Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
February 6, 2024NutrientsOpen Access

Appropriateness of responses ranged from 55.5% for sarcopenia to 73.3% for NAFLD, with only two recommendations contradicting guidelines, but limitations and contradictory advice emerged when handling overlapping conditions.

View Full Paper
Ask AI
Bookmark
Share

Why the study?

Patients increasingly use ChatGPT to seek health-related information, but prior studies have raised concerns regarding its accuracy in offering nutritional advice.

Is ChatGPT an effective and accurate tool for providing dietary advice for non-communicable diseases compared to established guidelines?

Population

Dietary advice queries for non-communicable diseases and a complex multi-disease case

Comparison

ChatGPT version 3.5 dietary advice vs nutritional guidelines

Design

Expert panel evaluation study

Authors

VPValentina PonzoIGIlaria GoitreEFEnrica Favaro

Discussion

Loading...

Member takes

Overview

Caution advised for ChatGPT in overlapping NCD dietary advice; leaves open its role for isolated conditions pending validation.

Structured PICO

Is ChatGPT an effective and accurate tool for providing dietary advice for non-communicable diseases compared to established guidelines?

P
Population
Queries regarding dietary advice for various non-communicable diseases (NCDs) and complex cases with overlapping conditions
I
Intervention
ChatGPT (version 3.5) generated dietary advice
C
Comparator
Established nutritional guidelines
O
Outcome
Appropriateness of dietary advice compared with guidelines, assessed by a panel of nutrition experts

While ChatGPT can provide reasonably accurate general dietary advice for single non-communicable diseases, it cannot safely manage complex cases with overlapping conditions and should not replace professional consultation.

Limitations

  • Efficacy decreased in complex situations necessitating customized strategies

Cite This Study

Ponzo et al. (2024) studied this question.

synapsesocial.com/papers/69da2a10ba6014a02e8366d2https://doi.org/10.3390/nu16040469
View Full Paper
Ask AI
Bookmark
Share