{
  "harness": "sj-harness 0.8.3",
  "job": "job:beb56a2b4963abda2575d830ae8d208b",
  "kind": "domain_review",
  "bundle": "sha256:c8974c199293e79d2d4c9b6161af672ad21710e8291727dbe6e5537d0e85e2db",
  "claims": [
    {
      "local_id": "C1",
      "claim_id": "claim:841b20b8e1d73ad8a2b5dfd676927368f00ca35ebf5a46fc1e33ef37e0d2c4fd",
      "verdict": "minor_issues",
      "significance": "minor",
      "reason": "The audit facts and restricted estimates match the primary reports, but the two strict cohorts keep time-zero misalignment and post-entry selection (Hariri) and referral differences (Adrichem) that would keep a below-null HR regardless, so severity restriction says little about the observational-randomized gap; earlier syntheses (Franke 2023, Abdelfattah 2024) are not cited, and the deferred-replacement point is largely in Adrichem's design and Majed's abstract."
    }
  ]
}
