Laboratory Execution Systems: Ticked Is Not Executed
A signed SOP says a procedure was followed. It cannot say when the third step happened or whether the fourth was skipped. That difference is the whole case for a laboratory execution system.
In brief
A laboratory execution system records a method as it runs: who did each step and when, against a procedure that cannot be edited while it executes. It sits between a LIMS, which tracks the sample, and a notebook, which holds the write-up. A lab needs one once an auditor asks how a result was produced, not only what it was.
Key takeaways
- Execution is evidence that a method ran. A signature on an SOP is an assertion that it did, made once, at the end.
- A method that can be edited while it runs is not the validated method, whatever the record says afterwards.
- The step, the calculation and the form belong in the same record, not in three places that have to be reconciled.
- A notebook is enough until someone asks how a result was produced and expects the record, not a person, to answer.
- The line between LIMS, LES and notebook is the unit of work: the sample, the run and the write-up, not the feature list.
Every regulated lab has a shelf of signed SOPs and a drawer of completed worksheets, and every one of those worksheets carries the same claim: this procedure was followed. An auditor who wants to test that claim has one question, and it is not whether the signature is genuine. It is when the third step happened, and whether the fourth was done at all.
A laboratory execution system, usually shortened to LES, exists to answer that question. It is the least glamorous of the laboratory informatics categories and the one whose purpose is easiest to state. This article explains what an LES records, why the record is different in kind from a signed procedure, why the method has to be locked while it runs, and when a lab has outgrown a notebook and an SOP. It does not re-argue LIMS against ELN; that comparison has its own article. The question here is narrower: what does execution prove?
What is a laboratory execution system?#
A laboratory execution system is software that walks an analyst through a method one step at a time and records each step as it is completed. The record it produces is not a description of the procedure. It is a log of a particular run of that procedure: which steps were performed, in what order, by whom, at what time, with what values entered along the way.
That places it between two systems most labs already know. A LIMS is organised around the sample: where it is, what tests it needs, what results came back. An electronic notebook is organised around the write-up: what was done and what it means. An LES is organised around the run. The unit of work is one execution of one method on one occasion, and everything it records hangs off that.
The idea is older than the acronym. Manufacturing has had execution systems for decades, because a batch record that says what actually happened is the difference between a release decision and a guess. Laboratories took longer to adopt the same idea because a bench method feels more like craft than production. Regulators have not shared that view for some time. Where a result depends on a procedure being followed, they want evidence that it was, and a signature at the bottom of a page is an assertion rather than evidence.
What does a ticked step prove that a signed SOP cannot?#
A paper SOP ticked with a pen says somebody followed the procedure. A step list says the third step happened at 09:22, by name, four minutes after the second. The first is a claim about the whole procedure made once, at the end. The second is a series of small facts recorded as they occurred, and each one can be checked against the others.
Consider what that makes possible. If a method requires a fifteen-minute incubation between steps three and four, the record shows whether fifteen minutes elapsed. If a step must be performed by a second trained analyst, the record shows whose name is on it. If the analyst went back and re-did step two after step five, the record shows that too, because the sequence is part of what was captured. None of those questions can be asked of a signature.
The difference matters most when something goes wrong. An out-of-specification result under a signed worksheet opens an investigation that starts with interviews, because the worksheet cannot say what happened. The same result under an execution record starts with the record, which either shows a deviation in the run or rules the run out and points the investigation at the sample, the instrument or the method. That is hours against days, and it is the case for the category in one paragraph.
Why must the method be locked while it runs?#
An execution record is only evidence if the procedure it records against was fixed before the run started. If an analyst can add, remove or reorder steps while executing, the record shows a procedure that was followed perfectly, because the procedure was edited to match what was done. The record is complete, signed and worthless.
So an LES locks the method. Steps in a locked method can be ticked but not edited, reordered or removed, so a validated method is the method that actually ran. Changing the method is a separate act, with its own approval and its own version, and a run records which version it was executed against. When the method changes, old runs keep pointing at the version they used.
This is the property that separates an execution system from a checklist feature bolted onto a notebook. A checklist the analyst can edit is a to-do list. A checklist the analyst can only tick, against a version the quality unit approved, is an execution record. Ask a vendor which one they mean. The words on the screen look the same; the evidential weight does not.
The lock also settles a question that otherwise consumes review time: was the right version of the method used? Under paper, that is checked by comparing the worksheet header against the document register. Under a locked execution record, it is a property of the record itself. The run could not have used any other version, because no other version was offered.
Where do the calculations and the forms belong?#
Most methods end in arithmetic. A titration produces a volume; the result is a concentration. A tensile test produces a force; the result is a strength and a pass or fail against a specification. Under paper and spreadsheet, that arithmetic happens somewhere other than the record: the analyst reads a number off the worksheet, types it into a spreadsheet, reads the answer back and writes it down. Three transcriptions, each a chance to differ, and the spreadsheet is usually nobody’s controlled document.
The calculation belongs in the record. When the formula lives with the step that produced the inputs, the result is judged at the bench, with its units, against the specification, and the pass or fail is recorded as part of the run rather than carried to another system. The same is true of the forms a regulator requires. A certificate or a batch record produced from the execution record cannot disagree with it, because it was not retyped.
None of that requires the analyst to see a formula. The method’s author defines it once, on the approved version, so every run computes the same way. That is what “the step, the calculation and the form belong in the same record” means in practice.
- Each step records who completed it and when, and the sequence is kept
- The method cannot be edited, reordered or shortened while a run is open
- Every run records which approved version of the method it executed
- Calculations are defined on the method and performed in the record, with units
- Pass or fail against the specification is recorded in the run, not carried elsewhere
- Required forms are produced from the record and cannot be edited afterwards
- A deviation during the run is recorded in the run, with a reason, at the time
When is a notebook enough, and when is it not?#
A notebook is enough for as long as the question anyone asks is what happened. Research groups, method development, exploratory work: the write-up is the product, the procedure changes with every run because that is the point, and an execution record would document a method that did not exist yet. Locking it would be absurd.
A notebook stops being enough the first time someone asks how a result was produced and expects the answer to come from a record rather than a person. That moment arrives with a regulated method, a customer audit, a contract that specifies a procedure, or an out-of-specification result that has to be investigated to a conclusion. From then on, every run of that method is a run that might be questioned, and a write-up cannot answer the question.
The practical test is to pick one method your lab runs every week and ask when its third step happened last Tuesday. If the answer is a signature, the method has outgrown the notebook. That does not mean everything the lab does has: the same lab can keep a notebook for development and run its released methods through execution. The line between LIMS, LES and notebook is the unit of work, not the feature list. The sample, the run and the write-up are three different things, and a lab that knows which one a given record is has already made most of the buying decision.
What execution carries in this work#
LabLynx has built execution into its automation suite for the labs that needed a batch record to stand up, and electronic signatures that bind a name to a step rather than to a page. The validation work that goes with it is the part most labs underestimate, because a locked method is only as good as the approval that locked it.
Frequently Asked Questions #
Is a LES the same thing as an ELN?
No. An electronic lab notebook is organised around the write-up: what was done and what it means, in whatever structure the scientist chooses. A laboratory execution system is organised around one run of an approved method, recording each step as it is completed, by whom and when, against a procedure that cannot change during the run. The records answer different questions.
Do I need a LES if I already have a LIMS?
Only if someone can ask how a result was produced and expect the record to answer. A LIMS tracks the sample and stores the result; it does not record the sequence and timing of the steps that produced it. If your methods are regulated, audited or investigated when they go wrong, an execution record fills that gap. If nobody asks how, the LIMS is enough.
What does a laboratory execution system record?
Each step of a method as it is completed: who performed it, when, in what sequence, and any values entered along the way. It records which approved version of the method the run executed against, any deviation raised during the run with its reason, and the calculated results with their pass or fail against the specification. The output is a record of one run.
Can a LES work without a full LIMS?
Yes. Execution is about the run, not the sample, so a lab can record method execution against records it already keeps in a notebook or a simple sample register. Many smaller regulated labs start there. What matters is that the method is locked while it runs and each step is attributed and timed. The sample tracking around it is a separate decision.
Which regulations expect execution records?
Any framework that requires written procedures to be followed and deviations recorded implies evidence of execution. In the United States, the current good manufacturing practice rules for finished pharmaceuticals require written procedures and a record of any deviation, and FDA's data integrity guidance expects records to be contemporaneous and attributable. Accreditation standards for testing laboratories ask for the same in different words.
Sources and references
- 21 CFR 211.100 Written procedures; deviations Electronic Code of Federal Regulations
- Data Integrity and Compliance With Drug CGMP: Questions and Answers U.S. Food and Drug Administration
