# K4 control – summary (30.10.2026)
552 rows checked; every printed number correct. No critical errors.

| Study | Rows | Confirmed | Critical | Major | Minor | Missing | Fit |
|---|---|---|---|---|---|---|---|
| bernstein2015 | 163 | 163 | 0 | 0 | 2 | 0 (S1–S5 n/a) | yes; _096 disputed (CI implausibly narrow, paper's own misprint) |
| knuppel2020 | 276 | 276 | 0 | 0 | 2 | 0 | yes; S5 = red meat, S7 = poultry verified; calibrated trend rows not main by rule |
| zouiouich2025 | 91 | 91 | 0 | 2 | 3 | 146 (figure numbers in PDF-derived text) | no – completed in a follow-up extraction |
| papadimitriou2022 | 22 | 22 | 0 | 0 | 4 | 4 (fish subtypes) | yes, as uncalibrated per-SD row |

Hamburger rule: Bernstein partial vs Zouiouich no → harmonised to partial. Papadimitriou: p_value holds FDR q; offal/liver includes_processed → unclear.