A DOT auditor does not need to be convinced that you intended to run a fair random program. They need to see what your eligible pool looked like on the day of the draw, how the selections were made, who was notified, and what happened next. That is how you build audit evidence that holds up when someone asks for the record months later.
For a transportation employer, random testing compliance is not one document. It is a connected chain of records. A spreadsheet with names and a clinic receipt may show activity, but it often cannot prove that your process met the standard at the time the selection occurred.
Build Audit Evidence From the Moment You Run the Pull
The strongest records are created as part of the work, not assembled after an audit notice arrives. Every manual correction, emailed roster, overwritten spreadsheet cell, and undocumented phone call creates a gap someone may have to explain.
Start with a clear question: could you recreate a specific random selection from its original records without relying on memory? If the answer is no, the process needs tighter controls.
A defensible random-testing record should establish five things: the correct employees were eligible, the selection was genuinely random, the required number of employees was selected, each selected employee was handled promptly, and the final outcome was recorded without rewriting history.
That does not mean every fleet needs a complicated enterprise system. It means your records must preserve the facts that matter. Small fleets are often especially exposed because one person may manage driver records, dispatch, notifications, and collection coordination. A simple process is fine. A process that depends on one person's recollection is not.
Start with the eligible pool snapshot
Your pool is the foundation of every selection. Before a random pull, identify the safety-sensitive employees who belong in that DOT testing pool at that point in time. Keep the pool separated by DOT agency when required. Do not combine employees subject to different agency rules simply because they work for the same company.
The record should show the names or unique employee identifiers in the pool, the pool count, the program type, and the date and time the pool was used. It should also preserve the status of the roster as it existed before the draw.
A locked snapshot matters because driver rosters change. A driver may be hired, terminated, placed on leave, or moved into a safety-sensitive role after the pull. If a live spreadsheet is edited later, it becomes difficult to prove whether that person was in the pool on selection day. The current roster may be accurate today and still be the wrong evidence for last quarter's selection.
Use a process that saves the original eligible-pool snapshot and prevents later edits from changing it. If a correction is necessary, add a documented correction record rather than silently replacing the original.
Document How the Random Selection Was Generated
Random means every covered employee has an equal chance of selection each time a draw is run. It is not enough to label a list "random" or use a manager's judgment about who has not tested recently.
Your evidence should identify the selection method, date and time of the draw, the population size, the number selected, and the random-number result or selection output. If software performs the draw, preserve documentation showing that the method is computer generated and not manually manipulated.
The details matter when the pool is small. Suppose a fleet has 18 CDL drivers and needs to make a drug selection. An auditor may ask how the number selected was calculated, whether all eligible drivers had the same chance, and whether selections across the year met the applicable annual testing rate. Your record should answer those questions directly.
Do not use predictable workarounds such as selecting the same people on a schedule, excluding difficult-to-reach drivers, or rerunning a selection because the first result was inconvenient. If an employee is legitimately in the eligible pool, that employee must have the same chance as everyone else.
There is a practical trade-off here. A basic spreadsheet can calculate a number of selections, but it rarely creates a reliable, timestamped record of the exact roster and randomization process. A purpose-built random-selection platform costs less than reconstructing a broken paper trail, especially when it removes the service fee and delay of routing routine pulls through an outside administrator.
Tie Each Selected Driver to a Notification Record
A selection record is only the first half of the story. Once employees are selected, document how and when they were notified. The record should connect the selected driver to the draw, show the notification date and time, and identify the person who issued or documented the notification.
The notification method can vary by operation. A dispatch office may notify a driver in person. A field-service fleet may use a documented phone call and a follow-up instruction. A passenger carrier may notify at the end of a shift to protect the integrity of the testing process. The right operational method depends on your work pattern, but the evidence must show that the employee was directed for testing without unnecessary delay.
Keep the notification record separate from casual communications. A text message buried in a supervisor's personal phone is weak evidence if it cannot be retrieved, tied to the selection, and placed in the program file. Use a consistent notice format or system-generated notification record instead.
If a driver cannot be reached, document the attempts and the operational facts. Do not quietly replace the person with another driver just because scheduling is inconvenient. A replacement decision can change the integrity of the selection record and should be handled under your written program procedures.
Track Collection, Exceptions, and Final Status
From notification through completion, each change should have a timestamp and a clear status. Examples include notified, sent for collection, collection completed, result pending, negative result recorded, positive result received, canceled with reason, or refusal handled under company policy and DOT requirements.
This is where fragmented records create problems. The roster may live in one spreadsheet, the notice in an email folder, the collection confirmation at a clinic, and the result with HR. When the audit request arrives, someone has to match those items manually and hope no document is missing.
A single status history gives the record continuity. It shows not only the final result but the sequence of actions that produced it. Write-once result records are particularly useful because they preserve accountability. If a result entry needs correction, the correction should be traceable, with the original record and change history retained.
Be careful with results handling. Drug and alcohol testing records contain sensitive information and must be stored, accessed, and retained in accordance with applicable DOT requirements. Audit-ready does not mean broadly available. It means the authorized employer representative can retrieve complete records when required while limiting access to people with a legitimate compliance role.
Record exceptions instead of hiding them
Real operations do not always follow the happy path. A selected driver may be on approved leave, dispatched far from the collection site, unavailable because of a medical emergency, or removed from safety-sensitive work before a collection can occur. The answer is not to make the event disappear.
Record the exception, the date it was identified, the action taken, and the person responsible for the decision. Your written policy and the circumstances will determine the next step. What matters for evidence is that the file tells the truth about what happened.
The same applies to late testing, canceled collections, and administrative errors. An audit trail that shows a problem and a documented correction is more credible than a file that appears artificially perfect.
Make Annual Rate Tracking Part of the Record
Each DOT agency sets annual random testing rates, and those rates can change. Your program must use the applicable rate for the period and maintain enough testing activity across the year to meet it. A quarterly pull schedule can be practical, but the number selected in each period may need adjustment as the average pool size changes.
Keep a running view of your average number of safety-sensitive employees, selections made, completed tests, and remaining tests needed for the year. Do not wait until December to discover the program is short.
A system such as DTSMS Online can preserve pool snapshots, document computer-generated selections, issue notifications, and maintain timestamped status histories in one place. You run the pull. The record stays connected. No sales call, no quote request, and nobody in between when routine random administration needs to happen.
A Record Should Answer the Auditor Before They Ask
The practical test is simple: select any completed draw from the past year. Can you produce the locked pool, selection output, notification history, collection status, and final record in a logical sequence? Can you explain an exception without searching through inboxes or asking a former dispatcher what happened?
If not, fix the workflow now, while the details are current. Build records that reflect the work as it happens, protect the original facts, and give your company control of the random-testing process. When DOT asks, you should be able to show the chain of evidence, not reconstruct it.
