Legion Diagnostics

Summary & Replicate Guidelines

Interference Summary (Dose Response)

The Interference Summary table reports the maximum interferent concentration at which the interference effect stays below the defined allowable limit. The module fits the dose-response data to the best regression model (linear or quadratic), then uses the upper bound of the 95% confidence interval to identify the threshold concentration — a conservative, worst-case estimate.

Interpretation: The value in the Interference Summary represents the maximum interferent concentration below which the interference effect is expected to remain within acceptable limits with 95% confidence. At concentrations above this value, the interference bias may exceed the clinically significant threshold. This concentration should be compared against the typical range of interferent concentrations seen in the laboratory's patient population.

Replicate Calculator

The module includes a replicate calculator that determines the minimum number of replicates required per sample condition based on CLSI EP07 statistical guidance. The required number of replicates depends on the precision of the measurement procedure (repeatability SD) and the allowable difference — the tighter the acceptance criterion relative to method imprecision, the more replicates are needed to achieve adequate statistical power.

Advanced settings allow control over:

  • Alpha risk (α) — The probability of falsely concluding that interference is significant when it is not (Type I error). Default is typically 0.05 (5%).
  • Beta risk (β) — The probability of failing to detect a true interference effect of the specified magnitude (Type II error). Default is typically 0.20 (80% power).
  • Test type — Two-sided (recommended default): used when the direction of interference is unknown and either positive or negative effects are possible. One-sided: used only when the direction of interference is known in advance (e.g., only negative effects are biologically plausible for this interferent). Using a one-sided test when the direction is uncertain increases the risk of missing interference in the opposite direction.