An audit rarely starts with a question about whether you intended to run a fair program. It starts with proof: Who was eligible on the date of the pull? How was the selection made? Who was selected? When were they notified? What happened next? DOT random selection audit records must answer those questions without forcing you to reconstruct the story from a spreadsheet, email inbox, clinic receipt, and somebody's memory.
For a transportation employer, the random-selection process is not complete when names appear on a screen. The pull, the roster behind it, the notifications, the collection status, and the final result record all belong to the same compliance trail. If one part is missing or editable after the fact, the file is harder to defend when DOT asks.
What DOT auditors need to see
A defensible file shows that your random testing program operated as a repeatable process, not an occasional administrative task. The strongest records establish four facts: the eligible pool was correct, the selection was random, selected employees moved through the required testing workflow, and the record history was preserved.
The eligible pool is the starting point. A driver who is not performing a DOT-covered, safety-sensitive function should not be included merely because they remain on payroll. Conversely, a covered driver should not disappear from the pool because a roster was not updated after a hire, return to duty, transfer, or status change. Pool management is where many spreadsheet programs become vulnerable. A current roster may look clean today while offering no evidence of who was eligible when last quarter's draw occurred.
Randomness also needs documentation. A list labeled “random pull” is not enough by itself. Keep the selection date and time, the pool used, the number of people eligible, the number selected, the testing type, and the method used to generate the selection. The record should make clear that managers did not choose names based on route, performance, availability, or convenience.
Then comes follow-through. A selected employee needs prompt direction to report for testing, along with a status trail showing notification, collection progress, result entry, and any exception that required action. DOT rules and retention requirements can vary by testing category and operating administration, so employers should apply the requirements that govern their specific program. Your system should still preserve the underlying event history in every case.
The record set behind a random pull
Think of each selection as a closed file with a beginning, middle, and end. The beginning is the locked eligible-pool snapshot. The middle is the documented random draw and employee notification. The end is the collection and result-status history.
Start with a locked pool snapshot
A pool snapshot captures the roster as it existed at the moment of selection. It should identify the company or testing pool, each eligible employee, a stable identifier such as employee ID or driver ID, and the total number of eligible drivers. It should also show the date and time the snapshot was created.
Why lock it? Because a live roster changes. Drivers are hired, removed, placed on leave, or moved between pools. Those changes may be appropriate, but they should not rewrite the historical population from which a prior random draw was made. A locked snapshot lets you show an auditor the exact population used on that date, even after the current roster has changed 20 times.
Avoid treating a saved spreadsheet as equivalent proof. A spreadsheet can be overwritten, formulas can change, and there may be no reliable indication of who edited it or when. If you use spreadsheets, preserve the original exported roster and document access controls. A purpose-built system is usually cleaner because it retains the snapshot automatically when you run the pull.
Document the random-number method
The selection record should state how the system generated the draw. Computer-generated random selection is practical and easy to document, provided the method is consistently applied and the pool is defined correctly before the draw.
Your record should capture the randomization event, not just its output. That means a timestamp, the selected testing category, the draw size, the pool size, and the resulting selected names or IDs. If the platform can retain a draw reference or randomization log, keep that with the record.
The number selected deserves attention. Annual minimum random testing rates are generally translated into periodic selections based on your pool size and testing schedule. A small fleet may not select the same number every month, especially when rounding and annual rate planning are involved. What matters is that the employer can show a rational, documented approach that meets the applicable annual requirement. Do not use a fixed monthly number without checking whether it actually produces the required annual testing rate for the pool.
Preserve notification and status history
A selected driver who does not receive a documented instruction to test can create a gap between the pull and the collection. Record when the employee was notified, who issued the notification, and the delivery method. If a driver is dispatched, unavailable, or otherwise cannot report immediately, document the operational facts and the next action taken.
This is not paperwork for paperwork's sake. Notification history can show that the employer acted promptly and did not give advance warning that compromised the random nature of the test. It can also explain why a collection occurred later than the selection event.
Use status changes that reflect the real workflow: selected, notified, scheduled or sent for collection, collected, result pending, result entered, and closed. If a status changes, the system should retain the prior status, the time of change, and the user or source responsible. A single field that simply says “complete” does not tell an auditor much.
Keep result records from being quietly rewritten
Result entry needs controls. Corrections sometimes happen, particularly when a result is entered from a laboratory, medical review officer, or alcohol testing record. The answer is not to make results impossible to correct. The answer is to preserve the original entry, record the correction, and show who made it and why.
Write-once result records or a clear amendment history prevent a common audit problem: a record that appears to have changed with no explanation. The same principle applies to canceled tests, refusals, and other exceptions. Capture the status and supporting documentation under your applicable procedures rather than deleting the event because it complicates the report.
A practical workflow your team can repeat
The cleanest process is the one a safety manager can run on a busy day without handing the work to a third party. Start by reviewing the active DOT-covered driver pool before each draw. Add eligible new hires, remove people no longer covered, and confirm that each driver is assigned to the right company and testing category.
Run the computer-generated selection and immediately retain the locked pool snapshot with the draw record. Issue notifications from the same workflow when possible, then update each selected driver's status as testing moves forward. Enter results from authorized documentation and retain the time-stamped status history. At the end of the period, export the record set and store it according to your retention schedule.
DTSMS Online is built around that chain: you upload and manage the pool, run the pull yourself, create driver notifications, and retain the snapshot, selection method, status history, and result record in one place. No sales call, no quote request, and nobody in between when a routine random selection needs to happen.
Where audit records usually break down
The most common failure is not a dramatic violation. It is a fragmented process. The roster is in one file, the randomizer is in another, notification is in email, and results sit in a clinic portal. Each piece may exist, but nobody has assembled a complete historical record.
Another weak point is editable history. If a manager can change a prior selection date, replace a selected driver, or remove a result without an audit trail, the program may look less credible even when staff acted in good faith. Limit permissions, log meaningful changes, and preserve original events.
Finally, do not confuse a current report with a historical record. An auditor may ask for a past period. Your report must reproduce what happened then, not display today's driver roster with old test dates attached.
Your goal is simple: when DOT asks for a random-selection file, produce a complete chain of evidence in minutes. Build the process so the record is created as the work happens. That is how you retain control, skip unnecessary service fees, and keep your testing program ready for scrutiny.
