Spike recovery

Spike recovery checks a fortified sample against a calibration curve you already have. The spike is a control. It does not build the curve, and it does not estimate the native amount from a line through unspiked and spiked aliquots.

Use it when calibration samples define the curve, and a Sample (or QC) is fortified with a known amount so you can see how much of that amount comes back. SpikeGroup pairs the unspiked vial with the fortified vial(s) so recovery can subtract the amount that was already there.

Spike recovery vs standard addition

Standard addition does not read SpikeGroup. See Standard addition.

What to set

1. Leave the Cal group Mode on Calibration. Put the calibration samples and the fortified sample in that same Cal group. Samples with an empty Cal group share the default group. The Cal group is the curve. It does not pair the spike.
2. Set the calibration samples to type Calibration and give each a Dose that is a point on the curve. Calibration and Blank are not part of spike pairing.
3. Set the fortified vial to type Sample or QC and give it a Dose. The dose is the amount you added, not a new sample type. There is no Spike type.
4. Set the unspiked vial to type Sample or QC and leave Dose empty.
5. Type the same SpikeGroup on the unspiked vial and on each fortified vial that belongs with it. The column is SpikeGroup, next to Cal group, on the Samples grid and on Add Samples. Empty means this sample is not in a spike group. A sample has at most one SpikeGroup. The name is a method catalog, edited under View → Spike groups, with a display colour. The Samples cell and the Add Samples cell use that colour. A new name takes the next palette colour, the same way a new matrix or cal group does. Change the colour on the spike-group row, or in Settings → Colors… under Method — spike group. Matching ignores case and surrounding spaces. Copy, paste, and fill work the same way as Cal group. Add Samples import mapping and name patterns can set SpikeGroup wherever they can read and write Cal group. Reports use Sample.SpikeGroup.
6. Enter level names under View → Doses. Enter each analyte’s concentration at those names under View → Analyte doses. Turn Use off on a level to keep the stored concentration out of calibration and out of the dose lookup. A level with Use off is not a dose, so that vial is unspiked.
7. Set Dil on the sample. The on-column dose is the analyte dose concentration divided by Dil. A dilution of 0 or below is treated as 1.
8. The reserved dose name IS is the spike for ISTD and Surrogate analytes. On a target, IS is not used as that target’s dose.
9. Choose the method calibration mode in Batch properties: Initial, Partial, Bracket, or Bracket with borrow. That chooses which calibration injections form the curve. It does not change the recovery formula. See Bracketing.

Quantitate reads the curve, then standard addition (for groups in that mode), then spike recovery, then blank-concentration correction.

How the numbers are calculated

Conc is the concentration read from the bracket curve (external standard, or the ISTD ratio when this quant uses an ISTD). It stays the on-curve value. SpikeGroup never changes Conc. A missing curve leaves it unset.

Dose on the result is the on-column amount: the matching Analyte doses concentration divided by the sample Dil. It is empty when the sample has no dose, the dose is 0, the level’s Use is off, or the name is IS on a target.

Inside one SpikeGroup, and separately for each quant:

• An unspiked vial is a Sample or QC with no on-column dose.
• A spiked vial is a Sample or QC with an on-column dose.
• The unspiked reference is the mean Conc of the used unspiked results. A rejected sample, a rejected result, or a rejected peak candidate is left out. The result that is not the selected quant is left out. The result has to have a Conc value. Zero counts.
• A different SpikeGroup is a different reference. A different quant is a different reference.

NConc for a spiked member is Conc − on-column dose, kept at or above 0. That includes a QC that has a dose and a SpikeGroup. FConc follows from NConc as usual. Unspiked members are not rewritten by spike recovery (blank correction can still adjust them).

RecPct for each spiked result is 100 × (Conc − unspiked reference) / on-column dose. It uses the curve concentration, not NConc. It is empty when the dose is missing or 0. Unspiked results get no RecPct. DevPct = RecPct − 100.

If the SpikeGroup has spiked vials and no usable unspiked vial, RecPct is left empty on those spiked results and Warnings gains: “Spike recovery needs an unspiked sample in the same SpikeGroup.” NConc still subtracts the dose.

Samples with no SpikeGroup keep the previous recovery: RecPct = 100 × Conc / on-column dose for an ISTD or Surrogate on any sample, and for a target that is QC, Calibration, a Sample with a dose, or a Solvent with a dose. A QC outside a SpikeGroup does not subtract the dose from NConc.

When Blank corr is on and the quant origin is not Blank offset, Sample and QC subtract the blank concentration toward NConc after the dose subtraction that applies. For a spiked SpikeGroup result, including QC, that is (Conc − blank) − dose. RecPct is calculated again from Conc and the unspiked mean Conc, so the blank is not part of the recovery percent. Blank, Calibration, and Solvent do not get that blank subtraction. See Quantitation.

CalcFConc = NConc × Dil × the limit group’s Final Conc multiplier (an empty multiplier is 1). A non-positive dilution is treated as 1. FConc is ManFConc when you have typed one, otherwise CalcFConc.

The recovery flag compares the stored RecPct with Min recovery and Max recovery on the sample’s limit group for that analyte name. After Analyze, that stored value is the SpikeGroup recovery when the sample is paired. Concentration flags for a normal Cal group use CalcFConc.

A Cal group whose mode is Standard addition is not spike-paired, even if those samples have a SpikeGroup.

How a spiked sample relates to an unspiked sample

They are paired only when they share a SpikeGroup. The same Cal group only means they share one calibration curve and one bracketing scope.

Each fortified vial uses its own curve concentration, the group’s unspiked reference for that quant, and its own on-column dose. An unspiked Sample or QC in that SpikeGroup is quantified on its own: with no dose, NConc equals its curve concentration (after blank correction when that applies), and RecPct stays empty.

Where the values show up

On the Samples grid, SpikeGroup is the column next to Cal group. The cell background is the colour of that spike group. View → Spike groups lists the names and colours.

On the result grid:

Report formulas read the same fields: Sample.SpikeGroup, Result.Concentration, Result.Dose (also Result.DosedConcentration), Result.NativeConcentration, Result.RecPct, Result.DevPct, Result.CalcFConc, Result.FConc, Result.BlankConcentration, Result.Warnings.

Example

These numbers are an illustration of the arithmetic above.

Cal group Matrix, Mode Calibration. Linear curve through the origin with slope 10 (response per concentration unit). Dilution 1 except where noted. Final Conc multiplier 1. No blank correction.

The unspiked reference for LotA is 3. The fortified Sample’s recovery is 100 × (25 − 3) / 20 = 110. Its NConc is 25 − 20 = 5. The QC in LotA uses the same reference: 100 × (23 − 3) / 20 = 100, and NConc is 23 − 20 = 3. The vial with no SpikeGroup still uses 100 × 25 / 20 = 125. The calibration vial is not part of the pairing.

If LotA has two used unspiked results, Conc 2 and Conc 4, the reference is 3 and the fortified recovery above is unchanged. A rejected unspiked result is not included in that mean.

With dilution 2 on the fortified Sample and the same response 250, the on-column dose is 20 / 2 = 10, Conc is still 25, the reference is still 3, NConc is 25 − 10 = 15, RecPct is 100 × (25 − 3) / 10 = 220, and CalcFConc is 15 × 2 = 30.

A SpikeGroup that has a fortified vial and no usable unspiked vial leaves RecPct empty, writes the warning above, and still sets NConc to Conc − dose.

Limitations

• A sample with no SpikeGroup does not subtract an unspiked aliquot. Its recovery stays 100 × Conc / dose.
• Calibration and Blank are not paired, even when they carry a SpikeGroup name. Solvent is not paired.
• QC outside a SpikeGroup does not subtract the dose from NConc. A QC with a dose inside a SpikeGroup does.
• NConc and CalcFConc are kept at or above 0. A curve reading below the dose becomes NConc 0, while RecPct still uses the curve reading and the unspiked reference.
• Spike recovery does not run inside a Standard addition Cal group.
• Turning a bracket point’s Use off changes the curve, and therefore changes every result read from that curve. Rejecting a sample, result, or peak candidate keeps that unspiked Conc out of the reference.
• The reserved name IS on a target does not produce a dose, a native subtraction, or a recovery.

The spike-recovery example batch is described in Example batches.