Content Uniformity: USP <905> and ChP Judgement
What Is Content Uniformity?
Content uniformity examines how evenly the active ingredient is distributed among individual units of a single-dose dosage form (tablets, capsules, single-dose injections), preventing individual units from deviating too far and causing efficacy or safety problems. It assesses each unit's individual content rather than the average and is a key indicator of dosage-form uniformity. Failure often traces to mixing, compression or filling variation and calls for process analysis; maintaining uniformity is the basis of therapeutic efficacy and medication safety.
USP <905> Judgement
Assay 10 units individually, compute the mean Xbar and standard deviation s, and calculate the acceptance value AV = |M - Xbar| + k*s (k = 2.2), where M depends on the position of the mean relative to 100% of the label claim. AV of 15.0 or less (L1) passes; between 15.0 and 25.0 (L2), assay 20 additional units for 30 total and pass if the AV of the 30 units is 15.0 or less. The tool automates the two-stage decision flow.
The Chinese Pharmacopoeia Criterion
The ChP content-uniformity criterion passes when A + 2.2S is 15.0 or less, where A = |100 - Xbar| and S is the standard deviation; between 15.0 and 25.0, double the assay with 20 more units and recheck against the 25.0 limit. The thinking is the same as USP but details differ (M determination, decision flow); the tool supports both criteria and outputs the conclusion for the selected pharmacopeia. Confirm which pharmacopeia version the product's registration standard adopts before choosing.
How to Use It
Enter the assay results of 10 units (or 30 if needed), choose the pharmacopeia (USP/ChP), and the tool auto-computes the mean, SD, AV (or A+2.2S), judges pass or retest, and outputs a report. Use the content-uniformity-specific assay method and make the label claim explicit; verify that unit sampling was proper if the data look abnormal. Evaluate uniformity together with the content mean - individual over-limit units and a low mean should be handled separately.