Norwood Distribution: a depot that reports nothing and injures regularly

Start with the pattern, not the list. Thirteen injuries in 14 months across roughly 250 people per day demands an explanation, and the blank near-miss book is the clue. Where thirteen injuries occur, many more near misses occur too; a nil return over six months indicates suppressed reporting, not good luck.

Deficiencies and their effect

• No risk assessments for agency picking work. Risk assessment is the foundation on which controls, training and supervision are built. Without it, controls are generic, and the nine manual handling injuries are consistent with untrained people meeting unfamiliar loads in an unfamiliar layout.
• A five-minute gate briefing for agency workers. These workers face the highest risk because they are new to the building, the stock and the equipment. A gate briefing transfers no task-specific knowledge, so the injury burden falls on the least protected group.
• Shift managers with no qualification and no defined safety responsibilities. Safety depends on individual inclination rather than on a system. Nobody owns the risk at the point where it arises.
• No safety committee and no consultation. The workforce, including agency staff, sees the problems first; with no channel to raise them, the organisation learns nothing.
• Annual, lagging-only board review. The board sees injuries a year after the event and never sees the controls that prevent them, so it cannot challenge or resource properly.
• A part-time external adviser. Two days a month can write documents but cannot supervise, coach or sustain ownership.

Judgement: this is a reactive, compliance-shaped arrangement. The injuries are symptoms of a system failure, not of careless individuals, and training alone will not fix it.

Plan. Set a written policy signed by the operations director, then profile risk by area and task. The agency picking operation should be assessed first because that is where the evidence points. Set objectives with dates: complete risk assessments for all picking tasks, extend inductions to a minimum supervised shift, introduce a reporting route that agency staff can use.

Do. Define responsibilities so each shift manager owns the risk in their area, and make that explicit in their job description and appraisal. Provide competence: a short, practical course for shift managers in risk assessment, incident investigation and supervision, not a general certificate. Design a proper induction for agency workers covering layout, vehicles, manual handling and reporting, with a signed record. Establish a safety committee with management attendance and agency representation.

Check. Replace the near-miss book with a simple card or app route usable by agency staff, and monitor reporting rates as a health check. Add monthly workplace inspections, quarterly audit against the plan, and investigation of every injury with root causes recorded.

Act. Review performance at board level quarterly, not annually, and act on the trends.

On the adviser question: the legal duty is to appoint competent advice proportionate to risk, not to buy a full-time post. This depot is moderate risk, not high hazard, so the gap is line ownership, not consultant days. The defensible answer is to retain the part-time adviser with a defined brief (audit, advice, training), strengthen shift manager competence, and revisit the full-time case only if the injury and reporting data do not improve within two quarters.

Choose measures that predict, not only record. Lagging measures (injuries, days lost, RIDDOR reports) tell the board what has already happened; they must be paired with leading measures that show whether controls are working.

Recommended set

• Leading: percentage of tasks with a current, suitable and sufficient risk assessment; percentage of agency workers receiving a full induction before unsupervised work; near-miss and hazard reports per 100 workers per month (with a target that rises at first - a rising count here is success, not failure); shift manager inspections completed against plan; percentage of corrective actions closed by the due date.
• Lagging: injury rate per 100,000 hours worked, split by employment status so agency risk is visible; days lost; number of RIDDOR reports; average days to close an investigation.
• Cultural: proportion of staff who say they know how to report a hazard (from a short quarterly pulse survey).

Presentation that drives decisions

1. Lead with two or three sentences of interpretation, not a data dump: what is improving, what is not, what decision is needed.
2. Show a trend over at least six points, because a single month is noise.
3. Separate permanent from agency figures; combining them hides the group most at risk.
4. For every red measure, put the proposed action, the owner, the cost and the date on the same page.
5. Ask for decisions explicitly - for example, approving supervision hours on the night shift.

Reasoning: boards act on comparisons (trend, target, peer site) and on consequences. Presenting a rising near-miss count alongside a falling injury rate tells a coherent story of improving reporting, which builds confidence that the numbers are real.