Application note · Pharmaceutical QC
A documentation-focused checklist for connecting sample preparation, measurement, review and report export in pharmaceutical QC — where the record, not the instrument, usually fails.
The short answer
Pharmaceutical quality control does not usually fail because a measurement is wrong. It fails because the link between the method, the instrument, the raw data and the calculation cannot be demonstrated. The instrument is often the strongest part of that chain; the record is the weakest.
Four links, each reconstructable without asking the analyst: the method that was used and its version, the instrument that produced the data and its state at the time, the raw data as acquired, and the calculation that turned raw data into a reported result.
| Link | What it must show | Where it is lost |
|---|---|---|
| Method | Version, parameters, acceptance criteria | Settings saved on the instrument rather than in a controlled document |
| Instrument | Identity, qualification status on the day | Qualification records kept apart from the result |
| Raw data | As acquired, unedited, with time stamps | Export to a spreadsheet that is not version controlled |
| Calculation | Formula, inputs, reviewer signature | Re-typed values, calculation outside the record |
Three breaks account for most findings. Weights are transcribed from the balance display into a worksheet. Raw data is exported and then edited in a spreadsheet that nobody controls. Methods live in the analyst’s saved settings, so the method actually used is the one on the instrument, not the one in the document.
Software can support the chain but cannot invent it. Decide first what the record has to contain and who signs each part; then choose the settings that produce it. A laboratory that starts from the software’s feature list usually ends up with a record that satisfies the software and not the reviewer.
The instrument contributes four things to the record: an identifier, a timestamp, the raw signal, and the parameters in force at the time. Everything else — the method, the calculation, the review — belongs to the workflow around it. Keeping that division in mind stops a laboratory from buying software features it does not need and from overlooking the ones it does.
It also explains why two laboratories with the same instrument can hold very different records. The instrument offers the same four contributions to both; the workflow decides how many of them survive.
For each batch: method reference and version, instrument identifier and its qualification status on the day, the raw data file, the calculation sheet, the review signature, and any deviation record. Six items, one folder, one naming rule. Everything else is optional.
Pick a result from three months ago and ask a colleague who did not run it to reconstruct it from the record alone. If they need to ask you a question, the chain has a break — and the break is cheaper to fix now than during an inspection.
The configuration described above, with the full specification of each model.
K Series / Double beam
A double-beam member of the K Series, with touchscreen operation for laboratory UV-Vis measurement workflows.
3 models
Functions include Photometric measurement, Quantitative measurement, Kinetics, Spectrum scan, DNA/Protein test, multi-wavelength test, etc
11 models
Supports PC connection.
The requirement is traceability, not medium. A paper record that carries the raw values, the method version and two signatures is traceable; a spreadsheet that anyone can overwrite is not, however modern it looks.
Transcription between instruments and worksheets, uncontrolled spreadsheets, and shared logins that make the analyst unidentifiable. All three are workflow design problems rather than instrument problems.
Enough that a competent analyst who has never run it could repeat the measurement: instrument settings, cell type, wavelength, calibration approach, calculation and acceptance criteria. Screenshots of a menu are not a method.
Application note
The specification points that decide whether a UV-Vis spectrophotometer is fit for assay work, and the documentation that…
Application note
The questions that separate two UV-Vis configurations for assay, identity and routine release work — wavelength, photometry, stray…
Application note
How balance performance, moisture method and sample identity fit into a defensible pharmaceutical QC chain — and where…
Let’s talk about your application
Send the sample, the concentration range and the reference method. We will confirm the instrument and the accessories.