The organization's safety effort is to create an injury-free workplace. This traditionally means writing programs, training employees, inspecting worksites, and managing the process so as to improve the loss outcomes. To manage the process information is needed. Traditionally the information to manage the process comes from historical loss data. The analysis of this information and accident reports becomes the basis for deploying improvement strategies. This has been the predominant practice for decades, and though there has been some improvement in safety outcomes, they have been less than stellar.
Safety traditionally is reactive in its posture, using historical loss data to deploy interventions so as to improve performance. There are some inherent weaknesses in this approach. Loss reports get their information from accident reports. The data is then reviewed and categorized by a claims person who does not necessarily have operational knowledge and may mischaracterize the nature of the injury or its cause. An example of this may be a worker carrying material tripping on debris on the floor and injuring their back. This injury may be characterized as a manual material handling accident. This would logically result in an ergonomic recommendation for improvement, whereas the true driver of that particular injury was poor housekeeping. Of course the quality of the accident reports are going to play a key role in the value derived from these loss reports. The better and more through the accident report, the better the quality of t he loss data.
Another thing to consider is that the future state is never exactly the same as the past. Therefore interventions deployed from historical data may not necessarily prove to be the "best" solution because of possible changed conditions, practices, procedures, people, goals, objectives, etc. etc. etc!
Back in the 1920s Heinrich did a study of a large body of accident reports, whose finding placed 88% of the reasons for those accidents on the unsafe actions of workers. Later studies verified these findings even to a greater extent. This eventually placed the focus of safety interventions on the worker. These practices include training, warning, protecting, observing, feedback, communication, punishment, and other worker focused interventions. When the expected results are not achieved the next step is to make greater effort in these same areas or make one of them a priority which usually does not result is substantial gains. This has been the accepted practice for decades; all do work to some extent but never ultimately result in an injury free workplace.