Article Text
Abstract
OBJECTIVES: To assess the reliability of a classification scheme and interview questions to be used for retrospective expose assessment in a study on reproductive disorders among hairdressers. Based on the presence of an air cleaning device (yes or no) and setting waves (yes or no), this scheme divides hairdressers into groups with potentially high and low exposure to solvents. The reliability of this and other schemes was assessed. Also; the reliability of self reports on other job characteristics was evaluated. METHODS: The monitored hairdressers were interviewed one or two years after measurements were performed. Based on the interview information, hairdressers were classified into exposure groups according to the original and other classification schemes. Measured ethanol concentrations were compared between the classified exposure groups. Furthermore, the interview answers were compared with the registered information one to two years ago. RESULTS: Using self reports, the original scheme resulted in mean ethanol concentrations (used as indicator variable) of 11.8 and 7.4 mg/m3 for the high and low exposure groups, respectively. The resolution was slightly less than for the original classification based on observations (15.0 and 7.1 mg/m3). Surprisingly, the self reported presence of any ventilation device resulted in more contrast in mean exposure concentration between the groups (17.4 and 7.5 mg/m3, respectively). Hairdressers reported validly on salon characteristics such as the type of salon, the number of hairdressers that worked in the salon, and the presence of ventilation devices, but could not make a distinction between different kinds of ventilation devices. The numbers of customers and tasks performed were largely overreported, but most variables correlated significantly with the information registered during the measurements. CONCLUSION: The self reported presence of any ventilation device is most predictive for the level of exposure to ethanol in the hairdressing salon. Questionnaire data on work characteristics should be treated with caution.