What question did this study set out to answer?

To examine the availability of social determinants of health (SDoH) documented in electronic dental records and evaluate natural language processing techniques for automated SDoH extraction.

May 21, 2025Applied Clinical InformaticsOpen Access

Extracting Social Determinants of Health from Dental Clinical Notes

View Full Paper

Ask AI

Bookmark

Authors

FPFarhana PethaniCommonwealth Scientific and Industrial Research Organisation ACAlec B. ChapmanUniversity of Utah MCMike ConwayThe University of Melbourne

Discussion

Member takes

Overview

Informatics evaluation demonstrates low prevalence and moderate extraction accuracy of social determinants of health in dental notes, indicating a need for standardized clinical documentation.

Key Points

To examine the availability of social determinants of health (SDoH) documented in electronic dental records and evaluate natural language processing techniques for automated SDoH extraction.
Sampled 1,000 dental clinical notes from 105,311 patient clinic visits and manually annotated mentions of sugar, tobacco, alcohol, methamphetamine, housing, and employment along with temporality, dose, and frequency.
Evaluated and compared information extraction performance using fine-tuned pretrained language models (PLMs) against a rule-based baseline via F1-scores.
The best-performing PLM achieved F1-scores of 0.75 for sugar, 0.69 for tobacco, 0.67 for alcohol, 0.42 for housing, and 0 for employment, compared to 0.70, 0.69, 0.53, 0.44, and 0 for the rule-based approach, with an overall PLM advantage of 0.04 (95% CI: -0.01 to 0.09).
SDoH documentation frequency in clinical notes was low: sugar appeared in 9.1%, tobacco in 3.9%, alcohol in 1.2%, housing in 1.2%, employment in 0.2%, and methamphetamine in 0%.

Cite This Study

Pethani et al. (2025) studied this question.

synapsesocial.com/papers/6aa939dbd7d677827e255edf https://doi.org/10.1055/a-2616-9858

View Full Paper

Ask AI

Bookmark

May 21, 2025Applied Clinical InformaticsOpen Access

Extracting Social Determinants of Health from Dental Clinical Notes

View Full Paper

Ask AI

Bookmark

Authors

FPFarhana PethaniCommonwealth Scientific and Industrial Research Organisation ACAlec B. ChapmanUniversity of Utah MCMike ConwayThe University of Melbourne

Discussion

Member takes

Overview

Informatics evaluation demonstrates low prevalence and moderate extraction accuracy of social determinants of health in dental notes, indicating a need for standardized clinical documentation.

Key Points

To examine the availability of social determinants of health (SDoH) documented in electronic dental records and evaluate natural language processing techniques for automated SDoH extraction.
Sampled 1,000 dental clinical notes from 105,311 patient clinic visits and manually annotated mentions of sugar, tobacco, alcohol, methamphetamine, housing, and employment along with temporality, dose, and frequency.
Evaluated and compared information extraction performance using fine-tuned pretrained language models (PLMs) against a rule-based baseline via F1-scores.
The best-performing PLM achieved F1-scores of 0.75 for sugar, 0.69 for tobacco, 0.67 for alcohol, 0.42 for housing, and 0 for employment, compared to 0.70, 0.69, 0.53, 0.44, and 0 for the rule-based approach, with an overall PLM advantage of 0.04 (95% CI: -0.01 to 0.09).
SDoH documentation frequency in clinical notes was low: sugar appeared in 9.1%, tobacco in 3.9%, alcohol in 1.2%, housing in 1.2%, employment in 0.2%, and methamphetamine in 0%.

Cite This Study

Pethani et al. (2025) studied this question.

synapsesocial.com/papers/6aa939dbd7d677827e255edf https://doi.org/10.1055/a-2616-9858

Also Consider

Synapse has enriched 4 closely related papers on similar clinical questions. Consider them for comparative context:

View Full Paper

Ask AI

Bookmark