RESEARCH PAPER 01COMPLETE
01
Ambient PM2.5 exposure and life expectancy across countries
Country-year evidence, dual-track design checks and an explicit boundary between association and causation.
COVERAGE200 countriesPERIOD1990–2023STATUSFrozen v2
LIVE RESEARCH PORTFOLIO
Every conclusion
has an evidence trail.
GHEAS opens the work before the answer: protocols, searches, contradictions, uncertainty and the exact point where a claim becomes ready.
先公开研究过程，再讨论研究结论。
Enter the active study ↓EVIDENCE SIGNAL LIVE
ACTIVE SEARCH7,091PubMed records
PROTOCOLSCREENINGSYNTHESISSTUDIES2 programs
COMPLETE1 release
ACTIVE1 review
01 — PORTFOLIOSeparate the result
Separate the result
from the work in progress.
Each study has its own evidence boundary. Completed work is frozen; active work reports process metrics, never premature findings.
META-RESEARCH 01 ACTIVE
02
Long-term ambient PM2.5 exposure and mortality
A systematic review and Meta-analysis asking not only what prior studies report, but how design, measurement and transparency shape the answer.
PUBMED7,091 recordsVALIDATION5 / 5 sentinelsCURRENTCalibration
See the live workflow ↓02 — ACTIVE WORKFLOW7,091 records.
7,091 records.
Zero invented conclusions.
The count describes the search, not the answer. Every record must pass a traceable decision path before any pooled estimate is calculated.
Protocol
Question, scope and decision rules versioned
Search
PubMed v1 frozen after sentinel validation
Calibration
105 boundary cases prepared for screening
Screening
Formal eligibility review not yet started
Extraction
Effects, design and provenance
Synthesis
No pooled result until the evidence is locked
CALIBRATION SET105
A deliberately balanced practice set tests the difficult boundaries before all 7,091 records are screened. It is not the number of included studies.
03 — EVIDENCE ROUTINGEqual entry.
Equal entry.
Different scientific roles.
Publication type does not change whether a record enters the topic search. It changes how that record can contribute after screening.
Same gate
Primary studies and published Meta-analyses enter under the same topic search.
Different roles
Primary estimates are synthesized; prior reviews are mapped and compared.
No double counting
A published pooled estimate is never mixed with its underlying studies.
Visible uncertainty
A missing result stays visibly missing until the required evidence exists.
ONE TOPIC SEARCHPM2.5 · mortality
→PRIMARY STUDIESEffect synthesis
META-ANALYSESCoverage + methods map
OTHER RECORDSExclusion register
THE GHEAS STANDARD