Pro Tier Module
Bioequivalence 90% Confidence Interval Calculator
Average bioequivalence 90% CI for a 2x2 crossover - exact, in-browser, against the FDA 80-125% window.
See it run - a worked example, 100% in this browser tab
The problem
Generic-drug teams must prove average bioequivalence by showing the 90% confidence interval on the geometric mean ratio falls entirely within the FDA 80.00-125.00% window, yet the ln-scale arithmetic and the Student-t lookup are easy to mis-apply across a study report.
The local-first solution
This plugin computes the 90% CI from your point estimate and intra-subject variance using exact ln-to-ratio transforms entirely in your browser, with every regulatory constant cited and nothing uploaded to anyone's cloud.
What it does
90% CI on the GMR via mean_diff +/- t * sqrt(MSE * 2 / n), exponentiated to the ratio scale
Accepts ANOVA MSE directly or converts intra-subject CV% via MSE = ln(1 + (CV/100)^2)
Verdict against the exact FDA 80.00-125.00% acceptance window (0.80 = 1/1.25)
Surfaces the Student-t critical value and degrees of freedom (df = n - 2) it actually used
Flags highly variable drugs (reference CV >= 30%) where reference-scaled ABE may apply
Optional unbalanced-arm override accepting n_TR and n_RT for the SE term
Honest scope
Exact here are the ln-to-ratio transforms, the CV-to-MSE identity, the SE and CI arithmetic, and the 80-125% comparison; the Student-t critical value is a table lookup you can override from your validated software, and df is always shown. The ANOVA itself, carryover effects, RSABE scaling, and higher-order designs are not modeled. Not medical, regulatory, or statistical advice.
Authorities cited
- 21 CFR 320.24 - Types of evidence to establish bioavailability or bioequivalence; the average-BE 90% confidence-interval criterion on the log-transformed pharmacokinetic measures.
- FDA Guidance for Industry: Statistical Approaches to Establishing Bioequivalence (2001) - log transformation of AUC/Cmax, the 80.00-125.00% limits, and the two one-sided tests (TOST) at alpha = 0.05.
- ICH M13A - Bioequivalence for Immediate-Release Solid Oral Dosage Forms (2024) - the 90% CI of the Test/Reference geometric mean ratio within 80.00-125.00%.
- FDA Guidance: Bioequivalence Studies with Pharmacokinetic Endpoints for Drugs Submitted Under an ANDA (2021) - 2x2 crossover design and analysis on the ln scale.
- Schuirmann, D. J. (1987). A comparison of the two one-sided tests procedure and the power approach for assessing the equivalence of average bioavailability. J. Pharmacokinet. Biopharm. 15(6), 657-680. DOI 10.1007/BF01068419.
- Chow, S.-C. & Liu, J.-P. (2009). Design and Analysis of Bioavailability and Bioequivalence Studies, 3rd ed. CRC Press. (2x2 crossover ANOVA; CI = exp(mean_diff +/- t*sqrt(MSE*2/n)); CV = sqrt(exp(sigma^2)-1).)
- FDA Guidance: Bioequivalence: Reference-Scaled Average Bioequivalence (RSABE) Approach for Highly Variable Drugs - the within-subject CV >= 30% (sigma_WR >= 0.294) scaling territory (flagged, not computed here).
- NIST/SEMATECH e-Handbook of Statistical Methods, 1.3.6.7.2 - critical values of the t distribution (one-sided t_{0.95, df}).
Run the BE interval
Enter your point estimate and variance and read the 90% CI verdict in the browser - nothing is uploaded. Route the result into a Sandbox workspace, a Worklog case, or a Gate client portal.