Conflated Trust, Misread Loyalty: Disentangling Artefact Trust and Provider Trust in Digital Health[Confiance confondue, fidélité trompeuse : Distinguer la confiance dans le logiciel et la confiance dans son éditeur en santé numérique]

This study examines trust in decentralized digital health by distinguishing two trust objects: the health information technology (HIT) artefact and the Health Software Provider (HSP) organization. Based on 25 semi-structured interviews with self-employed physicians, medical secretaries, and IT support staff in France, we conduct a theory-informed thematic analysis using system trust antecedents (functionality, helpfulness, reliability) and organizational trustworthiness dimensions (competence, benevolence, integrity). Findings show that physicians primarily ground trust in artefact performance and workflow fit, whereas provider trustworthiness is assessed more indirectly through support responsiveness, update strategies, pricing practices, and the translation of regulatory mandates into software defaults. We theorize inscription-based inference as the mechanism through which artefact evolution shapes attributions of provider trustworthiness under opacity and regulation. This explains why continued use can be misread as provider loyalty when artefact and provider trust are conflated, even as organizational trust erodes under high switching costs, informing governance and design choices.