Reliable ELISA results are determined by three core criteria: replicate precision, valid standard curve, and compliant control wells. These are the prerequisites for usable experimental data. Details are as follows:
- Acceptable replicate precision: assessed by coefficient of variation (CV). Formula: CV% = (Standard Deviation SD / Replicate Mean) × 100%. Normally CV < 10% indicates good repeatability. High CV may be caused by inaccurate pipetting, uneven washing, cross‑contamination or inconsistent sample preparation. Optimize operations and retest if needed.
- Valid standard curve: for quantitative assays, R² ≥ 0.99 is required. The curve should have no distortion or obvious outlier points, and the fitting model must match the assay type to ensure accurate concentration calculation.
- Qualified control wells: blank‑well OD should not be abnormally high; positive and negative controls should behave as expected without systematic false‑positive or false‑negative bias, confirming the whole reaction system works properly.
- Blank well: only substrate and stop solution are added; no sample, standard or antibody reagents. Used to subtract instrument background OD, calibrate baseline and eliminate absorbance interference from microplates or reagents.
- Negative control: matrix sample without target antigen (e.g. blank serum or buffer). Defines the negative baseline and excludes non‑specific matrix interference; critical for calculating Cutoff in qualitative assays.
- Positive control: sample with known concentration of target antigen. Validates assay performance, confirms proper reagents, incubation and color development, and prevents false‑negative outcomes.
Qualitative ELISA: only reports positive or negative results. No standard‑curve fitting is needed. Results are judged against the Cutoff value. Common for antibody screening and pathogen detection.
Quantitative ELISA: requires a standard curve to calculate exact analyte concentrations from OD readings. Used for measuring proteins, cytokines and other substances. It demands stricter standard‑curve quality; final reports should include measured values and dilution factors.
Linear fitting cannot be applied universally. Indirect and sandwich ELISA typically produce S‑shaped dose‑response curves; four‑parameter logistic (4PL) non‑linear fitting is preferred for minimal error. Linear fitting may only be used for narrow concentration ranges. Never force‑fit a bent curve linearly, as this will cause severe bias for high‑ and low‑concentration samples.
For quantitative ELISA, R² ≥ 0.99 is acceptable. R² < 0.99 indicates poor fitting and invalid data. Typical causes include standard dilution errors, degraded standards, inconsistent pipetting, unstable incubation time/temperature and insufficient washing. Remedies: re‑prepare standards, standardize pipetting, strictly control assay conditions, refit or repeat the standard curve.
High‑concentration outliers: hook effect (prozone effect from antigen excess), over‑calibrated standards, over‑developed color. Low‑concentration outliers: high background, non‑specific binding, uncalibrated instrument baseline. Curve distortion: repeated freeze‑thaw of standards, mismatched reagent batches, large incubation temperature fluctuation. Remove isolated outliers (do not arbitrarily discard many points); optimize assay conditions and rerun.
Extrapolation is strictly prohibited. The top standard defines the upper limit of quantitation; beyond this point curve relationships break down and calculated concentrations are unreliable. Correct workflow: serially dilute sample with matched diluent (e.g. 1:10, 1:100) so that diluted OD falls within the valid curve range, re‑measure, and multiply final result by dilution factor.
The analyte concentration is below the lower limit of quantitation (LLOQ). Do not plug values directly into the fitting formula, to avoid false‑positive or artificially low readings. Possible causes: intrinsically low analyte level, sample degradation from freeze‑thaw, excessive dilution, matrix incompatibility with the kit.
Data must be discarded and samples retested. Check for imprecise pipetting, bubbles during dispensing, incomplete mixing before loading, edge‑well evaporation, uneven temperature or insufficient washing. Use multichannel pipettes; avoid touching well walls to minimize manual error.
Elevated blank OD raises assay baseline and causes falsely inflated sample values. Sources: contaminated wash solution, residual wash liquid, spoiled substrate, over‑long incubation, high microplate background. Solutions: use fresh wash and substrate solutions, optimize wash cycles, control incubation duration strictly, replace defective microplates.
OD > Cutoff: positive; OD < Cutoff: negative. If OD lies very close to Cutoff (grey‑zone), do not assign a result directly. Repeat testing to rule out assay noise and prevent false‑positive / false‑negative misclassification.
Spike‑and‑recovery evaluates matrix interference and assay accuracy. Formula: Recovery % = (measured spiked concentration − native background concentration) ÷ theoretical spiked concentration × 100%. Acceptable range is typically 80%‑120%. Recovery < 80% suggests matrix suppression; recovery > 120% points to matrix enhancement or experimental bias. Adjust sample dilution or pre‑treat matrix.
Matrix effect means components in serum, cell supernatant, tissue lysate or other samples alter measurements via non‑specific binding or inhibition of antigen‑antibody interactions. Mitigation strategies: dilute samples with kit‑supplied diluent, optimize dilution ratios, include matrix controls, centrifuge turbid samples. For severe interference switch to a compatible kit.
Globally low OD: insufficient incubation time, low incubation temperature, expired reagents, over‑washing, insufficient substrate development, degraded standards.
Globally high OD: incomplete washing, prolonged incubation, excessive temperature, over‑developed substrate, blank‑matrix interference, contaminated reagents.
Sandwich ELISA: OD positively correlates with analyte concentration (higher concentration yields higher OD); 4PL fitting is commonly applied. Competitive ELISA: OD negatively correlates with analyte concentration (higher concentration yields lower OD); 4PL fitting is still used. Pay attention to concentration‑response trend to avoid reversed results.