Dilution verification

Run this study#

Your question: do diluted high samples recover their expected concentration after correction?

Bring: high specimens or pools, a documented expected undiluted concentration for each, the assay's validated measuring limits, and the intended diluent and factors. CLSI's public dilution guidance names these as the basis for checking an extended measuring interval.

Suggested starting plan: 2 high pools × 3 dilution factors × 2 separate preparations = 12 readings, six per pool. Choose pools across the high concentrations you intend to report, and factors that bring each reading inside the validated measuring interval. Adapt the count to your own dilution scheme.

1. Decide first. Set the acceptable recovery band (%), the expected neat concentration of each pool, the validated reportable range for measured (uncorrected) readings and, if you will report beyond it, the claimed reporting limits. Take the expected values from a source other than the readings being checked. Create the study and enter them on Set up (see Study setup).
2. Prepare and measure. Make each dilution with the intended diluent and measure it. Record the volumes or the decimal factor, and every step of a serial dilution.
3. Import your readings. Enter specimen ID, level, measured reading, preparation, diluent and replicate. Enter the reading before multiplying by the factor. On Data, choose Import a file and choose your file (start from Blank template (CSV) or dilution.csv). Answer any question the dialog asks, then choose Import into this study.
4. Calculate and review. Choose Calculate results; Results opens. Check each corrected value, recovery, whether it was measured in range, and its outcome. Investigate failing recoveries and readings outside the reportable range.
5. Print the report. On Report, choose Download PDF or Print.

What you get: recovery for each dilution and the verified range, the span of corrected results that passed. The verified range covers only the concentrations you tested, so test pools near every limit you intend to claim. Calculation minimum: two recoveries per level. Recovery needs a positive expected concentration on a ratio scale.

Purpose#

For each dilution, the reading × the dilution factor gives the corrected result, and the corrected result ÷ the expected neat concentration gives the recovery. The reading is also checked against your reportable range.

When to use it#

Use it when your laboratory dilutes high specimens and reports the corrected result. To check a straight-line response across prepared concentrations, use Linearity.

Study setup#

Choose New study. Enter Analyte and Unit, select Dilution verification under Measuring range, then enter Lowest reportable value before dilution, Highest reportable value before dilution and Expected value of the undiluted sample, which becomes the expected value for level 1. Choose Create study, or Create and import data if you already have results.

Each high specimen or pool is a level with its own expected neat concentration. Plan several dilutions of each, with readings inside your reportable range.

On Set up:

• Assigned values: one row per level with its expected neat concentration. When some levels in the data have no assigned value, Add assigned-value rows for N levels in the data adds the missing rows.
• Reportable range: Reportable range — lower limit and Reportable range — upper limit are the validated limits for readings before correction. Claimed lower limit and Claimed upper limit are optional: enter both or neither.
• Acceptance limits: Enter your own limit offers Dilution recovery with Lowest acceptable recovery (%) and Highest acceptable recovery (%). Leave one empty for a one-sided limit. Applies to limits it to one level. See Acceptance limits.

Entering data#

One row per dilution:

Give the factor one of three ways. If a row has more than one, they must agree.

• Volumes: (specimen + diluent) / specimen. 100 + 400 gives 5. 50 + 950 gives 20.
• Stated factor: used as written.
• Serial steps: multiplied. 5, 2 gives 10.

On Data, choose Import a file or Add row. See Importing a file.

A stated factor follows the study's number format: 1,000 is 1000 with a decimal point and 1 with a decimal comma. See Numbers, units and identifiers.

Correct or exclude a censored result such as >600. See Corrections and exclusions.

The calculation stops, with a message on the row or on Set up, when:

• a row has no volumes, stated factor or serial steps, only one of the two volumes, a factor below 1, two ways of giving the factor that disagree, or no level
• the reportable range is missing, or its lower limit is at or above the upper
• only one claimed limit is entered
• a level has no assigned value

Statistics#

• Corrected result = measured × overall factor.
• Recovery % = 100 × corrected / expected neat concentration. It needs a positive expected value and a ratio scale. Each recovery is a point estimate.
• Verified range = lowest to highest corrected value among at least 2 dilutions that were read inside the reportable range (limits included) and met every recovery limit. Without a recovery limit there is no verified range.
• Evaluated range = all corrected values with readings inside the reportable range, failed recoveries included. The report shows it as Corrected values evaluated when it differs from the verified range.
• Claimed range: a claimed limit beyond the verified range gets a warning that it was not evaluated.

How acceptance is decided.

• Each dilution gets its own outcome. One failing dilution fails the study.
• A dilution read outside the reportable range is Undecided. Its recovery is still shown.
• A level with fewer than 2 recoveries is Undecided.
• A limit for one level is Not applicable to the others.
• On an interval scale (Measurement scale under Calculation settings), recovery limits are Undecided and no range is verified.

Worked example#

The example project's study Glucose · dilution verification uses dilution.csv: pools HP-1 (level 1, three dilutions) and HP-2 (level 2, four dilutions), with factors from volumes, a stated factor and serial steps 5, 2.

Setup: analyte Glucose, unit mg/dL. Reportable range 2 to 600 mg/dL. Claimed range 2 to 3000 mg/dL. Expected neat 1500 mg/dL (level 1) and 2800 mg/dL (level 2). Limit: Dilution recovery, 90 to 110 %.

What the saved result shows#

Level 1 (expected neat 1500 mg/dL):

Level 2 (expected neat 2800 mg/dL):

Measured and corrected values are in mg/dL. Every reading is inside the reportable range, so Measured in range is Yes and every Outcome is Met.

The study status is Criteria met: all seven dilutions meet the limit. Each level carries a warning that the claimed lower and upper limits were not evaluated. For level 1, the lowest usable corrected result is 1486 mg/dL, far above the claimed 2 mg/dL.

If one dilution reads 750 mg/dL, above the 600 mg/dL limit, its recovery is still shown but its outcome is Undecided, and it is left out of the verified range. A level with only one usable dilution gets no verified range.

Reading the results#

Concentration level switches between specimens. Each level shows:

• Dilutions and recovery · level …, with the expected neat concentration beside the heading: specimen and row number, measured value, preparation (volumes, serial steps or stated factor), factor, corrected value, recovery and Measured in range (Yes or No). When every limit is a per-dilution recovery limit, this table also has each dilution's Outcome. Otherwise each limit's outcome is under Why.
• Reportable range limits: reportable, claimed and Verified range.

Recovery that falls at higher dilutions points to a matrix or diluent effect, even when every value passes.

When the study fails:

1. Recheck data entry. Enter the measured reading, before correction. Check that the volumes, factor or steps match what you pipetted, and check the decimal convention.
2. Check the expected value. If every dilution at one level is off in the same direction, recheck that level's expected neat value.
3. Retest. Repeat any dilution that read outside the reportable range at a factor that brings it inside.