# K9a control – summary (30.10.2026)
Scope: supplement rows by opus-S1 (bernstein2015, papier2025) and opus-S4 (mehta2020): 718 rows. All numbers were compared with the supplement texts by script, and the flags were checked by reading.

| Study | Rows checked | Confirmed (numbers) | Critical | Major | Minor | Missing | Verdict |
|---|---|---|---|---|---|---|---|
| bernstein2015 | 352 | 352 | 0 | 0 | 5 | 0 (2 printed values = existing _145/_146) | fit |
| papier2025 | 50 | 50 | 0 | 2 (_007 should be includes_processed=partial; calibrated trend rows as main need an explicit 4.2(b) ruling) | 3 | 0 | fit after flag fix |
| mehta2020 | 316 | 316 | 0 | 2 (opus-G _233/_234 duplicate S4 _495/_540) | 3 | 0 | fit; mark _233/_234 superseded |

**Wrong numbers found:** none.

**Printed anomalies in the paper** (recorded as printed and noted):
- Mehta T4 Q1 cases "4"
- Mehta T5 processed Q2 CI
- Mehta bacon median "028"

**Cross-study item for Fable:** R6 coding of hamburger/burger in "unprocessed/red meat" aggregates.
- Bernstein = partial
- Papier = no, but the source lists burger
- Mehta = no; FPED includes hamburgers