We appreciate Dr Jensen's comments on the Central American Survey on Working Conditions and Health (ECCTS) and his concern that our sampling methodology may have produced biased results. The ECCTS has broadly followed the methodological criteria of the European Working Condition Survey (EWCS). Of note is that, to estimate prevalence of exposures to different working conditions, the EWCS is applied every 5 years to a representative sample of only 1000 workers in the majority of European countries. The ECCTS had double the number, that is, 2000 per country. However, we acknowledge that a larger sample is better and, in fact, some European countries have started to increase their sample size. To achieve representativeness, a national population sample must be properly spread over geographic sub-areas and population sub-groups. The random selection of a large number of census segments, proportional to the respective populations of the departments or provinces, accounted for geographic regions and levels of urbanisation, similar to the stratified procedures in the EWCS. In addition, we applied weights for sex, age and economic sector to each individual in the national samples to correct for differences in the sample with the underlying national working population with regard to these key socio-demographic parameters. Finally, for regional comparisons we applied an additional weight to adjust for the population size of the different countries. However, unlike the EWCS, we did not adjust for type of industry and occupation because the latter information was not always available from the census. As Dr Jensen points out, this would have been a more ideal approach. The ECCTS is a first experience in central America and, as such, prone to improvements. Despite limitations, the results from the central American survey provide a first, baseline measure about specific working and employment conditions that allows for future monitoring across countries and serves as a useful tool for prevention. Further analysis, comparing prevalence of exposures among countries, has to be properly adjusted by economic activity and job category. All are available in the dataset upon request. In conclusion, the first ECCTS represents a key step towards the development of a permanent information system for informed regional and national planning of preventive policies. The experience gained in the first central American survey will allow improvements in future surveys.
Professor Fernando G Benavides, Center for Research in Occupational Health, Pompeu Fabra University, Doctor Aiguader, 80 Barcelona 8003, Spain