In pharmaceutical analysis, accuracy is defined as the "closeness of agreement" between the value found by the analytical procedure and an accepted reference value. While often used interchangeably with precision, accuracy specifically measures the systematic error or "bias" of your method. Calculating accuracy is a non-negotiable step in validating any Stability-Indicating Method (SIM) to ensure that the drug's potency and its impurities are quantified correctly.
The Recovery Study
The most common way to calculate accuracy is through a recovery study. This involves adding a known quantity of a pure standard (the analyte) to a "matrix" (usually a placebo containing all excipients but no drug) and then analyzing the mixture using your method.
1. The Three-Level Spiking Protocol
Regulatory guidelines like ICH Q2(R1) recommend that accuracy be assessed across a range of concentrations. Standard practice is to test at three levels, typically:
80% of the target concentration.
100% of the target concentration.
120% of the target concentration.
In some cases, such as early development, a wider range of 50%, 100%, and 150% may be used.
2. The Recovery Formula
The calculation for accuracy is straightforward:
% Recovery = (Measured Amount / Theoretical Added Amount) × 100
If you are spiking into a sample that already contains some drug (the Standard Addition Method), the formula adjusts to:
% Recovery = ((Total Observed Amount − Initial Amount in Sample) / Amount Spiked) × 100
Correcting Accuracy for Impurities: Relative Response Factors (RRF)
A major challenge in accuracy arises when quantifying unknown degradation products. Standard HPLC-UV detectors assume that the impurity and the parent drug absorb light identically, but this is rarely true. If the response factors differ significantly, your mass balance—and therefore your accuracy—will be flawed.
To solve this, you must calculate a Relative Response Factor (RRF) using a "universal" detector like a Chemiluminescent Nitrogen Detector (CLND) or Charged Aerosol Detector (CAD). The RRF is calculated as:
RRF = (UV Area/Mass)Impurity / (UV Area/Mass)Parent
Applying this factor corrects the UV peak area, providing a more accurate measurement of the impurity's true mass.
Acceptance Criteria
To declare a method accurate, the results must fall within pre-defined limits.
Mean recovery: For most drug substance assays, a mean recovery of 98.0% to 102.0% is expected. For impurities or complex dosage forms, 95% to 105% may be acceptable.
Statistical significance: You should use ANOVA to compare within-day and between-day variability to ensure the accuracy is consistent.
Precision coupling: Accuracy data is often reported alongside the Relative Standard Deviation (%RSD). An RSD of < 2.0% across the spiked samples is the standard indicator of a sensitive and reliable method.
Calculating accuracy is a proof of truth in measurement. By using a structured spiking protocol and accounting for Relative Response Factors, it is possible to ensure that the safety and efficacy data reported to regulatory agencies like the FDA or EMA are based on an absolute understanding of the drug's chemical reality.