Reprinted with permission from ASSE's Professional Safety Journal
Near-miss reporting, or the lack of it, is a controversial indicator of an organization’s safety culture. Over the years, safety professionals have heard concerns about the statistical validity of the many ratios published in the literature.
This article presents a practical process that should help overcome resistance to near-miss reports becoming a useful tool to help an organization reduce injuries. Limited research is referenced, not to statistically substantiate near-miss-to-injury ratios, but rather to show a long-standing interest in using this concept as another tool to focus on eliminating workplace incidents.
Let’s begin with this question: Does your organization receive about 50 near-miss reports for every minor injury suffered by workers? If not, several significant barriers within the organization’s culture may be preventing the organization from learning the lessons available from incidents that did not result in loss—at least not this time.
While building a power plant in Louisiana, a major construction company used an effective near-miss reporting program to trigger safety success. Eighteen months into the project, the site had worked 3.1 million hours without a lost-time injury, had an OSHA recordable rate of 0.68 and achieved Voluntary Protection Programs status. Additionally, the site worked the first 1 million project hours without a single OSHA recordable.
At the start of the near-miss reporting improvement project, the number of near misses reported averaged one or two per month (or about 0.005 per employee). Three months after initiation of the project, that number increased nearly 40 fold (to about 0.2 near-misses reported per employee). This level has continued to climb to a current level of about 230 near-misses per week (or about 0.6 per employee), which is more than 100 times the rate when the program was launched. This initiative has built trust, encouraged employee involvement, enabled the identification and control of previously unknown or unrecognized risks, and enhanced management credibility through visible, positive action.
While four main leading indicators (near-miss reports, near-misses resolved, supervisor audits, manager audits) were utilized to support this accomplishment, this article focuses on the near-miss indicators and the methods employed to overcome cultural barriers that typically inhibit near-miss reporting success.
Is Your Current Approach Working?
The management team knew that identifying and investigating near-misses were key elements to finding and controlling risks before workers were injured or property was damaged. The group also knew that near-miss reports were few and far between.
To cement organizational dissatisfaction, as well as determine the amount of improvement needed, the safety department turned to varying studies regarding incident ratios. Numerous studies can provide insight as to whether a near-miss reporting program is working. The statistical validity of the three references cited here has been widely debated. With that in mind, this organization’s team used the references not for statistical validation, but rather to show an ongoing interest in the concept of near-miss reporting being a tool that could help reduce injuries in the industrial workplace.
Heinrich’s (1931) accident triangle offers an early theory on incident probabilities. Heinrich proposed that for every major injury, there were 29 minor injuries and 300 no-injury incidents (near-misses). His ratios have been widely debated and refuted, but the concept that near-misses could be used to reduce injuries was the team’s focus. Bird and Germain (1969) completed a study to determine accident ratios as they occur in various industries. Their analysis of 1.75 million incident reports within 297 organizations and 21 different industries revealed that for every serious or major accident, there were 10 minor injuries, 30 property damage events and 600 no-loss incidents. Health and Safety Executive (1993) researchers concluded that for every lost-time injury (more than 3 days in length), there were seven minor injuries (first-aid only in this study) and 189 noninjury cases.