Insiders’ perspectives: Are leading indicators reliable measures of safety performance?

By Dave Johnson, Phylmar Group newsletter editor

Not one person on an expert panel of six global safety and health professionals could come up with an example of a company that made a significant positive pivot in safety and health performance based on a leading indicator, according to a July, 2021, article in Industrial Health and Safety Review. This apparent absence of evidence that leading indicators correlate (with cause) to lagging results such as injury rates comes despite a growing emphasis on leading safety and health metrics in recent years. Some recent developments:

• OSHA held a stakeholder meeting on leading indicators in 2019;
• A proposed ANSI Z.16 standard on safety and health metrics is currently out for ballot;
• ISO is developing a metrics standard (ISO 45006), though its work lags behind Z.16;
• The National Safety Council’s center of excellence, the Campbell Institute, has issued several excellent reports on leading indicators (“Practical Guide to Leading Indicators” and “Transforming EHS Performance Measurement Through Leading Indicators” are two).
• The Phylmar Regulatory Roundtable (PRR), Occupational Safety and Health (OSH) Forum, held a focused discussion on practical, experience-based strategies to develop valuable leading indicators that correlate with lagging indicators at PRR’s fall conference on November 3rd, 2021. One take-away was a multi-faceted approach and authentic communication between management and the field is key. A component of successful communication is how the EHS expert effectively communicates with senior leaders who do not have a safety background. The event was facilitated by PRR Director Helen Cleary and attended by approximately 70 PRR members who are Environmental Health and Safety Professionals. For more information on the PPR OSH Forum and how to participate visit https://www.phylmar.com/regulatory-roundtable/

None of a dozen safety and health professionals contacted for this article (former OSHA officials, standards committee members, veteran pros and consultants) were aware of solid statistical proof that leading indicators actually do what the theory says they will do. It remains an assumption without significant research backing that using and improving leading indicators improves lagging indicators such as the OSHA total incident rate (TRIR), the days away/restricted work activity and/or job transfer (DART) rate, first aid incident rate, fatality rate, incident-free hours over time, days since last lost-time incident and other “rear view mirror” past history measures.

The importance of linking leading indicator performance to changes in lagging indicators was indicated in a Campbell Institute survey: the ability to tie changes in leading indicators to stronger or weaker performance on injury rates was very important to 64% and extremely important to 21% of respondents.

Hard evidence of a correlation may be missing, but conventional wisdom in occupational safety and health holds that companies should track both leading and lagging measures. Leading indicators (management system components, management inspections, leadership safety meetings, job observations, employee safety suggestions, inspection percentage completed, risk assessment, number of hazards found and fixed, safety and health training hours, near miss reports, and scores of others) by themselves have limited value, says Steve Newell, National Safety Council consultant and vice chair of the ANSI Z.16 standard. “What really matters is whether the action being tracked by the leading indicator is achieving the desired result,” he says.

Professionals contacted for this article all agreed with the view expressed by John Dony, Senior Director, Thought Leadership, for the National Safety Council: “While the academic research on the efficacy of leading indicators, particularly their inverse correlation with lagging indicators, may be lacking, there do exist many company-level efficiency studies that show these correlations exist.” These anecdotes, unknown to the panel described at the beginning of this article, emerge only with a company’s blessing. Many companies keep data on leading and lagging indicators for “internal eyes” only.

The Campbell Institute has reported these case studies:

• Cummins found a very strong negative correlation indicating that an increase in safety and health training hours was associated with a decline in the incident rate (in particular, specific training in risk assessment and job safety analysis).
• Honeywell’s Building Solutions business unit reported more than 82,000 safety observations in one year, and the correlation of safety observations with injury rate was clear – the business unit reduced the number of recordable injuries from 108 to 54.
• As far back as the 1990s, United States Gypsum USG saw vast improvements in safety operations receiving two- to three-day audits compared to facilities that did not receive audits. Use of USG’s Safety Activity Rating document was linked to reduction in injury rate and heightened awareness of safety among employees.

More anecdotal evidence:

• Allergan (subsequently acquired by AbbVie) did an analysis that showed a correlation between “Good Observations” (employee notifications of hazards to be corrected, potential safety improvements that could be made) and OSHA rates. When Good Observations went up, OSHA rates declined. Employees were thinking more of safety and becoming more vigilant.
• Three companies that organizational management consultant Dr. John Kello has worked with generated data roughly correlating the impact of safety sessions/briefings on bottom line safety outcomes. One company substantially reduced incident rates from greater than five to less than two. Another tracked cost savings from its reduced incident rates and lost-time incidents and estimated annual savings of above $700,000.
• Using leading indicators such as an Action Item Matrix, Continuous Improvement teams resolving issues, monthly Safety Steering Team (SST) metric monitoring meetings, and training for SST team members were effective in helping a Fortune 20 food manufacturer reduce recordable injury (RFI) frequency from 20+ to 1.12 in three years; reduce the RIF from five to less than one for an international paving contractor company; and decrease the RIF by 58% at a global heavy equipment company, writes consultant Dr. Mike Williamsen in his 2021 book, “Developing Safety Excellence: Engagement Culture at Every Level.”

Two surveys also point to a safety-improving correlation between leading and lagging indicators. A study of 60 companies reported in the July, 2015, International Journal of Occupational Safety Ergonomics found that companies adopting occupational safety and health management rules at a level higher than the median (“high performing companies”) had a lower number of incidents, number of sickness absence days, severity rate of accidents, cost of accidents, non-conformities with OSH management system requirements and also legal requirements compared to “poor performing” companies with below median adoption of OSH management rules.

The “2021 Safety Performance Report” issued by Associated Builders and Contractors uncovered these connections:

• Safety training, new hire safety orientation and behavior-based safety observations leads to a 64% reduction in TRIR and DART rates;
• Companies conducting in-depth indoctrination of new employees into the safety culture and processes experience 52% to 54% lower incident rates;
• Substance abuse programs and policies with provisions for drug and alcohol testing where permitted lead to a 69% reduction in TRIR and a 61% reduction in DART rates;
• Companies that conduct daily toolbox talks reduce TRIR by 76% and DART rates by 78% compared toto companies that hold them monthly;
• Companies ensuring supervisory personnel have an in-depth, well-versed understanding of the company’s safety policies, supervisory safety expectations and other specific competencies reduce TRIR and DART by more than 50%.

Metrics experts list numerous caveats to these apparent linkages. The leading indicator intervention must be of high quality. It’s the quality of leading indicators that matters, not the quantity. For example, instead of merely measuring total hours of safety training, dig deeper into the types of training and analytics on training effectiveness. (For a recent Phylmar article on training effectiveness, visit https://www.phylmar.com/2021/08/are-you-throwing-away-training-dollars/)

A diverse portfolio of leading indicators must be used. Leading indicator portfolios evolve through trial and error. They have a shelf life and many over time lose their effectiveness and need to be retired. Leading indicators are unique and as numerous and varied as the companies generating them. There is no “silver bullet” leading indicator for universal adoption. Many companies are too quick to try to find correlations, and some are too quick to try to classily incidents as non-recordable or non-lost workdays to keep rates down.

Context is also key. “Even with the best early indicators and related activities, every organization needs leaders who create a culture that is supportive, caring, transparent, and accountable,” says David J. Sarkus, MS, CSP, speaker, consultant and coach.

These caveats and others are what make establishing a definitive link between leading indicators and better safety and health outcomes difficult. That could change with the coming voluntary metric dashboards from ANSI and ISO that can standardize adoption of metrics across companies for better comparisons and possible correlations. But for the foreseeable future it will continue to be individual companies that analyze potential correlations to establish which customized leading indicators most significantly impact safety and health lagging outcomes.