What Your Agency Actually Pays for Miles Nobody Verified

There is a line on your P&L labeled something like "mileage and travel reimbursement." Most operators I talk to can tell me the number. Almost none of them can tell me whether it is correct.

That is not a criticism. It is a structural problem. In home health, the vehicle belongs to the clinician, the fuel card does not exist, and the odometer reading arrives on a piece of paper two weeks after the fact. You are reimbursing a cost you cannot observe, reported by the person receiving the reimbursement, on a form nobody has time to check.

This article is about how much that costs, how to find out in a single pay period, and what to fix first.

Start with the rate, because it moved twice this year

The IRS standard business mileage rate has changed more in the last eighteen months than it usually does in five years:

• 2025: 70 cents per mile
• January 1 – June 30, 2026: 72.5 cents per mile
• July 1 – December 31, 2026: 76 cents per mile

That July increase is worth pausing on. Mid-year rate revisions are rare - the IRS does not normally reopen the number halfway through a tax year. It did this time.

Two things follow from that. First, if your reimbursement rate is still set where you put it in 2025, you are now paying your field staff roughly six cents a mile below the federal figure. On a clinician driving 900 miles a month, that is about $54 a month out of their pocket, quietly. Second, if you did update the rate but you are paying against unverified logs, you just increased the cost of every inflated mile by the same six cents.

Both of those are usually true at the same time in the same agency. You are underpaying the honest ones and overpaying on the miles nobody checked.

The five leaks, and what each one looks like on a real visit sheet

None of this is fraud in the way people imagine fraud. Almost all of it is the predictable result of asking a busy clinician to remember a number at the end of a long day.

1. Estimated odometer readings

The clinician forgot to write down the reading and reconstructs it from memory that evening. A log says 47 miles between two addresses that are twelve minutes apart. Nobody lied. Somebody guessed, and guesses round upward.

2. Rounded-up trips

A 6.3-mile leg becomes "about 8." Every leg, every day, every clinician. This one is invisible because each individual instance is defensible. It only appears in aggregate.

3. The same trip claimed twice

Agencies running a paper log alongside an app, or a mileage field in the EMR alongside a spreadsheet the office keeps, will eventually pay for the same drive out of two systems. Not often. But nobody is looking, so nobody knows how often.

4. Logs submitted three weeks late

Late submission is the leak that enables all the others. Once a log arrives outside the pay period it describes, verification becomes impossible - the schedule has changed, the visit sequence is fuzzy, and the person reviewing it has no basis to push back on anything.

5. No receipts on supply reimbursements

Mileage is the big number, but it is not the only one. Gloves, wound-care supplies, parking, tolls, and phone stipends usually run on an honor system with no documentation requirement. Small amounts, high frequency, zero visibility.

The scale, so you can size this for your own agency

Home healthcare workers in the United States drove approximately 7.9 billion miles to deliver 718 million patient visits in a single measured year. That is roughly eleven miles per visit, on average, across the industry.

Run that against your own volume. An agency doing 2,000 visits a month is looking at something in the neighborhood of 22,000 reimbursable miles a month. At 76 cents, that is about $16,700 a month, or $200,000 a year, per branch.

Now assume a conservative 8% of those miles are estimated, rounded, or duplicated - a figure that is difficult to argue is aggressive given the reporting method. That is $16,000 a year, per branch, going out the door on miles nobody drove.

I am not going to pretend that number is precise for your agency. It is an order of magnitude, and the point of the next section is that you can replace it with your actual figure in about two hours.

How to measure it this month, with what you already have

You do not need software to find out whether this is a problem. You need one pay period and an afternoon.

1. Pick one completed pay period and one discipline. Skilled nursing is usually the cleanest place to start because visit addresses are well documented.
2. Pull the submitted mileage per clinician for that period. Whatever form it arrives in.
3. Pull the actual visit sequence from your EMR for the same people and dates - addresses in the order they were seen.
4. Calculate the real driving distance for each sequence. Any mapping tool will do this. You are looking for the distance between consecutive visit addresses, excluding the trip from home to the first visit and from the last visit home.
5. Compare the two totals per person and look at the spread.

What you are looking for is not one bad actor. You are looking at the shape of the distribution. If most people land within 5% of the calculated distance and one person is 40% over, you have a conversation to have. If everyone is 15–20% over, you do not have a people problem - you have a process that produces inflated numbers regardless of who is filling it in.

In my experience it is almost always the second one.

Fix it in this order

Do not start with software. Start with the definition, because software applied to an undefined policy just produces faster ambiguity.

First: define what is compensable, in writing

The general rule is that travel between patient visits during the workday is reimbursable, and the commute from home to the first visit and from the last visit back home is not. Agencies get into trouble when this has never been written down, because then every clinician has their own version of it and none of them are wrong.

Write it down. One page. Include what counts, what does not, what documentation is required, and when it is due.

One caution: a handful of states impose reimbursement obligations beyond the federal floor, and the rules on how those interact with wage aggregation vary. Check your state's labor guidance directly rather than a vendor's summary of it - I have seen more than one software marketing page carrying a stale rate or an oversimplified state rule.

Second: require submission inside the pay period

This single change does more than any other. A log submitted while the week is still legible can be checked. A log submitted three weeks later cannot. Make it a condition of that period's reimbursement, and enforce it the first time so you never have to enforce it again.

Third: verify against the schedule, not against the clinician

The comparison you ran above becomes a monthly spot check on a sample. Not an audit of everyone, every time - that costs more than it saves. A visible sample is enough to change reporting behavior across the whole team.

Fourth, and only then: automate

Once the policy is written and the timing is enforced, automating capture is straightforward and worth doing. Automating before that just moves the ambiguity into a database.

The part of this that is not about money

There is a second cost here that does not show up on the expense line, and it is the more expensive one.

When fuel prices rise and the reimbursement rate does not, clinicians start declining cases - specifically the far ones, the rural ones, the ones that require crossing the service area. They do not announce this. It shows up as an admission you could not staff, and it gets recorded as a capacity problem rather than a reimbursement problem.

This is what makes home health different from every generic expense-management article you will read on this topic. In a sales organization, an under-set mileage rate is an employee-satisfaction issue. In home health, where the clinician is paid per visit and drives their own vehicle, it is a census issue. You are not just managing an expense. You are managing whether the referral you fought to win can actually be seen.

That is the real argument for getting this right. Not the $16,000. The referrals you turned down without ever deciding to.

Where to go next

If you run the pay-period comparison and find a spread you did not expect, the two adjacent problems are worth looking at in the same pass. Your clinicians driving their own cars creates an insurance exposure most agencies do not know they carry, and the routes they are driving are usually longer than they need to be. All three are the same underlying issue: field activity you are paying for but cannot see.

Fixing the process is the hard part, and it is the part we do. Keeping it fixed after the project ends is why we built our own platform - you can read about where HHAEdge fits in an engagement if that is useful.