A delivery firm measuring the wrong things
A parcel delivery firm employs 30 drivers and 8 warehouse staff. Over three years it has recorded three injuries: a twisted ankle in the yard, a deep cut from a broken cage, and a back strain lifting a heavy parcel at a customer's door. None was reported under RIDDOR. The near-miss book has four entries, all written by the same driver. Monthly toolbox talks have been cancelled for the last five months because of volume. No incident has ever been investigated beyond a note in the accident book. The manager displays a board reading "142 days without an accident" and believes the firm's safety performance is good because the number is rising. Drivers say they do not report near misses because "you get a reputation".
1. Assess the manager's '142 days without an accident' measure. What would you measure instead, and why?
Read the worked answer
The days-without-an-accident count is a lagging indicator that only moves after someone has already been hurt, and it rewards the wrong behaviour: the easiest way to keep the number rising is to not report. With a near-miss book filled in by one person and an accident book that may be incomplete, the low numbers reflect reporting, not safety.
Better measures include:
- Leading indicators of the work being done: risk assessments completed and reviewed, toolbox talks delivered (and attended), vehicle and cage inspections done, near misses reported and closed out, training completed, safety issues raised and resolved.
- Quality of reporting: ratio of near misses to injuries - a healthy site generally reports many more near misses than injuries. Here it is four to three, which is a warning sign, not a good one.
- Severity and repeat causes: are the same tasks producing the injuries again?
The board should show both kinds. A number that can only improve when nobody speaks up is not a measure of performance.
2. Set out how you would investigate the back strain at the customer's door, without turning it into a blame exercise.
Read the worked answer
Steps:
- Respond and record - get the person appropriate care, record the facts promptly: parcel weight and size, what the driver was told about the delivery, the access at the door, the driver's training and equipment, time pressure on the round.
- Gather evidence before conclusions - speak to the driver and the customer, look at the parcel, check whether a trolley or sack truck was available and whether the round schedule allowed it to be used.
- Look for root causes, not just the immediate act. Was the load assessable in advance? Are heavy items flagged by the sender? Is the round timed so tightly that lifting alone feels quicker? Is manual handling training practical or a video watched once?
- Involve the person who was hurt - they know how the job really works, and blame closes that route off.
- Agree corrective actions with owners and dates - for example, flagging and limiting parcel weights, issuing and requiring a light trolley, adjusting round times, refresher handling training including awkward doorways.
- Check the action worked, feed back to the team, and record the outcome.
The reasoning: an investigation that ends with "driver should have lifted properly" leaves every underlying condition unchanged, so the next driver has the same injury.
3. Explain why near-miss reporting is so low and what the firm could do to raise it.
Read the worked answer
"You get a reputation" is the key evidence: drivers believe reporting creates risk for them, so the safe behaviour is silence. Contributing factors are an empty book that no one talks about, no feedback loop (nothing visibly happens after a report), cancelled toolbox talks that have removed the main face-to-face channel, and a manager whose headline measure improves when reports fall.
To change it:
- Make reporting easy and immediate - a short form or phone-based report, no essay required.
- Close the loop visibly - tell people what happened to their report and what changed. Reports that disappear are the last ones you get.
- Respond to reports without blame - thank the reporter, separate the person from the condition.
- Use the reports as data - review them monthly alongside the toolbox talks, and restart those talks so reporting is discussed face to face.
- Leadership behaviour - the manager should ask for near misses and welcome them, and stop using accident-free days as the headline.
The test of success is not a flood of reports for their own sake, but reports that lead to fixed conditions - for example, damaged cages being taken out of service before someone is cut again.