12.2 Health and Safety Performance Measurement, Statistics & Trend Analysis

1. Why Measure Performance

Health and safety performance measurement tells management whether the system described in Section 12.1 is working. It supports the Check stage of the Plan-Do-Check-Act cycle and gives evidence for decisions on resources and corrective action. On a construction project, the CHSM uses measurement to:

• report to the client and its agent, who audit the principal contractor at least once every 30 days under CR 5(1)(o);
• give the health and safety committee information on incidents, which it must discuss under section 20(1)(b) of the OHS Act;
• identify contractors, activities or areas that need intervention before a serious incident occurs; and
• feed the annual management review.

2. Lagging and Leading Indicators

Lagging indicators measure outcomes after harm has occurred:

• fatalities, lost-time injuries, medical treatment cases and first aid cases;
• days lost and restricted work days;
• occupational disease cases, such as noise-induced hearing loss shifts or silicosis findings;
• reportable incidents under section 24 of the OHS Act and property damage.

Leading indicators measure the inputs and controls that prevent harm:

• percentage of planned inspections, audits and safety observations completed;
• NCRs closed out on time, and the average age of open corrective actions;
• percentage of workers inducted, trained or holding valid Annexure 3 certificates;
• number of task-based risk assessments reviewed, permits audited and toolbox talks held;
• near-miss and hazard reports submitted, and the time taken to act on them.

A good dashboard uses both. Lagging indicators show results; leading indicators show whether controls are being maintained. Leading indicators must be meaningful: counting toolbox talks says little unless their quality is also checked.

3. Standard Rates and Formulas

Raw injury counts cannot be compared between projects of different sizes, so they are normalised by hours worked. The base must always be stated:

$\text{LTIFR} = \frac{\text{Number of lost-time injuries} \times \text{Base}}{\text{Total hours worked}}$

$\text{Severity rate} = \frac{\text{Days lost} \times \text{Base}}{\text{Total hours worked}}$

$\text{Total recordable injury rate} = \frac{(\text{LTIs} + \text{restricted work cases} + \text{medical treatment cases}) \times \text{Base}}{\text{Total hours worked}}$

The two common bases are 200 000 hours (about 100 workers working 2 000 hours a year) and 1 000 000 hours. A rate calculated on one base is not comparable with a rate on the other unless converted. Definitions also vary: a client may define a lost-time injury as one that prevents the person returning to work on the next scheduled shift. The CHSM should state the definitions in the project's reporting procedure and apply them consistently to every contractor.

Worked example

A project records the following over 12 months:

• 600 workers on average, working 2 200 hours each: $600 \times 2\,200 = 1\,320\,000$ hours;
• 4 lost-time injuries with 96 days lost in total;
• 10 medical treatment cases.

Then:

• LTIFR (200 000 base) $= 4 \times 200\,000 \div 1\,320\,000 = 0.61$
• LTIFR (1 000 000 base) $= 4 \times 1\,000\,000 \div 1\,320\,000 = 3.03$
• Severity rate (200 000 base) $= 96 \times 200\,000 \div 1\,320\,000 = 14.5$ days lost per 200 000 hours
• Total recordable injury rate (200 000 base) $= 14 \times 200\,000 \div 1\,320\,000 = 2.12$
• Average days lost per lost-time injury $= 96 \div 4 = 24$

If one more injury causes 180 days lost, the LTIFR rises only from 0.61 to 0.76, but the severity rate more than doubles to 41.8. Both figures are needed to see the full picture.

4. Reading Trends Correctly

Construction workforces change month by month, and a single injury on a small project makes the monthly rate jump. Sound trend analysis uses:

• Rolling 12-month rates: each month, recalculate the rate over the last 12 months to smooth random variation.
• Pareto analysis: rank incidents by cause, activity or contractor; typically a few categories (falls from height, struck by plant, hand injuries) account for most harm, which shows where to focus.
• Control charts: plot monthly figures against a mean and control limits to separate normal variation from a real change.
• Exposure context: compare against the hours worked in each phase; earthworks, steel erection and roofing phases carry different risk.
• Leading-to-lagging links: if inspection completion falls in a month and injuries rise in the next, the link is worth investigating.

5. The Limits of Accident Triangles

Heinrich (1931) proposed a ratio of serious injuries to minor injuries to no-injury accidents, and Bird (1969) published a larger ratio of 1 serious injury to 10 minor injuries, 30 property damage incidents and 600 near misses. These triangles are useful to show that minor events are common and worth reporting. However, they do not prove that reducing minor injuries will reduce fatalities in proportion. Serious injuries and fatalities often come from different, high-energy hazards (falls from height, collapse, crush by plant, electrocution) that minor-injury counts do not track. Good practice therefore uses serious injury and fatality (SIF) precursors, such as work at height without fall protection, lifts without a lifting plan, or entry into unsupported excavations, as leading indicators.

6. Under-Reporting and Perverse Incentives

Low injury rates are only meaningful if reporting is honest. Bonuses tied only to "zero LTIs", or pressure to bring injured workers back on light duty so that a lost-time injury is not recorded, hide risk and can breach section 24 reporting and COIDA claim duties. The CHSM should:

• reward reporting of near misses and hazards, not only the absence of injuries;
• reconcile site incident records with first aid records, clinic visits and COIDA claims;
• check that every incident covered by section 24 has been reported to the provincial office; and
• audit restricted-work cases to confirm that the work offered is genuine and medically appropriate.

7. Reporting Performance

A monthly health and safety report to the client and management typically contains:

• hours worked and headcount per contractor;
• lagging and leading indicators against targets, with rolling 12-month rates;
• significant incidents, investigation status and corrective actions;
• audit and inspection results and open NCRs;
• upcoming high-risk activities and the controls planned for them.

The report should be short, use the same definitions every month and lead to decisions, not just display numbers.