Define an Analytical Target Profile (ATP), run a formal FMEA risk assessment (RPN = S × O × D) with live traffic-lights, screen robustness against the Method Operable Design Region (MODR), and generate an audit-ready, SHA-256-stamped risk-control-strategy report — with an optional AI draft for expert review. Free, in your browser; nothing is stored. Decision support under ICH Q14/Q9, not a substitute for formal method validation.
1Analytical Target Profile (ATP)
The ATP is the criteria-based objective the procedure must meet — the yardstick against which risk and robustness are judged (ICH Q14 §5.1). How to fill each field:
Procedure name — the method's title, e.g. Aspirin Assay (RP-HPLC).
Method intent — what the procedure measures. This sets the expected acceptance criteria: an assay typically targets 98–102% recovery, whereas impurity limit/quantitative tests use wider windows.
Target analyte — the substance or attribute quantified (e.g. Aspirin (acetylsalicylic acid)).
Range low / high — the working range you report over, in your reporting units (e.g. 50–150% of nominal, or µg/mL). Leave blank if not applicable.
Accuracy (recovery) low / high % — the acceptable mean-recovery window; low must be below high (assay default 98–102%).
Precision limit (%RSD) — the maximum precision you accept. The same bound is reused to judge robustness, so set it to your real SST limit (often ≤ 2.0% for an assay).
Analyte pKa — the pKa nearest the working pH. Optional, but it powers the in-silico auto-scoring of the mobile-phase pH parameter (retention is most sensitive when the working pH sits near the pKa). Aspirin ≈ 3.5.
Specificity requirement & Decision rule — the selectivity statement (e.g. no interference at the analyte retention time) and the rule that makes a result reportable (e.g. only when all system-suitability criteria pass).
2FMEA — method-parameter risk (RPN = S × O × D)
Parameter
Cls
Unit
Nominal
Low
High
Failure mode
S
O
D
RPN
Risk
Each row is one method parameter that could push the procedure off its ATP. How to fill each column:
Parameter — the variable under control, e.g. Mobile-phase pH, Column temperature, Flow rate, Gradient slope.
Unit / Nominal / Low / High — the set-point and the deliberate excursions you would probe for robustness (e.g. pH nominal 2.8, varied 2.6–3.0). Low/High are the method-operable edges you test.
Failure mode — how that parameter makes the method fail its ATP (e.g. retention shift / coelution of the critical pair).
S — Severity (1–10) — impact on the reported result if it fails: 1 = negligible, 4–6 = result questionable / re-test, 8–10 = wrong result or patient-safety impact.
O — Occurrence (1–10) — how likely the excursion is in routine use: 1 = almost never, 4–6 = occasional, 8–10 = frequent / expected.
D — Detection (1–10) — how hard it is to catch before reporting: 1 = system-suitability catches it immediately, 10 = effectively undetectable.
RPN = S × O × D updates live as you type. A parameter is flagged High — requiring multivariate robustness (DoE / MODR) — when its RPN ≥ the High threshold or its Severity ≥ the severity gate (a severe failure is high-risk even if rare). Adjust the three thresholds to match your site's QRM policy.
3Robustness / MODR screen (optional)
Optional, but this is what turns a risk score into robustness evidence. Add one study per parameter you actually varied:
Parameter name — match the FMEA row you're supporting (e.g. Mobile-phase pH).
SST metric — the system-suitability response you measured, e.g. Resolution (critical pair) or %RSD of replicate injections.
Observations — one level, value per line, e.g. low, 1.9 / nominal, 2.4 / high, 2.6. You need at least two.
Lower / Upper acceptance — set a lower bound for "≥" metrics (e.g. resolution ≥ 2.0) and/or an upper bound for "≤" metrics (e.g. %RSD ≤ 2.0).
Every observation must stay inside the bound for that edge of the MODR to be supported; a single breach flags the parameter as not robust across the studied range.
Ready.
This tool applies ICH Q9 quality-risk-management arithmetic (FMEA / RPN) and a univariate robustness screen to values you enter, in support of ICH Q14 analytical procedure development, entirely in your browser. Results — and any AI-drafted narrative — are decision support for a qualified reviewer and do not replace formal method validation (ICH Q2(R2)) or a multivariate design-space study. The control strategy and MODR remain the sponsor's responsibility. Generated via the MolWard Platform.
