Application of data on compliance to epidemiological assessment of exposure-response: the case of data on exposure of United States coal miners.
Attfield MD; Hearl FJ
Occup Hyg 1996 Jan; 3(1-3):177-184
The extent of bias among data contained in the main database on hazardous exposures in coal mines was studied. Full shift samples collected during unannounced visits by the Mine Safety and Health Administration Special Inspection Program were extracted and analyzed. Small mines with fewer than 120 employees were selected as tampering with data obtained by the coal mine operators themselves for compliance purposes was suspected. These were also mines with poor records of compliance with health and safety regulations. A considerable difference arose between data collected by the operators and data collected by the special inspection teams. The authors conclude that special account must be taken of dust concentrations in excess of 0.1mg/m3 when deriving estimates of bias, and the low concentrations which had been reported by the operators of the mines may preferentially represent situations where there was a lack of compliance with standards. This suggests that simple proportional reallocation of the excess low data across the remainder of the range of data may not be appropriate. The authors conclude that while there was poor compliance in the small mines and there was evidence of bias in sampling, data for larger mines provides no evidence that the correction should be larger than 13%.
NIOSH-Author; Epidemiology; Coal-miners; Mining-industry; Air-quality-monitoring; Coal-dust; Airborne-particles; Dust-exposure; Occupational-exposure