ACMG Education Workgroups-July 23 Evaluation Report

July 23 Evaluation Report

Pdf Summary

The document presents feedback from a survey about leveraging Large Language Models (LLMs) for decision support in personalized oncology. The results show a mix of interest, skepticism, and logistical challenges in adopting LLMs within the field.<br /><br />A significant proportion of respondents (97.73%) expressed interest in the potential applications of LLMs in personalized oncology, with 100% acknowledging their potential role in clinical decision support. However, respondents noted limitations such as LLM unreliability, lack of development, and the absence of institutional resources to support their application. A majority (75%) of survey respondents mentioned that they are currently unable to incorporate LLMs due to these constraints.<br /><br />Some respondents are optimistic about integrating LLMs into their practices. They envision their use in creating patient summaries, variant interpretation, and enhancing patient communication. Respondents indicated an interest in utilizing LLMs for precision medicine tools in diagnostics and integrating them for efficient patient data analysis. <br /><br />Despite initial hesitations, educational content was well-received, with an average satisfaction score of 4.39 out of 5. Participants appreciated the informative nature of the sessions, noting them as enlightening and highlighting the potential impact of LLMs in precision medicine. Suggestions for future topics included a focus on more practical applications such as implementing LLMs in normalizing genetic nomenclature and employing chatbots for clinical interpretation.<br /><br />Overall, while there's enthusiasm about the transformative potential of LLMs in personalized oncology, practical barriers and skepticism about their readiness remain significant hurdles. Addressing these challenges requires improvements in technological maturity, infrastructure support, and clarity on LLM capabilities and limitations.

Keywords

Large Language Models

personalized oncology

clinical decision support

LLM adoption challenges

precision medicine

patient data analysis

educational content

genetic nomenclature

chatbots

technological maturity