Evidence for the effectiveness of occupational health interventions

Article type
Authors
Ruotsalainen J, Verbeek J, Salmi J, Jauhiainen M, Laamanen I, Pasternack I, Husman K
Abstract
Background: It is often argued that the occupational health field is different from the rest of medicine. Randomisation of studies for example can be impossible or unethical. Consequently, it would be difficult to generate the equivalent level of evidence regarding the effectiveness of interventions.

Objectives: We aimed to obtain a representative overview of the current quality and quantity of evidence in the biomedical literature for the effectiveness of occupational health intervention (OHI) studies.

Methods: Two volumes of 16 occupational health, general and specialised medical journals were searched by hand to locate articles that reported evaluating OHI studies. Studies were classified according to research design, type of OH outcome, type of intervention and the ICD 10 disease category involved. The relations between these four features were then examined and the results compared with reviews of the evidence-base of other medical fields like orthopaedics and mental health.

Results: Out of the total 8687 original articles searched we found 148 OHI studies. The prevalence of OHI studies varied from zero in the New England Journal of Medicine to 12.5% in the American Industrial Hygiene Association Journal. In 21% of the OHI studies the study design was a RCT, and in 28% it was a CT, an interrupted time-series in 7% and some other evaluation design in 44%. The targeted occupational health outcome was exposure in 27% of the studies, worker behaviour in 12%, disease symptoms or signs in 30%, disability or sickness absence in 24%, injuries in 4% and quality of care in 3%. The most common ICD 10 disease categories were XII, XIII, XIX, XX and XXI together accounting for 85% of the studies. Interventions aimed at groups were significantly more often preventive in nature than curative. RCTs were significantly more often targeted to individual participants rather than groups.

Conclusions: High quality evaluation studies are conducted in all areas of occupational health. However, a large part of evaluation studies is of low quality and needs improvement. The situation in occupational health is therefore not too different from other comparable fields of medicine.