Effectiveness of interventions in reducing pesticide overexposure and poisonings.

OBJECTIVE The objective of this paper was to review the effectiveness of interventions to reduce pesticide overexposure and poisonings in worker populations. METHODS We used the Cochrane Collaboration search strategy to search the following databases for articles that tested the effectiveness of interventions in reducing human pesticide exposure or poisonings: MEDLINE, EMBASE, and Occupational Safety and Health (NIOSHTC). Interventions considered included comparisons of pesticide application methods, pesticide mixing methods, worker education, biological monitoring programs, personal protective equipment (PPE) use, pesticide substitutions, and legislation. The outcomes of interest included biological monitoring measures or personal exposure monitoring indicating a reduction of pesticide exposure, observed increased use of PPE, reduction in lost workdays, and where possible, evidence of changes in pesticide poisoning rates as identified by registries and population surveys. Studies were reviewed in depth with special attention to size and study design. RESULTS Most studies evaluated exposure during differing configurations of PPE or during different mixing or handling methods. Most studies were small field tests of protective equipment involving less than 20 workers. Some studies examined biological indices of exposure such as cholinesterase or urinary metabolites. Studies showed that PPE was effective in reducing exposure. No controlled studies were found that addressed reducing pesticide poisonings. CONCLUSIONS Changes in application procedures, packaging, mixing, use of personal protective equipment, and biological monitoring reduced pesticide exposure under controlled conditions. Cholinesterase monitoring can identify workers with a higher risk of overexposure. Most techniques were not tested in actual worksite programs. Interventions should be examined for their ability to reduce pesticide overexposure in actual working populations. No controlled evaluations of large legislative initiatives were found.