Abstract
The TruthCert meta-analysis engine is a rigorous filter, using a suite of six statistical estimators—including Robust Variance Estimation. This checks clinical data against small-study effects and heterogeneity. This does not rely on simple significance but uses a twelve-point "Threat Ledger" for evidentiary fragility. This then automatically assigns a binding stability verdict (Stable, Moderate, Exposed, or Uncertain) for all downstream decisions.
This clinical verdict is then used for the economic analysis in S14-HTA+, which acts against against false precision. Under the protocol’s "Refusal of Precision" rule, the system suppresses definitive cost-effectiveness metrics (and adoption claims) if the evidence tier is low. The "Africa-First" principle ensures that the decision pathways prioritise resource protection over speculative modeling (which is often the case in other HTA) whenever data quality is compromised.