[Reliability and validity of a generic job exposure matrix applied on a small-business].

Objective: to evaluate the reliability and validity of a generic job exposure matrix (JEM) applied in a small business.

Methodology: procedures to evaluate a JEM integrated by six sections: the number of exposed workers per area, frequency of exposure, time of exposure time, level of exposure, safety controls, and proximity to source of exposure, was evaluated. The JEM also obtains information about possible health effects from exposure to occupational/environment agents. Two observers estimated the risk of exposure to epoxy resins on 31 workers of an epoxy resin facility in Mexico City. The rater agreements between the two observers were assessed through percent agreement (PA), weighted kappa (kappa(w)) and the intraclass correlation coefficient (ICC).

Results: disagreements were greater for the number of exposed workers (PA = 61.3, kappa(w) = 0.24, ICC = 0.33), level of exposure (PA= 66.7, kappa(w) = 0.25, ICC= 0.56), and safety controls (PA = 54.8, kappa(w) = 0.23, ICC = 0.69) sections. Percent agreement and kappa(w) were 64% and 0.58, respectively. In accordance with Landis and Koch, Altman, Fleiss, and Byrt classifications for the interpretation of kappa value, the weighted kappa (0.58) ranged from moderate to a fair good level.

Conclusions: despite the discordance in some sections, the JEM proved to be useful to identify the risk of exposure in this type of small business.