Evidence-guided reassessment of hazard bands in engineered nanomaterial control banding

Control banding (CB) supports precautionary management of engineered nanomaterials when quantitative hazard and exposure data are incomplete. However, CB methods differ in how they treat missing information and incorporate new nanoform-specific evidence. In this work, the hazard inputs, default rules, and reassessment procedures of 13 CB and CB-like methods were verified against original publications or official sources. Three nanoforms were evaluated under a fixed occupational scenario at three cumulative information states: limited information (K0), additional physicochemical information (K1), and additional nanoform-specific hazard evidence (K2). For 10% Fe-doped ZnO, the 2008 CB Nanotool score decreased from 52.5 (High; Risk level 3 [RL3]) to 50.0 (Medium; RL2), whereas the LLNL 2.0 classification remained High (RL3). For short multiwalled carbon nanotubes and cellulose nanofibers, additional morphological information lowered the European Commission (EC) concern category, but the final risk level remained unchanged at the fixed High exposure level. ANSES permits expert-supported reassessment, whereas GoodNanoGuide does not specify a quantitative threshold for determining when new evidence is sufficient to change the unknown-properties category. These results show that additional hazard information does not necessarily lower the CB outcome. Across the worked cases, an outcome changed when new evidence could validly update the input, default, analogue assumption, or decision rule driving the current result. Therefore, herein, the driver-linked reassessment rule (D-RULE) is introduced as an evidence-admissibility procedure that determines whether new evidence can validly update one or more drivers of the current CB result before the original method is rerun.