SocialLM: Social Signal Processing of Patient-Provider Communication using LLMs and Contextual Aggregation
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914560129105920 |
|---|---|
| author | Bedmutha, Manas Satish Chen, Feng Hartzler, Andrea Cohen, Trevor Weibel, Nadir |
| author_facet | Bedmutha, Manas Satish Chen, Feng Hartzler, Andrea Cohen, Trevor Weibel, Nadir |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2505_04152 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | SocialLM: Social Signal Processing of Patient-Provider Communication using LLMs and Contextual Aggregation Bedmutha, Manas Satish Chen, Feng Hartzler, Andrea Cohen, Trevor Weibel, Nadir Computation and Language Computers and Society Human-Computer Interaction H.5.2; H.1.2; I.2.7; I.2.m; J.3 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. |
| title | SocialLM: Social Signal Processing of Patient-Provider Communication using LLMs and Contextual Aggregation |
| topic | Computation and Language Computers and Society Human-Computer Interaction H.5.2; H.1.2; I.2.7; I.2.m; J.3 |
| url | https://arxiv.org/abs/2505.04152 |