Validation & Qualification
Basic requirements as design drivers
Instrument and data system requirements often appear straightforward (throughput, reproducibility, accuracy, MTBF) but become complex when translated into design constraints. Examples include:
- High throughput (e.g., 1000 samples/day) and 24/7 operation
- Retention time reproducibility (e.g., CV 0.1%)
- Compatibility with fast separations (e.g., ~1 minute or less)
- Mass accuracy requirements (e.g., a few ppm at a specified m/z)
These requirements imply constraints on sampling rates, timing, calibration, and the full chain of hardware/software components affecting measured values.
System Suitability Test (SST)
To verify critical design parameters, the instrument must be confirmed to be operating appropriately. For example, if an autosampler is malfunctioning, retention time reproducibility and peak area accuracy can be misleading. An SST should verify each critical quality item.
- Verify injection behavior (e.g., 1 µL vs 2 µL) using a validated detector method
- Verify mass chromatogram peak area/height linearity
- Verify absolute mass stability
- Verify chromatographic parameters (retention time, plates, tailing, etc.)
Design Qualification (DQ)
DQ develops design from requirements by listing critical items and verifying that each item is met. Examples include alignment of business plan and quality plan, sample throughput capacity, and precision of data relative to instrument specifications.
Algorithm qualification (mass accuracy and retention time)
Before discussing instrument accuracy, computational methods should be qualified using mathematically generated signals (with and without simulated noise). For example, centroid accuracy depends on sampling topology around peak apex; retention time accuracy depends on both algorithm validation and absolute timing.
In fast separations, timing details (e.g., injection signal handling and sampling intervals) can become critical to meeting stringent reproducibility targets.
Verification and validation of results (QC monitoring)
Verification can be performed using QC samples inserted into acquisition sequences. Long-term monitoring of QC variance provides context for interpreting differences between control and test samples.