How many samples?

As I’ve discussed in a previous post, there are many factors which lead to a wide variation in exposure for workers carrying out the same job. One implication of this is that it is very dangerous to draw conclusions from one or two samples. But how many samples need to be taken to ensure that exposure is adequately characterised? In practice, it’s difficult to draw up general rules which can be applied in every situation.

Over the last couple of years the British Occupational Hygiene Society in conjunction with their sister organisation in the Netherlands, Nederlandse Vereniging van Arbeidsdeskundigen, have been working on preparing some guidance on sampling strategy for occupational hygienists. This has now been published and can be viewed or downloaded from here. Recommendations are made on the number of samples that should be taken during a two stage approach to be used specifically for testing compliance with exposure limits.

1. Take 3 representative personal exposure measurements from random workers in the SEG. If all three exposures are <0.1xOEL, it can be assumed that the OEL is complied with.
2. Do a group compliance test. Take at least 6 more samples from the SEG, at least 2 per worker from workers picked at random. Then carry out a statistical analysis on all 9 results

I think that the science on which the guidance is based is sound, but I doubt that the approach set out is practicable for most organisations. The problem isn’t so much the number of samples (although the analytical costs could be a problem in some cases), but the time required to collect them. Following this guidance would probably involve four site visits and many companies wouldn’t be able to justify the cost involved.

So deciding on how many samples to collect can be one of the most difficult aspects of devising a sampling strategy. Although a better picture of exposure will be obtained if a large number of samples are collected,  a little common sense needs to be used, balancing the value of the results against the cost of the survey.

Advertisement