. No studies reported adverse effects of the interventions.

Effects of extensive instructions versus simple instructions for fit testing HPDs

Fit testing of foam earplugs with extensive instructions probably increases PAR (MD 8.62 dB, 95% CI 6.31 to 10.93; moderate‐certainty evidence; Table 2) and probably improves PAR pass rates (RR 1.75, 95% CI 1.44 to 2.11) at immediate follow‐up compared to simple instructions (moderate‐certainty evidence; Table 2).

RCTs on the short‐term and long‐term effects of fit testing HPDs with extensive instructions versus fit testing HPDs with simple instructions are missing. No studies reported adverse effects of the interventions.

Overall completeness and applicability of evidence

We did not find RCT studies that evaluated the effects of fit testing ear plugs with extensive instructions on short‐ or long‐term follow‐up. Two UBA studies suggested an effect could be maintained for up to a year, even when instructions are not repeated during that time period (Kim 2019; Liu 2018). However, from other UBA studies, it seems that the interventions achieve lower PAR values at longer follow‐up times. UBA studies indicated that repeated fit testing of HPDs with extensive instructions over time may maintain the achieved attenuation from use of HPDs. RCTs would be needed to evaluate the effects of periodic testing and training. It might be that refreshing only the training component is sufficient to maintain or improve the PAR. Three UBA studies reported that when the same participants received extensive instructions periodically (either weekly, biannually, or annually), mean PAR improved incrementally by an average of 2.4 dB (95% CI 1.04 to 3.75) per training period (Assunção 2019; Kim 2019; Martin 2019).

Studies did not report environmental noise levels very well. The measurement of noise levels in the work environment can be challenging and biassed by many factors, such as the worker, the task, and operational and environmental variables. We believe studies should strive to better measure and describe noise exposure, as it will determine the needed attenuation of HPDs.

The issues of previous training experience or experience of using earplugs were not reported very well in studies. Across studies, we noted very different initial attenuation values at the initial testing (Federman 2021; Murphy 2007; Salmani 2014), which may be due to different experience in use or training. Future studies should report this. Studies also did not report age and gender very well. In general, elderly people and women are reported to be less risk‐taking and better at following guidance (Rolison 2014). Future studies should better report on participants' gender and age.

None of the included studies measured or reported any adverse effects. Even though it is difficult to conceive adverse effects from this intervention, user satisfaction can indicate if there are any problems. We recommend that this should be measured in future studies.

Certainty of the evidence

Given the difficulty of conducting randomised studies in workplaces, we were happy to see three properly conducted RCTs in this field. It shows that it is possible to conduct high‐quality research in this field.

Since individual factors, such as the workers' attitude towards HPD use (Smith 2014), personal preference, and comfort (Samelli 2018), can have an important effect on the outcome, it is important to minimise the possibility of baseline differences between groups. Randomisation is the only way to ensure this equivalence. Authors we contacted indicated that a main obstacle to implementing randomisation was obtaining agreement from companies to have workers fit tested without giving the results to a subgroup of those workers. This confirms both that substantial effort is needed to implement a randomised design and that it is still possible to do so.

Publication of an a priori study protocol differed between studies. One RCT had a published study protocol (Murphy 2007). Federman 2021 did not report a study protocol, but when contacted, the authors indicated that they did have a protocol which was submitted to obtain funding, and that there were no differences in study design between the protocol and the study. Salmani 2014 did not report sufficient detail on the randomisation process and apparently lacked a study protocol. We recommend publishing study protocols for future studies.

Potential biases in the review process

Even though we made significant efforts to search databases that would contain grey literature (such as NIOSHTIC-2), and to contact the Open Researcher and Contributor IDentifier (ORCID), Google Scholar, and ResearchGate pages of principal investigators identified from relevant publications, we did not have access to a few conference proceedings. Therefore, it is possible that we missed some studies on HPD fit testing. Some eligible study authors did not provide information we requested, which could have led to selection bias.

Agreements and disagreements with other studies or reviews

Comparison with UBA studies

We located 17 UBA studies which measured PAR, PAR pass rates, or both, before and after the intervention for workers exposed to noise in different industries or military settings (Table 5).

3 Uncontrolled before‐and‐after studies

HPD	Intervention	Before situation (control)	Follow‐up	Study	MD PAR value / RR PAR pass rate	
1. Change in PAR after fit testing of various HPDs with extensive instructions at immediate, short‐ or long‐term follow‐up. Before: no instructions.	
Foam earplugs	Extensive instructions: individualised training	No instructions	Immediate	Copelli 2021;Federman 2021; Gong 2019; Gong 2021a; Kim 2019	PAR MD 11.5 dB higher, P < 0.00001	
Immediate	Chiu 2020; Federman 2016; Gong 2019	PAR pass rate RR 1.59, P = 0.08	
6‐month	Gong 2019	PAR MD 10 dB higher, P < 0.00001
PAR pass rate RR 2.08, P < 0.00001	
12‐month	Kim 2019	PAR MD 2.8 dB higher, P = 0.002	
Premoulded earplugs	Extensive instructions: individualised training	No instructions	Immediate	Gong 2021a; Liu 2020; Zhong 2023	PAR MD 12.7 dB higher, P < 0.00001	
6‐month	Liu 2020	PAR MD 8 dB higher, P < 0.00001	
12‐month	Liu 2020	PAR MD 6 dB higher, P < 0.00001	
Multi‐earplugs	Extensive instructions: individualised training	No instructions	Immediate	Hayes 2022;	PAR MD 15.3 dB higher, P < 0.00001	
Gong 2021a; Hayes 2022; Murphy 2016	PAR pass rate RR 1.49, P < 0.0001	
Earmuffs	Extensive instructions: individualised training	No instructions	Immediate	Liu 2018	PAR MD 5 dB higher, P < 0.00001
PAR pass rate RR 1.13, P = 0.001	
6‐month	Liu 2018	PAR MD 2 dB higher, P = 0.008	
12‐month	Liu 2018	PAR MD 3 dB higher, P = 0.006	
2 Change in PAR after fit testing of HPDs with extensive instructions at immediate follow‐up. Before: simple instructions.	
Foam earplugs	Extensive instructions: manufacturers' package instructions, expert demonstration, and self‐training	Simple instructions: manufacturers' package instructions
Simple instructions: manufacturers' package instructions	Immediate	Assunção 2019; Takada 2020	PAR MD 3.23 dB higher, P < 0.0001	
PAR pass rate RR 1.39, p < 0.0001	
3‐week	Park 1991	MD 12 dB higher, P= 0.01	
6‐month	Assunção 2019	PAR MD 1 dB higher, P = 0.55	
PAR pass rate RR 0.94, P = 0.80	
Premoulded earplugs	Extensive instructions: manufacturers' package instructions, expert demonstration, and self‐training (premoulded earplugs)	Simple instructions:
Manufactures' package instructions	Immediate	Assunção 2019; Takada 2020	PAR MD 2.59 dB higher, P <0.0001	
PAR pass rate RR 1.29, P = 0.0001	
3‐week	Park 1991	PAR MD 7.7 dB higher, P = 0.02	
6‐month	Assunção 2019	PAR MD 0.3 dB higher, P = 0.86	
PAR pass rate RR 0.80, P = 0.30	
Unknown type of earplugs	Extensive instructions: individualised training	Simple instructions: group training	Immediate	Tsukada 2008	PAR MD 4.4 dB higher, P = 0.007	
PAR pass rate RR 1.58, p = 0.003	
2‐month	Tsukada 2008	PAR pass rate MD 2.81 dB higher, P = 0.03	
PAR pass rate RR 1.35, P = 0.06	
Earmuffs	Manufactures' package instructions, expert demonstration, and self‐training	Manufactures' package instructions and self‐training (ear muffs)	3‐week	Park 1991	MD 1.60 dB higher, P = 0.39	
Canal caps	Manufactures' package instructions, expert demonstration, and self‐training	Manufactures' package instructions and self‐training (ear muffs)	3‐week	Park 1991	MD 3.5 dB higher, P = 0.29	
3 Change in PAR after fit testing of HPDs with periodically repeated extensive instructions. Before: initial one‐time extensive instructions	
Foam earplugs	Periodical extensive instructions	Initial extensive instructions	Weekly	Martin 2019	PAR MD 1.79 dB higher, P < 0.0001	
6‐month	Assunção 2019	PAR pass rate RR 1.05, P = 0.77	
12‐month	Kim 2019	PAR MD 7.39 dB higher, P = 0.001	
Premoulded earplugs	Periodical extensive instructions	Initial extensive instructions	Weekly	Martin 2019	PAR MD 1.21 dB higher, P = 0.005
RR 0.92, P = 0.52	
6‐month	Assunção 2019	
4 Change in PAR after fit testing of HPDs without instructions at short‐ or long‐term follow‐up. Before: fit‐testing of HPDs with extensive instructions	
Foam earplugs	Follow‐up without instructions	Initial extensive instructions	6‐month	Gong 2019	PAR MD 4 dB lower, P < 0.00001	
12‐month	Kim 2019	PAR MD 1.57 dB lower, P = 0.12	
Premoulded earplugs	Follow‐up without instructions	Initial extensive instructions	6‐month	Liu 2020	PAR MD 4 dB lower, P < 0.00001	
12‐month	Liu 2020	PAR MD 7 dB lower, P < 0.00001	
Unknown type earplugs	Follow‐up without instructions	Initial extensive instructions	2‐month	Tsukada 2008	PAR pass rate RR 0.86, P = 0.21	
Multi‐earplugs	Follow‐up without instructions	Initial extensive instructions	12‐month	Murphy 2016	PAR pass rate RR 0.53, P = 0.0005	
Earmuffs	Follow‐up without instructions	Initial extensive instructions	6‐month	Liu 2018	PAR MD 2 dB lower, P = 0.001	
12‐month	Liu 2018	PAR MD 2 dB lower, P = 0.03	
HPD: hearIng protection device; MD: mean difference; PAR: personal attenuation rating; RR: risk ratio

There were 2923, 440, and 159 workers who attended the initial test, short‐term follow‐up fit tests, and long‐term follow‐up tests, respectively. Fifteen of 17 evaluated immediate effects (Assunção 2019; Chiu 2020; Copelli 2021; Federman 2016; Federman 2021; Gong 2019; Gong 2021a; Hayes 2022; Kim 2019; Liu 2018; Liu 2020; Murphy 2016;Takada 2020; Tsukada 2008Zhong 2023). Seven studies evaluated the short‐term effects (Assunção 2019; Gong 2021a; Liu 2018; Martin 2019; Park 1991; Takada 2020; Tsukada 2008). Four studies evaluated long‐term effects (Kim 2019; Liu 2018; Liu 2020; Murphy 2016). We could deduce three interesting PICOs (population, interventions, comparators and outcomes) from the UBA studies that both compared well to the RCTs and provided information that was not available from the RCTs.

The first comparison was the evaluation of the change in PAR after fit testing with extensive instructions, when previously no instructions had been given. In the immediate follow‐up, in all studies, this statistically significantly increased the mean PAR from foam earplugs, with MDs ranging from 2.2 to 17.9 dB (Copelli 2021; Federman 2021; Gong 2019; Gong 2021a; Kim 2019), and also increased the PAR pass rates, with RRs ranging from 1.07 to 2.56 (Chiu 2020; Federman 2016; Gong 2019). The same results were found for premoulded earplugs, with MDs ranging from 12 to 15 dB (Gong 2021a; Liu 2020; Zhong 2023). Liu 2018 also observed an improvement in mean PAR values from earmuffs of 5 dB (95% CI 3.96 to 6.01 dB). This compares well to the difference in PAR values in the RCT included in this review (Salmani 2014), which found an MD of 8.34 dB (95% CI 7.32 to 9.36) from the comparison between extensive and no instructions.

The second interesting comparison was the evaluation of the change in PAR immediately after workers received extensive instructions, when previously they had received only simple instructions. This improved the mean PAR for foam earplugs, with MDs ranging from 2.5 to 3.6 dB, and for premoulded earplugs, with MDs ranging from 2 to 3 dB (Assunção 2019; Takada 2020; Tsukada 2008). Another study reported a change in PAR after fit testing with extensive instructions amongst workers who previously had only received simple instructions after short‐term follow‐up (Park 1991). The mean change in PARs was 12 dB, 7.7 dB, 1.6 dB, and 3.5 dB for foam earplugs, premoulded earplugs, earmuffs, and canal caps, respectively. This compares well to the difference in PAR values in the RCT included in this review (Federman 2021), with MD 8.62 dB (95% CI 6.31 to 10.93) from the comparison between extensive and simple instructions.

The third interesting comparison that was not available in the RCTs included in this review was the change in PAR after workers received periodic, extensive instructions, where previously they had had one initial session of fit testing with extensive training. Three UBA studies reported a change in mean PAR from foam and premoulded earplugs of 2.4 dB (95% CI 1.04 to 3.75; I2 = 66%) after weekly‐, biannually‐, or annually‐repeated fit testing with extensive instructions (Assunção 2019; Kim 2019; Martin 2019). This stresses the need for repetition of the instruction because the effect may wear off. That the effect wears off over time is demonstrated by six UBA studies that reported that mean PAR values generally decreased after the initial fit test with extensive instruction when the instructions were not repeated (Gong 2019; Kim 2019; Liu 2018; Liu 2020; Murphy 2016; Tsukada 2008).

Comparison with other reviews

A 2017 Cochrane review on the effectiveness of interventions to prevent work‐related hearing loss evaluated the noise attenuation effects of HPDs (Tikka 2017). Two RCTs were included and reported that fit testing improved the noise attenuation from HPDs. However, one of the studies was interpreted incorrectly and did not randomise workers to fit testing with instructions (Park 1991). We could only use this study as a UBA study after receiving further information from the authors.

A narrative HPD review summarised at least 23 available studies providing measurements of real‐world HPD attenuation (Berger 1996). More recently, Berger and Voix updated the analysis by removing results considered to be no longer relevant and by adding two more recent studies (Berger 2022). The two reviews compared the difference between the noise attenuation received in real workplaces and the NRR values on the labels of the HPDs. They found that 84% of workers received only approximately a 7‐dB to 15‐dB attenuation from their HPDs, much less than the labelled NRR values, ranging from mid‐20 dB to low 30 dB. The objectives and inclusion criteria of these reviews were different from ours because we evaluated the effects of fit testing with training.

Authors' conclusions

Implications for practice

We found that hearing protection device (HPD) fit testing associated with simple instructions probably does not improve the personal attenuation rating (PAR) for foam and premoulded earplugs in the short and long terms, compared to no instructions. Therefore, in the workplace, simply providing HPDs with no or with simple instructions probably does not ensure that noise‐exposed workers achieve the needed levels of protection from the HPDs they wear.

Fit testing earplugs accompanied by extensive instructions probably improves the PAR values from foam and premoulded earplugs by 8.6 decibels (dB) (95% confidence interval 6.3 to 10.9 dB) compared to no instructions at immediate follow‐up. This is a relevant increase in protection: every 3 dB decrease in noise exposure level halves the energy entering the ear. Workers who are using foam and premoulded earplugs whose fit is insufficient probably need extensive instructions on how to wear their earplugs correctly to achieve sufficient noise reduction from the earplugs they wear.

There were no studies that evaluated workers' adherence to wearing HPDs, and there was no evidence on how often the test and training would need to be repeated to maintain the results.

Implications for research

To increase the certainty of the evidence, more randomised controlled trials (RCTs) of fit testing HPDs with instructions on how to improve their fit are needed. Preferably, first the fit test should be performed, and then workers who have insufficient attenuation should be randomised to receive simple instruction, extensive instruction, or no instruction. This will avoid a ceiling effect from workers who already have sufficient attenuation, as noted in one of the included RCTs (Murphy 2007). In addition, RCTs that evaluate if and how often fit tests should be repeated are needed. It is conceivable – and supported by uncontrolled studies – that the effect of instructions wears off and should be repeated. This review shows that RCTs are possible and feasible. Given that RCTs have a lower risk of bias than other study designs, updates of the current review should only include RCTs.

Better characterisation of noise exposure conditions of the participants would allow the extrapolation of findings across studies. Noise levels and work tasks can vary widely, which suggests different needs in terms of noise attenuation. Earplugs can become "loose" with time and require periodic re‐inserting (Nélisse 2012), but, so far, few field studies have examined possible fluctuations in PAR values over work shifts (Gong 2023; Nadon 2021; Wu 2016). In theory, current technologies could address this issue by providing continuous monitoring of an individual’s actual noise exposure with HPD. In practice, however, this approach has had limited implementation so far (Voix 2022).

To effectively facilitate the implementation of HPD fit‐testing interventions, studies analysing the cost‐effectiveness of HPD fit testing associated with training, as well as evaluating the HPD fit‐testing integration into hearing loss prevention programmes are especially needed. Studies on the effects of fit‐test interventions should not necessarily be restricted to hearing outcomes. One study reported that using three HPD fit‐test frequencies (500, 1000, and 2000 Hz) was sufficient for accurately estimating workers' PAR (Federman 2016), and some studies reported that different PARs were estimated using different test methods (Kabe 2012; Neitzel 2006; Samelli 2015). Therefore, studies examining the specifics and suitability of test protocols are also needed.

Research is needed to address the quality concerns identified in the present review (randomisation, detailed description of the study population, of their exposures, testing procedures, and publication of a protocol before data collection). High‐quality studies would offer greater certainty on implementation strategies that could lead to the protection of workers from noise exposure.

Currently, adverse effects are generally not recorded in the literature. This may be because investigators lack a mechanism to identify them. Future studies should report any adverse effects if they happen.

History

Protocol first published: Issue 10, 2021

Risk of bias

Risk of bias for analysis 1.1 Personal attenuation rating (PAR; dB) (immediate, short‐term, long‐term)

Study	Bias	
Randomisation process	Deviations from intended interventions	Missing outcome data	Measurement of the outcome	Selection of the reported results	Overall	
Authors' judgement	Support for judgement	Authors' judgement	Support for judgement	Authors' judgement	Support for judgement	Authors' judgement	Support for judgement	Authors' judgement	Support for judgement	Authors' judgement	Support for judgement	
Subgroup 1.1.1 Simple instructions versus no instructions, earplugs, short‐term follow‐up (1 and 4 months)	
Murphy 2007	Low risk of bias	All subjects were assigned to group A, B, or C by a random card draw.	Low risk of bias	Participant were aware of their assigned intervention because of different intervention methods. It was unlikely to have deviation from the intended intervention.	Low risk of bias	Date from nearly all randomized participants.	Low risk of bias	The Fitchck technical method was used in the study.	Low risk of bias	The data reported were accordance with the prespecified protocol, published within the authors' organization	Low risk of bias	It was RCT study. Group B was received written instruction after the first visit, Group C didn't get instruction.	
Murphy 2007	Low risk of bias	All subjects were assigned to group A, B, or C by a random card draw.	Low risk of bias	Participant were aware of their assigned intervention because of different intervention methods. It was unlikely to have deviation from the intended intervention.	Low risk of bias	Date from nearly all randomized participants.	Low risk of bias	The Fitchck technical method was used in the study.	Low risk of bias	The data reported were accordance with the prespecified protocol, published within the authors' organization	Low risk of bias	It was RCT study. Group B was received written instruction after the first visit, Group C didn't get instruction.	
Subgroup 1.1.2 Simple instructions versus no instructions, earplugs, long‐term follow‐up	
Murphy 2007	Low risk of bias	All subjects were assigned to group A, B, or C by a random card draw.	Low risk of bias	Participant were aware of their assigned intervention because of different intervention methods. It was unlikely to have deviation from the intended intervention.	Low risk of bias	Date from nearly all randomized participants.	Low risk of bias	The Fitchck technical method was used in the study.	Low risk of bias	The data reported were accordance with the prespecified protocol, published within the authors' organization	Low risk of bias	It was RCT study. Group B was received written instruction after the first visit, Group C didn't get instruction.	
Subgroup 1.1.3 Extensive instructions versus no instructions, earplugs, immediate follow‐up	
Salmani 2014	Some concerns	Random number chart was used. It was not reported if the allocation sequence concealed before assigned to intervention.	Low risk of bias	All participants used premolded earplugs with different NRR value. It was unlikely to have deviation from the intended intervention.	Low risk of bias	Each group had 50 participants and no reported on missing participants.	Low risk of bias	Same REAT measurement methods applied for all participants and the outcome assessors (the audiologist) was blinded.	Some concerns	No protocol was available.	Some concerns	Lack of information on concealment of allocation and lack of a protocol led to some concerns.	

Risk of bias for analysis 2.2 Earplugs, PAR pass rate (%) (immediate)

Study	Bias	
Randomisation process	Deviations from intended interventions	Missing outcome data	Measurement of the outcome	Selection of the reported results	Overall	
Authors' judgement	Support for judgement	Authors' judgement	Support for judgement	Authors' judgement	Support for judgement	Authors' judgement	Support for judgement	Authors' judgement	Support for judgement	Authors' judgement	Support for judgement	
Federman 2021	Low risk of bias	Training recruits who failed to achieve the minimum target PAR of 25.0 dB (n=344) were randomized assigned to one of three training formats based on the penultimate digit in their SSN.	Low risk of bias	Participants and people delivering the intervention were aware of their assigned intervention during the trial due to different intervention methods. It was unlikely to have deviation from the intended intervention.	Low risk of bias	Data for this outcome available for 321 participants out of 344 randomized subjects.	Low risk of bias	It is unlikely because the measurement is fairly objective.	Some concerns	No protocol was available	Some concerns	This RCT was designed to compare the PAR value or PAR pass rates using different training methods. Nearly all participants attended interventions. Using the penultimate digit in their SSN to assign groups has some concern on the randomization process.	
Federman 2021	Low risk of bias	Training recruits who failed to achieve the minimum target PAR of 25.0 dB (n=344) were randomized assigned to one of three training formats based on the penultimate digit in their SSN.	Low risk of bias	Participants and people delivering the intervention were aware of their assigned intervention during the trial due to different intervention methods. It was unlikely to have deviation from the intended intervention.	Low risk of bias	Data for this outcome available for 321 participants out of 344 randomized subjects.	Low risk of bias	It is unlikely because the measurement is fairly objective.	Some concerns	No protocol was available	Some concerns	This RCT was designed to compare the PAR value or PAR pass rates using different training methods. Nearly all participants attended interventions. Using the penultimate digit in their SSN to assign groups has some concern on the randomization process.	

Acknowledgements

We thank Mr Erik Kateman, from Solcon, the Netherlands, for his help with identifying the scope of this protocol. We thank Jessica R. Thomas, Managing Editor, Julitta Boschman, former Managing Editor, and Jan Hoving, Co‐ordinating Editor, from Cochrane Work Review Group for their help in the early stages of the protocol. We thank Joanna Taliano, Information Specialist from the US Centers for Disease Control and Prevention library for her advice on the search strategy and for running all electronic searches. We thank the peer reviewers of the protocol and the review, as well as the copy editors, Lisa Winer and Faith Armitage, Cochrane Central Production Service. Finally, we thank Capt. William J. Murphy for his input regarding the calculation of PARs for the included studies.

The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the National Institute for Occupational Safety and Health, Centers for Disease Control and Prevention. The mention of any company or product does not constitute endorsement by the National Institute for Occupational Safety and Health, Centers for Disease Control and Prevention.

Appendices

Appendix 1. Search strategy

Database	Strategy	Records
Inception –
10/07/2021	Records
10/07/2021
–08/77/2022	Records
08/77/2022
–07/13/2023	Records
07/13/2023–02/29/2024	Records
in total	
MEDLINE
(Ovid)
1946‐	1. Noise, Occupational/ OR ((noise OR hearing OR sound exposure*) AND (occupation* OR work‐related OR worker* OR job* OR industry OR industrial OR employment)).ti,ab,kf.
2. exp Ear Protective Devices/ OR (((noise* OR hearing OR sound) ADJ5 (attenuation OR dampen* OR suppress* OR cancel* OR reduc* OR diminish*)) OR protective equipment OR protective device* OR protector* OR control* OR (loss ADJ3 prevent*) OR (hearing ADJ3 conserv*) OR (hearing ADJ3 protect*) OR ear muff* OR earmuff* OR ear plug* OR earplug* OR ear defender*).ti,ab,kf.
3. (field attenuation* OR personal attenuation* OR sound attenuation OR attenuation rating* OR attenuation estimate* OR attenuation threshold OR PARs OR PAR OR noise dosimetry OR real ear* OR REAT OR FAES OR fit test* OR (field ADJ3 test*) OR field microphone* OR (field ADJ3 attenuation) OR fit check* OR fitcheck OR validation system* OR E‐A‐Rfit OR estimation system* OR evaluation OR assessment* OR protection level* OR training).ti,ab,kf.
4. 1 and 2 and 3	884	43	74	18	1019	
Embase
(Ovid)
1974‐	1. Industrial noise/ OR ((noise OR hearing OR sound exposure*) AND (occupation* OR work‐related OR worker* OR job* OR industry OR industrial OR employment)).ti,ab,kw.
2. exp Ear Protective Device/ OR (((noise* OR hearing OR sound) ADJ5 (attenuation OR dampen* OR suppress* OR cancel* OR reduc* OR diminish*)) OR protective equipment OR protective device* OR protector* OR control* OR (loss ADJ3 prevent*) OR (hearing ADJ3 conserv*) OR (hearing ADJ3 protect*) OR ear muff* OR earmuff* OR ear plug* OR earplug* OR ear defender*).ti,ab,kw.
3. (field attenuation* OR personal attenuation* OR sound attenuation OR attenuation rating* OR attenuation estimate* OR attenuation threshold OR PARs OR PAR OR noise dosimetry OR real ear* OR REAT OR FAES OR fit test* OR (field ADJ3 test*) OR field microphone* OR (field ADJ3 attenuation) OR fit check* OR fitcheck OR validation system* OR E‐A‐Rfit OR estimation system* OR evaluation OR assessment* OR protection level* OR training).ti,ab,kw.
4. 1 and 2 and 3
5. limit 4 to dc="20230713‐20240229"
6. limit 5 to "pubmed/medline"
7. 5 NOT 6	720‐
duplicates =
367 Unique
items
49‐
duplicates =
24 Unique
items	49‐
duplicates =
24 Unique
items	40‐
duplicates =
34 Unique
items	13‐
duplicates =
0 Unique
items	425	
PsycINFO
(Ovid)
1806‐	1. exp "Noise Levels (Work Areas)"/ OR ((noise OR hearing OR sound exposure*) AND (occupation* OR work‐related OR worker* OR job* OR industry OR industrial OR employment)).ti,ab,hw.
2. (((noise* OR hearing OR sound) ADJ5 (attenuation OR dampen* OR suppress* OR cancel* OR reduc* OR diminish*)) OR protective equipment OR protective device* OR protector* OR control* OR (loss ADJ3 prevent*) OR (hearing ADJ3 conserv*) OR (hearing ADJ3 protect*) OR ear muff* OR earmuff* OR ear plug* OR earplug* OR ear defender*).ti,ab,hw.
3. (field attenuation* OR personal attenuation* OR sound attenuation OR attenuation rating* OR attenuation estimate* OR attenuation threshold OR PARs OR PAR OR noise dosimetry OR real ear* OR REAT OR FAES OR fit test* OR (field ADJ3 test*) OR field microphone* OR (field ADJ3 attenuation) OR fit check* OR fitcheck OR validation system* OR E‐A‐Rfit OR estimation system* OR evaluation OR assessment* OR protection level* OR training).ti,ab,hw.
4. 1 and 2 and 3	152‐
duplicates =
61 Unique
items	12‐
duplicates =
6 Unique
items	9‐
duplicates =
8 Unique
items	7‐
duplicates =
7 Unique
items	82	
Cochrane Library	1. [mh "Noise, Occupational"] OR ((noise OR hearing OR (sound NEXT exposure*)) AND (occupation* OR work‐related OR worker* OR job* OR industry OR industrial OR employment)):ti,ab
2. [mh "Ear Protective Device"] OR (((noise* OR hearing OR sound) NEAR/5 (attenuation OR dampen* OR suppress* OR cancel* OR reduc* OR diminish*)) OR "protective equipment" OR "protective device" OR "protective devices" OR protector* OR control* OR (loss NEAR/3 prevent*) OR (hearing NEAR/3 conserv*) OR (hearing NEAR/3 protect*) OR "ear muff" OR "ear muffs" OR earmuff* OR "ear plug" OR "ear plugs" OR earplug* OR "ear defender" OR "ear defenders"):ti,ab
3. ((field NEXT attenuation*) OR (personal NEXT attenuation*) OR "sound attenuation" OR (attenuation NEXT rating*) OR (attenuation NEXT estimate*) OR "attenuation threshold" OR PARs OR PAR OR "noise dosimetry" OR (real NEXT ear*) OR REAT OR FAES OR (fit NEXT test*) OR (field NEAR/3 test*) OR (field NEXT microphone*) OR (field NEAR/3 attenuation) OR (fit NEXT check*) OR fitcheck OR (validation NEXT system*) OR E‐A‐Rfit OR (estimation NEXT system*) OR evaluation OR assessment* OR (protection NEXT level*) OR training):ti,ab
4. #1 AND #2 AND #3
August 2022 – February 2024	125‐
duplicates =
78 Unique
items	10‐
duplicates =
7 Unique
items	24‐
duplicates =
22 Unique
items	7‐
duplicates =
7 Unique
items	114	
CINAHL
(Ebsco)	S1. ((noise OR hearing OR "sound exposure*") AND (occupation* OR work‐related OR worker* OR job* OR industry OR industrial OR employment))
S2. (MH "Ear Protective Device") OR (((noise* OR hearing OR sound) N5 (attenuation OR dampen* OR suppress* OR cancel* OR reduc* OR diminish*)) OR "protective equipment" OR "protective device*" OR protector* OR control* OR (loss N3 prevent*) OR (hearing N3 conserv*) OR (hearing N3 protect*) OR "ear muff*" OR earmuff* OR "ear plug*" OR earplug* OR "ear defender*")
S3. ("field attenuation*" OR "personal attenuation*" OR "attenuation rating*" OR "attenuation estimate*" OR "attenuation threshold" OR "sound attenuation" OR PARs OR PAR OR "noise dosimetry" OR "real ear*" OR REAT OR FAES OR "fit test*" OR (field N3 test*) OR "field microphone*" OR (field N3 attenuation) OR "fit check*" OR fitcheck OR "validation system*" OR E‐A‐Rfit OR "estimation system*" OR evaluation OR assessment* OR "protection level*" OR training)
S4 S1 AND S2 AND S3
Exclude Medline records
August 2022 – February 2024	378‐
duplicates =
278 Unique
items	40‐
duplicates =
27 Unique
items	41‐
duplicates =
32 Unique
items	36‐
duplicates =
29 Unique
items	366	
Scopus	TITLE‐ABS((noise OR hearing OR "sound exposure*") W/10 (occupation* OR work‐related OR worker* OR job* OR industry OR industrial OR employment)) AND TITLE‐ABS(((noise* OR hearing OR sound) W/5 (attenuation OR dampen* OR suppress* OR cancel* OR reduc* OR diminish*)) OR "protective equipment" OR "protective device*" OR protector* OR control* OR (loss W/3 prevent*) OR (hearing W/3 conserv*) OR (hearing W/3 protect*) OR "ear muff*" OR earmuff* OR "ear plug*" OR earplug* OR "ear defender*") AND TITLE‐ABS("field attenuation*" OR "personal attenuation*" OR "attenuation rating*" OR "attenuation estimate*" OR "attenuation threshold" OR "sound attenuation" OR PARs OR PAR OR "noise dosimetry" OR "real ear*" OR REAT OR FAES OR "fit test*" OR (field W/3 test*) OR "field microphone*" OR (field W/3 attenuation) OR "fit check*" OR fitcheck OR "validation system*" OR E‐A‐Rfit OR "estimation system*" OR evaluation OR assessment* OR "protection level*" OR training)
August 2022 – February 2024	1398‐
duplicates =
537 Unique
items	59‐
duplicates =
26 Unique
items	72‐
duplicates =
32 Unique
items	38‐
duplicates =
29 Unique
items	624	
NIOSHTIC‐2	Noise OR hearing OR sound exposure
AND
attenuation OR dampen* OR suppress* OR cancel* OR reduc* OR diminish* OR protect* OR control* OR prevent* OR conserv* OR ear muff OR earmuff OR ear plug OR earplug OR ear defender
AND
'field attenuation' OR 'personal attenuation' OR 'attenuation rating' OR 'attenuation estimate' OR 'attenuation threshold' OR PARs OR PAR OR 'noise dosimetry' OR 'real ear' OR 'real ears' OR REAT OR FAES OR 'fit test' OR 'fit tests' OR 'field test' OR 'field tests' OR 'field microphone' OR 'field microphones' OR 'fit check' OR 'fit checks' OR fitcheck OR 'validation system' OR 'validation systems' OR E‐A‐Rfit OR 'estimation system' OR 'estimation systems'
August 2022 – February 2024	444‐
duplicates =
367 Unique
items	51‐
duplicates =
4 Unique
items	3‐
duplicates =
0 Unique
items	8‐
duplicates =
7 Unique
items	378	
ClinicalTrials.gov	"Noise exposure" OR "hearing loss" OR "sound exposure" OR "ear plugs" OR earplugs OR earmuffs OR "ear muffs"
AND
(
("hearing protection" OR fitcheck OR "fit check" OR attenuation OR "fit test" OR "fit testing" OR "fit tests" OR "validation system" OR "validation systems")
OR
("noise dosimetry" OR "real ear" OR "real ears" OR "field test" OR "field testing" OR "field tested" OR "field tests" OR "field microphone" OR "field microphones" OR E‐A‐Rfit OR "estimation system" OR "estimation systems")
)
August 2022 – February 2024	 	 	3	0	3	
WHO ICTRP	(
Condition: "Noise exposure" OR "hearing loss" OR "sound exposure"
OR
Title: "ear plugs" OR earplugs OR earmuffs OR "ear muffs"
)
AND
(
Intervention: ("hearing protection" OR fitcheck OR "fit check" OR attenuation OR "fit test" OR "fit testing" OR "fit tests" OR "validation system" OR "validation systems" OR "noise dosimetry" OR "real ear" OR "real ears" OR "field test" OR "field testing" OR "field tested" OR "field tests" OR "field microphone" OR "field microphones" OR E‐A‐Rfit OR "estimation system" OR "estimation systems")
)
Results :
August 2022 – February 2024	 	 	0	0	0	

[Notes: Duplicates were identified using the EndNote automated 'find duplicates' function with preference set to
match on title, author, and year, and removed from EndNote library. There will likely be additional duplicates that EndNote was unable to detect.]

Appendix 2. Risk of bias assessment

The risk of bias Excel sheet can be found here: https://docs.google.com/spreadsheets/d/11Q1ss-4ZUOvu7tKbV_BN13IM13378-zn/edit?usp=sharing&ouid=108595225591542539022&rtpof=true&sd=true

Data and analyses

Comparison 1 Fit testing of HPDs with instructions versus fit testing of HPDs without instructions for improving noise attenuation (immediate, short‐term, long‐term)

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size	
1.1 Personal attenuation rating (PAR; dB) (immediate, short‐term, long‐term)	2	 	Mean Difference (IV, Random, 95% CI)	Totals not selected	
1.1.1 Simple instructions versus no instructions, earplugs, short‐term follow‐up (1 and 4 months)	1	 	Mean Difference (IV, Random, 95% CI)	Totals not selected	
1.1.2 Simple instructions versus no instructions, earplugs, long‐term follow‐up	1	 	Mean Difference (IV, Random, 95% CI)	Totals not selected	
1.1.3 Extensive instructions versus no instructions, earplugs, immediate follow‐up	1	 	Mean Difference (IV, Random, 95% CI)	Totals not selected	

1.1 Analysis

Comparison 1: Fit testing of HPDs with instructions versus fit testing of HPDs without instructions for improving noise attenuation (immediate, short‐term, long‐term), Outcome 1: Personal attenuation rating (PAR; dB) (immediate, short‐term, long‐term)

Comparison 2 Fit testing of HPDs with extensive instructions versus fit testing of HPDs with simple instructions for improving noise attenuation (immediate)

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size	
2.1 Personal attenuation rating (PAR; dB) (immediate)	1	 	Mean Difference (IV, Random, 95% CI)	Subtotals only	
2.1.1 Earplugs	1	321	Mean Difference (IV, Random, 95% CI)	8.62 [6.31, 10.93]	
2.2 Earplugs, PAR pass rate (%) (immediate)	1	321	Risk Ratio (M‐H, Random, 95% CI)	1.75 [1.44, 2.11]	

2.1 Analysis

Comparison 2: Fit testing of HPDs with extensive instructions versus fit testing of HPDs with simple instructions for improving noise attenuation (immediate), Outcome 1: Personal attenuation rating (PAR; dB) (immediate)

2.2 Analysis

Comparison 2: Fit testing of HPDs with extensive instructions versus fit testing of HPDs with simple instructions for improving noise attenuation (immediate), Outcome 2: Earplugs, PAR pass rate (%) (immediate)

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Federman 2021

Study characteristics	
Methods	RCT	
Participants	Training recruits
N = 321
Marine Corps
USA	
Interventions	A total of 821 randomly selected recruits were screened for the fit of their hearing protection devices. They completed a baseline earplug fit test and a live, large group instruction. Of these 821, 344 recruits did not achieve the target 25 dB PAR pass rate and were randomly assigned to one of three HPD fit‐training (current, experiential, integrated) formats. Three were somehow not randomised, and after randomisation, 20 participants were excluded: 11 used a different earplug, 7 could not achieve a PAR of 25 dB even after fit by an expert; and 2 did not complete the training. These losses to follow‐up seemed similar across trial arms but there were no clear figures in the report.
Intervention group 1: experiential training; n = 105 analysed
Intervention group 2: integrated training (small group video training followed experiential training); n = 105 analysed
Control group: current training (small group video training); n = 111 analysed
We compared the outcomes measured between intervention group 1 and control group, and between intervention group 2 and control group, for the effect of extensive instructions versus simple instructions.
For both intervention groups, we also compared the within‐group before and after training outcomes as a UBA study in the discussion section. The results were used to compare with the findings from RCTs for the effects of extensive instruction verses no instructions.	
Outcomes	PAR (dB) and PAR pass rate (%)	
Notes	Immediate follow‐up	

Murphy 2007

Study characteristics	
Methods	RCT	
Participants	Automotive workers 
N = 285
An automobile stamping factory
USA	
Interventions	Baseline test: total N = 285
Intervention group: multi‐earplug (foam and premoulded) users with written instructions; n = 153
Control group: multi‐earplug (foam and premoulded) users without written instructions; n = 132
Follow‐up test:
1‐month: intervention group n = 100, control group n = 109

4‐month: intervention group n = 103, control group n = 94

12‐month: intervention group n = 49, control group n = 54

	
Outcomes	PAR (dB)	
Notes	Short‐term and long‐term follow‐up	

Salmani 2014

Study characteristics	
Methods	RCT	
Participants	Occupations not reported
N = 150
Age: 28.3 ± 5.4 (range: 19 to 39) years
Gender: 42% male
Iran	
Interventions	Intervention: earplugs with training in correct methods of wearing and inserting plugs; n = 50
Control group 1: earplugs without training; n = 50
Control group 2: earplug with higher noise attenuation without training; n = 50
We compared the outcomes measured between intervention group and control group 1 for the effect of extensive instructions versus no instructions.	
Outcomes	PAR (dB)	
Notes	Immediate follow‐up	
dB: decibel; HPD: hearing protection device; PAR: personal attenuation rating; RCT: randomised controlled trial; UBA: uncontrolled before‐and‐after study

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion	
Biabani 2017a	Study design: no measurements both before the intervention and after the intervention	
Biabani 2017b	Study design: no measurements both before the intervention and after the intervention	
Byrne 2011	Study design: no measurements both before the intervention and after the intervention	
Cassano 2013	Study design: no measurements both before the intervention and after the intervention	
Davis 2011	Study design: no measurements both before the intervention and after the interventionthey	
De Vito 2017a	Study design: no measurements both before the intervention and after the intervention	
De Vito 2017b	Study design: no measurements both before the intervention and after the intervention	
Edwards 1978	Study design: no measurements both before the intervention and after the intervention	
Edwards 1979	Study design: no measurements both before the intervention and after the intervention	
Gaudreau 2008	Study design: no measurements both before the intervention and after the intervention	
Giardino 1996	Study design: no measurements both before the intervention and after the intervention	
Gibbs 2017	Study design: no measurements both before the intervention and after the intervention	
Gong 2020	Study design: no measurements both before the intervention and after the intervention	
Green 1989	Study design: no measurements both before the intervention and after the intervention	
Kabe 2012	Study design: no measurements both before the intervention and after the intervention	
Lee 2020	Study design: no measurements both before the intervention and after the intervention	
Lempert 2010	Study design: no measurements both before the intervention and after the intervention	
Mlynski 2014	Study design: no measurements both before the intervention and after the intervention	
Murphy 2013	Population and study design: one case was not for noised‐exposed workers; another case was no intervention at the workplaces	
Neitzel 2006	Study design: no measurements both before the intervention and after the intervention	
Pfeiffer 1992	Study design: no measurements both before the intervention and after the intervention	
Rocha 2016	Study design: no measurements both before the intervention and after the intervention	
Samelli 2015	Study design: cross‐sectional study	
Samelli 2018	Study design: cross‐sectional study	
Smoorenburg 1982	Study design: no measurements both before the intervention and after the intervention	
Ullman 2021	Study design: no measurements both before the intervention and after the intervention	
Upegui‐Rincon 2019	Study design: no measurements both before the intervention and after the intervention	
Wu 2016	Study design: no measurements both before the intervention and after the intervention	

Differences between protocol and review

We made minor changes to the review's title, objectives, background, and methods sections to improve readability and clarity, as recommended by the editorial team and Cochrane Methods. These adjustments did not deviate from our original intentions.

In the protocol, we did not include adverse effects as an outcome. However, in the review, we included adverse effects as a primary outcome. Since we are not aware of any specific reported adverse effects of the intervention of interest, we planned to include any adverse effects that were mentioned by the trial authors. However, no studies reported adverse effects.

Contributions of authors

Conceived the protocol: TM, CT, WG, AS

Commented on drafts of protocol and review: all authors

Co‐ordinated the protocol and review: TM, JV

Designed the search strategies: all authors

Conducted eligibility screening, quality assessment of studies, data extraction: WG, TM, CT, AS

Data analysis: WG, JV, TM, CT

Wrote the protocol and review: TM, WG, CT

Provided general advice on the protocol and the development of the review: JV

Secured funding for the protocol: TM

Performed previous work that was the foundation of the current study: WG, AS, CT

Sources of support

Internal sources

National Institute for Occupational Safety and Health, USA

Contract #75D30120P07102 with Synergy America for expertise and co‐authorship for Ms Wei Gong.

External sources

Cochrane Work, formerly known as Cochrane Occupational Safety and Health Review Group, Netherlands

Provided support in kind.

Finnish Institute of Occupational Health, Finland

Provided support in kind.

Universidade de São Paulo, Brazil

Provided support in kind.

National Institute for Occupational Safety and Health, USA

Provided support in kind.

Declarations of interest

Three review authors (TM, WG, and AS) are authors on papers related to the topic of this review and are specified below. Two different review authors who were not authors of the included papers conducted data extraction and quality assessment for this review, ensuring a neutral assessment of the studies.

The author team has no affiliations with or involvement in any organisation or entity with a direct financial interest in the subject matter of the review (e.g. employment, consultancy, stock ownership, honoraria, or expert testimony). No financial or non‐financial conflicts of interests have influenced the submitted work. All authors checked and approved the final version.

Thais Morata (TM): is a senior research audiologist at the National Institute for Occupational Safety and Health (NIOSH) and has engaged in research activities related to the topic of this review as a hearing expert. She was a co‐author on Gong 2021b, which was mentioned but not included in this review. She was not involved in the data analysis decisions for that article. TM is a member of the Cochrane Work Editorial Board but was not involved in the editorial process. The views expressed are her own and do not necessarily represent the official position of NIOSH, Centers for Disease Control and Prevention (CDC). Mention of any company or product does not constitute endorsement by NIOSH, or the CDC.

Wei Gong (WG): was awarded the independent contract #75D30120P07102 with Synergy America for completion of this review from January 2020 to January 2022. Synergy America received funds from NIOSH for this contract, which she held. She became a research fellow at NIOSH in the final phases of completing this review (26 September 2022). She was a co‐author on Gong 2023, which was mentioned but not included in this review. She has published four uncontrolled before‐and‐after studies in audiology and industrial medicine journals. WG was involved in data collection, analysis, and preparing the reports for these papers before she worked as a contractor at NIOSH. Three of these studies were included in this review (Gong 2019; Gong 2021a; Liu 2020). The studies were supported by the Jiangsu Provincial Outstanding Medical Academic Leader and Innovation Team (CXTDA2017029), the Natural Science Foundation of the Jiangsu Province (Grant No. BK20151594) and Personal Safety Division of 3M China Ltd. In this review, WG had no involvement in the assessment or data extraction for the studies she co‐authored.

Christina Tikka (CT): none known.

Alessandra Samelli (AS): has published four book chapters in Portuguese which detail the use of fit testing within hearing conservation programs. All chapters were part of technical‐scientific books, which serve as a basis for the guidance of audiologists and were produced by scientific societies. AS has also been involved with seven studies on topics concerning the use of fit testing as a way to verify the attenuation of different hearing protectors or comparing different forms of training for proper placement of hearing protectors, related to this review in audiology or public health journals. She was involved in data collection, analysis, and preparation of the reports. No external financial support was obtained for those studies. Two of these studies were included in this review (Assunção 2019; Takada 2020). For this review, AS had no involvement in the assessment or data extraction for the studies she co‐authored.

Jos Verbeek (JV): Advisor and Member of the Editorial Board of Cochrane Work. JV has excluded himself entirely from the editorial process of this review. In the past five years, he has worked as a consultant methodologist for the World Health Organization (WHO) on a project to synthesise evidence on the adverse health effects of radiofrequency waves.
==== Refs
References

References to studies included in this review

Federman 2021 {published data only}

Federman J , Karch SJ , Duhon C . How hearing conservation training format impacts personal attenuation ratings in U.S. Marine Corps Training Recruits. International Journal of Audiology 2021;60 (2 ):151-9. [DOI: 10.1080/14992027.2020.1811407]
Murphy 2007 {published and unpublished data}

Davis RR , Murphy WJ , Byrne D , Franks JR . Comfort and personal NRR in a longitudinal study of highly experienced earplug users. Spectrum: the National Hearing Conservation Association Newsletter 2006;Suppl 1 :21-2.
Franks JR , Davis RR , Murphy WJ . Field measurements of hearing protection device performance. International Congress on Noise Control Engineering, 2005;3 :2094-103.
Murphy WJ , Davis RR , Byrne DC , Franks JR . Advanced hearing protector study conducted at: General Motors Metal Fabricating Division Flint Metal Center - Flint, Michigan, January 2004 - February 2005. Cincinnati (OH): Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health; 2007 May. Report No.: EPHB Report 312-11a, 42 pp. www.cdc.gov/niosh/surveyreports/pdfs/312-11a.pdf.
Salmani 2014 {published data only}

Salmani NM , Mehrparvar AH , Torab JM , Safaei S , Mollasadeghi A . Training in using earplugs or using earplugs with a higher than necessary noise reduction rating? A randomized clinical trial. International Journal of Occupational & Environmental Medicine 2014;5 (4 ):187-93. [PMID: 25270008]25270008
References to studies excluded from this review

Biabani 2017a {published data only}

Biabani A , Aliabadi M , Golmohammadi R , Farhadian M , Shahidi R . Evaluation of noise reduction rating of current earmuffs based on field microphone in real ear (FMIRE) in workplaces. Iran Occupational Health 2017;14 (4 ):118-27.
Biabani 2017b {published data only}

Biabani A , Aliabadi M , Golmohammadi R , Farhadian M . Individual fit testing of hearing protection devices based on microphone in real ear. Safety and Health at Work 2017;8 (4 ):364-70. [DOI: 10.1016/j.shaw.2017.03.005]
Byrne 2011 {published data only}

Byrne DC , Davis RR , Shaw PB , Specht BM , Holland AN . Relationship between comfort and attenuation measurements for two types of earplugs. Noise and Health 2011;13 (51 ):86-92. [DOI: 10.4103/1463-1741.77193]
Cassano 2013 {published data only}

Cassano F , Aloise I , Labianca G , Gaccione V , Mazzotta C , Cardascia F , et al. Measurement of real personal noise attenuation using earplugs with the E-A-Rfit system. Medicina del Lavoro 2013;104 (3 ):213-23. [PMID: 23879065]23879065
Davis 2011 {published data only}

Davis RR , Murphy WJ , Byrne DC , Shaw PB . Acceptance of a semi-custom hearing protector by manufacturing workers. Journal of Occupational and Environmental Hygiene 2011;8 (12 ):D125-D130. [DOI: 10.1080/15459624.2011.626262]
De Vito 2017a {published data only}

De Vito C . Comparison of personal attenuation ratings attained by agricultural and industrial workers for four hearing protection types using the 3M E-A-RFIT dual validation system [Masters thesis]. Iowa City, IA: The University of Iowa, 2017.
De Vito 2017b {published data only}

De Vito C , Gibbs JL , Anthony TR . Personal attenuation ratings of farmers: how well do earplugs really fit? In: Best practices for new horizons: 16th Annual Midwest Rural Agricultural Safety & Health (MRASH) Conference, 2017 November 14-15, Pella, Iowa. 2017.
Edwards 1978 {published data only}

Edwards RG , Hauser WP , Moiseev NA . Effectiveness of earplugs as worn in the workplace. Sound and Vibration 1978;12 (1 ):12-22.
Edwards 1979 {published data only}

Edwards RG , Hauser WP , Moiseev NA , Broderson AB , Green WW . A field investigation of noise reduction afforded by insert-type hearing protectors. Cincinnati, Ohio: Centers for Disease Control and Prevention, National Institutes for Occupational Safety and Health; 1970 November. Report No.: 79-115.
Gaudreau 2008 {published data only}

Gaudreau MA , Laville F , Nélisse H , Voix J . Méthode de mesures terrain de l'atténuation F-MIRE de protecteurs auditifs durant un quart de travail. Canadian Acoustics - Acoustique Canadienne 2008;36 (2 ):3-10.
Giardino 1996 {published data only}

Giardino DA , Durkt GJ . Evaluation of muff-type hearing protectors as used in a working environment. American Industrial Hygiene Association Journal 1996;57 (3 ):264-71. [DOI: 10.1080/15428119691014990]
Gibbs 2017 {published data only}

Gibbs J , DeVito C , Anthony TR . Personal attenuation ratings (PAR) of agricultural workers: how well do earplugs really fit? In: 2017 Expanding Research Partnerships: State of the Science, 2017 June 21-23, Denver, Colorado. Denver, Colorado: Center for Health Work & Environment, Colorado School of Public Health, 2017.
Gong 2020 {published data only}

Gong W , Morata TC . Verifying the attenuation of earplugs in an auto parts factory in China. In: Advances in Noise and Vibration Control Technology, the 49th International Congress and Exposition on Noise Control Engineering, 2020 August 23-26, Virtual Event. Reston, VA: Institute of Noise Control Engineering. 2020:2662-8.
Green 1989 {published data only}

Green WW , Edwards RG , Broderson AB . Effectiveness of insert-type hearing protectors (earplugs) in the workplace. In: 1989 industrial hearing conservation conference, 1989 April 12-14; Cincinnati (OH). Centers for Disease Control, National Institute for Occupational Safety and Health, 1998.
Kabe 2012 {published data only}

Kabe I , Kochi T , Tsuruoka H , Tonegawa T , Denda I , Nonogi M , et al. Noise attenuation of earplugs as measured by hREAT and F-MIRE methods in a Japanese metal manufacturing plant. Journal of Occupational Health 2012;54 (4 ):310-5. [DOI: 10.1539/joh.11-0217-fs]
Lee 2020 {published data only}

Lee KH , Benke G , McKenzie DP . The efficacy of earplugs at a major hazard facility. Internal Medicine Journal 2020;50 (Suppl 1) :13.
Lempert 2010 {published data only}

Lempert BL , Edwards RG . Field investigations of noise reduction afforded by insert-type hearing protectors. American Industrial Hygiene Association Journal 1983;44 (12 ):894-902. [DOI: 10.1080/15298668391405913]
Mlynski 2014 {published data only}

Mlynski R , Kozlowski E . Assessment of the impulse noise attenuation by earplugs in metalworking processes. Medycyna Pracy 2014;65 (2 ):197-207. [PMID: 25090849]25090849
Murphy 2013 {published data only}

Murphy WJ , Themann CL , Stephenson MR , Byrne DC . Two case studies for fit testing hearing protector devices. Journal of the Acoustical Society of America 2013;132 (3 ):2. [DOI: 10.1121/1.4755450]
Neitzel 2006 {published data only}

Neitzel R , Somers S , Seixas N . Variability of real-world hearing protector attenuation measurements. Annals of Occupational Hygiene 2006;50 (7 ):679-91. [DOI: 10.1093/annhyg/mel025]
Pfeiffer 1992 {published data only}

Pfeiffer BH . Real-world effectiveness of hearing protection devices in German industry. In: 1992 Hearing Conservation Conference, 1992 April 1-4. Lexington (KY). University of Kentucky, 1992:21-24.
Rocha 2016 {published data only}

Rocha CH , Longo IA , Moreira RR , Samelli AG . Evaluation of the hearing protector in a real work situation using the field-microphone-in-real-ear method. Codas 2016;28 (2 ):99-105. [DOI: 10.1590/2317-1782/20162014146]
Samelli 2015 {published data only}

Samelli AG , Rocha CH , Theodosio P , Moreira RR , Neves-Lobo IF . Training on hearing protector insertion improves noise attenuation. Codas 2015;27 (6 ):514-9. [DOI: 10.1590/2317-1782/20152014128]
Samelli 2018 {published data only}

Samelli AG , Gomes RF , Chammas TV , Silva BG , Moreira RR , Fiorini AC . The study of attenuation levels and the comfort of earplugs. Noise and Health 2018;20 (94 ):112-9. [DOI: 10.4103/nah.NAH_50_17]
Smoorenburg 1982 {published data only}

Smoorenburg GF , Mimpen TM . Assessment of personal hearing protection in practice. Scandinavian Audiology 1982;Suppl 16 :13-22. [PMID: 6962507]6962507
Ullman 2021 {published data only}

Ullman ED , Smith LM , McCullagh MC , Neitzel RL . Hearing loss as a predictor for hearing protection attenuation among miners. Occupational and Environmental Medicine 2021;78 (5 ):371-6. [DOI: 10.1136/oemed-2020-106838]
Upegui‐Rincon 2019 {published data only}

Upegui-Rincon S , Araque-Munoz L , Lizarazo-Salcedo C , Berrio-Garcia S , Guarguati-Ariza J . Verifying the attenuation levels of earmuff hearing protectors in oil and gas workers. Revista de Salud Publica 2019;21 (2 ):195-201. [DOI: 10.15446/rsap.V21n2.70989]
Wu 2016 {published data only}

Wu M , Guffey SE , Takacs BC . Comparison of noise reduction of earplugs between fit test and coal mine work and observed variability of real-time noise reduction in coal mine work. Work 2016;55 (3 ):531-8. [DOI: 10.3233/WOR-162439]
Additional references

ACGIH 2018

American Conference of Governmental Industrial Hygienists. Notice of intended change. In: 2018 TLVs® and BEIs®. Cincinnati, OH: ACGIH, 2018:74.
ANSI 2018

American National Standard Institute. Performance criteria for systems that estimate the attenuation of passive hearing protectors for individual users (ANSI S12.71). Washington, DC: ANSI, 2018.
Assunção 2019

Assunção CHC , Trabanco JC , Gomes RF , Moreira RR , Samelli AG . Longitudinal evaluation of a hearing protector fit training program. Medicina del Lavoro 2019;110 (4 ):304-11. [DOI: 10.23749/mdl.v110i4.8214]
Basner 2014

Basner M , Babisch W , Davis A , Brink M , Clark C , Janssen S , et al. Auditory and non-auditory effects of noise on health. Lancet 2014;383 (9925 ):1325-32. [DOI: 10.1016/S0140-6736(13)61613-X] [PMID: 24183105]24183105
Berger 1983

Berger EH . Assessment of the performance of hearing protectors for hearing conservation purposes. Journal of the Acoustical Society of America 1983;74 (S1 ):S94. [DOI: 10.1121/1.2021231]
Berger 1996

Berger EH , Franks JR , Lindgren F . International review of field studies of hearing protector attenuation. In: Axlesson A , Borchgrevink H , Hamernik RP , Hellstrom P , HendersonD, Salvi RJ , editors(s). Scientific Basis of Noise-induced Hearing Loss. New York: Thieme Medical Publishing, 1996:361–77.
Berger 2008

Berger EH , Voix J , Hager DL . Methods of fit testing hearing protectors, with representative field test data. In: Griefahn B , editors(s). Ninth Congress of the International Commission on the Biological Effects of Noise (ICBEN): Noise as a Public Health Problem. Mashantucket, CT: International Commission on Biological Effects of Noise, 2008:70-8. [URL: www.icben.org/2008/PDFs/ICBEN_Proceedings_2008.pdf]
Berger 2022

Berger EH , Voix J . Hearing protection devices. In: Meinke DK , Berger EH , Neitzel RL , Driscoll DP , Bright K , editors(s). The Noise Manual. 6th edition. Fairfax (VA): American Industrial Hygiene Association (AIHA), 2022:255-301. [ISBN: 978-1-935082-98-9]
Byrne 2018

Byrne D , Morata TC . Recognizing and preventing disease and injury [Noise Exposure and Hearing Disorders]. In: Levy BS , Wegman DH , Baron S , Sokas RK , McStowe HL , editors(s). Occupational and Environmental Health- Recognizing and preventing disease and injury. 7th edition. New York (NY): Oxford University Press, 2018:243-58. [ISBN: 0190662670]
Casali 1991

Casall JG , Park MY . Real-ear attenuation under laboratory and industrial test conditions as provided by selected hearing protectors. Proceedings of the Human Factors Society 1991;2 :1110-4. [DOI: 10.1177/154193129103501519]
Cassano 2015

Cassano F , Aloise I , Labianca G , Gaccione V , Mazzotta C , Cardascia F , et al. The role of information and training for workers on the correct use of earplugs in assessing real noise attenuation with E-A-Rfit system. Medicina del Lavoro 2015;106 (4 ):284-93. [PMID: 26154471]26154471
Chiu 2020

Chiu CC , Wan TJ . Individual fit testing of hearing-protection devices based on microphones in real ears among workers in industries with high-noise-level manufacturing. International Journal of Environmental Research and Public 2020;17(9) :3242. [DOI: 10.3390/ijerph17093242]
Cohen 1997

Cohen AL . Elements of Ergonomics Programs: A Primer Based on Workplace Evaluations of Musculoskeletal Disorders. Publication 97-117 edition. Cincinnati, OH: US Department of Health and Human Services, Public Health Service, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, 1997.
Copelli 2021

Copelli F , Behar A , Ngoc LT , Russo FA . Field attenuation of foam earplugs. Safety and Health at Work 2021;12 (2 ):184-91. [DOI: 10.1016/j.shaw.2020.09.006]
Covidence [Computer program]

Covidence. Version accessed prior to 27 April 2024. Melbourne, Australia: Veritas Health Innovation, 2023. Available at covidence.org.
Deeks 2021

Deeks JJ , Higgins JP , Altman DG (editors), Cochrane Statistical Methods Group. Chapter 10: Analysing data and undertaking meta-analyses. In: Higgins JP, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA (editors). Cochrane Handbook for Systematic Reviews of Interventions version 6.2 (updated February 2021). Cochrane, 2021. Available at training.cochrane.org/handbook/archive/v6.2/chapter-10.
Dzhambov 2017

Dzhambov A , Dimitrova D . Occupational noise exposure and the risk for work-related injury: a systematic review and meta-analysis. Annals of Work Exposures and Health 2017;61 (9 ):1037-53. [DOI: 10.1093/annweh/wxx078]
Elbourne 2002

Elbourne DR , Altman DG , Higgins JP , Curtin F , Worthington HV , Vail A . Meta-analyses involving cross-over trials: methodological issues. International Journal of Epidemiology 2002;31 (1 ):140-9. [DOI: 10.1093/ije/31.1.140]
EPA 1979

Environmental Protection Agency. Noise Labeling Requirements for Hearing Protectors. Federal Regulation 44 (190), 40CFR 211 Part 211 1979;44 (190 ):130–56. [CODE OF FEDERAL REGULATIONS LINK: www.ecfr.gov/current/title-40/chapter-I/subchapter-G/part-211]
EPOC 2012

Cochrane Effective Practice and Organisation of Care (EPOC) Review Group. The data collection checklist. www.epoc.cochrane.org (accessed 27 August 2021). [URL: epoc.cochrane.org/sites/epoc.cochrane.org/files/public/uploads/datacollectionchecklist.pdf]
Federman 2016

Federman J , Duhon C . The viability of hearing protection device fit-testing at navy and marine corps accession points. Noise and Health 2016;18(85) :303-11. [DOI: 10.4103/1463-1741.195806]
GBD 2019 Hearing Loss Collaborators

GBD 2019 Hearing Loss Collaborators. Hearing loss prevalence and years lived with disability, 1990–2019: findings from the Global Burden of Disease Study 2019. Lancet 2021;397 (10278 ):996-1009. [DOI: 10.1016/S0140-6736(21)00516-X] [PMID: 33714390]33714390
Gong 2019

Gong W , Liu X , Liu Y , Li L . Evaluating the effect of training along with fit testing on foam earplug users in four factories in China. International Journal of Audiology 2019;58 (5 ):269-77. [DOI: 10.1080/14992027.2018.1563307]
Gong 2021a

Gong W , Zhang J , Liu X , Ding E , Zhu B . Verifying earplug attenuation and evaluating the effectiveness of one-on-one training along with earplug fit testing at nine facilities in China. American Journal of Industrial Medicine 2021;64 (9 ):771-80. [DOI: 10.1002/ajim.23270]
Gong 2021b

Gong W , Zhao L , Li L , Morata TC , Qiu W , Feng HA , et al. Evaluating the effectiveness of earplugs in preventing noise-Induced hearing loss in an auto parts factory in China. International Journal of Environmental Research and Public Health 2021;18 (13 ):7190. [DOI: 10.3390/ijerph18137190]
Gong 2023

Gong W , Murphy WJ , Meinke DK , Feng HA , Stephenson MR . Evaluating earplug performance over a 2-hour work period with a fit-test system. Seminars in Hearing 2023;1 (44 ):470-84. [DOI: 10.1055/s-0043-1769586] [PMID: 37818150]37818150
GRADEpro GDT [Computer program]

GRADEpro GDT. Version accessed prior to 27 April 2024. McMaster University (developed by Evidence Prime), 2024. Available from gradepro.org.
Green 2021

Green DR , Masterson EA , Themann CL . Prevalence of hearing protection device non-use among noise-exposed US workers in 2007 and 2014. American Journal of Industrial Medicine 2021;64 (12 ):1002-17. [DOI: 0.1002/ajim.23291]
Groenewold 2014

Groenewold MR , Masterson EA , Themann CL , Davis RR . Do hearing protectors protect hearing? American Journal of Industrial Medicine 2014;57 (9 ):1001-10. [DOI: 10.1002/ajim.22323]
Hayes 2022 

Hayes ME , Hammond S , Montgomery AP , Stephenson L . Improving hearing protection device noise attenuation through fit-testing in an occupational health clinic. Workplace Health & Safety 2022;70 (4 ):196-204. [DOI: 10.1177/21650799211067927]
Higgins 2023

Higgins JP , Li T , Deeks JJ (editors). Chapter 6: Choosing effect measures and computing estimates of effect. In: Higgins JP, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA (editors). Cochrane Handbook for Systematic Reviews of Interventions version 6.4 (updated August 2023). Cochrane, 2023. Available from training.cochrane.org/handbook.
ISO 2013

International Organization for Standardization. Acoustics - Estimation of Noise-induced Hearing Loss. ISO 1999:2013. 2013 edition. Geneva, Switzerland: ISO, 2013.
ISO 2018

International Organisation for Standardization. Acoustics - Hearing Protectors - Part 1: subjective method for the measurement of sound attenuation. ISO 4869-1:2018. Geneva, Switzerland: ISO, 2018.
Johnson 2011

Johnson C , Berger EH , Logan P , Ramachandran G . Fit Testing of Hearing Protection in a Large Occupational Survey [Masters thesis]. Minneapolis, MN: University of Minnesota, 2011.
Kerr 2017

Kerr MJ , Neitzel RL , Hong O , Sataloff RT . Historical review of efforts to reduce noise-induced hearing loss in the United States. American Journal of Industrial Medicine 2017;60 (6 ):569-77. [DOI: 10.1002/ajim.22627]
Kim 2019

Kim JW , Yang S , Chung I , Lee MY . The effect of earplug training on noise protection. Annals of Occupational and Environmental Medicine 2019;31 :e34. [DOI: 10.35371/aoem.2019.31.e34]
Kunz 2014

Kunz J . Hearing Protection Attenuation Values Over Time in a Population of Manufacturing Workers Using the E-A-Rfit Validation System [Masters thesis]. Minneapolis, MN: University of Minnesota, 2014.
Leigh 1999

Leigh J , Macaskill P , Kuosma E , Mandryk J . Global burden of disease and injury due to occupational factors. Epidemiology 1999;10 (5 ):626-31. [PMID: 10468442]10468442
Liu 2018

Liu Y , Yang M . Evaluating the effect of training along with fit testing on earmuff users in a Chinese textile factory. Journal of Occupational and Environmental Hygiene 2018;15(6) :518-26. [DOI: 10.1080/15459624.2018.1456662]
Liu 2020

Liu Y , Gong W , Liu X , Li L . Evaluating the effect of training along with fit testing on premolded earplug users in a Chinese petrochemical plant. Ear and Hearing 2020;41 (4 ):838-46. [DOI: 10.1097/AUD.0000000000000803]
Martin 2019

Martin L , Negrini A , Gaudreau MA , Sgard F , Berbiche D , Doutres O . Earplug personal attenuation rating (PAR) in noise-exposed workers: evolution over a five weeks follow-up. In: 26th International Congress on Sound and Vibration, ICSV; 2019 July 7-11; Montreal, Canada. 2019.
Masterson 2016a

Masterson, EA, Themann CL , Luckhaupt SE , Li J , Calvert GM . Hearing difficulty and tinnitus among U.S. workers and non-workers in 2007. American Journal of Industrial Medicine 2016;59 (4 ):290-300. [DOI: 10.1002/ajim.22565]
Masterson 2016b

Masterson EA , Bushnell PT , Themann CL , Morata TC . Hearing impairment among noise-exposed workers - United States, 2003-2012. MMWR. Morbidity and Mortality Weekly Report 2016;65 (15 ):389-94. [DOI: 10.15585/mmwr.mm6515a2]
McKenzie 2021

McKenzie JE , Brennan SE . Chapter 12: Synthesizing and presenting findings using other methods. In: Higgins JP, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA (editors). Cochrane Handbook for Systematic Reviews of Interventions version 6.2. Cochrane, 2021. Available from training.cochrane.org/handbook.
Morata 2005

Morata TC . Health effects of noise interactions at work, leisure and home. In: NTER-NOISE and NOISE-CON Congress and Conference; 2005 Aug 06-10; Rio de Janeiro, Brazil. Rio de Janeiro, Brazil: Institute of Noise Control Engineering, 2005:23-33.
Murphy 2011

Murphy WJ , Stephenson MR , Byrne DC , Witt B , Duran J . Effects of training on hearing protector attenuation. Noise and Health 2011;13 (51 ):132-41. [DOI: 10.4103/1463-1741.77215]
Murphy 2016

Murphy WJ , Themann CL , Murata TK . Hearing protector fit testing with off-shore oil-rig inspectors in Louisiana and Texas. International Journal of Audiology 2016;55(11) :688-98. [DOI: 10.1080/14992027.2016.1204470]
Nadon 2021

Nadona V , Bonneta F , Bouserhala RE , Berniera A , Voixa J . Method for protected noise exposure level assessment under an in-ear hearing protection device: a pilot study. International Journal of Audiology 2021;60 (1 ):60-9. [DOI: 10.1080/14992027.2020.1799082]
Neitzel 2006

Neitzel R , Somers S , Seixas N . Variability of real-world hearing protector attenuation measurements. Annals of Occupational Hygiene 2006;50 (7 ):679-91. [DOI: 10.1093/annhyg/mel025]
Nelson 2005

Nelson DI , Nelson RY , Concha-Barrientos M , Fingerhut M . The global burden of occupational noise-induced hearing loss. American Journal of Industrial Medicine 2005;48 (6 ):446-58. [DOI: 10.1002/ajim.20223]
NIOSH 1998

National Institute for Occupational Safety and Health. Criteria for a recommended standard- occupational noise exposure - 1998 [Criteria for a Recommended Standard- Occupational Noise Exposure - 1972]. Centers for Disease Control and Prevention, Department of Health and Human Services 1998;2nd (Publication Number 98-126 ):106 pp.
Nélisse 2012

Nelisse H , Le Cocq C , Boutin J , Laville F , Voix J . Estimation of noise levels and hpd attenuation in the workplace using microphones located in the vicinity of the ear. Canadian Acoustics/Proceedings of the Acoustics Week in Canada 2012;40 (3 ):122-3.
Park 1991

Park MY , Casali JG . A controlled investigation of in-field attenuation performance of selected insert, earmuff, and canal cap hearing protectors. Human Factors 1991;33 (6 ):693-714. [DOI: 10.1177/001872089103300606]
Prince 1997

Prince MM , Stayner LT , Smith RJ , Gilbert SJ . A re-examination of risk estimates from the NIOSH Occupational Noise and Hearing Survey (ONHS). Journal of the Acoustical Society of America 1997;101 (2 ):950-63. [DOI: 10.1121/1.418053]
Ramsay 2003

Ramsay CR , Matowe L , Grilli R , Grimshaw JM , Thomas RE . Interrupted time series designs in health technology assessment: lessons from two systematic reviews of behavior change strategies. International Journal of Technology Assessment in Health Care 2003;19 (4 ):613-23. [DOI: 10.1017/s0266462303000576]
Reeves 2021

Reeves BC , Deeks JJ , Higgins JP , Shea B , Tugwell P , Wells GA . Chapter 24: Including non-randomized studies on intervention effects. In: Higgins JP, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA (editors). Cochrane Handbook for Systematic Reviews of Interventions version 6.2. Cochrane, 2021. Available from training.cochrane.org/handbook.
RevMan 2024 [Computer program]

Review Manager (RevMan). Version 7.2.0. The Cochrane Collaboration, 2024. Available at revman.cochrane.org.
Rimmer 1997

Rimmer TW , Ellenbecker MJ . Hearing protector attenuation measurement by bone conduction loudness balance compared with real ear attenuation at threshold in a sound field. Applied Occupational and Environmental Hygiene 1997;12 (1 ):62-8. [DOI: 110.1080/1047322X.1997.10389457]
Rolison 2014

Rolison JJ , Hanoch Y , Wood S , Liu PJ . Risk-taking differences across the adult life span: a question of age and domain. Journals of Gerontology, Series B: Psychological Sciences and Social Sciences 2014;69 (6 ):870-80. [DOI: 0.1093/geronb/gbt081]
Samelli 2018

Samelli AG , Gomes RF , Chammas TV , Silva BG , Moreira RR , Fiorini AC . The study of attenuation levels and the comfort of earplugs. Noise and Health 2018;20 (94 ):112-9. [DOI: 10.4103/ nah.NAH_50_17]
Sayler 2018

Sayler SK , Rabinowitz PM , Cantley LiF, Galusha D , Neitzel RL . Costs and effectiveness of hearing conservation programs at 14 US metal manufacturing facilities. International Journal of Audiology 2018;57 (Suppl 1 ):S3-11. [DOI: 10.1080/14992027.2017.1410237]
Schulz 2011

Schulz TY . Individual fit-testing of earplugs: a review of uses. Noise and Health 2011;13 (51 ):152. [PMID: 21368441]21368441
Smith 2014

Smith PS , Monaco BA , Lusk SL . Attitudes toward use of hearing protection devices and effects of an intervention on fit‐testing results. Workplace Health and Safety 2014;62 (12 ):491-9. [DOI: 10.3928/21650799-20140902-01]
Social Security Bulletin 1982

Social Security Board, Social Security Administration, United States Office of Research, Evaluation, and Statistics. Social Security Bulletin: Meaning of the social security number. Vol. 45 (11) . Washington, DC: Social Security Board, 1982.
Stephenson 2011

Stephenson CM , Stephenson MR . Hearing loss prevention for carpenters: part 1 - using health communication and health promotion models to develop training that works. Noise and Health 2011;13 (51 ):113-21. [DOI: 10.4103/1463-1741.77207]
Sterne 2016

Sterne JA , Hernán MA , Reeves BC , Savović J , Berkman ND , Viswanathan M . ROBINS-I: a tool for assessing risk of bias in non-randomised studies of interventions. BMJ (Clinical Research Ed.) 2016;355 :i4919. [DOI: 10.1136/bmj.i4919]
Suter 2017

Suter AH . Occupational hearing loss from non-Gaussian noise. Seminars in Hearing 2017;38 (3 ):225-62. [DOI: 10.1055/s-0037-1603726] [PMID: 28740322]28740322
Suter 2022

Suter AH . Standard and regulations for occupational settings. In: Meinke DK , Berger EH , Driscoll DP , Neitzel RL , Bright K , editors(s). The Noise Manual. 6th edition. Falls Church (VA): American Industrial Hygiene Association (AIHA), 2022:485-508. [ISBN: 978-1-935082-98-9]
Takada 2020

Takada MM , Rocha CH , Neves-Lobo IF , Moreira RR , Samelli AG . Training in the proper use of earplugs: an objective evaluation. Work 2020;65 (2 ):401-7. [DOI: 10.3233/WOR-203092]
Teixeira 2021

Teixeira LR , Pega F , Dzhambov AM , Bortkiewicz A , da Silva DT , Andrade CA de , et al. The effect of occupational exposure to noise on ischaemic heart disease, stroke and hypertension: a systematic review and meta-analysis from the WHO/ILO Joint Estimates of the Work-Related Burden of Disease and Injury. Environment International 2021;154 :106387. [DOI: 10.1016/j.envint.2021.106387]
Themann 2013a

Themann CL , Suter AH , Stephenson MR . National occupational research agenda for the prevention of occupational hearing loss - part 1. Seminars in Hearing 2013;34 :145-207. [DOI: 10.1055/s-0033-1349351]
Themann 2013b

Themann CL , Suter AH , Stephenson MR . National occupational research agenda for the prevention of occupational hearing loss - Part 2. Seminars in Hearing 2013;34 (3 ):208-52. [DOI: 10.1055/s-0033-1349352]
Themann 2019

Themann CL , Masterson EA . Occupational noise exposure: a review of its effects, epidemiology, and impact with recommendations for reducing its burden. Journal of the Acoustical Society of America 2019;146 (5 ):3879. [DOI: 10.1121/1.5134465]
Tikka 2017

Tikka C , Verbeek JH , Kateman E , Morata TC , Dreschler WA , Ferrite S . Interventions to prevent occupational noise-induced hearing loss. Cochrane Database of Systematic Reviews 2017, Issue 7 . Art. No: CD006396. [DOI: 10.1002/14651858.CD006396.pub4]
Tsukada 2008

Tsukada T , Sakakibara H . A trial of individual education for hearing protection with an instrument that measures the noise attenuation effect of wearing earplugs. Industrial Health 2008;46(4) :393-6. [DOI: 10.2486/indhealth.46.393]
Ullman 2021

Ullman ED , Smith LM , McCullagh MC , Neitzel RL . Hearing loss as a predictor for hearing protection attenuation among miners. Occupational and Environmental Medicine 2021;78 :371-6. [DOI: 10.1136/oemed-2020-106838] [PMID: 33436381]33436381
Voix 2022

Voix J , Smith P , Berger EH . Field fit-testing and attenuation estimation procedures. In: Meinke DK , Berger EH , Neitzel RL , Driscoll DP , Bright K , editors(s). The Noise Manual. 6th edition. Fairfax (VA): American Industrial Hygiene Association (AIHA), 2022:309–28. [ISBN: 978-1-935082-98-9]
Wells 2024

Wells LL , Schulz TY , Saleem M , Dantscher S , Borst B , Giguere C , Fackler C , Murphy W . Standards and Regulations for Hearing Protector Fit Testing – Outcomes of the International Hearing Protector Fit-Testing Symposium. Proceeding of Meetings on Acoustics 2024;53 :1-12. [DOI: https://doi.org/10.1121/2.0001843]
WHO 1997

World Health Organization. The World Health Report 1997: conquering suffering; enriching humanity. World Health Forum 1997;18 (3-4 ):248-60. [PMID: 9478137]9478137
Wu 2016

Wu M , Guffey SE , Takacs BC . Comparison of noise reduction of earplugs between fit test and coal mine work and observed variability of real-time noise reduction in coal mine work. Work 2016;55 :531-8. [DOI: 10.3233/WOR-162439]
Zhong 2023

Zhong X , Fu SJ , Zhu JW , Li X , Shi B , Su SB . Research on the protective effect of hearing protective device for workers exposed to noise in a motor manufacturing enterprise. Zhonghua Lao Dong Wei Sheng Zhi Ye Bing Za Zhi 2023;41 (11 ):814-18. [DOI: 10.3760/cma.j.cn121094-20220815-00402] [PMID: 38073207]38073207
References to other published versions of this review

Morata 2021

Morata TC , Gong W , Tikka C , Samelli A , Verbeek JH . Effects of hearing protection field attenuation estimation systems and associated training on the level of noise attenuation in workers exposed to noise. Cochrane Database of Systematic Reviews 2021, Issue 10 . Art. No: CD015066. [DOI: 10.1002/14651858.CD015066]
