Implementation determinants of a national hub-and-spoke tele-oncology service: a CFIR 2.0 framework analysis of the VA National TeleOncology service

Tele-oncology implementation remains uneven and theoretically underdeveloped. Existing literature catalogs barriers and facilitators without systematic mapping to established implementation science frameworks, and without a replicable method for applying CFIR to generate transferable implementation guidance. This paper addresses both gaps by operationalizing a dual-stream theory-based framework analysis design and anchoring CFIR 2.0 construct mapping in the VA National TeleOncology (NTO) service, the largest national tele-oncology service in the United States, while extending construct coverage through targeted systematic searching. We conducted a theory-based conceptual framework analysis using CFIR 2.0 as a pre-specified deductive analytic lens operationalized through two complementary evidence streams. Stream 1 comprised eight peer-reviewed NTO papers (2021–2026) purposefully selected through expert familiarity with the NTO literature, including input from co-authors with program-level knowledge. Stream 2 comprised five broader tele-oncology implementation papers (2022–2026) identified through systematic searching of four databases. Papers were coded using CFIR 2.0 construct definitions and a structured extraction template. Reporting followed the Standards for Reporting Implementation Studies. Database searches retrieved 1,261 records. Following removal of 673 duplicates, 588 unique records were screened at title and abstract level; 48 advanced to full-text review, of which 44 were excluded with reasons documented in Supplementary File 3. Five papers met Stream 2 inclusion criteria, including one [1] added via an updated search conducted after initial Stream 2 screening to capture newly published tele-oncology literature (Table 1). Across all 13 included papers, findings mapped to all five CFIR 2.0 domains and 22 constructs. Three significant factors impacting implementation emerged. First, the VHA’s 2018 federal telehealth rule and single-payer structure functioned as necessary outer-setting preconditions rather than contextual moderators, removing licensure and reimbursement barriers that inner-setting or process-level factors cannot compensate for in their absence. Second, unequal technology access and digital literacy operated as a cross-domain barrier spanning outer-setting infrastructure, individual self-efficacy, and implementation process engagement; observational evidence suggests differential modality patterns by race and ethnicity within NTO, though causal analyses of these disparities remain ongoing. Third, high implementation urgency in the THRIVE (Telehealth Research and Innovation for Veterans with Cancer) pragmatic trial is proposed to have paradoxically constrained evaluation rigor: because NTO, as a clinical service, expanded on a timeline independent of and faster than THRIVE’s research funding, site-level randomization was no longer feasible by the time funding was secured, prompting a shift from a stepped-wedge cluster-randomized design to a quasi-experimental difference-in-differences approach. This analysis offers a CFIR 2.0-based conceptual framework for tele-oncology implementation in rural and underserved settings and demonstrates a replicable dual-stream synthesis method. Outer-setting regulatory reform appears to be the implementation lever with the greatest potential for population-level impact in non-integrated health systems. Digital equity requires coordinated action across multiple CFIR domains.