# Verification checks (P4)

| check | outcome |
|---|---|
| Clean re-run (deleted results/table/figure/exclusions, re-ran analysis.py): results.json byte-identical | PASS |
| baseline_metric telescoping identity: 100·(y_last − y_first)/2497 = 100·(1.68−2.75)/2497 = −0.042851 | PASS (matches −0.0428514) |
| Exclusion ledger: 2498 − 1 = 2497 changes; 2497 − 2497 = 0 in estimand windows; n_analysed null (none used) | PASS |
| n_treated = count(post_cut==1) = 0; consistent with sample_end 2019-12-31 < 2020-03-04 | PASS |
| Chronology sample_start (2010-01-04) < sample_end (2019-12-31) | PASS |
| table1: 10 years, n_days sum = 2498 = n_input_rows | PASS (250×8 + 249×2 = 2498) |
| table1 note: 2010 mean_change_bp uses 249 changes (first row has no change) | noted |
| Effect / CI / RMSE: null because inestimable; CI-contains-estimate and positive-SE checks not applicable | N/A (documented) |
| conclusion code premise_invalid ⇔ feasibility infeasible, estimand keys null | PASS |
| magnitude: baseline ≈ −0.04 bp/day, yearly means in [−0.46, 0.30] bp, plausible | PASS |