C0X

Population Health Triple Store

SPARQL / FHIR queries over fhir-intake indexes — measure IPP vs NUMER cohort presets (heuristic proxies, not official CQL). Dialect: VALUES / UNION / MINUS / FileMan value-set FROM / FILTER — see docs/SPARQL_SUBSET.md.

Context

Index stats: —

Candidates ≠ official CQL. Population scans and SPARQL presets are heuristic proxies for discovery. After Use results in measure, open the quality dashboard and Re-evaluate CQL for official IPP/DENOM/NUMER.

Population scans should use intake. cache is for single-patient SPARQL after warming FHIR; cohort mode will not warm cache per patient.

Measure cohort presets (population)

Heuristic SPARQL proxies — not official CQL. Separate IPP (age/sex/encounter/condition) vs NUMER (outcome evidence) presets for CMS165, CMS122, CMS125, CMS130, CMS138, CMS2, CMS131, CMS22.

FHIR query → SPARQL

Rendered SPARQL


    


    

      
SPARQL editor

      
Examples: VALUES ?code { "44249-1" "44261-6" } · VALUES ?code FROM oid "2.16…" · FROM measure "CMS2" · FROM abbr "AMI33" · brace UNION. Value-set status: /c0x/valueset/status.

      
      

        
      

      
Population scan returns up to Max patients. A LIMIT n line in the editor sets Max when you run the scan.

      

        
        
        
      

    


    

      
Research / Linked data

      
Live federation demos (browser → open endpoints, no PHI leaves the page: only codes and drug names are sent out). See docs/LINKED_DATA_STRATEGY.md (Vista-on-FHIR) and the metformin Phase-0 walk.


      
Graph exports (target role)

      

        
        
      


      
Federation walk (searcher role)

      

        
        
      

      

        
      

      
Local graph → RxNav → Wikidata → ClinicalTrials.gov. Ready.


      
Trial finder (ClinicalTrials.gov)

      

        
        
      

      

        
      

      
    


    

      
Results

      
Ready.