Occupational noise-induced hearing loss: What is the contribution of personal noise exposures? A case study of mineworkers at a large-scale platinum mine in South Africa

We conducted a cross-sectional analysis of retrospective data from two datasets, containing records of underground mineworkers employed by a platinum mine in Limpopo province, South Africa. The two datasets comprised records of hearing screening and occupational noise measurements for the period 2014–2018. and included demographic data, area noise exposure measurements, personal noise measurements, and calculated STS values. The 648 records represented 5% of the mine’s population that was exposed to 5% of the individual pieces of equipment that emitted excessive noise, as per the Mine Health and Safety Act.^7,20,21

The hearing screening dataset contained bilateral hearing screening records for individual mineworkers, from 2014 to 2018. Regardless of when a mineworker was initially employed at the mine, we used the 2014 - 2016 hearing thresholds as the baseline values; the 2018 hearing thresholds were used to calculate changes from baseline for each mineworker.¹⁴ The records were accessed from the audiometry medical examinations conducted by a qualified audiometrist (occupational health nurse) according to the South African National Standard (SANS) 10,083:2013.⁷ This dataset contained hearing thresholds at the frequencies of 0.5, 1, 2, 3, 4, 6, and 8 kHz, for each ear and the ear-specific standard thresholds (ST) and calculated bilateral STS.²²

The occupational noise dataset contained personal noise exposure measurements and area noise measurements. The personal noise exposure measurements for individual mineworkers were collected using noise dosimeters, positioned close to the ear, as a function of time, in 2018. The area noise measurements were collected using a sound level meter for various job areas. The data were captured in MS Excel by the mining company. All data were collected by qualified occupational hygienists and included mineworkers’ age, sex, workplace area, HEG, job title, and work shift (hours) in 2018.

The two datasets, containing 648 mineworkers’ records were merged in STATA (version 15.1), using the mineworker’s assigned employee number, age, and sex. After removing duplicate and incomplete records, the merged dataset contained 521 records. The data were categorised according to the miners’ personal noise measurements and corresponding risk rankings, as follows: ≤81.99 dB(A) (Group 0), 82–84.99 dB(A) (Group 1), 85–104.99 dB(A) (Group 2), and ≥105 dB(A) (Group 3). Hearing function (as hearing sensitivity accessed from the audiometry screening dataset) was classified as within normal limits (-10 – 25 dBHL), mild hearing loss (26–45 dBHL), moderate hearing loss (46–55 dBHL), moderate-severe hearing loss (56–70 dBHL), severe hearing loss (71–90 dBHL), and profound hearing loss (>90 dBHL).²³

The demographic, audiometry and work-specific characteristics of the mineworkers are summarised as numbers and frequencies (Table 1). Pearson Chi-square test was used to describe associations between personal noise measurements, age, sex and shift duration, and hearing deterioration (STS) in the left and right ears. Locally weighted scatterplot smoothing (LOWESS) was used to assess the relationships between personal and area noise measurements (Figure 1) and between personal noise exposure levels and hearing function (Figure 2). Multinomial logistic regression analysis was used to predict hearing deterioration, measured as STS (dependent variable), based on the mineworkers’ age, personal noise measurements, and work shift duration (independent variables), which were selected as predictor variables. These predictor variables were also controlled for in the model. The risk ratios are reported with 95% confidence intervals. Statistical significance was set at 5% (p-value = .05).

Permission to use the mineworkers’ data was granted by the mine management. Ethical clearance was obtained from the University of the Witwatersrand Human Research Ethics Committee (clearance certificate no. M180273).

Risk-based audiometry assessment in HCPs uses sampled records (5%) of integrated data from the personal noise measurements, audiometry and the HEGs (demographic and medical data) to identify risks associated with ONIHL.⁷ Data used for the current study were aligned with the CoP that guides the risk-based audiometry assessment since mineworkers’ personal noise measurements and audiometry screening datasets were used to predict risk for ONIHL. In addition, the two datasets used in our study consisted of the 5% sampled population accessed from the audiometry medical surveillance records of the mineworkers in line with the CoP for risk-based medical surveillance for HCPs to monitor mineworkers at risk of developing ONIHL.¹⁷ Medical conditions of the mineworkers could not be accessed from the mine, thus, were not included in our study, and this is noted as a limitation. However, the findings from our study could be used by the study mine to benchmark their CoP for their HCPs and measurable outcomes towards the prevention of ONIHL.

Worldwide, outcomes for effective HCP include strategies used by the mines to reduce occupational noise exposure levels to mitigate risk for ONIHL. From the data we analysed, personal and area noise measurements were used to describe noise exposure levels for the mineworkers in our study. Our findings indicated a positive association between the area and personal noise measurements at levels below 85 dB(A) and no association at high noise exposure levels (>85 dB(A)). Notably, for those mineworkers who were exposed to personal noise measurements higher than 85 dB(A), high noise exposure levels were attributed to localised sources of noise, rather than general area workspace occupational noise. This finding is imperative, indicating that the mineworkers’ noise source (job-specific equipment used by the mineworker) was the main contributor, and movements between various job activities contributed to higher personal noise measurements. Our study findings indicated targeted noise sources and were in agreement with the MHSC strategy of quietening specific equipment that emits high noise exposure levels to meet their noise reduction target by December 2024.⁵ However, it must be noted that profiling noise exposure levels using dosimeters was restricted to only five percent (5%) of the mineworkers who were profiled as at risk and were identified according to their HEGs by the occupational hygienist. Therefore, accurate recording of personal noise measurements may be used as a context-specific risk-based assessment strategy by the South African mines. In addition, personal noise measurements should be used to track progress made regarding buying quiet equipment, in turn, showing progress made towards reducing occupational noise exposure levels.

Profiling of mineworkers at risk for ONIHL includes monitoring noise exposure levels and tracking hearing function using STS, to monitor progress made by the South African mines towards meeting the set NIHL milestone targets.¹⁴ Firstly, most of the mineworkers in our study presented with STS within normal hearing (≤25 dBHL) in both ears. While the STS may fall within normal hearing limits, it is important to note that a mean shift of 8 dBHL was reported in our findings. A similar finding was reported by Ntlhakana et al. (2021). Although this shift was less than 10 dBHL, which is a recommended reportable shift according to the NIHL Regulations, any shift, however small, should still be reported for monitoring purposes.¹⁴ Therefore, we posit that the STSs within normal hearing limits do not imply a no shift but rather a risk-based surveillance marker for early intervention. In addition, a shift may not meet the threshold for clinical or regulatory significance, but in the case of the South African mineworkers who are at risk of exposure to high levels of occupational noise, any shift should be documented accordingly. Secondly, our findings indicated that personal noise exposure measurements were not a significant predictor of STS for the mineworkers. This holds true since most of the sampled mineworkers were exposed to noise levels that were less than 85 dB(A). Ideally, similar findings would be reported in contexts where an effective HCP is in place, particularly through consistent use of PPE, engineering and administrative controls which reduce noise levels below 85 dB(A), but this was contrary to previous study findings in the South African mines.^3,6,13 However, age was found to be significantly associated with the mineworkers’ STS, this finding correlates with Grobler et al. (2020) and Strauss et al. (2014) but contradicts Ntlhakana et al. (2021). The fact that the current study used personal noise measurements as opposed to area noise measurements meant that differences in findings should be noted. Specific to our case, mineworkers’ changes in hearing thresholds were mostly associated with age and not necessarily occupational noise exposure levels.^2,19 Profiling of mineworkers using personal noise measurements to determine risks for ONIHL was imperative. Without minimising hearing health risks associated with noise exposure levels, especially in a South African context, our case study highlighted that mineworkers’ age was a significant risk factor for ONIHL. Thus, in the absence of personal noise exposure measurements, mineworkers’ age should be used as an outcome measure for ONIHL.

A clearly defined audiometry surveillance protocol intended as a risk assessment tool for the South African mines could be a useful measure to accurately track the hearing function of mineworkers and to identify early those at risk for ONIHL. Our study showed that the mine conducted a risk-based audiometry surveillance, using personal noise measurements on a sample of mineworkers. The regression model we used revealed that the sampled mineworkers’ age was significantly associated with hearing deterioration for mineworkers with STS at mild, moderate, and moderate-severe hearing thresholds. The current study findings were in agreement with the study we conducted previously, which predicted STS, that age was a significant risk factor associated with hearing deterioration.^15,19 However, the frequency range selected to calculate the STS was 2000 Hz, 3000 Hz and 4000 Hz, as outlined in the revised South African NIHL Regulations, was restrictive and does not offer a broad high-frequency spectrum sufficient to identify early signs of ONIHL.^14,15 This offers an opportunity to review the STS definition and to redefine the audiometry surveillance protocol as a risk assessment tool specific for the South African mines to accurately track mineworkers’ hearing deterioration and predict STS as a risk for ONIHL.

Age and sex are well documented risk factors associated with ONIHL,^3,24 thus, susceptibility to hearing loss cannot be separated from the two risk factors.²⁵ Our findings from our regression model indicated that age was a significant risk associated with hearing deterioration for the mineworkers who presented with STS at mild and moderate hearing sensitivity. Our findings agree with previous study findings,^3,26,27 but our findings went further to show the severity of hearing deterioration for these mineworkers. In addition, we have provided evidence-based findings of the risk assessment tools (personal noise measurements and audiometry records) used by the mine to track the mineworkers’ hearing function.^18,19 Therefore, a concerted effort by the study mine has been acknowledged in our findings, and other mines in South African mines may reflect on their outcomes for risk-based audiometry surveillance and the effectiveness of their HCPs using our study as a guide.

Duration of noise exposure (hours/weeks/year) in any environment, either at work or elsewhere, increases the risk for other health conditions such as heart conditions, discomfort and sleep among adults,^28,29 and, specific to our case, the risk for ONIHL. In South Africa, the legislated work shift is 8 hours per day and 40 hours per week.³⁰ Our study findings indicated that most (n = 445; 85.4%) mineworkers worked more than 8 hours per shift. However, our study findings indicated that shift duration was not significantly associated with changes in STS. But it must be noted that most of the sampled mineworkers were exposed to personal noise levels less than 85 dB(A), hence, duration of exposure of more than 8 hours was not a factor as indicated in our findings. Although our findings provided evidence that most of the mineworkers worked for more than 8 hours per shift, future studies could investigate the cumulative effect caused by extended shift duration for mineworkers exposed to noise levels greater than 85 dB(A), to account for the associated risk of ONIHL for the South African mineworkers. The South African mines could use datasets drawn from sampled personal noise measurements since they contain duration of noise exposure (shift/hour) for the mineworkers to guide future decision-making towards the prevention of ONIHL.