Blog > October 10, 2026

Preventive indicators for Vision Zero: measuring action before an accident

Leading indicators for Vision Zero: measure prevention before an accident.

An accident-free period does not, by itself, explain how prevention is working. To manage it, you also need to know whether risks are reviewed, controls implemented and people’s observations answered. Preventive indicators help examine that work before harm reveals a problem that requires attention.

In 2020, ISSA published a guide to proactive leading indicators for Vision Zero containing 14 indicators connected to its seven rules. It addresses areas including visible management commitment, learning and training. The guide is an official reference. The operational formulas below are our own examples for designing follow-up appropriate to an organisation.

Start with the decision you need to make

Before choosing a number, ask a question: which controls need intervention, which owners need support, or which changes remain unreviewed? A useful indicator connects the answer to someone able to act. If no one knows what they would do with the result, reconsider its purpose.

The Explora entry on OSH indicators explains the importance of calculation, source and review frequency. Add the covered population, included sites and exclusions. Keep underlying records available so a result can be checked without relying entirely on an aggregated chart or management dashboard.

Three examples of an indicator definition

Actions verified within their deadline

As an original example, divide actions whose verification was due in the period and completed on time by all actions scheduled for verification in that period, then multiply by one hundred. Preserve the reference schedule so that rescheduling does not retrospectively remove a missed commitment from the calculation.

The source is action and verification records. The owner reviews overdue items, priorities and obstacles. This indicator measures follow-up discipline; it does not establish that every measure is technically appropriate. The verification evidence and its acceptance criterion should remain available for examination alongside the percentage.

Responses to people’s observations

Select observations whose response deadline falls within the period. Calculate the proportion receiving a documented response within the agreed time. A response can explain an action, request information or justify a decision. An automatic receipt notification is different from a substantive review of the observation.

Read the percentage alongside outstanding observations and recurring reasons. More reports may reflect greater trust in the channel as well as changed working conditions. The Explora entry on prevention culture helps interpret participation within its organisational context rather than treating every increase as deterioration.

Risk reviews before changes are introduced

Identify changes requiring preventive review before implementation under the organisation’s procedure. Check how many have the necessary review and decisions before work begins. The denominator should come from the change register, not only from assessments already completed by the safety department or known to its team.

This avoids a measure that appears complete because it counts only familiar cases. The quality of the inventory is part of measurement. If information from one department is missing, record that limitation and assign corrective work instead of treating the missing entries as evidence that no review was needed.

Distinguish quantity, quality and outcomes

Counting inspections describes activity. Checking whether they identified relevant issues and whether resulting actions were verified adds another perspective. Accident, disease and other harm outcomes remain necessary for understanding events. Read the different measures together, with their scope made explicit and their limitations visible to reviewers.

The Explora entry on safety leadership connects the dashboard with management decisions. A meeting can review a small set of indicators and still be useful if it resolves priorities, resources and ownership. Additional figures without capacity to act create reporting effort without necessarily producing learning.

Avoid incentives that weaken the data

Do not make fewer reports a target that discourages people from raising concerns. Examine changes alongside activity levels, reporting coverage and recording habits. When comparing sites, use consistent definitions and identify differences in risk or scope that may affect interpretation. Keep changes to formulas documented so trends remain understandable.

Connect the indicators with Vision Zero’s seven Golden Rules and the Cultura 5Z action plan, retaining appropriate measures for each area. Software can support traceability between data and decisions. When assessing Sabentis OSH software, trace an indicator back to its source records and establish who will act when performance deviates from the agreed expectation.

A practical checklist for testing indicator quality

Before presenting an indicator to the review committee, test whether its calculation can be reconstructed and identify the decision that would follow a change in the reported result.

  • Recalculate a small reporting period manually and compare it with the dashboard to identify differences in definitions.
  • Check that outstanding cases remain in the denominator where appropriate, even when closure evidence is not yet available.
  • Distinguish a zero value from missing data and identify sites or periods with incomplete reporting coverage.
  • Test the effect of a late correction and retain the explanation for the difference from the previous report.
  • Connect each internal threshold to an action and an owner; avoid alert colours that nobody knows how to interpret.

Frequently asked questions about preventive indicators

What should be shown when the denominator is zero?

If there were no eligible cases during the period, the proportion cannot be calculated. Show “not applicable” with an explanation, rather than defaulting to one hundred per cent or zero. Distinguish this situation from lacking the data needed to carry out the calculation.

Does an increase in reports mean prevention is getting worse?

That conclusion cannot be drawn from the count alone. It may reflect greater exposure, increased confidence in reporting or a change in recording practices. Examine the reports, working conditions and responses before attributing the change to a single cause or judging preventive performance.

Must every official Vision Zero indicator be used?

The guide provides a reference for selecting and developing measurement. Choose indicators the organisation can interpret and use, explaining their scope. If you adapt a formula or create your own indicator, identify the change so that it is not presented as an unchanged official indicator.

Can sites with different activities be compared?

Only after reviewing definitions, exposure, size and data coverage. Comparing execution of the same process may be useful, but a single ranking can conceal important differences. Provide context alongside the figures and use the results to decide where support or improvement is needed.

Sources and methodology

  • ISSA: Proactive Leading Indicators: ISSA’s international guide to proactive indicators; supports measurement without replacing contextual analysis.
  • ISSA: Vision Zero resources: the seven-rule framework with which indicators can be connected. The article’s operational formulas and checks are original practical proposals.

Sources consulted on 6 October 2026. Checklists and examples are practical implementation proposals.

Request a Demo

Discover all that Sabentis can do for your organization.

Try Sabentis

request a demo
stars 5
GetApp Software Advice Capterra