THE SMART MOLE INSTITUTE — JAPAN RESEARCH EVIDENCE
Derived public-data tables, version1.0, prepared19October2026. No individual records or clinician list.
Run python verify.py using Python3.10+ (standard library only). The independent arithmetic verifier checks the headline summaries and retains opposite/null patterns. It does not certify causal validity. Full original-data pipelines and large cached NDB files remain in the internal research archive.
Source URLs, retrieval dates and hashes are in sources.json. Source definitions and periods are preserved in tables/dictionaries; dictionaries may be in Spanish. MIC age/sex is national, not prefecture-specific. NDB counts are provider-located acts, not unique patients. Suppression is not zero. Mortality is age-standardised all-age, not age75+ mortality.
Figures: static PNG/SVG. Prefectural cartography is schematic geoBoundaries2017 / OpenStreetMap contributors, Wambacher, ODbL1.0. See article for attribution and links; geometry is not redistributed in this package. No SMA boundary-dependent cross-source mapping is claimed.
All associations are descriptive. Spatial dependence, survey uncertainty and causal confounding are not resolved. Do not interpret as a validated clinical access-gap score.
