Berg 2009.
Methods | Cluster‐randomised controlled study | |
Participants | Agricultural students involved in farm work n = 753 USA 34 schools |
|
Interventions | Intervention: hearing test yearly, instruction once, 11 mailings at home, free hearing protection plus replacements, use of sound meter Control: yearly hearing tests plus questionnaires |
|
Outcomes | STS with ≥ 10 dB loss at 2, 3, 4 kHz in either ear Median and mean thresholds at 0.5, 1, 2, 3, 4, 5, 6 kHz High‐frequency average (3, 4, 6 kHz) and low‐frequency average (0.5, 1, 2 kHz) thresholds Bulge depth 3‐year and 16‐year follow‐up |
|
Notes | Long‐term Comparability ‐ intervention/control: mean age: intervention 14.5 years, control 14.6 years hearing: max. threshold (R or L) at 0.5 kHz intervention md 10 dB/control md 5 dB, at 1, 2, 3, 4 kHz intervention/control md 5 dB, at 6 kHz intervention md 15 dB/control md 10 dB |
|
Risk of bias | ||
Bias | Authors' judgement | Support for judgement |
14. Blinding (subjects) | High risk | no blinding |
15. Blinding (outcome assessors) | Unclear risk | not reported |
16. Retrospective unplanned subgroup analyses | Low risk | no subgroup analysis |
17. Follow‐up | Low risk | RCT |
18. Statistical tests | Low risk | multilevel analysis |
19. Compliance | Low risk | increase in hearing protector use |
20. Outcome measures | Low risk | hearing thresholds, STS |
21. Selection bias (population) | Low risk | RCT |
22. Selection bias (time) | Low risk | RCT |
23. Randomization | Unclear risk | randomisation stated but no method reported |
24. Allocation concealment | Unclear risk | no information provided |
25. Adjustment for confounding | Low risk | adjustment for baseline differences |
26. Incomplete outcome data | Low risk | < 20% |