evidence: source inventory for oral-health-access-global

[cschanhniem]

Task

• Task ID: source-inventory
• Agent role: literature-scout
• Pack: public-health/oral-health-access-global
• Safety risk: low
• Reviewer needed: domain-reviewer

Claim

Available oral health data sources are sufficient for country-level framing of disease burden and workforce density in low-income countries, but no source provides sub-national oral health service access or dentist distribution data for any low-income country. Sub-national gap mapping is not currently feasible with public data.

Kill condition

A reviewer identifies a public dataset that provides sub-national oral health service access or dentist distribution for at least one low-income country, or finds that WHO country profiles list more than 30 LMICs with national oral health surveys using the 2013 protocol.

Files changed

• evidence.json — added 4 new evidence records (WHO GHO dentistry workforce, WHO oral health survey methods, IOMT fluoridation systematic review, World Bank HNP statistics)
• datasets.md — replaced placeholder content with real source classification table (6 sources: 4 usable, 1 limited, 1 rejected)
• evidence.md — added notes for all 5 evidence records
• claims.json — new file with dry-lab-verified claim validated against claim.schema.json
• agent-radar.json, docs/wiki/Agent-Radar.md — regenerated by pnpm build

Evidence added

ID	Source	Type	Status
who-oral-health-2022	WHO Global Oral Health Status Report 2022	primary-source	Usable (existing)
who-global-health-observatory-dental-workforce	WHO GHO Dentistry Workforce Density	dataset	Usable
who-oral-health-survey-methods-2013	WHO Oral Health Surveys Basic Methods 5th ed	primary-source	Usable (proxy)
iomt-fluoride-dental-caries-systematic-review-2020	IOMT water fluoridation systematic review	peer-reviewed-study	Limited
world-bank-health-nutrition-population-stats	World Bank HNP Statistics	dataset	Usable
DHS/MICS oral health modules	(not added as evidence record — rejected)	—	Rejected

Validation method

pnpm validate passes (schemas, formatting, links, wiki freshness, task index, agent radar). pnpm reproducibility:check passes (516 tasks). pnpm build passes (100 packs).

Known limitations

• Country-level dentist density data is self-reported by health ministries and may exclude private and informal providers.
• The inference about fewer than 30 LMICs having national oral health surveys depends on WHO country profile completeness; some surveys may exist but not be indexed.
• The fluoridation systematic review evidence base is from high-income countries; transferability to LMICs is uncertain.
• The claim is about data availability, not about oral health outcomes or access quality.

Protocol Notes

• The problem.json why_it_matters field is identical to the summary field — a likely copy-paste from pack generation that should be replaced with a distinct justification. Flagging for a future pack-quality PR.
• The tasks.json inputs field for this task lists generic categories ("primary survey data", "satellite data") that do not match the actual oral health domain. This is a pack-quality issue, not a schema issue.
• The validator does not check whether claims.json exists in a problem pack directory. The claim schema is compiled and validated against examples, but there is no pack-level check that claims.json (if present) validates against claim.schema.json. A future validator rule could add this.