Advanced noise-induced deafness among workers in singapore - what has changed?

CONTEXT: Noise-induced deafness (NID) contributes a significant disease burden internationally, and is a leading occupational disease in Singapore. AIMS: This study profiles the epidemiological characteristics of advanced noise-induced deafness (NID(A)) cases and explores whether these have evolved with time. SETTINGS AND
DESIGN: A quantitative retrospective analysis of prior NID(A) cases was performed. National records of NID(A) cases from 2001 to 2010 were obtained, with permission from the Ministry of Manpower, consisting of worker audiograms, case records, and noise monitoring reports. METHODS AND MATERIAL: Comparison was made with data from a prior study (127 cases, 1985-1994) to identify shifts in NID(A) epidemiology; 71 out of 73 NID(A) case records (2001-2010) from the national data were reviewed. STATISTICAL ANALYSIS USED: Statistical Package for Social Sciences (IBM) was used for statistical analysis.
RESULTS: Mean noise exposure (24-29.6 years) and age at diagnosis (48-54.6 years) have risen. Total case numbers (127 to 73 cases), average hearing loss levels (61.5 to 56.0 A-weighted decibels), and delayed reporting of NID(A) cases (82.7% to 47.9%) have decreased. Metal manufacturing and marine industries remain top contributors (69.1%, from 68.5%); 31% were noncompliant with hearing protector (personal protective equipment, PPE) use and 38% did not use PPE properly.
CONCLUSIONS: NID(A) case profiles have shifted over time, with reduced case numbers, lower hearing loss levels at diagnosis, and earlier case notification, possibly linked to improvements in legislative coverage and NID prevention programs. Changes in age and average duration of noise exposure may be related to these improvements. Early case notification, PPE compliance, and proper usage should be focus areas for NID prevention and hearing conservation programs.

INTRODUCTION

Prolonged exposure to excessive noise, such as that generated via industrial processes and activities in the workplace, may lead to noise-induced deafness (NID). This is a condition in which permanent loss in hearing sensitivity may result from irreversible damage to cochlear hair cells, causing communication difficulties, social isolation, and degradation in quality of life. Workers may experience a variety of problems, ranging from tinnitus to difficulty in discerning sounds in the presence of background noise.[1] In the workplace, these problems may lead to loss of recognition of warning signals and hamper the ability to localize sound sources or discriminate between different sound frequencies.[2]

NID has a characteristic pattern of hearing loss that is initially worse in the higher frequencies. Cochlear damage initially occurs in the segment that detects sound at 4000 Hz, as the resonance characteristics of the outer ear amplify acoustic energy in the higher frequencies by up to 10 decibels (dB). The next affected areas are often in the 3000 and 6000 Hz regions, as the hair cells in this region are more sensitive to acoustic damage. This accounts for the pattern of sensorineural hearing loss seen in NID, which typically shows a “dip” in audiograms at the 4000 Hz range, signifying hearing loss at those frequencies.

The rate of hearing loss due to noise is greatest in the first 2 decades of exposure, contrasting with age-related loss, which accelerates over time. As such, given its insidious onset and the fact that hearing deficit initially occurs in the higher nonspeech frequencies, the bulk of irreversible damage has often occurred by the time hearing loss is clinically detected. Workers often remain functionally normal until their hearing has deteriorated to the point where hearing loss impinges upon the audible frequencies of normal conversation and environmental/workplace sounds. It is thus important to study and analyze the contributory factors behind the development of NID and its effect on workers.

Scope of problem

International estimates show that approximately 16% of the worldwide hearing loss disease burden is attributable to occupational noise exposure.[3] Approximately 4.1 million disability-adjusted life-years per year are attributed to occupational NID, affecting males more than females, and with a greater proportional burden in the developing world.[4]

NID has traditionally been the most prevalent occupational disease in Singapore, comprising a large portion (53%) of occupational disease cases confirmed by the Ministry of Manpower (MOM) in 2017, with a total of 329 cases.[5] Manufacturing, marine, and construction represent the major industries contributing to the NID burden.

NID in Singapore is classified as early NID (NID(E)) when the exposure duration is equal to or greater than 5 years, and the average hearing loss (AHL) is equal to or less than 50 A-weighted decibels (dBA) in the better ear (averaged across the hearing threshold levels at 1000, 2000, and 3000 Hz).[6] advanced noise-induced deafness (NID(A)) is when the occupational exposure is equal to or greater than 10 years, and the AHL is equal to or greater than 50 dBA in the better ear.

Apart from occupational noise exposure, the high prevalence is also, in part, due to a robust legal structure, with the Workplace Safety and Health Act providing a framework for the mandatory surveillance and detection of NID cases. Workers exposed to excessive noise in the workplace have to undergo mandatory audiometric examinations[7] to detect NID, and it is compulsory for companies with high noise exposure levels to develop, implement, and monitor noise control programs.

Current workplace hearing conservation programs in Singapore include, among other measures, conducting noise level monitoring for high noise exposure workplaces, provision of hearing protectors, and worker education.[8] Additionally, NID is a compensable disease under the Work Injury Compensation Act.[9]

NID(A) in Singapore

A previous study by Tay (data covering 1985–1994)[10] demonstrated that the bulk of NID(A) cases were from shipyards (44.9%), manufacturing (23.6%), and transport (7.2%). These are heavy industries traditionally known to have high noise exposure levels. This is in contrast to other Asian countries where agriculture, foresting, and mining activities are significant contributors to the burden of occupational NID.[11] Additionally, the mean age at diagnosis was 48 Standard deviation (SD 8) years, signifying that many workers actually have to spend more than a decade of productive working life with the effects of NID(A).

The mean duration of noise exposure at time of NID(A) diagnosis was 24 (SD 9) years, consistent with the long exposure duration necessary to cause severe NID, and the mean AHL was significantly higher than the diagnosis threshold of 50 dBA, at 61.5 (SD 4.3) dBA. Also, the mean average workplace noise level was 90 dBA over 8 hours, again significantly exceeding the permissible exposure limit of 85 dBA. Finally, 82.7% of workers actually fulfilled NID(A) diagnostic criteria when they were first notified to the MOM, signifying that a large proportion of cases were actually diagnosed late.

Objective and Aims

Despite strong regulatory support and the presence of well-developed hearing conservation programs in a majority of workplaces with high noise exposure over the last few decades, many workers continue to develop NID, with contributory factors ranging from the nonmodifiable, such as genetic susceptibility,[12] to the modifiable, such as noncompliance with hearing protectors.

This study aims to explore whether the epidemiological characteristics of NID(A) cases have evolved over time, in response to a changing industrial landscape and broadened legislative coverage in Singapore, and, through this, to identify shifts in the epidemiology of NID(A) over time, pinpoint risk factors and occupations at risk, correlate noise exposure levels with NID(A), correlate the use of personal protective equipment (PPE) with NID(A), and profile NID notification trends in Singapore.

Subjects and Methods

A quantitative retrospective analysis of prior NID(A) cases was performed. National database records of NID(A) cases for the period 2001 to 2010 were obtained, with permission from the MOM. This consisted of worker audiograms, case records, and noise monitoring reports.

Cases were tagged with unique anonymous identifiers, and extracted data was analyzed using Statistical Package for Social Sciences. A retrospective data analysis was conducted to identify epidemiological factors, including participant demographics, occupational history/noise exposure history/noise monitoring levels, and usage of hearing protectors. All data was securely stored, and institutional review board approval was obtained for the purpose of this study.

RESULTS

A total of 73 workers were on record as having been diagnosed with NID(A) during the period 2001 to 2010. Of these, 71 cases were included for the analysis. One case was excluded because of missing case records, and one case was excluded as the criteria for NID(A) was not strictly met. All 71 included cases were male, with 89% Chinese, 7% Malays, and 4% Indians. No significance testing was performed for this because of the inability to obtain an accurate racial breakdown of workers in noise-exposed industries in Singapore.

A breakdown of the industries of the workers with NID(A) is provided in Table 1. Wherever possible, the workers’ industries were grouped in a similar nature to that used in the previous comparable study (NID(A) cases from 1985 to 1994), so as to provide comparison over time.[10] There are several differences; first, given the shifts in Singapore’s industrial landscape over time, stone quarrying and wooden furniture manufacturing are no longer significant. Next, there has been an increase over time in other types of manufacturing, necessitating the creation of an “other manufacturing” category. This includes products such as polymers, biofuels, refined petroleum products, or waxes/polishes.

Table 1

Industry breakdown

Industry	1985–1994		2001–2010
	Number	Percentage (%)	Number	Percentage (%)
Building and repair of ships and boats	57	44.9	19	26.8
Basic metal industries, metal stamping and forging, manufacturing of metal products, manufacturing and repairing of machinery, and manufacturing of storage batteries	30	23.6	30	42.3
Transportation and allied services	9	7.2	6	8.5
Manufacturing of food and drinks including dairy products	7	5.5	2	2.8
Manufacturing of glass, cement, and structural clay products	5	3.9	5	7.0
Other manufacturing	–	–	8	11.3
Construction	–	–	1	1.4
Stone quarrying	7	5.5	–	–
Manufacturing of wooden furniture	5	3.9	–	–
Others	7	5.5	–	–
Total	127		71

The job descriptions of affected workers were similar in nature across the industries. Sources of noise were also similar, mainly coming from work processes such as hammering, grinding, or welding, and heavy machinery such as air compressors, metal parts fabrication, or engines/generators. A breakdown of job scopes in the main industry groups is provided in Table 2.

Table 2

Job types at risk

Industry group	Job description	Other noise sources
Building and repair of ships and boats	Welder, grinder, hammerer, plumber, fitter	Air compressors, engine noise, fabrication/repair of parts, grit blasting
Basic metal industries, metal stamping and forging, manufacturing of metal products, manufacturing and repair of machinery, and manufacturing of storage batteries	Welder, grinder, hammerer, plumber, fitter, machine operator, production line operator	Air compressors, engine/machinery noise, machining of metal parts, drilling
Transportation and allied services	Technician, machine operator, hammerer	Air compressors, aircraft noise, fabrication/repair of parts
Manufacturing of food and drinks including dairy products	Machine operator, production line operator	Machinery repair, welding/hammering
Manufacturing of glass, cement, and structural clay products	Welder, grinder, hammerer, fitter	Air compressors, engine noise
Other manufacturing	Production line operator, forklift driver, technician, fitter	Air compressors, engine noise
Construction	Crane rigger	Machinery repair, welding/hammering/grinding

For the purposes of comparison, the workers were divided into three groups: shipyard workers, metal industries workers, and other workers. This provided three groups of comparable size, and also divides the top three contributing industries into three separate groups. Due to the small sample size, further division into more than three groups for greater granularity was not viable from a statistical standpoint. The results of all analyzed parameters are presented in Table 3 and Figure 1.

Table 3

Comparative analysis across industries

		All	Shipyards	Metal industries	Others	P valuea
Mean age at NID(A) diagnosis (years (SD))		54.6 (7.0)	53.7 (5.3)	55.2 (8.8)	54.6 (5.6)	0.758
Mean noise exposure duration at NID(A) diagnosis (years (SD))		29.6 (10.7)	32.4 (14.2)	29.2 (9.0)	27.7 (9.2)	0.365
Mean dosimetry (dBA (SD))		89.3 (4.3)	91.1 (7.5)	88.8 (3.8)	89.6 (4.0)	0.545
Mean AHL when first notified (dBA (SD))		49.0 (10.9)	49.2 (9.6)	48.5 (10.2)	49.4 (12.3)	0.961
Mean AHL 5 years pre-NID(A) diagnosis (dBA (SD))		47.2 (10.2)	49.1 (7.5)	45.3 (9.4)	47.4 (13.6)	0.643
Mean AHL 3 years pre-NID(A) diagnosis (dBA (SD))		53.3 (8.9)	49.8 (6.7)	55.4 (6.0)	55.2 (12.9)	0.203
Mean AHL at NID(A) diagnosis (dBA (SD))		56.0 (6.6)	54.7 (3.9)	54.8 (3.6)	58.8 (10.1)	0.058
Proportion meeting NID(A) AHL criteria on first notification to MOM (%)		47.9	42.1	40.0	63.6	0.251
Proportion with tinnitusb (%)		66.2	63.2	66.7	68.2	0.942
PPE compliance (%)	100%	69.0	63.2	63.3	81.8	NAc
	>50%	26.8	31.6	33.3	13.6	NA
	<50%	4.2	5.3	3.3	4.6	NA
Adequacy of PPE technique (%)		62.0	52.6	66.7	63.6	0.604

P values for all means were tested using Analysis of Variance (ANOVA one-way) and for all proportions were tested using Pearson’s Chi-square test. b Proportion with tinnitus, PPE compliance levels, and adequacy of PPE technique were assessed at the point of NID(A) diagnosis. c No tests of significance were carried out for PPE compliance in view of the small subgroup sizes. AHL = average hearing loss; MOM = Ministry of Manpower; NA = not applicable; NID(A) = advanced noise-induced deafness; PPE = personal protective equipment; SD = Standard deviation.

Figure 1

Distribution of personal protective equipment (PPE) types used.

DISCUSSION

All measured parameters were comparable across the three industry groups, with no statistically significant differences. All P values were greater than 0.05, and all estimates were close to one another. There is therefore no industry that stands out from the others when considering NID(A).

The AHL levels are noted to worsen over time, from 5 years pre-diagnosis to the point of diagnosis, which is within expectations. However, the results should not be used to further comment on the rate of change per year, as most of the audiograms at the 5 years and 3 years pre-diagnosis points were performed at the workplace in less ideal conditions, compared to the “gold standard” audiogram performed at the point of diagnosis in hospital. Moreover, because of many workers being notified late, or the unavailability of old audiogram records from the company, a significant portion of audiogram results were missing for the 5 years and 3 years pre-diagnosis periods.

Next, tinnitus is a significant symptom, with 66.2% of workers experiencing tinnitus by the time of NID(A) diagnosis. In comparison, a previous study by Phoon et al. [13] found that up to 23.3% of Singaporean workers with both early/advanced NID experience tinnitus, and of these workers, 42.4% had bilateral symptoms and 30% reported that tinnitus significantly affected their sleep and activities of daily living. Given the higher prevalence of tinnitus among workers diagnosed with NID(A), tinnitus could possibly be used as an early screen preceding hearing loss. Also, hearing conservation programs and workplace risk assessments could take this into account when screening and planning for workers with NID, as tinnitus may have effects separate from those of hearing loss, such as the inability to discern alarm bells from background tinnitus and ambient noise.

Workers would have been reviewed several times at the point of NID(A) diagnosis by the Designated Workplace Doctors and other medical specialists, and would have been aware of their diagnosis. Additionally, they would have had prior education and counselling on NID, the importance of PPE compliance, and on proper PPE usage techniques. Despite this, 31% of workers are still not 100% compliant with PPE usage, and 38% were still assessed to have inadequate technique for wearing PPE, at the point of NID(A) diagnosis. This potentially represents an area to focus upon as part of improving hearing conservation programs.

Epidemiological shifts over time

In comparison with the previous study by Tay,[10] the distribution of industries with NID(A) cases has changed over time, with stone quarrying and wooden furniture manufacturing being phased out in Singapore. The top three contributing industries remain unchanged, although shipyard workers now take up a lower proportion of cases (26.8% from 44.9%), with metal manufacturing having a corresponding increase (42.3% from 23.6%). The contribution from transportation services remains almost the same (8.5% from 7.2%).

The increase over time in other types of manufacturing, such as polymers, biofuels, or refined petroleum products, may be either related to the development of advanced manufacturing over time or an increased pickup of cases in these industries because of the increase in regulatory coverage to encompass all workplaces. This may be the tip of the iceberg, with more cases to come from nontraditional industries over time.

Table 4 highlights several changes in the epidemiology of NID(A) cases over time. The mean noise exposure duration and age at NID(A) diagnosis have risen compared to the previous study. This is possibly because of better workplace safety and health management, such as improved effectiveness of noise control measures or hearing conservation programs, so it takes longer for sufficient cumulative noise exposure to develop hearing loss. Another reason for an increased age at NID(A) diagnosis may also be the onset of presbycusis after 50 years, when hearing loss due to age contributes to causing AHL levels to rise above 50 dBA to meet the NID(A) diagnosis criteria.

Table 4

Comparison between studies

	Then (1985–1994)	Now (2001–2010)
Noise exposure (years)	24	29.6
Age at NID(A) diagnosis (years)	48	54.6
AHL at diagnosis (dBA)	61.5	56.0
Late notifications (%)	82.7	47.9
Total case numbers	127	73

AHL = average hearing loss; NID(A) = advanced noise-induced deafness.

The AHL at diagnosis is also lower. Possible reasons include a greater proportion of workers being diagnosed earlier as they were notified as potential cases before reaching NID(A) criteria, better surveillance and detection for NID cases, and improved noise control measures leading to overall lower noise exposure.

The proportion of cases that already met NID(A) criteria on first notification to MOM has decreased significantly with time, likely because of the implementation of noise regulations requiring regular audiometric monitoring; therefore, cases are detected and notified earlier. However, the percentage remains relatively high at 47.9%. This is possibly because of late notifications, with cases being notified to MOM at a later stage of hearing loss. Also, workers may change their employment status with a possibility of becoming lost to follow-up, and may not be notified. The doctor who sees such workers may lack access to old records and may treat it as the first abnormal audiogram, delaying notification. Next, it may also be due to the broadening of legislative coverage from the Factories Act to the Workplace Safety and Health Act in 2006, leading to workers from hitherto unknown workplaces with existing severe hearing loss being picked up at their first audiogram. Finally, this could also be due to missed audiograms, when contract or temporary workers miss out on having their audiograms done, or regular workers who may miss their yearly audiograms appointments, for varied reasons, such as being posted overseas. Detection and reporting of NID cases remains an area that could be improved upon, and which can be achieved through education or enforcement on responsible parties to notify cases at an earlier stage.

Finally, the total NID(A) case load is lower for a similar 10-year period. A possible explanation is that because of better noise regulations, workers are picked up earlier at the NID(E) stage instead, which correlates with the decreased percentage of late notifications over time.

CONCLUSION

In conclusion, shifts in the epidemiological profiles of NID(A) cases over time appear promising. Workers now require a longer duration of noise exposure before being diagnosed with NID(A); the average AHL magnitude at diagnosis is now lower and total case numbers are down despite expansion of the Workplace Safety and Health Act to cover more workplaces. Although these changes could be linked to improvements in legislation, regulatory enforcement, and workplace hearing conservation programs, further study is needed to confirm this and to identify other possible contributory factors.

Additionally, a greater proportion of workers are now notified to the MOM before they reach NID(A) criteria, although the proportion of workers who were notified late remains high, at 47.9%. Although this has improved over time, but still remains an area that could be improved upon, for example, through education or enforcement on responsible parties to notify cases at an earlier stage.Next, tinnitus is a significant symptom in NID(A), with 66.2% of workers experiencing it. This bears further study as a potential symptom for prompt detection of severe NID cases as part of hearing conservation programs, before clinical hearing loss has developed.

The results also suggest that no industry-specific differences exist with regard to NID(A) case profiles. However, with the proliferation of nontraditional high-technology manufacturing, newer industries, such as solvents, may present additional occupational risks for hearing loss.[14] Going forward, these industries may potentially contribute a growing proportion of hearing loss cases, which warrants closer attention.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.