How It Works
Study Design
CLSI EP17-A2 specifies the following minimum experimental design for LoB evaluation:
Limit of BlankVerificationCLSI EP17-A2≥ 2 reagent lots
| Reagent lots | Minimum 2 (analyzed separately); 4+ lots enables combined analysis |
| Blank samples | Minimum 5 unique blank samples |
| Low-level samples | Minimum 5 unique low-level concentration samples |
| Days | Minimum 3 test days |
| Replicates | Minimum 4 replicates per sample per reagent-day combination |
| Total blank measurements | Minimum 60 per reagent lot |
| Total low-level measurements | Minimum 60 per reagent lot |
Blank samples contain no analyte (or a concentration at least an order of magnitude below the lowest level of interest). Low-level samples contain analyte near the expected LoB/LoD region. Both sample types are required — blank samples drive the LoB calculation while low-level samples characterize performance just above the detection threshold.
Reagent lot strategy: With 2–3 reagent lots, the LoB is calculated independently for each lot. The maximum LoB across all lots is reported as the final result — a conservative, worst-case approach. With 4 or more lots, combined analysis across lots is supported. Testing multiple lots is essential because lot-to-lot differences in reagent composition can meaningfully shift the blank measurement distribution.