Purpose: This study evaluates the efficacy of two advanced Large Language Models (LLMs), OpenAI’s ChatGPT 4 and Google’s Gemini Advanced, in providing treatment recommendations for head and neck oncology cases. The aim is to assess their utility in supporting multidisciplinary oncological evaluations and decision-making processes. Methods: This comparative analysis examined the responses of ChatGPT 4 and Gemini Advanced to five hypothetical cases of head and neck cancer, each representing a different anatomical subsite. The responses were evaluated against the latest National Comprehensive Cancer Network (NCCN) guidelines by two blinded panels using the total disagreement score (TDS) and the artificial intelligence performance instrument (AIPI). Statistical assessments were performed using the Wilcoxon signed-rank test and the Friedman test. Results: Both LLMs produced relevant treatment recommendations with ChatGPT 4 generally outperforming Gemini Advanced regarding adherence to guidelines and comprehensive treatment planning. ChatGPT 4 showed higher AIPI scores (median 3 [2–4]) compared to Gemini Advanced (median 2 [2–3]), indicating better overall performance. Notably, inconsistencies were observed in the management of induction chemotherapy and surgical decisions, such as neck dissection. Conclusions: While both LLMs demonstrated the potential to aid in the multidisciplinary management of head and neck oncology, discrepancies in certain critical areas highlight the need for further refinement. The study supports the growing role of AI in enhancing clinical decision-making but also emphasizes the necessity for continuous updates and validation against current clinical standards to integrate AI into healthcare practices fully.

Background: The Delphi method is increasingly used in otolaryngology to develop consensus in subjects lacking robust evidence. In these contexts, consensus documents play a dual role: they provide structured expert guidance while determining shared principles adaptable to the individual patient. Nevertheless, the methodological rigor and reporting consistency of Delphi studies remain variable, raising concerns about transparency, reproducibility, and potential bias. Methods: A systematic review was conducted according to PRISMA guidelines. MEDLINE, Embase, and Web of Science were searched for Delphi-based consensus studies in otolaryngology. Fully published studies in English, Italian, German, French, or Spanish were included. Each article was assessed using established methodological frameworks for consensus development, bias domains described in the methodological literature, and the DELPHISTAR reporting checklist. Reliability and completeness scores were calculated to enable comparisons. Results: Out of 3168 unique records, 86 studies were included. Most defined their purpose and consensus criteria, but transparency regarding panel selection, anonymity, feedback, and criteria for stopping the consensus process was often missing. Based on methodological bias domains, only 9 studies (10.5%) reached a “good” reliability score, while the majority were rated as “fair.” According to DELPHISTAR, just 3 studies (3.5%) showed good levels of completeness. Reporting completeness and risk of bias proved heterogeneous across subspecialties. Conclusions: Delphi-based studies are increasingly shaping clinical practice in otolaryngology, but persistent methodological and reporting limitations undermine their reliability. Wider adoption of standardized frameworks is essential to improve transparency, reproducibility, and clinical impact, ensuring that consensus statements support both evidence-informed practice and personalized patient care.

Background/Objectives: Chronic graft-versus-host disease (cGVHD) and oral lichen planus (LPO) are chronic inflammatory conditions with similar oral manifestations. This study aimed to assess whether serum and salivary cytokines (IL-1α, IL-6, IL-17) could serve as reliable biomarkers for cGVHD. Methods: A prospective cohort study was conducted involving cGVHD patients, LPO patients, and healthy controls. Cytokine levels in serum and saliva were measured by ELISA and compared across the groups using the Kruskal–Wallis test. Results: IL-17 levels were significantly elevated in the serum of cGVHD patients compared to LPO patients and controls (p < 0.05). However, IL-6 and IL-1α did not show significant differences among the groups. A comparison of salivary samples between the three groups did not reach statistical significance (p > 0.05). Conclusions: This study suggests that IL-17 could be a potential biomarker for cGVHD-related inflammation, warranting further investigation. Salivary samples do not seem to be a reliable biological marker for the diagnosis of cGVHD. The findings underline the need for larger studies to validate these preliminary results.

Objective. With the rise in life expectancy and increased age-related hearing loss (ARHL), audiological assessment is essential to enable prevention of cognitive decay associated with social isolation. This study evaluates audiological services in Italy to support future research, prevention strategies, and public health policies. Methods. We analysed national audiological service data in Italy from 2021 to 2023, focusing on the use of pure-tone and speech audiometry. Service provision was examined in relation to population size, estimated ARHL prevalence, and availability of specialised personnel.Results. The number of pure tone audiometry tests rose from 634,996 in 2021 to 790,245 in 2023; speech audiometry tests increased from 179,279 to 230,008, respectively. The national prevalence of pure-tone and speech audiometry tests performed in 2023 was 1340.1 x 100,000 and 390 x 100,000, respectively. Delivery of audiometry tests varied by area, with the North-East exhibiting the highest prevalence of speech audiometry tests in 2023, whereas pure tone tests were more more prevalent in the North-West of the country. The availability of audiology professionals was limited, with a just 2.16 consultant medical audiologists and 1.82 technicians per 100,000 residents.Conclusions. The supply of audiology tests perfomed in Italy during 2021-2023 was insufficient to meet the growing demand for hearing assessments among the aging population. However, the effectiveness of any screening audiological programme depends not only on the availability of well-trained specialists, but also on effective education programmes for the general population.

Introduction: Acceptable Noise Level (ANL) is defined as the most comfortable level (MCL) intensity for speech and is calculated by subtracting the maximum noise tolerable by an individual. The ANL test has been used over time to predict hearing aid use and the impact of digital noise reduction. This study analyzes this impact by using different masker babble spectra when performing the ANL test in both hearing-impaired and healthy subjects in three different languages (Dutch, French, and Italian). Materials and Methods: A total of 198 patients underwent the ANL test in their native language using a standardized protocol. The babble file was speech-weighted to match the long-term spectrum of the specific ANL language version. ANL was proposed in three different masking conditions: with multitalker Matched babble speech noise, with the same masking signal with the spectrum reduced from 2 kHz onwards (High Cut), and with the spectrum increased from 2 kHz onwards (High Boost). Results: In all of the comparisons among the three languages, ANL with High Boost noise gave significantly higher (worse) scores than ANL with Matched noise (p-value S1: <0.0001, S2: <0.0001, S3: 0.0003) and ANL with High Cut noise (p-value S1: 0.0002, S2: <0.0001, S3: <0.0001). The ANL values did not show any significant correlation with age and gender. In French, a weak correlation was found between ANL with High Cut noise and the Fletcher index of the worst ear. In Italian, a weak correlation was found between both ANL with Matched and High Boost noise and the Fletcher index of the best ear. Conclusions: ANL with High Boost added to noise stimuli was less acceptable for all patients in all of the languages. The ANL results did not vary in relation to the patients’ characteristics. This study confirms that the ANL test has potential application for clinical use regardless of the native language spoken.