ABSTRACT
When we intervene to prevent occupational injury, we are taking action in workplaces. For these interventions to be durable within the competing imperatives of the workplace, their effectiveness must be demonstrable. This requires evaluation: ‘the collection and analysis of information in order to facilitate informed decision making’ (Lambert and Owen, 1992). Traditionally, OHS researchers have used accident data as the primary, often even the sole, evaluation criterion. In this chapter, we argue that reliance on such a single outcome measure cannot provide adequate evaluation information nor indeed support successful interventions. An injury-prevention intervention in a meatworks illustrates this argument. In this meatworks, the OHS committee had determined to change working arrangements in the work area where sheep are killed and processed in order to address manual-handling risks associated with pelting (removing the skin of the beast). In introducing these changes, there were a number of industrial sensitivities. The company was also concerned that the changes did not cause damage to the skins, which are sold as, of course, sheepskins. The committee therefore decided to implement the changes on a trial basis. They then faced the problem of how to decide whether the trial had worked. The committee developed a series of indicators:
Number of injuries: collected according to the type of injury (e.g. cuts, sprains and strains) and whether it was a new injury or a recurrence of an existing injury. Rather than calculate frequency rate, which the committee found difficult to use, the number of injuries was matched against the number of people working in the area. This information was collected by the supervisors and the OHS officer.
