How is a recovery test designed and calculated when evaluating accuracy during IVD immunoassay development? Guide

At its core, a recovery test is a spike-and-measure experiment that answers a critical question: does my immunoassay give back what I put in?
To evaluate accuracy when a reference standard isn’t available, you add a known, highly concentrated amount of the pure target analyte into patient samples and then measure how much of that spike the assay actually detects. The design requires spiking into multiple base samples with different starting concentrations, using multiple spike levels, and running a parallel buffer-only control spike to isolate true recovery from matrix or dilution artifacts. The percentage recovery is calculated using the formula:

% Recovery = [(Assayed Concentration – Endogenous Concentration) / Concentration of Added Analyte] × 100

A well‑executed recovery test does more than confirm accuracy—it isolates proportional systematic bias. The real goal is to reveal whether the assay’s calibration slope is shifted relative to native clinical specimens. When the measured increase in concentration matches the spike, you have confidence the assay performs equivalently on calibrators and real patient samples. If not, you’ve uncovered a matrix‑driven scale error that must be corrected before the assay can be trusted.

The Purpose of Recovery Testing in Immunoassay Accuracy

Beyond Simple Accuracy – Detecting Proportional Bias

Recovery testing specifically probes for proportional systematic error.
That’s the kind of error that grows larger as the analyte concentration increases—mathematically equivalent to a slope deviation in method‑comparison studies.
A constant bias (intercept error) is a separate problem and must be evaluated with interference or blank‑matrix experiments.

Why a Properly Designed Recovery Test Matters

Immunoassays rely on antibodies and reagents that can behave differently in calibrator matrices versus real patient samples.
Without recovery data, you risk releasing a test that is precise but systematically inaccurate at clinically relevant concentrations.
A robust design ensures you detect subtle matrix effects early, saving time and preventing costly rework during clinical validation.

Designing a Rigorous Recovery Experiment

Choosing the Right Base Samples

Select at least two to three individual patient samples or unpooled matrices with endogenous analyte concentrations that span the clinical range.
Do not rely on overly treated or stripped pooled serum—these homogenized materials can mask true sample‑to‑sample matrix variability.
Using unaltered biological specimens reveals the kind of variation that real diagnostic samples will introduce.

Preparing and Adding the Spike

Prepare a highly concentrated pure analyte solution so the spiked volume remains small.
Aim for a final sample dilution no greater than 1:10 to avoid disrupting the matrix environment.
The actual concentration you add is calculated by accounting for the dilution:

Concentration Added = (Spike‑Solution Concentration × Volume of Spike) / (Volume of Sample + Volume of Spike)

This corrected value is your “Predicted Increase.”

The Critical Role of Control Spikes

Always run a parallel control where you spike matrix buffer without analyte into the same base samples.
This control accounts for any non‑specific dilutional effects or matrix shifts unrelated to the analyte.
Subtracting the control’s apparent increase from your spiked‑sample increase isolates the true analytic contribution.

Spiking Levels and Replicates

For reliable statistics, spike each base sample at two to three analyte levels that evenly bracket the clinical decision points.
Perform multiple replicates (at least three) at each spike level to average out volumetric and measurement errors.
This setup yields enough data points to assess whether recovery is consistent across the analytical range.

The Calculation – From Raw Data to Percent Recovery

The Core Recovery Formula

Recovery is never simply measured concentration divided by expected total concentration.
That approach confounds the endogenous baseline with the spike. The correct formula is:

% Recovery = [(Concentration of Spiked Sample – Endogenous Baseline Concentration) / Concentration of Added Analyte] × 100

This pairs the increase measured by the assay with the true amount you introduced, giving a clean view of proportional recovery.

Handling Dilutional Effects

If the volume of your spiking solution alters the sample matrix significantly, you must apply a dilutional correction before calculating recovery.
The most robust method is to directly compute the added concentration as shown above.
When the endogenous sample is diluted by the spike, don’t forget to multiply the baseline reading by the appropriate dilution factor so that both terms refer to the same matrix volume basis.

Interpreting the Numbers (Acceptable Range)

An acceptable percent recovery typically falls between 95% and 105%.
Values consistently above or below this window indicate a proportional bias—the assay is over‑ or under‑recovering the analyte.
Plotting observed concentration against added concentration can reveal trends; a regression slope that deviates from 1.00 signals a calibration‑to‑matrix mismatch.

Understanding the Trade-offs and Common Pitfalls

Pooled vs. Individual Matrices

Using pooled, stripped, or processed serum simplifies logistics but hides sample‑specific matrix effects.
Individual patient samples expose the assay to real‑world interferences, making your validation truly representative.
The trade‑off is increased variability, which demands more replicates and careful statistical analysis.

Spike Volume and Matrix Integrity

Keeping spike volumes small (≤10% of the total sample volume) preserves the native matrix environment.
Larger dilutions can alter protein content and binding kinetics, creating an artificial recovery artifact.
If your spike solution forces a bigger dilution, you must diligently apply dilutional corrections—but even then, the matrix may no longer behave as in undiluted clinical specimens.

Confusing Proportional and Constant Bias

A recovery test deliberately investigates the slope of your calibration relationship.
It cannot detect a constant offset (intercept error).
Developers sometimes misinterpret poor recovery as a simple offset and try to adjust it by re‑assigning calibrator values, when the root cause is a matrix‑driven scale problem requiring buffer or antibody‑pair reformulation.

Making the Right Choice for Your Goal

Tailor your recovery study design to the decision you need to make.

• If your primary focus is verifying calibration accuracy across the clinical range: Use at least three base samples spanning the low, mid, and high range, each spiked at three levels, and demand recovery between 95% and 105% for every combination.
• If you need to detect subtle matrix interferences from individual patients: Design the study with unpooled, native patient samples (not stripped or pooled materials), and evaluate recovery for each sample individually rather than averaging across them.
• If you are optimizing raw material or reagent lots during development: Perform recovery testing with the specific lots you plan to scale up, and include a buffer‑only control spike to confirm that any observed bias is truly analytic and not a formulation artifact.
• If you are troubleshooting a failed accuracy study: Separate proportional bias from constant bias by running a parallel interference experiment with potential interferents; use recovery data to guide matrix‑formulation adjustments, not calibrator reassignment.

When you make recovery testing a disciplined habit—with rigorous controls and un‑pooled clinical matrices—you transform it from a simple check‑box exercise into one of the most powerful tools for ensuring your immunoassay will perform accurately on real patient samples.

Summary Table:

Accelerate Your Immunoassay Development with CamelBio

Optimizing assay accuracy and overcoming matrix interference requires high-performance reagents and expert experimental design. CamelBio provides diagnostic manufacturers, labs, and research institutes with one-stop access to premium IVD raw materials, technical services, and consulting—covering every stage from concept to clinic.

Need guidance on assay optimization, raw material selection, or accuracy validation? Contact us today to learn how our tailored solutions can ensure your diagnostic assay delivers reliable, clinical-grade accuracy.