Author: Chad Hoxie, CPCU, ARM

Most employers review their workers’ compensation results as one number. Claims this year against claims last year, total incurred, the experience mod. Those numbers matter, and they are also where problems go to hide.

A company wide average blends your best run location with your worst one and reports something in between. The location where every injury is reported the same shift and every employee is back on modified work within days gets averaged together with the one where reports arrive a week late and nobody has a light duty list. The average looks acceptable. One of those locations is carrying the program and the other is costing it, and the summary cannot tell you which.

Why the spread is the real finding

When the same company, with the same carrier, the same policies and the same handbook, gets very different results at different sites, the difference is not the policy. It is execution. Some location managers built the habits and some did not. Some supervisors were shown what to do and some were handed a binder.

That is good news, because it means the fix is already working somewhere in your own organization. You do not need to import a best practice from an industry report. You need to find the location that is already doing it and make the others run the same way.

Why most employers never see it

The data usually exists. Loss runs carry a location field, and so do most incident reports. What is missing is the habit of cutting the numbers by location and putting each location’s results in front of the person who runs it.

Without that, the conversation about workers’ compensation happens once a year at renewal, at the corporate level, in aggregate, among people who were not on the floor for any of the injuries. Nobody at the location level ever sees their own numbers, so nobody at the location level is accountable for them.

The protocol

Cut every metric by location

Report lag, lost days per claim, percentage of claims with a written transitional offer, claim frequency per hundred employees. Every one of them by location, every quarter.

Normalize before you compare

A large location will have more claims than a small one. Compare rates, not counts: claims per hundred employees, or per hundred thousand hours worked. Otherwise the review becomes an argument about headcount.

Give each location an owner

Each location manager owns that location’s numbers. Not the safety department and not HR at headquarters. The person who can change what happens on the floor.

Review it on a schedule

A short quarterly review, location by location, with the same few numbers in the same order. The point is not to punish the worst location. It is to ask what the best one does differently, and to make that the standard.

What to measure

• Each core metric by location, as a rate
• The gap between your best and worst location on each one
• Whether that gap is closing quarter over quarter

The gap is the number to watch. It is usually larger than the gap between you and your industry, and unlike your industry, it is entirely within your control.

Where to start

Take your most recent loss run and sort it by location. Count claims and lost days for each, and divide by headcount. It takes an afternoon, and most employers who do it find their most expensive location is not the one they expected.

Ask almost any employer whether they bring injured employees back on light duty, and the answer is yes. Ask to see the list of light duty tasks, and most cannot show you one. The willingness is real. The system behind it usually is not.

That gap is expensive. Every day an injured employee stays home is a day of indemnity, a day of reserve growth, and a day of distance from the job and the people they work with. The longer that distance grows, the more likely the claim stops being about an injury and starts being about something else. Return to work is the largest lever an employer has on claim cost, and it is almost entirely in the employer’s hands.

Why return to work efforts stall

The first reason is that the work has not been identified. When a supervisor is asked on a busy afternoon whether there is anything an employee with a lifting restriction could do, nothing comes to mind. Given a week and a blank page, that same supervisor can list a dozen tasks. The work exists. The list does not, until it is needed, and by then the employee is already at home.

The second is that the treating provider is deciding blind. A provider who knows nothing about the job has one safe answer, which is to keep the employee off work. A provider who receives a written description of a specific transitional assignment, with its actual physical demands, has a real choice, and usually makes it in favor of getting the employee back.

The third is that nobody owns the outcome. Return to work sits between operations, HR and whoever handles claims, which means it sits nowhere in particular. When it belongs to everyone, it happens when someone remembers.

The protocol

Four components, all built before the next injury.

1. A transitional duty inventory. Every department writes down the tasks that can be done under the common restriction types: no lifting over a set weight, seated only, one hand, limited hours. Each department has a named owner and a deadline for producing it.
2. A written transitional job offer. When someone is injured, the offer goes to the treating provider in writing: the assignment, its physical demands, its hours and its expected length. A phone call asking whether light duty is possible does not count.
3. A contact cadence. A named person reaches the employee on day one, day three and weekly after that, and logs each contact. An employee who hears nothing concludes that nobody wants them back.
4. An end date on every assignment. Transitional duty without an end date slowly becomes a permanent accommodation, which is a different question and a worse outcome for everyone.

Multi state employers

Where an employer may direct initial care, and how much influence it has over the treating provider, varies by state. The inventory and the written offer work everywhere, which is exactly why they are the core of the system. Build them once, and adjust only the provider step state by state.

What to measure

• Days from injury to the first written transitional offer
• Percentage of lost time claims that received a written offer at all
• Lost days per claim, by location and by supervisor
• Percentage of transitional assignments that ended on their planned date

Where to start

If you do one thing this quarter, build the inventory. Everything else depends on it, it costs nothing but time, and it turns return to work from a favor your supervisors may or may not grant into an instruction your system issues.

Once an injury is reported, a lot of capable people start working on it. The adjuster, the carrier’s nurse, the defense attorney if it comes to that, and your broker. Every one of them is working from the same thing: whatever the employer captured on the day.

That is the part of the claim nobody downstream can fix. An adjuster cannot interview a witness who was never identified. Nobody can photograph a workstation that has since been cleaned up and moved. Nobody can recover the employee’s first account of what happened once three weeks of conversations have reshaped it. The quality of everything that follows is capped by the quality of the first few hours, and those hours belong entirely to the employer.

What goes wrong in the first hours

Three things show up again and again.

The report is late. Not because anyone hid it, but because the supervisor wanted to see whether it was serious first, or the employee mentioned it at the end of the shift, or the form sat in a tray until the HR manager was back. Every hour between the injury and the report is an hour where facts fade and the employee forms a view about how this is going to go.

The facts are thin. The first report says the employee hurt their back lifting a box. Which box, how heavy, from what height, who was nearby, what the floor was like, and what the employee said in their own words are all missing, and all impossible to rebuild later.

The employee goes to the wrong place. With no designated provider and no one to call, the default is the nearest emergency room, which treats the injury well and knows nothing about your operation or the modified work you could offer.

The protocol

Report the same shift, every time

The rule is simple: every injury is reported internally before the end of the shift it happened on, however minor it looks. Supervisors do not decide whether something is serious enough to report. That decision is the cause of most late reports, and taking it away from them is the fix.

Capture the scene before it changes

Photographs of the location and anything involved. The task being performed. Conditions at the time. Names of everyone present. The employee’s account in their own words, written down as they said it. Ten minutes on the day, and the single most valuable thing in the file.

Witness statements the same day

Written, dated and in each witness’s own words. A statement taken the same day is evidence. One taken a month later is a summary of a conversation about a memory.

A designated first call

Supervisors should never be deciding where an injured employee gets treated. They need one number to call, or one decision tree with three branches: emergency, same day care at the designated provider, or first aid with a check in. Each branch has one action.

A named owner for the first report

One person, with a backup, is responsible for getting a complete first report to the carrier. Not whoever is around. The most common failure in this window is two people each assuming the other one sent it.

What to measure

• Hours from injury to internal report
• Days from internal report to carrier report
• Percentage of first reports with photographs, a written employee account and witness names
• Percentage of injuries directed to the designated provider
• All of the above, by location

Where to start

Pull your last ten claims and read the first reports side by side. Count how many have a written employee account and named witnesses. That count is your baseline, and it usually makes the case for the protocol better than anything written here.

Your broker and your carrier can advocate hard on a claim. They can only advocate with the facts you give them.