Beginner's Guide to Accuracy & Recovery Studies — % Recovery, %RSD & ICH Q2

🧪 Accuracy Calculator 📘 Beginner Tips & Guide

📘 The Basics

What is an Accuracy (Recovery) Study?

Accuracy tells you how close a measured result is to the true amount of drug actually present. It's checked by spiking a known, pre-weighed amount of the active ingredient into the sample matrix (a placebo, or a real product), analyzing it exactly like a normal test sample, and then comparing what the method "recovers" against what was actually added.

This is usually done at three levels — 80%, 100% and 120% of the label claim — with three replicate preparations at each level (a "3 × 3" design), so you can see whether accuracy holds across the working range of the method, not just at one point.

🎯

Accuracy vs Precision

Accuracy = how close to the true value. Precision (%RSD) = how close repeated results are to each other. A method can be precise but inaccurate (consistently wrong) — that's exactly what a recovery study is designed to catch.

📐

Why 80/100/120%?

Real products vary a little batch to batch, and stability testing needs the method to work even as potency drifts. Testing three levels across this range shows the method behaves linearly and reliably, not just at the label claim. The calculator opens with these three levels pre-filled — every % is a plain editable field, so you can change any of them to whatever your method SOP actually calls for.

Need a different spread? Click ➕ Add Spike Level and the new level automatically suggests +20% above your last level (e.g. after 80/100/120 it suggests 140%) — just a starting point you can overtype with any value, including one below 80% or between two existing levels.

🧮 The Formula

How % Recovery is Calculated

Step 1 — Recovered Amount

Recovered (mg) = (Sample Area ÷ Avg. Standard Area) × Std Weight (mg) × Dilution Factor

Step 2 — % Recovery

% Recovery = (Recovered Amount ÷ Spiked Amount) × 100

The Avg. Standard Area comes from your standard injections — usually 5 replicate injections of a reference standard solution, averaged. This is the same "against a known reference" logic used in the assay method itself.

The Dilution Factor accounts for every dilution step between weighing your standard/sample and injecting it. Multiply each stepwise ratio together — e.g. if you dilute 1→200, then 200→1 (draw 200, make to 1), then 50→5, the combined factor is 1/200 × 200/1 × 50/5 = 10.

Tip: The Conc./Target % field for each level is just a label for your report — it doesn't feed into the math. What actually drives % Recovery is the Area and the Spiked Amount (Wt. of Spike) you enter for that replicate.

✅ Acceptance Criteria

What Counts as a Pass?

There's no single universal number — your method's validation protocol or pharmacopeial monograph sets the real limit — but typical ranges used across the industry are:

Criterion	Typical Range	When it's used
Overall Mean % Recovery	98% – 102%	Tight-spec assay methods, ICH Q2(R2)-style validations
Overall Mean % Recovery	95% – 105%	Common general assay specification
Overall Mean % Recovery	90% – 110%	Wide-spec or trace/impurity-level methods
%RSD per level	≤ 2.0%	Standard spec for replicate precision at each level

Both conditions usually need to be true: the overall mean recovery across all levels must sit inside the chosen range, and the %RSD within each individual level must be at or below the limit. The calculator above checks both automatically and gives you a combined pass/fail verdict.

🛠️ Troubleshooting

Common Reasons Recovery Fails

📉 Consistently Low Recovery (e.g. < 95%)

Often points to incomplete extraction from the matrix, adsorption onto filters/glassware, degradation during sample prep, or an error in the dilution factor (a missed step under-counts the true dilution).

📈 Consistently High Recovery (e.g. > 105%)

Check for interference from excipients/placebo co-eluting with the analyte peak, an overstated dilution factor, or a standard weighed/prepared incorrectly (too dilute a standard inflates the ratio).

📊 High %RSD Within a Level

Usually a sample-preparation consistency issue rather than an instrument issue — check spike weighing accuracy, mixing/extraction time, and whether all replicates were diluted the same way.

🔀 One Level Passes, Another Doesn't

Points to a non-linear response — check the method's linearity data over this range, and confirm the failing level's dilution and spike weight were prepared correctly (a single mis-pipetted replicate can skew a whole level).

🚀 Using the Calculator

Quick Start

Enter your 5 (or however many) standard injection areas first — this sets the reference area used everywhere below.
Enter the Weight of Standard and Dilution Factor exactly as used in your method.
Pick (or customize) your acceptance range and %RSD limit.
The three spike levels open pre-filled at 80% / 100% / 120% — edit any level's % field if your SOP uses different values; editing one row auto-syncs the rest of that level.
For each level, enter the Area of Sample and Wt. of Spike for every replicate — add or remove levels/replicates freely with the ➕ buttons. Adding a new level auto-suggests +20% above the last one, fully editable.
Click Calculate % Recovery to see the verdict, per-level breakdown, recovery chart and full math log.
Use Print Report to generate a fillable calculation sheet with your lab's letterhead for filing.

Understanding Accuracy & Recovery Studies