Special Topics in Safety Management

Safety Metrics: Count on Them

To receive management support, safety programs must be objectively evaluated in quantifiable terms. Here are effective ways to do it.

It’s often been said that business runs on numbers. Profit. Loss. Return on investment. It’s the language senior management speaks and understands. And it’s the reason many safety managers who talk exclusively in terms of human behavior often don’t get the hearing they should. Wouldn’t it be great if safety also had its own metrics?

In fact, safety does. These were recently explained in an article in our sister print publication, OSHA Compliance Advisor (OCA), and other sources. Here’s a digest of what these articles had to say:

OSHA’s way the least effective

First, OSHA’s way of measuring safety may be the least effective way to do it. For years, the agency has relied on the 300 Log program, which basically counts accidents after they occur. This is considered a trailing measure of safety, in that it tells only what has already happened, with no guarantee that future experience will repeat the past.


Need to investigate an accident? Learn to do it the right way in BLR’s February 28 audio conference on the subject. Can’t attend? Pre-order the CD. Satisfaction assured. Click for details.


OSHA admits that this approach is problematic in another way. “While it’s nice to know what the bottom line performance is—i.e.—accident rates—overemphasis on rates … typically only drives accident reporting under the table,” explains an agency document. “What’s more, it’s too easy to manipulate accident rates.”

What form of metrics, then, are effective? The consensus leans toward leading indicators, measures that count either the ingredients of safety or those that create hazards, before any actual accidents occur. If a pattern that spells danger can be spotted, preventive measures can be taken so that those mishaps are avoided. As a corollary, patterns that create safer results can be strengthened and repeated.

One such measure, covered by safety expert Don Groover writing in Occupational Health & Safety magazine, is numbers of exposures … how many times employees are placed in a risky situation.

Simply put, the more often workers are in proximity to danger, the greater chance of harm. If your number is high, your aim then is to find out whether the job could have been done in another way. As Groover explains it, “when an exposure is identified, we must then ask what encouraged or discouraged an employee to tolerate or accept the level of exposure.”

Another leading indicator noted in OCA is the level of safety-related activities being carried out. Examples of such activities are:

  • Number of hazards (not accidents), reported or corrected

  • Number of inspections and equipment safety checks

  • Number of job safety analyses conducted

  • Data from employee opinion surveys asking about safety, and how the results change over time

  • Closure rates on safety issues, i.e. hazards removed or fixes put in place


    Keep your investigations legal. Attend BLR’s February 28 accident investigation audio conference. If you can’t attend, pre-order the CD. Satisfaction assured or your fee refunded. Learn more.


    Quality, too, must be measured

    In looking at closure rates, it’s vital to examine quality along with quantity, says Paul Esposito, a certified safety professional with the risk management company, ESIS. Quoted in the OCA article, he notes that, “if an inspection revealed that equipment was in need of repair, the fact that maintenance signed off on the fix doesn’t really tell you if it was a ‘patch’ or a significant repair based on root causes.”

    We’ll talk more about safety metrics—and the importance of getting accurate answers to drive those metrics—in tomorrow’s Safety Daily Advisor.

  • Print