Icon Legend

This session is not in your schedule.

This session is in your schedule. Click again to remove it.

Favorite

Scientific Oral Presentations

Professional and Personal Development

Session: Scientific Oral Presentations 1 - Professional and Personal Development

Assessing the Use of Generative AI Tools Among Otolaryngology Residency Applications

Monday, October 13, 2025

9:42 AM - 9:48 AM EDT

Location: 235-236

Disclosure(s):

Brett I. Deng, BSc:

faculty for this accredited education activity has no relationship(s) with ineligible companies to disclose.

Introduction: To determine the prevalence of AI tool use in personal statements among otolaryngology residency applicants and identify any associations between applicant characteristics and their use of AI tools.

Methods: This is a retrospective cross-sectional study analyzing personal statements from a cohort of otolaryngology residency program applicants. Applicant demographic and educational data was collected from applicant records. Personal statements (n = 190) were analyzed using three different AI detection tools, each generating an AI detection score on a scale of 0-100 that was subsequently averaged to create a composite score. Summary statistics were calculated from the composite scores. Pearson correlation was used to assess relationships between AI scores and continuous variables (age, step 2 CK scores). One-way ANOVA and t-test used to determine differences in AI scores across different demographic characteristics.

Results: The average composite AI-detection score was 23.35% (SD = 23.18). Of the 190 personal statements, 88 (46%) had an average AI detection score of ≥ 30%. Regression analysis showed no significant correlation between AI detection scores and Step 2 CK scores (r² = 0.00022, p = 0.03) or age (r² = 0.097, p = 0.714). Furthermore, no significant associations were found between AI detection scores and ethnicity (p = 0.22) or degree type (MD vs. DO, p = 0.29). However, significant differences in AI detection scores were observed based on FMG status (FMG = 49.46%, non-FMG = 29.4%, p < 0.05), gender (female = 24.67%, male = 33.94%, p < 0.02), and AOA status (AOA = 32.85%, non-AOA = 24.10%, p < 0.03).

Conclusions: A substantial proportion of applicants (46%) demonstrated a moderate likelihood of AI assistance in their personal statements (AI detection score ≥30%). Significant differences in AI use were associated with FMG status, AOA membership, and gender, while no associations were observed for age, Step 2 CK scores, ethnicity, or degree type. These findings underscore the growing role of AI in residency applications and highlight the need for further research into the implications of its use in this context.

Presenting Author(s)

Brett I. Deng, BSc

Medical Student
Baylor College of Medicine
Colleyville, TX, United States

Disclosure(s):
faculty for this accredited education activity has no relationship(s) with ineligible companies to disclose.

Co-Author(s)

Kevin L. Li, MD

Resident
Baylor College of Medicine
Houston, TX, United States

Disclosure information not submitted.

Senior Author(s)

Kelly Gallagher, MD

Associate Professor
Baylor College of Medicine
Houston, TX, United States

Disclosure information not submitted.

Co-Author(s)

AP

Angela S. Peng, MD

Assistant Professor
Baylor College of Medicine
Houston, TX, United States

Disclosure information not submitted.
Angela D. Haskins, MD

Assistant Professor
Baylor College of Medicine
Houston, Texas, United States

Disclosure information not submitted.
Meha G. Fox, MD

Assistant Professor
Baylor College of Medicine
HOUSTON, TX, United States

Disclosure information not submitted.
KO

Krista L. Olson, MD

Assistant Professor
Baylor College of Medicine
Houston, Texas, United States

Disclosure information not submitted.