Nemt quality control: the questions that matter

What to take away

An operator who asks a facility whether the service is satisfactory will be told it is fine until the day they are replaced. The information that would have helped was available months earlier, in a load that took too long and a driver who had stopped mentioning it.

So the checks below are things you look at, not things you ask about.

Check one: the observed ride

Once a month, ride along for a full shift with a different driver each time. Sit in the back where you can see the load.

Record for each trip the arrival-to-departure time, split when the passenger reaches the vehicle. Note whether access notes matched what was there, and whether securement followed procedure without improvisation.

Also record how the driver spoke to the passenger at the point of most effort, and whether the record was completed at delivery or left for later.

Two rules make this useful. Tell the driver in advance what you are recording, so it is not a trap. And record what you saw rather than what you concluded, because the conclusions change and the observations do not.

Run it in this order so the day is not disrupted:

What an observed ride usually reveals

The last row matters. Most quality reviews only look for faults, and the practices worth spreading are found the same way.

Check two: punctuality against the appointment

Measure lateness against the appointment time, not against the pickup window.

A trip that collected the passenger inside the agreed window and delivered them after their appointment started has failed, and a pickup-window measure will score it as a success. In this trade the appointment is the deadline: a treatment slot missed is a session rescheduled and, for a repeating patient, a whole week disturbed.

Report it as the share of trips delivered before the appointment time, by account and by trip type. Where the share is poor for one account, look at their access conditions before you look at your drivers. The scheduling arithmetic behind this is set out in the material on building a schedule that holds.

Check three: complaints, classified by cause

Log every complaint, including the ones made verbally to a driver, and classify by cause rather than by who complained.

Useful categories: late against the appointment, wrong vehicle or equipment for the passenger, assistance level not delivered, driver manner, record or billing error, communication failure when something changed. Count them monthly.

The point of classifying by cause is that one cause usually produces complaints at several accounts, and fixing it fixes all of them. Classifying by account produces a list of difficult customers and no improvements.

The near miss is the most valuable record

A load that nearly went wrong, a securement that did not feel right, a doorway that was almost impassable, a passenger who was almost left waiting. None of these produce a complaint and all of them predict one.

Make it easy and consequence-free to report them. An operation where near misses are never reported is not one where they do not happen; it is one where nobody has been given a way to say so.

A securement that did not feel right deserves particular attention. It usually signals a chair, vehicle and tie-down system combined in a way none of their manufacturers described.

Check the manufacturers' own instructions, not the federal cargo securement rules, which govern freight and are cited here far more often than they apply.

What to do with the results

Fix the process before the person. Where a load consistently overruns, change the allowance. Where access notes are wrong, change the intake questions. Where a securement was improvised, put the instructions in the vehicle and retrain on that specific chair.

Where a finding is genuinely about a person, it belongs in the competency review described in the hiring and training material. Where it is about equipment that makes work harder than needed, it belongs in the fleet and premises guide.

The daily design that all of this is testing is set out in the operations and workflow guide.

Records, and what they have to support

A payer may audit the same records that support quality review: times at both ends, mileage, assistance level delivered, and your contract's required confirmation.

Keep documents that support what was billed, as the Internal Revenue Service describes a general expectation. Trip-level requirements come from your program, and the Centers for Medicare and Medicaid Services page explains they vary by state.

Common questions

How often should I ride along?

Monthly is enough to see patterns and rare enough not to be intrusive. Rotate drivers and rotate the type of work, because a morning of standing orders shows you something different from an afternoon of will-calls.

Should I survey passengers?

You can, and expect little from it. Passengers in this trade are frequently unwell, dependent on the service and reluctant to complain. The observed ride tells you more.

What if a driver behaves differently because I am watching?

They will, and the timings still tell you something: if the load takes too long even on a good day, the allowance is wrong. Behavior you cannot observe is better addressed by making the near-miss report easy than by watching harder.

How do I raise a finding with a facility?

With the record, at a scheduled review, and framed as a shared problem. A facility told that its own entrance costs eleven extra minutes per trip usually wants to help, because those minutes cost them too.