Measuring NEMT driver performance: on-time, safety, complaints, and inspections

Start with what your brokers already measure

Your brokers grade the company on numbers that are really the sum of your drivers’ days. The simplest driver scorecard uses the same definitions, so a driver’s number and the company’s number move together.

Two current MTM handbooks show what that grading looks like:

The Virginia handbook (approved August 10, 2026) covers the Medicaid fee-for-service rides MTM Health brokers from October 1, 2026, and a provider that falls short of it risks liquidated damages, a corrective plan, or losing its network contract. Rhode Island’s handbook (last updated July 1, 2026) uses green, yellow, and red bands, and timeliness misses of 7 percent or more land in the red. The miles-versus-minutes and overlapping-claims checks exist to catch records that cannot be true, which is why they belong on a driver scorecard too. A driver whose trip times do not match the miles is either keeping sloppy records or something worse.

Your company-level measures, the ones in the NEMT KPIs guide, sit on top of these. This page is about the driver level underneath.

The driver scorecard

Keep it to measures a driver controls and your records capture without extra work. Eight measures do the job:

Set each window from your contracts, not from habit. In MTM Health’s Virginia handbook, a driver may reach the first (A leg) pickup up to 15 minutes early or 15 minutes late and still count as on time; that allowance does not extend to drop-offs, and the rider must reach the appointment by its start time. Will-call returns get 45 minutes from the rider’s ready call, and hospital discharges get three hours. Rhode Island’s MTM handbook also allows 45 minutes for will-call returns, and it treats a pickup that lands over 15 minutes off the time the rider was given as untimely. When a driver works for several payers, score each trip against that payer’s pickup window.

Some useful conversions: 95 percent on-time means no more than 5 late pickups in every 100. A complaint standard of 0.1 percent is 1 substantiated complaint per 1,000 trips. At 20 trips a day on a five-day schedule, a driver finishes about 1,000 trips in 10 weeks, so a single complaint moves that driver’s rate a lot. Use a rolling window long enough to be fair.

Making the numbers fair

A scorecard drivers do not trust makes things worse. Four rules keep it fair.

Give every late trip a cause. Tag each late trip as driver-caused (left late, long break, wrong turn), schedule-caused (no drive time between pickups, a trip added mid-route), or outside (a crash on the highway, a facility that held the rider). Only driver-caused lateness goes on the driver’s score. Schedule-caused lateness goes on dispatch’s. The on-time performance guide covers the causes in more depth.

Count only substantiated complaints. Virginia’s standard counts substantiated complaints for a reason. Brokers take complaints seriously: CareOregon’s manual assumes a grievance or incident may well be valid and preventable until the records say otherwise. That is why a driver’s clean trip record and camera footage matter, and why an unreviewed complaint should not hit a score.

Score crashes on preventability, not fault. The OSHA, NETS, and NHTSA guidelines tell employers to have every crash reported, however minor, and to review each one for its cause and whether it could have been prevented, whoever was at fault. The police report tells you who got the ticket. Your review tells you whether a safer choice was available to your driver.

Require a minimum number of trips. Percentages on small samples swing wildly. Set a floor, such as 60 trips in the period, before a rate counts, and use a rolling 90 days for complaints and crashes.

Example: one driver’s month

This scorecard is an illustration with made-up numbers.

The follow-up is short: a strong month, three missing signatures to fix, and a question for dispatch about why one afternoon stacked five will-call returns on one van.

Sharing results without shaming

Give each driver their own numbers privately, with the trips behind them. A driver who can see which pickups were late, and why, can fix something. A driver who sees only a percentage, or a ranking posted on the break-room wall, learns to argue.

• Private, regular, and specific. Send each driver their scorecard every week or two and include the trip list for anything flagged.
• Fleet averages without names. Drivers want to know where they stand. Show the fleet average and the target, not who is last.
• Recognition in public, correction in private. Name the drivers who hit every target. Handle the misses one on one.
• Explain the tracking. OSHA’s advice on in-vehicle monitoring is to tell workers up front why the system is going in and what it will do. The same goes for GPS scoring and cameras.
• A way to challenge a number. A driver who thinks a trip was tagged wrong gets it reviewed with the record on screen.

Brokers treat providers the same way. In MTM Health’s Virginia handbook a field monitor, and in MTM’s Rhode Island handbook a vendor account manager, goes over each metric with the company on a regular basis. Mirror that with your drivers.

Coaching with timestamps and GPS

Trip records turn a vague complaint into a specific fix. A coaching session built around them follows a pattern:

1. Pick one measure. The weakest one, or the one that matters most to your contracts this month.
2. Pull three to five trips that show the pattern, with scheduled and actual times.
3. Walk the timeline together. When the trip was assigned, when the driver left, when they arrived, when the rider boarded, when they reached the destination. The GPS track shows where the minutes went: a long stop before the first pickup, a detour, a wait at the wrong entrance.
4. Ask the driver first. Often the driver knows the cause (a building whose main entrance is not the one on the address, a rider who is never ready on Tuesdays), and the fix is a trip note or a schedule change, not a lecture.
5. Agree on one change and a date to look at the numbers again, usually two weeks out.
6. File a short note of what you agreed, so the next review starts from it.

Coaching with data works. OSHA’s employer page points to NIOSH research: drivers who got feedback from monitoring equipment in the cab, plus supervisor coaching based on video facing both the driver and the road ahead, cut risky driving significantly and buckled up more often than a comparison group. Cameras help here too: CareOregon’s provider manual strongly recommends dash cameras that film both the cabin and the road, and the video answers questions a timestamp cannot. The telematics guide covers the equipment.

Hold the conversation at the yard or in a parked van. OSHA notes that dispatch and rerouting can distract a driver at the wheel, and a coaching call is no different.

Linking the numbers to reviews, bonuses, and discipline

Metrics earn their keep when they feed the decisions drivers care about. The OSHA, NETS, and NHTSA guidelines note that positive results follow when driving performance becomes part of the overall evaluation of job performance.

• Reviews. Use the scorecard as the factual half of a quarterly review, with observations from a ride-along or road test as the other half. The driver evaluation form has a layout.
• Bonuses. Pay for thresholds on the same measures, such as on-time above your contract standard with no preventable crash in the period. The driver bonus guide covers the design and the overtime rules that come with it.
• Discipline. A pattern on the scorecard, such as three months of driver-caused lateness after coaching, is the trigger for the next progressive step. The driver discipline guide covers the steps.
• Broker incentives. Some brokers pay providers for punctuality. Under MTM Health’s Virginia handbook, providers that hit at least 95 percent on time for a quarter share an incentive pool funded from liquidated damages. Driver scorecards show which drivers earned the company its share.

Review the scorecard itself twice a year. Drop a measure nobody acts on, and change a definition only at the start of a period.

Pulling driver numbers from HealthRide

In HealthRide, the trip log lists every leg with its scheduled and actual times, the driver and vehicle, and GPS-verified miles. Punctuality is tracked automatically on every leg, overall and per driver. Late warnings on the live map flag trouble during the day, so a coaching talk can start while the trip is still fresh. See reports for the exports.