{
  "schema_version": 1,
  "builder_plan": {
    "scenario": "An agent looks good in demos, but you cannot tell which product tasks it fails.",
    "change": "Use the study's evaluation idea to inspect a small set of tasks with human-checked answers.",
    "baseline": "Your current evaluator and a manually reviewed reference set.",
    "steps": [
      "Locate the paper's implementation and check its license, model support, data needs, and hardware requirements. Stop here if you cannot obtain the required artifacts.",
      "Save a small set of representative product tasks and expected outcomes. Keep evaluation tasks out of any training or tuning, and choose quality and cost limits before running.",
      "Run the baseline and the proposed change on identical inputs. Keep other settings fixed and repeat paired runs to expose variation.",
      "Save per-task outputs, failures, timings, and all usage costs. Review regressions before expanding the test."
    ],
    "measure": "Agreement with human review, missed failures, false alarms, and evaluation cost per task.",
    "decision": "Keep the change only if it meets your preselected quality and latency limits and improves the metric you care about after setup and operating costs. Otherwise retain the baseline. This is an exploration plan, not a validated recommendation."
  },
  "builder_review": null,
  "arxiv_id": "2609.30027",
  "title": "Synthetic Hospital: An Open, Verifiable, Physician-Validated Longitudinal EHR Benchmark",
  "summary": "Replaces inaccessible real EHR data and unverifiable chart-derived ground truth with a fully synthetic, ontology-grounded (ICD-10-CM, SNOMED CT, LOINC) longitudinal benchmark of 1,268 patients served through a simulated EHR system with standard APIs and function calling. It costs only benchmark adoption effort and API or inference spend for evaluation, since no training or infrastructure is required, but it is relevant only if the reader builds clinical AI. It can fail by misdirecting development toward synthetic-education-material distributions that differ from real clinical documentation, and near-chance physician discrimination does not guarantee that strong benchmark scores transfer to production on real charts.",
  "rationale": "It is a high-quality evaluation resource rather than a deployable technique, so it matters only to teams building or evaluating clinical LLM systems and otherwise does not affect production practice.",
  "technique": "Synthetic, ontology-grounded longitudinal EHR benchmark",
  "family": "avaliacao_de_agente",
  "recommendation": "observar",
  "published": "2026-09-24",
  "as_of": "2026-10-18",
  "editorial_status": "indexed",
  "source_url": "https://arxiv.org/abs/2609.30027",
  "independent_implementations": 1,
  "total_implementations": 2,
  "stars": 136,
  "citations": 0,
  "claims": [
    {
      "claim_id": "claim-01",
      "statement": "Replaces inaccessible real EHR data and unverifiable chart-derived ground truth with a fully synthetic, ontology-grounded (ICD-10-CM, SNOMED CT, LOINC) longitudinal benchmark of 1,268 patients served through a simulated EHR system with standard APIs and function calling. It costs only benchmark adoption effort and API or inference spend for evaluation, since no training or infrastructure is required, but it is relevant only if the reader builds clinical AI. It can fail by misdirecting development toward synthetic-education-material distributions that differ from real clinical documentation, and near-chance physician discrimination does not guarantee that strong benchmark scores transfer to production on real charts.",
      "basis": "inferred",
      "result": "",
      "baseline": "",
      "conditions": "",
      "source_url": "https://arxiv.org/abs/2609.30027",
      "source_page": null,
      "source_excerpt": ""
    },
    {
      "claim_id": "claim-02",
      "statement": "Best of 10 models achieves severity-weighted F1 of 0.73 on longitudinal problem-list reconstruction (vs. 0.89 best physician) and misses roughly half of clinically relevant findings in summarization; physicians distinguished synthetic from real charts at near-chance rates (53%).",
      "basis": "inferred",
      "result": "",
      "baseline": "",
      "conditions": "",
      "source_url": "https://arxiv.org/abs/2609.30027",
      "source_page": null,
      "source_excerpt": ""
    }
  ],
  "exposure_map": [
    {
      "dimension": "quality",
      "basis": "not_evaluated",
      "finding": "",
      "interpretation": "",
      "limitation": "",
      "next_check": "",
      "source_url": "",
      "source_page": null,
      "source_excerpt": ""
    },
    {
      "dimension": "compute",
      "basis": "not_evaluated",
      "finding": "",
      "interpretation": "",
      "limitation": "",
      "next_check": "",
      "source_url": "",
      "source_page": null,
      "source_excerpt": ""
    },
    {
      "dimension": "latency",
      "basis": "not_evaluated",
      "finding": "",
      "interpretation": "",
      "limitation": "",
      "next_check": "",
      "source_url": "",
      "source_page": null,
      "source_excerpt": ""
    },
    {
      "dimension": "operations",
      "basis": "not_evaluated",
      "finding": "",
      "interpretation": "",
      "limitation": "",
      "next_check": "",
      "source_url": "",
      "source_page": null,
      "source_excerpt": ""
    },
    {
      "dimension": "compatibility",
      "basis": "not_evaluated",
      "finding": "",
      "interpretation": "",
      "limitation": "",
      "next_check": "",
      "source_url": "",
      "source_page": null,
      "source_excerpt": ""
    },
    {
      "dimension": "security",
      "basis": "not_evaluated",
      "finding": "",
      "interpretation": "",
      "limitation": "",
      "next_check": "",
      "source_url": "",
      "source_page": null,
      "source_excerpt": ""
    },
    {
      "dimension": "data_and_training",
      "basis": "not_evaluated",
      "finding": "",
      "interpretation": "",
      "limitation": "",
      "next_check": "",
      "source_url": "",
      "source_page": null,
      "source_excerpt": ""
    },
    {
      "dimension": "reproducibility",
      "basis": "not_evaluated",
      "finding": "",
      "interpretation": "",
      "limitation": "",
      "next_check": "",
      "source_url": "",
      "source_page": null,
      "source_excerpt": ""
    }
  ],
  "risks": [],
  "minimum_test": [],
  "open_questions": [],
  "independent_tests": [],
  "lab_reproductions": [],
  "report_available": false,
  "validation_tier": "unknown",
  "evidence_tier": "unknown",
  "infrastructure_basis": "unknown",
  "equations": [],
  "equations_status": "not_fetched",
  "core_kind": "",
  "equations_fetched_at": "",
  "equations_source_url": "",
  "equations_source_sha256": "",
  "exposures_reviewed_at": "",
  "exposures_source_sha256": ""
}