William Pennington-FitzGerald
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
William Pennington-FitzGerald has a total of 11 co-authors. Here are the top co-authors by number of shared publications:
Akshay Warrier
Department of Otolaryngology-Head and Neck Surgery, Rutgers New Jersey Medical School, Newark, USA
1 shared publication- Ana Hae‐Ok Kim
Department of Otolaryngology–Head and Neck Surgery New York‐Presbyterian/Columbia University Irving Medical Center and Columbia University Vagelos College of Physicians and Surgeons New York New York USA
1 shared publication Francesca Carlino
Department of Otolaryngology-Head and Neck Surgery, Rutgers New Jersey Medical School, Newark, USA
1 shared publicationIbraheem Sharaf
Department of Otolaryngology-Head and Neck Surgery, Rutgers New Jersey Medical School, Newark, USA
1 shared publicationJean Anderson Eloy
Department of Otolaryngology-Head and Neck Surgery, Rutgers New Jersey Medical School, Newark, USA
1 shared publicationJessica Levi
Department of Otolaryngology-Head and Neck Surgery, Boston University Medical Center, One Boston Medical Center Pl, Boston, USA
1 shared publication
Latest Publications
- 13 May 2026
- Published by Springer Nature in European Archives of Oto-Rhino-Laryngology
- Vol. 283 (7) , 1-8
Diagnostic accuracy and citation integrity of four large language models on otolaryngology vignettes
Objectives This study aimed to compare the diagnostic accuracy and citation integrity of four large language models (LLMs) including one general (ChatGPT-4) and three intended for clinical and research use (OpenEvidence, Perplexity, and Pathway), using standardized otolaryngology clinical vignettes. Methods One hundred validated otolaryngology clinical vignettes were presented to each LLM with a prompt requesting both a diagnosis and supporting citations. Diagnostic accuracy was determined against reference answers, and errors were categorized as logical, informational, or explicit. Citation number, source type, hallucination rate, and journal CiteScores were also compared. Results All models demonstrated high diagnostic accuracy (82.0%–91.0%), with ChatGPT-4 achieving the highest numerical accuracy (91.0%), though differences between models were not statistically significant (p = 0.057). Logical errors were most frequent across all models. OpenEvidence and Perplexity generated the most citations per response, while ChatGPT-4 produced the fewest and had the highest hallucination rate (23.0%). Source preferences varied, with OpenEvidence and Pathway favoring narrative reviews and Perplexity favoring government/public health websites. OpenEvidence had the highest mean CiteScore for journal citations. Conclusion This is the first study to assess both diagnostic accuracy and citation integrity of LLMs in otolaryngology. While ChatGPT-4 was most accurate, it had the highest rate of citation hallucinations, suggesting a trade-off between accuracy and source reliability. OpenEvidence, though slightly less accurate, provided more consistent and verifiable references, demonstrating the ability to prioritize citation integrity alongside diagnostic performance for clinical integration.
- 13 May 2026
- Published by Springer Nature in European Archives of Oto-Rhino-Laryngology
- Vol. 283 (7) , 1-8
Diagnostic accuracy and citation integrity of four large language models on otolaryngology vignettes
Objectives This study aimed to compare the diagnostic accuracy and citation integrity of four large language models (LLMs) including one general (ChatGPT-4) and three intended for clinical and research use (OpenEvidence, Perplexity, and Pathway), using standardized otolaryngology clinical vignettes. Methods One hundred validated otolaryngology clinical vignettes were presented to each LLM with a prompt requesting both a diagnosis and supporting citations. Diagnostic accuracy was determined against reference answers, and errors were categorized as logical, informational, or explicit. Citation number, source type, hallucination rate, and journal CiteScores were also compared. Results All models demonstrated high diagnostic accuracy (82.0%–91.0%), with ChatGPT-4 achieving the highest numerical accuracy (91.0%), though differences between models were not statistically significant (p = 0.057). Logical errors were most frequent across all models. OpenEvidence and Perplexity generated the most citations per response, while ChatGPT-4 produced the fewest and had the highest hallucination rate (23.0%). Source preferences varied, with OpenEvidence and Pathway favoring narrative reviews and Perplexity favoring government/public health websites. OpenEvidence had the highest mean CiteScore for journal citations. Conclusion This is the first study to assess both diagnostic accuracy and citation integrity of LLMs in otolaryngology. While ChatGPT-4 was most accurate, it had the highest rate of citation hallucinations, suggesting a trade-off between accuracy and source reliability. OpenEvidence, though slightly less accurate, provided more consistent and verifiable references, demonstrating the ability to prioritize citation integrity alongside diagnostic performance for clinical integration.
- 6 October 2024
- Published by Wiley in Otolaryngology–Head and Neck Surgery
- Vol. 172 (3) , 960-966
Impact of Cochlear Implant Electrode Array Design on Post‐Op Speech Perception
Objective: Electrode array design may impact hearing outcomes in patients who receive cochlear implants. The goal of this work was to assess differences in post operative speech perception among patients who received cochlear implants of differing designs and lengths.Study Design: Retrospective chart review.Setting: Tertiary Care Hospital.Methods: Patients (n = 129) received 1 of 9 electrode arrays, which were categorized by design: Lateral wall electrodes (n = 36) included CI522, CI622 (Cochlear Americas), Flex24, and Flex28 (Med El). Midscala electrodes (n = 16) included HiRes Ultra 3D (Advanced Bionics). Perimodiolar electrodes (n = 77) included CI512, CI532, CI612, and CI632 (Cochlear Americas). Speech perception was evaluated using consonant‐nucleus‐consonant (CNC) tests and at 3, 6, 12, and 24 months postimplantation.Results: Perimodiolar electrodes showed significantly higher CNC scores compared to lateral wall electrodes at 6 and 24 months. Perimodiolar electrodes also outperformed midscala electrodes at 12 months. An inverse relationship was observed between electrode length and CNC scores noted at 6, 12, and 24 months.Conclusion: Perimodiolar electrode arrays, which tend to be shorter, demonstrated better speech perception outcomes compared to the longer lateral wall and midscala arrays at some timepoints. These findings suggest a potential advantages of perimodiolar electrodes for optimizing hearing outcomes.
- 6 October 2024
- Published by Wiley in Otolaryngology–Head and Neck Surgery
- Vol. 172 (3) , 960-966
Impact of Cochlear Implant Electrode Array Design on Post‐Op Speech Perception
Objective: Electrode array design may impact hearing outcomes in patients who receive cochlear implants. The goal of this work was to assess differences in post operative speech perception among patients who received cochlear implants of differing designs and lengths.Study Design: Retrospective chart review.Setting: Tertiary Care Hospital.Methods: Patients (n = 129) received 1 of 9 electrode arrays, which were categorized by design: Lateral wall electrodes (n = 36) included CI522, CI622 (Cochlear Americas), Flex24, and Flex28 (Med El). Midscala electrodes (n = 16) included HiRes Ultra 3D (Advanced Bionics). Perimodiolar electrodes (n = 77) included CI512, CI532, CI612, and CI632 (Cochlear Americas). Speech perception was evaluated using consonant‐nucleus‐consonant (CNC) tests and at 3, 6, 12, and 24 months postimplantation.Results: Perimodiolar electrodes showed significantly higher CNC scores compared to lateral wall electrodes at 6 and 24 months. Perimodiolar electrodes also outperformed midscala electrodes at 12 months. An inverse relationship was observed between electrode length and CNC scores noted at 6, 12, and 24 months.Conclusion: Perimodiolar electrode arrays, which tend to be shorter, demonstrated better speech perception outcomes compared to the longer lateral wall and midscala arrays at some timepoints. These findings suggest a potential advantages of perimodiolar electrodes for optimizing hearing outcomes.