Key points
- Automatic integration with fixed, method-specific parameters should be the rule; manual integration the documented exception.
- Every reintegration needs a reason, the original and the final integration kept, and an audit trail showing who changed what and when.
- Standards, QCs and study samples are integrated the same way. Integration must never be chosen to make a run pass.
Why integration gets inspected
The peak area is the raw result of a chromatographic method - every concentration in the report is calculated from it. Moving a baseline by a few seconds can change a QC from 84% to 87% accuracy, and that is exactly the difference between a failed and an accepted run. This is why inspectors open the audit trail of the chromatography data system, filter it for reintegration, and ask the analyst to explain each entry. Integration practice is one of the most frequent data integrity observations in chromatography laboratories.
What ICH M10 expects
ICH M10 expects chromatogram integration and reintegration to be described in an SOP, and any deviation from that SOP to be discussed in the report. Integration parameters, and for any reintegration the initial and the final integration data, should be documented and available on request, together with the reason for the change. In a 21 CFR Part 11 environment the computerised system has to make this traceable: a secure, time-stamped audit trail that records the user, the old and new values and the reason, and that cannot be switched off or edited.
When reintegration is justified
Reintegration is a correction of how the software drew the peak, not of the result. Typical justified cases:
- The baseline was drawn to the wrong point - for example across a noise spike or a small neighbouring peak.
- The retention time drifted outside the expected window and the peak was missed or a different peak was picked.
- A partly resolved interference or a shoulder was included in the analyte peak (or split from it) in one injection only.
- At the LLOQ the peak was not detected because of noise, although it is clearly visible with the same pattern as the other low samples.
In each case the reviewer should be able to look at the original and the reintegrated chromatogram side by side and agree, without knowing the concentration, that the new integration describes the peak better.
What is not acceptable
- Peak shaving or peak enhancing - moving the baseline or the start and end points to reduce or increase the area.
- Selective reintegration - reintegrating only the failing QCs or standards, or only samples near a decision point, while similar peaks stay as they were.
- Different rules for different sample types - for example drop-line integration for QCs and valley-to-valley for study samples.
- Reprocessing until it passes - repeated changes of processing method or parameters after the result is known, without a documented scientific reason.
- Hiding the history - deleting, overwriting or not saving the original integration, or processing outside the audited system.
What the SOP should define
| Topic | What to write down |
|---|---|
| Default parameters | Integration parameters fixed during validation and stored in the processing method, version-controlled |
| Allowed manual actions | Which changes are permitted (baseline points, peak window) and which are not |
| Consistency | Same approach for calibration standards, QCs and study samples within a run |
| Reasons | A short list of predefined reasons plus a free-text comment |
| Timing | Integration finalised before the run is evaluated against acceptance criteria wherever possible |
| Approval | Every manual integration reviewed and approved by a second person before results are reported |
Reviewing integration
A good integration review does not start with the numbers. Open the overlay of all standards and QCs, check that the peaks are integrated the same way, then filter the audit trail for manual changes and check each one: reason, original versus final chromatogram, and whether the change was made before or after the run result was known. Count them per run and per analyst - a run with twenty reintegrations or an analyst who reintegrates far more often than colleagues is a signal to look at the method or at training. Keep this assessment with the run; it is the first thing an inspector will ask for.
Educational summary for chromatographic (LC-MS/MS) bioanalysis. Always apply the current ICH M10 text, the expectations of the regulator concerned and your laboratory's SOPs. Ligand-binding assays have wider limits in ICH M10.
ARISENTO Insights is published by Autofai Technologies Private Limited, the makers of ARISENTO bioanalytical software.