Backing out before start often sits with a drug mention

When a candidate backs out before the start date, the write-up is often not just a change of heart. A drug or alcohol mention sits in the same file far more often than chance: a pre-employment test that never happened, a last-minute never started, sometimes a ticket already bought. Treat that pairing as one screen, not two unrelated problems.

Among the reviews carriers left on CDLScan, 11.7% mention drugs or alcohol. That figure describes those reviews, not drivers in general: carriers write after a bad run, so the share tells you what shows up once someone bothers to report it.

Backing out before the start date is a smaller slice: 1.6% of reviews mention it. Flip the pair and read the denominator. Inside that backing-out slice, 25.8% also mention drugs or alcohol. That is 25.8% of changed-mind reviews, not 25.8% of all reviews, and not 25.8% of drugs-or-alcohol reviews. Same overlap, two bases. Always name which one you are standing on.

That pairing is not random. Drugs or alcohol are 11.7% of all reviews. If backing out and a drug mention showed up together only by chance, you would expect about 11.7% of backing-out reviews to also mention drugs or alcohol. The observed share is 25.8% — about 2.2 times the chance rate. Lift is that multiplier. It is not a probability that the next person who says they changed their mind failed a test.

Last minute, never started

The write-ups talk about a last minute change, a changed mind, a drug test, never started, a plane ticket or a flight ticket in the same paragraph, never showed, another company, a booked flight, orientation they never came to. That is the incident register.

Stay away and waste your time are a different register. Those are the author’s recommendation to the next recruiter, not a description of what the candidate did. In 39.3% of drugs-or-alcohol reviews the carrier also writes a verdict. In 42.8% of backing-out reviews they do the same. That is a fact about how these reviews end, not a fact about the driver. Read the scene. Do not hire off the temperature of the last sentence.

From the drugs-or-alcohol side, the louder companions are still the empty chair and the unused ticket. Paid travel with no arrival sits in 16.0% of drugs-or-alcohol reviews. A no-show sits in 14.5%. Going unreachable sits in 12.2%. An abandoned truck sits in only 10.3% — this is mostly not the walk-off population. Aggression or threats sit in 7.5%. Theft sits in 7.0%. A crash or damage sits in 6.4%.

Backing out is the quieter cousin of that same week. From the other side of this pair, paid travel sits in 15.4% of backing-out reviews. A no-show sits in 8.8%. Going unreachable sits in 5.1%. They accepted. You scheduled the pre-employment test. They changed their mind. Sometimes the ticket was already bought. Sometimes they never started and the next line is another company.

I keep the test next to this pair without turning the article into a clinic manual. Official reports are built to hold a result. They are not built to hold a start date that never happened. The backing-out itself is the spend. The test that never happened is the same spend told twice. The no-show side of that story is in a drug mention that often arrives with a no-show. The ticket side is in paid travel and a drug mention. The drug-mention trend itself is in a drug mention that is often a missed test.

The share of reviews mentioning drugs or alcohol ran 13.6% in 2018, 12.1% in 2019, 13.0% in 2020, 14.7% in 2021, 11.8% in 2022, 10.9% in 2023, 10.2% in 2024, and 11.2% in 2025. I stop that line at 2025. 2026 is still open. 2021 is the high you should not smooth over, 14.7%. Then the share stepped down and 2025 is 11.2% — lower than 2018, not gone.

Backing-out mentions ran 1.2% in 2018, 0.8% in 2019, 1.4% in 2020, 1.4% in 2021, 1.8% in 2022, 1.5% in 2023, 1.5% in 2024, and 1.9% in 2025. The drug-and-alcohol line eased. The last-minute never-started line did not. How these shares are counted is on the methodology page. Compare percentages inside each year, not year totals.

What the pairing is not

It is not proof that every backing-out is a drug story. Most backing-out reviews do not also mention drugs or alcohol. 25.8% is a heavy overlap relative to chance, not a majority.

It is also not the same population as the driver who takes the truck and leaves it. Only 13.0% of backing-out reviews also mention an abandoned truck. Only 10.3% of drugs-or-alcohol reviews do. The person who never starts and the person who arrives, takes a unit, and walks are mostly different write-ups. Screening for one will not catch the other. That split is the point of two failure populations that barely overlap.

Going unreachable sits in 12.2% of drugs-or-alcohol reviews and in 5.1% of backing-out reviews. That is the quieter version of the same week: the test is scheduled, then the phone dies, and you find out they changed their mind when the clinic asks who is paying for the empty slot.

Why the packet you already buy stays quiet

PSP, an MVR, and the Clearinghouse hold what someone filed. A last-minute backing out after a scheduled test is an accounts-payable problem until a result exists. A Clearinghouse hit can sit next to it — when a test was actually taken and reported. The empty start date itself never becomes a query result.

Previous-employer forms are not much better here. The last carrier may never have onboarded the person. There is no eligible-for-rehire box for someone who never got a truck. The story lives in the write-up the last recruiter left when they were tired of paying for seats and clinic slots that stayed empty.

A clean packet on Monday morning does not tell you whether the candidate you just scheduled will still want the job on Wednesday. Those are different claims.

I am not going to pretend this is an independent study. CDLScan is the source, and I write from the data they share. Drugs or alcohol are common enough in what carriers bother to write down, backing out rides along far above chance, and the packet you already run is designed for the result, not for the absence.

What was visible before you spent

The pattern I’d watch for is urgency around the start date and vagueness around the last test. They want orientation this week. The last carrier is a short stint or a hole. References are slow. Official reports are dull, because there is nothing to file yet.

Typical sequence, built from how these reviews cluster, not from one person’s file: you schedule the pre-employment test. They confirm. They change their mind the day before. They never start. Sometimes a flight was booked on the same invoice. Sometimes they take another company after you already paid. Sometimes they stop answering and you only later hear they backed out.

None of that is a reason to stop testing. It is a reason to run the name before the clinic slot and the ticket are paid.

What to do before the start date

Look the name up in what previous carriers already wrote. If you see a drug test and a last-minute changed mind in the same history, treat the next clinic slot as a decision, not a courtesy.

Ask when they can start without you buying anything. People who are choosing between two real seats can usually wait a few days. People who need the ticket or the test as the job will push the purchase forward and go quiet on the last employer.

If you already booked the slot, freeze the hotel and the advance until someone has looked. Check the name on CDLScan first. It takes less than a minute, and it is where that pairing actually sits.

Frequently asked questions

Does 25.8% mean a backing-out is usually a drug mention?

No. 25.8% is the share of backing-out reviews that also mention drugs or alcohol, among reviews carriers left on CDLScan. Most backing-out write-ups do not include a drug mention. The pairing is far above chance. It is not a majority.

Why talk about 11.7% if the pairing is 25.8%?

Because those are different bases. 11.7% of all reviews mention drugs or alcohol. 25.8% of backing-out reviews also mention drugs or alcohol. Always name the denominator.

What does lift 2.2 mean here?

Drugs or alcohol are 11.7% of all reviews. Seeing them in 25.8% of backing-out reviews is about 2.2 times the chance rate. Lift is that multiplier, not the odds on your next hire.

Will the Clearinghouse warn me?

It will warn you about a reported result. It will not warn you about a start date that never happened. A last-minute backing out after a scheduled test is not a query result until something was filed.

Is this the same problem as an abandoned truck?

Usually not. Only 13.0% of backing-out reviews also mention an abandoned truck. The person who never starts and the person who leaves a unit are mostly different populations.

What should I ask before I schedule the test?

When they can start if the clinic waits two days, and what the last carrier will say about why they left. Then read the write-ups. Do not let the appointment become the background check.

Where do I check the name?

On CDLScan, before the slot is paid. Previous carriers already wrote this pairing down. You should read it while the money is still yours.