Scand J Work Environ Health 2009;35(6):403-413    pdf | Published online: 01 Oct 2009, Issue date: 00 Dec 2009

A systematic review of occupational safety and health business cases

by Verbeek J, Pulliainen M, Kankaanpää E

Objective Business cases are commonly developed as means to rationalize investment. We systematically reviewed 26 reported cases on occupational safety and health (OSH) interventions to assess if health and productivity arguments make a good business case.

Methods To be included in the review, studies had to analyze the costs and benefits, including productivity, of an OSH intervention at the enterprise level. We searched Medline and Embase for studies and used Google search in addition. Two reviewers independently selected studies and extracted data. The intervention profitability was calculated in euros (€ in 2008) as the first year’s benefits minus the total intervention costs per worker. The payback period was calculated as the intervention costs divided by the first year’s benefits.

Results We found three ex-ante and 23 ex-post cases. In 20 cases, the study design was a before–after comparison without a control group. Generally a 100% reduction of injuries or sickness absence was assumed. In two cases, productivity and quality increases were very large. The main benefit was avoided sick leave. Depreciation or discounting was applied only in a minority of cases. The intervention profitability was negative in seven studies, up to €500 per employee in 12 studies and more than €500 per employee in seven studies. The payback period was less than half a year for 19 studies. Only a few studies included sensitivity analyses.

Conclusions Few ex-ante business cases for management decisions on OSH are reported. Guidelines for reporting and evaluation are needed. Business cases need more sound assumptions on the effectiveness of interventions and should incorporate greater uncertainty into their design. Ex-post evaluation should be based preferably on study designs that control for trends at a time different from that of the intervention.

This article refers to the following texts of the Journal: 2007;33(2):122-130  2007;33(2):81-83  2007;33(3):161-164