thebench.bio #1 Ecosystem for the Bio-Tech Industry
All Quality Control questions

Question 11 of 20 Intermediate Troubleshooting

How would you distinguish instrument drift from sample instability in a QC assay?

Short answer

Look at what the standards do. If bracketing standards drift in the same direction as the samples, the instrument is moving. If standards stay steady while samples fall over time, the sample is changing. Reinjecting retained solutions and a fresh preparation confirms which.

Why

Both problems show the same symptom, results that change through a sequence, so the evidence has to come from controls that respond to one cause and not the other. Bracketing standards are the main tool. The FDA's OOS guidance notes that reference standard solutions injected through a chromatographic run measure drift, noise and repeatability, and that data from a period where they show the system is not working should not be used.

A detector lamp losing energy or a slowly changing column temperature moves the standards and the samples together. A sample that degrades in the autosampler, adsorbs to vials or evaporates moves on its own while standards hold steady. A useful test is to reinject the first sample at the end of the sequence.

If the late injection of that same vial is lower and standards are flat, the solution is unstable. Preparing a fresh sample and injecting it at once separates solution instability from a problem in the solid sample. Q2(R2) says robustness work should cover stability of sample preparations for the duration of the procedure, so check the validated solution stability time and whether the sequence exceeded it.

Confounders matter. A cooled autosampler that failed overnight can cause both effects at once, and standards dissolved in a different diluent may not share the sample's instability.

How to work it through

  1. Plot standard and sample responses against injection time on the same chart.
  2. Check whether the standard responses meet their drift criteria.
  3. Reinject an early sample vial at the end and compare with its first result.
  4. Prepare and inject a fresh sample solution immediately.
  5. Check autosampler temperature logs and the validated solution stability time.
  6. Conclude on the evidence and invalidate only data from a proven suspect period.

An example

An overnight sequence of 40 injections shows sample potency falling from 100.4 to 97.1 percent. Bracketing standards stay within 0.4 percent of each other. Reinjecting sample 1 at the end gives 97.0 percent, 3.4 points below its first result.

The autosampler log shows 25 C instead of the 5 C setting from midnight. Validation had shown solution stability of 24 hours at 5 C only. The cause is sample instability from a failed cooler, the late results are not reportable, and the cooler failure becomes a deviation.

Where the answer stops holding

Standards and samples behave differently if they differ in matrix, so stable standards do not prove a stable instrument response for the sample. Some drifts are too small to separate from normal variability without more replicates. If retained solutions were discarded, the cause may stay unproven.

A strong answer covers

  • Bracketing standards as the discriminating control
  • Reinject early sample vials at the end
  • Fresh preparation versus aged solution
  • Validated solution stability time
  • Check autosampler temperature and confounders

Common mistakes

  • Assuming the instrument drifted without checking whether the standards moved.
  • Discarding solutions before the cause is established.
  • Ignoring that standards in a different diluent may not share the sample's instability.