Checks

Checks run when you Quantitate. A failure is a flag on the result, collected in the Outliers column, and it can color the result in the grids and in FastPositive review tiles. Flags use three levels for most limits: Too low, OK, and Too high. Concentration uses a wider set that includes in range and above the top of the curve.

Empty limits do not fail. A check runs only when the limit is filled in.

Where the numbers live:

• Analyte-level limits (concentration too low / too high, recovery window, final-concentration multiplier) are on Limit groups for the sample’s group and the analyte name. See Limit groups and Continuing calibration verification.
• Quant-level limits (response, qualifier counts, minimum Q, minimum R², unexpected peak height) are on the same limit row, optionally for one quant key.
• Qualifier tolerances are on the qualifier and on Method → Qualifier tolerances…. See Qualifiers.

ISTD checks

An ISTD analyte uses the recovery window as an ISTD response window. The minimum recovery is also the allowed deviation of the ISTD response from 100 percent (the window is stored as 100 + deviation).

If the ISTD response percent falls outside that window, the recovery flag is Too low or Too high. The same flag lights the ISTD tile in review.

An ISTD that is assigned but not ready blocks the target concentration (the target is not read from its own curve).

Surrogate checks

A surrogate is calculated like a target that has a known spike. Recovery percent is 100 × calculated concentration / dose. The limit group’s minimum and maximum recovery flag Too low or Too high. With no recovery limits, a recovery that is a valid number is not flagged.

Calibration checks

Calibration points are not QC. They are included in the fit.

• Recovery of a calibration point uses the same recovery window. A point outside it sets the recovery flag. Review treats a calibration analyte as failed when either the recovery flag or the R² flag is not OK.
• Minimum R² on the limit group sets the R² flag to Too low when this sample’s bracket R² is missing or below the cutoff.
• Relative residuals are stored on the curve points for you to read. They are not a separate pass/fail flag unless you put a limit on R² or recovery.

Failed calibration points stay on the curve. Bracketing does not drop a standard because its check failed. See Bracketing.

CCV checks

A continuing calibration verification vial is a known concentration injected through the sequence. It is quantitated on its Cal group curve. RecPct is 100 × Conc / dose and DevPct is RecPct − 100. RecFlag is too low or too high against the limit group’s recovery window. CcvFlag marks Sample and QC results that follow a failed CCV, or that sit in a stretch longer than the group’s maximum injections between CCVs. See Continuing calibration verification.

QC checks

QC samples are calculated and kept off the fit. The check is recovery against the limit-group window (recovery flag). Review treats QC as passed when every QC result’s recovery flag is OK. If the batch has no QC sample, that roll-up passes.

R² is not the QC check. A QC result can still show an R² flag when a minimum R² is set, because the flag is about the curve used for that sample.

Sample checks

For Sample (unknown) and Solvent rows, the usual flags are:

• Concentration too low, in range, too high, or above the highest calibration level.
• Recovery, when the sample has a dose (a spiked unknown). The flag reads the stored RecPct. With a SpikeGroup that is 100 × (Conc − unspiked reference) / dose. With no SpikeGroup, on a normal Cal group, it is 100 × curve concentration / dose. On a Standard addition group it is 100 × (Conc − NConc) / dose. See Spike recovery and Standard addition.
• Response below the minimum or above the maximum set on the quant’s limit.
• Unexpected peak height when a point outside the accepted peak is higher than the absolute limit, or higher than the highest calibration response when that rule is in use.
• Q value too low when the qualifier match is below the minimum Q on the limit. This is separate from the Q confirmation threshold, which decides identity.
• Passed and failed qualifier counts against the minimum passed and maximum failed counts.
• Identity. The result fails identity when qualifier ratios miss the tolerance and the Q value is below the quant’s confirmation threshold. Identity failure is written into Outliers even when the count limits are empty. A result that fails identity is not auto-selected for the sample-key report.
• Retention time outside the analyte time window is shown with the ion-ratio failure on the review tile.

Sample type Blank is used as a blank (including blank offset on a curve). It is not given the unknown’s on-curve recovery path.

How a failure is shown

The result grid column Outliers lists the triggered names, for example a recovery flag that is too high, or an identity failure with the qualifier match. Review tiles use color for too low, OK, and too high. Reports can read the same fields (Result flag fields and Outliers) and can hide rows with the show/hide formulas. See Report templates.