The Annex C fallacy: why unscreened databases are usually preferable for comparison of industrially exposed groups.

One can study occupational noise exposure by comparing hearing thresholds of exposed people to control data from national or international standards. ANSI S3.44 (1996) offers Annex C--thresholds for people without occupational noise exposure--as appropriate control data for such comparisons. Annex C is based on the false assumption that people who have had occupational noise exposure are similar in all other important ways to those without such exposures. In fact, people with noisy jobs are more likely than others to be smokers, diabetics, poorly educated, white and exposed to non-occupational noise. Taking these other risk factors into account, the appropriate thresholds for comparison to industrial study populations are closer to those of the unscreened population than to an 'Annex C' population that simply excludes occupationally noise-exposed persons.