Governance, scale, and integration: building community health worker systems ready for artificial intelligence
The Lancet Primary Care · 2026-03 · 2 (3) · 100144
DOI: 10.1016/j.lanprc.2026.100144 ↗Kurz gefasst
Fachartikel in der Zeitschrift The Lancet Primary Care, 2026-03. Verfasst von James O’Donovan und 39 weiteren Autor:innen. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 8 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Governance, scale, and integration: building community health worker systems ready for artificial intelligenceArtificial intelligence (AI) is rapidly entering communitybased primary health care in areas such as diagnostics, logistics, and clinical decision making.Community health worker (CHW) programmes operate within these broader systems, and their ability to adopt AI meaningfully depends on whether the surrounding layers of care-clinics, supervision, supply chains, data systems, and governance-are ready to integrate it.Within CHW programmes, the promise of AI is not to replace human care but to assist it by reducing administrative burden, improving consistency, and supporting safer decisions.However, this promise will not be realised through technology alone.Previous waves of digital health innovation outpaced government institutionalisation, producing disconnected tools and pilots. 1,2 Although some innovations delivered short-term gains, fragmentation undermined sustainability, accountability, and equitable learning at scale.AI now risks following the same trajectory, not because the tools are inherently flawed, but because the associated systems are not yet designed to own, govern, and scale them.Beyond these implementation barriers lies a bigger risk: starting with the technology before understanding what the system requires.An AI cancer-screening tool deployed in areas without oncologists, imaging tools, or treatment pathways, yields poor results.Success is often measured by pilot launches rather than by improved health outcomes.3 Without systematic safety checks-assessing bias, error propagation, and contextual validity-AI might be considered successful while burdening CHWs, entrenching inequity, or fragmenting systems.The real danger is not failure, but scaling on terms that undermine care, trust, and professional judgement.We surveyed 28 non-governmental organisations (NGOs) operating CHW programmes across 18 countries through Community Health Impact Coalition (appendix pp 1-2).Among the NGOs, 17 (61%) are piloting AI-enabled solutions, yet only one (4%) reports meaningful integration within local or national public health-care systems.Furthermore, 13 (48%) of the existing AI initiatives are led by NGOs or private entities with government approval, five (19%) operate with no
Zitieren
APA
O’Donovan, J., Ballard, M., Hope, R., Wamae, J., Miranda, A. V., DeRenzi, B., Kabanda, R., Chen, N., Bazira, L., Raghavan, M., Finnegan, K., Yegon, E., Razafinjato, B., Johnson, A., Aranda, Z., Juma, M., Johnston, J. S., Napier, H., Nair, D., …, & Ravinutala, S. (2026). Governance, scale, and integration: building community health worker systems ready for artificial intelligence. The Lancet Primary Care, 2(3), 100144. https://doi.org/10.1016/j.lanprc.2026.100144
Vancouver
O’Donovan J, Ballard M, Hope R, Wamae J, Miranda AV, DeRenzi B, et al. Governance, scale, and integration: building community health worker systems ready for artificial intelligence. The Lancet Primary Care. 2026;2(3):100144. doi:10.1016/j.lanprc.2026.100144
BibTeX
@article{ODonovan2026,
author = {O’Donovan, James and Ballard, Madeleine and Hope, Rebecca and Wamae, Jane and Miranda, Adriana Viola and DeRenzi, Brian and Kabanda, Richard and Chen, Nan and Bazira, Lennie and Raghavan, Mallika and Finnegan, Karen and Yegon, Erick and Razafinjato, Bénédicte and Johnson, Ari and Aranda, Zeus and Juma, Mary and Johnston, Jamie Sewan and Napier, Harriet and Nair, Divya and Nagar, Ruchit and Savanur, Pranav and Nagraj, Shobhana and Harris, Matthew and Palazuelos, Daniel and Nardella, James and Miller, Nathan and Ramanathan, Nithya and Stansert-Katzen, Linnea and Mudhune, Sandra and Tomar, Namrata and Brockman, Ben and Scotney, Soleine and Ward, Victoria and Blaj, Andra and Kok, Maryse and Akram, Ali and Tschida, Scott and Rogers, Ash and Chambert, Emilie and Ravinutala, Sid},
title = {Governance, scale, and integration: building community health worker systems ready for artificial intelligence},
journal = {The Lancet Primary Care},
year = {2026},
volume = {2},
number = {3},
pages = {100144},
doi = {10.1016/j.lanprc.2026.100144}
}
EUVIMED