How to Measure Health and Safety Performance Beyond Injury Rates

Health and safety leading indicators measure preventive work and the condition of a safety programme before an injury outcome tells you something went wrong. Lagging indicators describe events already recorded, such as injuries or illness cases. A useful dashboard combines both and asks whether important controls are actually working.

This guide helps managers and HSE coordinators choose a small, explainable set of measures. It offers an illustrative dashboard and calculation, not a universal compliance score or proof that a workplace is safe. US and Canadian guidance below informs the method; it does not establish Pakistani legal reporting requirements.

OSHA's explanation of leading indicators distinguishes proactive measures from historical injury and illness outcomes. The practical lesson is to avoid asking one number to describe the whole system.

Measure type	Question it can help answer	What it cannot establish alone
Lagging outcome	What harm or events have been recorded?	Whether unreported hazards or long-latency health risks are controlled
Leading activity	Was planned preventive work performed?	Whether that work was relevant or effective
Control verification	Did a defined check find the expected control condition?	Whether the sample represents every task, shift and location

For example, counting training sessions tells you about delivery. Checking whether people understood role-specific restrictions provides different information. Neither measure replaces a competent assessment of the work.

A dashboard should make the next management decision clearer. Before adding a metric, write the concern in plain language: “We do not know whether overdue maintenance actions are being resolved” is more actionable than “We need more safety KPIs”.

Choose a meaningful scope

Specify the location, activities and people covered. If contractor work is excluded, state that exclusion beside the result. A site-wide percentage can conceal a weakness in a small but important work group. Review the underlying records before interpreting the headline.

Define the evidence and owner

Record the data source, numerator, denominator, period, exclusions and person responsible. Decide what counts as completion. An action marked closed after an email is different from an action whose result has been checked at the workplace.

Write the response before setting a target

Ask what management will do if the measure deteriorates. If nobody can name a response, the metric may be decorative. Set local objectives from the risks, baseline and resources; do not copy an unexplained target from another organisation.

The following measures are SIP teaching examples. Select or adapt them to your own programme. None is a regulator-prescribed target or a guarantee of injury prevention.

Candidate indicator	Suggested definition	Interpretation check
Verified action completion	Actions due in the period completed and verified by deadline, divided by all actions due in that period	Show overdue high-priority actions separately; do not let minor items dominate
Induction understanding checks	New starters with the defined understanding check recorded before relevant work, divided by starters requiring that check	Sample the quality of checks; an attendance tick is insufficient
Reporting response time	Elapsed time from a worker's concern to recorded initial response	Separate initial acknowledgement from resolution and effective control
Planned verification coverage	Relevant checks completed against a risk-based schedule	Document missed shifts and locations, and what the checks found
Recurring findings	Issues returning after a previous action was recorded as complete	Investigate the earlier decision and evidence rather than only recounting the issue

Pair these with suitable outcome information and explanatory notes. The Canadian Centre for Occupational Health and Safety discusses the complementary roles of both types in its leading and lagging indicators guidance.

Imagine a hypothetical workshop with 20 corrective actions due during October. Sixteen were completed and independently checked against the agreed evidence before their deadlines. Two were completed late, one is awaiting verification and one remains open.

The on-time verified completion measure is 16 ÷ 20 × 100 = 80%. The two late actions do not enter the numerator. The item awaiting verification does not count as verified. Keep all four outside the numerator visible in the supporting action list.

If management reports 19 ÷ 20 because 19 items have some activity recorded, that is a different definition. It cannot be presented as the same measure. Changing definitions between months makes the trend unreliable.

Now suppose the single open action concerns an important unresolved control. The 80% result does not authorise work or reduce the significance of that gap. Management still needs the responsible competent people to address the issue through the site's established arrangements.

A falling incident count could reflect improvement, but it could also reflect incomplete reporting or a change in work exposure. An increase in reported concerns may reflect greater trust in the reporting system. Investigate the explanation before awarding success or assigning blame.

• Do not reward silence: give workers a safe way to raise concerns and report what happened after they did so.

• Keep denominators visible: ten checks out of ten planned differs from ten checks out of fifty relevant tasks.

• Review quality: sample completed records and discuss findings with workers rather than accepting every tick as evidence.

• Preserve a stable definition: annotate changes to scope, workforce, working hours or reporting systems.

• Show exceptions: keep important overdue items visible even when the aggregate result is improving.

Do not compare injury rates between organisations unless their definitions and exposure bases are compatible. Small numbers can fluctuate sharply, and a short period with no injury is limited evidence about the condition of controls.

OSHA's programme evaluation guidance connects monitoring with worker participation and programme improvement. At your review, ask what changed, what evidence supports the explanation and which action now needs an owner.

Use a short decision log: finding, interpretation, action, owner, due date and later effectiveness check. For training-related measures, SIP's HSE training matrix guide helps structure the underlying learning records. Broader system responsibilities are covered in the ISO 45001 implementation checklist.

Managers exploring structured study can review SIP's occupational health and safety management diploma information. Check its current specification and suitability with SIP; a qualification is not evidence that a site's performance measures are effective.