Can workers' reports of their tasks help estimate daily noise exposure as well as direct observation?

Details

Noise exposure can be assessed using full-shift dosimetry or short-term task-based measurements. Task-based measurements offer opportunities to identify high-noise tasks directly, target control measures and estimate full-shift exposure, but little empirical evidence has compared the two methods.

The National Institute for Occupational Safety and Health conducted an observational method-comparison study at three industrial facilities, assessing noise exposure with both dosimetry and task-based methods to examine their agreement and the causes of disagreement. Researchers created eight indices of task-based full-shift exposure. They used three ways to assess time spent on tasks—direct observation by an industrial hygienist, workers' estimates at the end of the shift and supervisors' estimates—and three ways to assign task noise levels: direct measurement, arithmetic mean (AM) and geometric mean (GM). They evaluated precision, bias and absolute agreement using Bland-Altman plots and the concordance correlation coefficient (CCC).

Overall, task-based methods worked fairly well, with mean biases less than ±2.8 dBA and precision ranging from 3.3–4.4 dBA. Estimates using supervisors' assessments of time spent on tasks showed the poorest agreement with dosimetry by every measure. Estimates using workers' assessments of time spent on tasks generally agreed with dosimetry as well as those using direct observation. For assigning task noise levels, direct measurements produced the best agreement with dosimetry, while agreement for indices using task AM or GM noise levels varied.

The combination of directly observed task time and directly measured task noise levels achieved the best agreement with dosimetry: CCC was 0.84, 95% of the differences were within 7.4 dBA, and 56% of the differences were <3 dBA. This index had high accuracy (accuracy coefficient = 0.96); the main source of disagreement was a lack of precision (precision coefficient = 0.88). When measurements were grouped by job characteristics, agreement improved significantly in categories with low job mobility, low job complexity and low job variability.

The findings suggest that task-based and dosimetry-based estimates of full-shift exposure can achieve a high degree of absolute agreement. The task-based approach combining worker reports with task AM or GM noise levels was similar to the more time-intensive direct observation method for estimating full-shift exposure.