Assay Run with a Deleted Injection
A contract HPLC laboratory assays a 10 mg tablet for a small-molecule drug. The method uses a 150 × 4.6 mm, 3.5 µm C18 column, 1.0 mL/min, 30 °C, 254 nm, 10 µL injection, and a phosphate buffer pH 3.0/acetonitrile mobile phase. System suitability requires five replicate standard injections with RSD ≤2.0%, tailing ≤2.0, and resolution ≥2.0. The run report shows RSD 3.8%, tailing 1.9, resolution 2.1, and assay 98.5% (specification 95.0–105.0%). The audit trail shows that one of the five suitability injections was deleted after acquisition and that the main-peak integration baseline was manually adjusted at 12.4 min after the sequence ended. A blank injected after the highest calibrator contains a peak at the analyte retention time equal to 0.8% of the standard peak area. The analyst asks the reviewer to approve the run because the assay is within specification.
1. Identify the data-integrity and system-suitability failures and decide whether the assay result can be approved.
Read the worked answer
The run cannot be approved. System suitability fails on injection precision: RSD 3.8% exceeds the 2.0% limit. Tailing at 1.9 is at the edge of the limit, and carryover at 0.8% of the standard peak is above the usual 0.5% acceptance threshold. These are analytical failures independent of the assay value.
The data-integrity problems are separate and serious. Deleting one of the five suitability injections after acquisition removes data that may explain the RSD failure; deletion is only acceptable for a documented, scientifically justified invalid injection, with the original record retained and the reason recorded. Manually adjusting the main-peak integration baseline after the sequence ended, without a contemporaneous justification, changes the reported result and must be reviewed against the audit trail. The fact that 98.5% is within 95.0–105.0% does not cure either failure. The run should be invalidated, the original data preserved, and the assay repeated after a proper investigation.
2. What are the likely root causes of the RSD failure and the carryover, and what experiments should be run?
Read the worked answer
For the RSD failure, likely causes include autosampler precision (needle, seat, loop, or syringe), bubbles in the injection path, variable injection volume, column temperature fluctuation, mobile-phase degassing, detector lamp instability, and inconsistent integration. The tailing of 1.9 also points to column or extra-column effects: a void at the column inlet, a mismatched sample solvent, or mobile-phase pH outside the method target. The deleted injection prevents knowing whether one outlier drove the RSD; restore and review it.
For carryover, the blank after the highest calibrator indicates retention in the needle, injection port, or early column. Run a carryover study: inject blank, high standard, blank, and repeat with extended needle wash and different wash solvent. Check injection volume and sample concentration. For RSD, run ten replicate injections of the same standard, monitor peak area and retention time, check system pressure, replace the needle/seal if needed, and verify mobile-phase pH and column equilibration. Establish whether the failure is injector-related or column-related before repeating the assay.
3. Draft the CAPA and reporting decision, including what to do with the 98.5% result.
Read the worked answer
Do not report 98.5% as a valid assay result. Record the run as invalid, retain the original chromatograms and audit trail, and initiate a laboratory investigation. The reporting decision is to repeat the assay with a fully passing system suitability, including five valid injections, tailing and resolution within limits, and a carryover blank below the acceptance threshold.
The CAPA should address data integrity and instrumentation. Data-integrity actions: retrain analysts that injections may not be deleted without a documented invalid-injection rationale, require audit-trail review before approval, define integration-change authorization, and have QA review the deleted injection and manual baseline adjustment. Instrument actions: autosampler maintenance, needle/seal replacement, carryover wash optimization, column evaluation, and mobile-phase preparation checks. If the repeat run passes and the assay remains within 95.0–105.0%, report the new result and reference the invalidated run. If the repeat fails, escalate to method robustness or column lot investigation. Notify QA and evaluate whether other runs by the same analyst have similar audit-trail events.