The Australian Hepatitis C Point-of-Care Testing Program, a nationwide effort to expand point-of-care testing across 125 community and prison settings, demonstrated that while early adoption succeeded through portable and rapid diagnostics, long-term routine delivery and sustainment encountered significant structural obstacles. In a comprehensive multilevel qualitative process evaluation, researchers analyzed how diverse implementation determinants influenced initial program uptake, everyday service provision, local adaptations, and anticipated sustainment across these complex environments. The investigation identified 571 distinct implementation determinants, revealing that the practical reality of scaling decentralized diagnostic technologies depends heavily on navigating multi-layered organizational and system-level challenges.
To capture the multi-tiered dynamics of the national rollout, the research team conducted their process evaluation in alignment with Medical Research Council guidance, structuring their inquiry around the updated Consolidated Framework for Implementation Research. The empirical dataset drew from three extensive information streams, incorporating 28 official program documents, formal records from 124 administrative meetings, and 28 semi-structured interviews conducted with 30 key program stakeholders. Using a rigorous deductive-inductive thematic analysis strategy, the authors mapped qualitative material directly to established CFIR 2.0 constructs before inductively deriving specific determinants within those constructs, while calculating evidence strength scores to summarize the breadth and triangulation of supporting evidence across diverse roles, jurisdictions, and source materials.
A central methodological component of the study involved quantifying the practical influence of each identified determinant through structured human scoring and automated natural language processing. Two independent researchers rated each determinant on a five-point scale spanning from minus two for major barriers to plus two for major enablers, resolving initial scoring differences through formal consensus procedures. To explore complementary analytical techniques, the investigators deployed an offline transformer model to generate automated sentiment scores directly from the qualitative text. The team then statistically compared the algorithmic sentiment scores against the human barrier and enabler ratings using Pearson correlation coefficients and Cohen’s kappa statistics to measure concordance.
Across the 571 identified determinants, the qualitative data clustered most heavily within three primary CFIR domains: the Outer Setting, the Innovation, and the Individuals involved in program delivery. During the initial adoption phase of the national initiative, uptake was powerfully propelled by the clear relative advantage of point-of-care testing over conventional diagnostic methods. Stakeholders praised the rapid turnaround of test results, the physical portability and operational flexibility of the equipment, and the strong compatibility of the devices with outreach-oriented clinical models. This early momentum was reinforced by productive collaborative partnerships, decisive program leadership, and peer-supported engagement strategies that actively encouraged community participation.
However, as the program shifted from initial adoption into routine delivery, local workflow adaptation, and forward-looking sustainment planning, critical operational constraints emerged within institutional and systemic environments. Furthermore, participating sites struggled with chronic workforce instability, severe clinical workflow pressures, unpredictable long-term funding arrangements, and pervasive social and institutional stigma regarding hepatitis C virus infection, creating substantial friction for clinical staff attempting to maintain uninterrupted diagnostic testing.
The comparative analysis between algorithmic sentiment and human qualitative assessment demonstrated moderate statistical concordance, with the offline transformer pipeline matching human-assigned ratings for 80.9 percent of determinant records. This computational alignment yielded a Cohen’s kappa value of 0.64, accompanied by a 95 percent confidence interval spanning from 0.59 to 0.69. By comparison, independent human raters achieved an initial agreement rate of 91.5 percent prior to consensus discussions, reflecting a Cohen’s kappa of 0.83 with a 95 percent confidence interval between 0.79 and 0.87. An analysis of discordant cases demonstrated that divergence occurred primarily when the practical implementation impact of a determinant differed fundamentally from the linguistic tone in which stakeholders discussed that factor.
Ultimately, the study concluded that successful national scale-up of decentralized hepatitis C point-of-care testing requires dedicated, sustained interventions to address workflow integration, electronic data connectivity, workforce staffing capacity, dedicated financing, and anti-stigma measures across participating organizations.
Subject of Research: Implementation determinants and sentiment analysis of Australia's national hepatitis C point-of-care testing scale-up
Article Title: National scale-up of decentralised hepatitis C point-of-care testing in Australia: a multilevel qualitative process evaluation with exploratory sentiment analysis
Article References: Rudge, T., Nicholls, W., Taylor, N., Torres-Robles, A., Marshall, A. D., Mazariego, C. G., Lafferty, L., Treloar, C., Kingsland, M., Wolfenden, L., Pedrana, A., Matthews, S., Markus, C., Milat, A., Grebely, J., & Fontaine, G. (2026). National scale-up of decentralised hepatitis C point-of-care testing in Australia: a multilevel qualitative process evaluation with exploratory sentiment analysis. Implementation Science. https://doi.org/10.1186/s13012-026-01545-9
Image Credits: AI Generated
DOI: 10.1186/s13012-026-01545-9
Keywords: hepatitis C virus, point-of-care testing, process evaluation, Consolidated Framework for Implementation Research, sentiment analysis, natural language processing, transformer models, implementation determinants, prison settings, community health, workflow integration, data connectivity, workforce capacity, health service delivery, qualitative research, interoperability
Cite Scienmag News
Kristina Jarvis. (November 16, 2026). Australian Hepatitis C Testing Evaluation Identifies Workflow and System Constraints. Scienmag. https://scienmag.com/australian-hepatitis-c-testing-evaluation-identifies-workflow-and-system-constraints/
Kristina Jarvis. "Australian Hepatitis C Testing Evaluation Identifies Workflow and System Constraints." Scienmag, 16 November 2026, https://scienmag.com/australian-hepatitis-c-testing-evaluation-identifies-workflow-and-system-constraints/. Accessed 16 November 2026.
Kristina Jarvis. "Australian Hepatitis C Testing Evaluation Identifies Workflow and System Constraints." Scienmag. November 16, 2026. https://scienmag.com/australian-hepatitis-c-testing-evaluation-identifies-workflow-and-system-constraints/








