SocialLM: Social Signal Processing of Patient-Provider Communication using LLMs and Contextual Aggregation

Manas Satish Bedmutha, Feng Chen, Andrea L Hartzler, Trevor Cohen, Nadir Weibel
Proceedings of the 7th Conference on Health, Inference, and Learning, PMLR 333:754-777, 2026.

Abstract

Effective patient-provider communication is difficult to assess at scale. We examine whether large language models (LLMs) can track 20 social behaviors from clinical transcripts without fine-tuning. Across three model families and multiple prompting strategies, LLMs reliably detect social signals, though performance varies by patient race and visit segment. To address this variability under query-only API constraints, we introduce an agreement-weighted ensemble using group-level agreement patterns. This approach improves both accuracy and stability over the best individual model, demonstrating a practical pathway for scalable social signal tracking in clinical conversations.

Cite this Paper

BibTeX
@InProceedings{pmlr-v333-bedmutha26a, title = {SocialLM: Social Signal Processing of Patient-Provider Communication using LLMs and Contextual Aggregation}, author = {Bedmutha, Manas Satish and Chen, Feng and Hartzler, Andrea L and Cohen, Trevor and Weibel, Nadir}, booktitle = {Proceedings of the 7th Conference on Health, Inference, and Learning}, pages = {754--777}, year = {2026}, editor = {Healey, Elizabeth and Fries, Jason and Pollard, Tom and Tang, Shengpu and Zink, Anna and Hartvigsen, Tom and Agrawal, Monica and Finlayson, Sam and Glicksberg, Benjamin and Beaulieu-Jones, Brett and Wang, Kai and Fontalvo, Daseyra and Sarker, Tasmie and Chen, Irene and Alsentzer, Emily}, volume = {333}, series = {Proceedings of Machine Learning Research}, month = {29--30 Jun}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v333/main/assets/bedmutha26a/bedmutha26a.pdf}, url = {https://proceedings.mlr.press/v333/bedmutha26a.html}, abstract = {Effective patient-provider communication is difficult to assess at scale. We examine whether large language models (LLMs) can track 20 social behaviors from clinical transcripts without fine-tuning. Across three model families and multiple prompting strategies, LLMs reliably detect social signals, though performance varies by patient race and visit segment. To address this variability under query-only API constraints, we introduce an agreement-weighted ensemble using group-level agreement patterns. This approach improves both accuracy and stability over the best individual model, demonstrating a practical pathway for scalable social signal tracking in clinical conversations.} }
Endnote
%0 Conference Paper %T SocialLM: Social Signal Processing of Patient-Provider Communication using LLMs and Contextual Aggregation %A Manas Satish Bedmutha %A Feng Chen %A Andrea L Hartzler %A Trevor Cohen %A Nadir Weibel %B Proceedings of the 7th Conference on Health, Inference, and Learning %C Proceedings of Machine Learning Research %D 2026 %E Elizabeth Healey %E Jason Fries %E Tom Pollard %E Shengpu Tang %E Anna Zink %E Tom Hartvigsen %E Monica Agrawal %E Sam Finlayson %E Benjamin Glicksberg %E Brett Beaulieu-Jones %E Kai Wang %E Daseyra Fontalvo %E Tasmie Sarker %E Irene Chen %E Emily Alsentzer %F pmlr-v333-bedmutha26a %I PMLR %P 754--777 %U https://proceedings.mlr.press/v333/bedmutha26a.html %V 333 %X Effective patient-provider communication is difficult to assess at scale. We examine whether large language models (LLMs) can track 20 social behaviors from clinical transcripts without fine-tuning. Across three model families and multiple prompting strategies, LLMs reliably detect social signals, though performance varies by patient race and visit segment. To address this variability under query-only API constraints, we introduce an agreement-weighted ensemble using group-level agreement patterns. This approach improves both accuracy and stability over the best individual model, demonstrating a practical pathway for scalable social signal tracking in clinical conversations.
APA
Bedmutha, M.S., Chen, F., Hartzler, A.L., Cohen, T. & Weibel, N.. (2026). SocialLM: Social Signal Processing of Patient-Provider Communication using LLMs and Contextual Aggregation. Proceedings of the 7th Conference on Health, Inference, and Learning, in Proceedings of Machine Learning Research 333:754-777 Available from https://proceedings.mlr.press/v333/bedmutha26a.html.

Related Material