Deep dive — chart 1

Chart 1 — rolling mean eligibility-exclusion score by patient group

The chart the sponsor's VP of Regulatory Affairs should see first. Where the silent P3 divergence begins, in one visual.

7-day rolling mean eligibility-exclusion score per patient group over 90 days

7-day rolling mean eligibility-exclusion score per patient group over the 90-day audit period. P3 divergence begins Day 44 and never recovers.

What you are looking at

X-axis: Day 1 through Day 90 of the audit period
Y-axis: 7-day rolling mean of the AI's exclusion score (0-100 scale)
Each line: one patient group (P1 = young low-comorbidity, P2 = middle-aged moderate-comorbidity, P3 = elderly multimorbid + minority representation, P4 = pediatric/AYA, P5 = prior-line-heavy)
Baseline window: the first 30 days establishes what "normal" looks like per patient group

What the eye sees

Four patient-group lines stay roughly parallel across the 90 days. One line — P3 — climbs. The climb begins around Day 44, and it does not recover. The line separates from the pack.

This is the pattern a clinical-development lead recognises instantly: something in the P3 stream changed, and the AI is scoring it as more excludable than baseline. The 53 percentage-point exclude-lane shift is the operational follow-on consequence.

What the sensor fires on

The sensor is not looking at the eye's pattern. It is running four quantitative thresholds against each 7-day window per patient group:

Mean-shift: is the 7-day mean > 1.5 standard deviations from baseline mean?
Variance-ratio: is the 7-day variance > 1.5x baseline variance?
KL-divergence: is the shape divergence between baseline and current distribution > 0.2?
Representative-population differential: is the max cross-patient-group shift > 12%?

Day 44 fires P3 KL-divergence 0.784 (~4x threshold). Day 51 fires P3 KL-divergence 1.126 (~5x threshold). The pattern the eye sees is the pattern the sensor quantifies.

What FDA + EMA reviewers care about

Reviewers do not read charts — they read Type C briefing books, scientific-advice responses, and BIMO Form 483 workups. But when a submission is questioned, the chart is the exhibit. FDA BIMO inspectors, EMA scientific-advice rapporteurs, ICH E6(R3) essential-documents reviewers, DSMB safety-signal chairs, and IRB continuing-review boards all recognise the "one patient group drifts silently" pattern instantly.

The chart also tells the counterfactual story: had this monitoring been in place operationally, the drift window would have been ~7-14 days rather than 45+.

The next chart

The rolling-mean chart shows when. The baseline-vs-drifted chart shows how the P3 distribution shape moved.

Chart 2 — baseline vs recent distribution →

$499 Snapshot. 3 business days.

Same 3-chart set for your sponsor's actual AI surface + full analysis report.

Buy $499

Back to pharma sample hub.