Assessing Retrieval-Augmented Large Language Model Performance in Emergency Department ICD-10-CM Coding Compared to Human Coders.

Background: Accurate medical coding is essential for clinical and administrative purposes but complicated, time-consuming, and biased. This study compares Retrieval-Augmented Generation (RAG)-enhanced LLMs to provider-assigned codes in producing ICD-10-CM codes from emergency department (ED) clinical records. Methods: Retrospective cohort study using 500 ED visits randomly selected from the Mount Sinai Health System between January and April 2024. The RAG system integrated past 1,038,066 ED visits data (2021-2023) into the LLMs' predictions to improve coding accuracy. Nine commercial and open-source LLMs were evaluated. The primary outcome was a head-to-head comparison of the ICD-10-CM codes generated by the RAG-enhanced LLMs and those assigned by the original providers. A panel of four physicians and two LLMs blindly reviewed the codes, comparing the RAG-enhanced LLM and provider-assigned codes on accuracy and specificity. Findings: RAG-enhanced LLMs demonstrated superior performance to provider coders in both the accuracy and specificity of code assignments. In a targeted evaluation of 200 cases where discrepancies existed between GPT-4 and provider-assigned codes, human reviewers favored GPT-4 for accuracy in 447 instances, compared to 277 instances where providers' codes were preferred (p<0.001). Similarly, GPT-4 was selected for its superior specificity in 509 cases, whereas human coders were preferred in only 181 cases (p<0.001). Smaller open-access models, such as Llama-3.1-70B, also demonstrated substantial scalability when enhanced with RAG, with 218 instances of accuracy preference compared to 90 for providers' codes. Furthermore, across all models, the exact match rate between LLM-generated and provider-assigned codes significantly improved following RAG integration, with Qwen-2-7B increasing from 0.8% to 17.6% and Gemma-2-9b-it improving from 7.2% to 26.4%. Interpretation: RAG-enhanced LLMs improve medical coding accuracy in EDs, suggesting clinical workflow applications. These findings show that generative AI can improve clinical outcomes and reduce administrative burdens. Funding: This work was supported in part through the computational and data resources and staff expertise provided by Scientific Computing and Data at the Icahn School of Medicine at Mount Sinai and supported by the Clinical and Translational Science Awards (CTSA) grant UL1TR004419 from the National Center for Advancing Translational Sciences. Research reported in this publication was also supported by the Office of Research Infrastructure of the National Institutes of Health under award number S10OD026880 and S10OD030463. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The funders played no role in study design, data collection, analysis and interpretation of data, or the writing of this manuscript. Twitter Summary: A study showed AI models with retrieval-augmented generation outperformed human doctors in ED diagnostic coding accuracy and specificity. Even smaller AI models perform favorably when using RAG. This suggests potential for reducing administrative burden in healthcare, improving coding efficiency, and enhancing clinical documentation.

Navigating the vestibular maze: text-mining analysis of publication trends over five decades.

Introduction: The field of vestibular science, encompassing the study of the vestibular system and associated disorders, has experienced notable growth and evolving trends over the past five decades. Here, we explore the changing landscape in vestibular science, focusing on epidemiology, peripheral pathologies, diagnosis methods, treatment, and technological advancements. Methods: Publication data was obtained from the US National Center for Biotechnology Information (NCBI) PubMed database. The analysis included epidemiological, etiological, diagnostic, and treatment-focused studies on peripheral vestibular disorders, with a particular emphasis on changes in topics and trends of publications over time. Results: Our dataset of 39,238 publications revealed a rising trend in research across all age groups. Etiologically, benign paroxysmal positional vertigo (BPPV) and Meniere's disease were the most researched conditions, but the prevalence of studies on vestibular migraine showed a marked increase in recent years. Electronystagmography (ENG)/ Videonystagmography (VNG) and Vestibular Evoked Myogenic Potential (VEMP) were the most commonly discussed diagnostic tools, while physiotherapy stood out as the primary treatment modality. Conclusion: Our study presents a unique opportunity and point of view, exploring the evolving landscape of vestibular science publications over the past five decades. The analysis underscored the dynamic nature of the field, highlighting shifts in focus and emerging publication trends in diagnosis and treatment over time.

ChatGPT's adherence to otolaryngology clinical practice guidelines.

OBJECTIVES: Large language models, including ChatGPT, has the potential to transform the way we approach medical knowledge, yet accuracy in clinical topics is critical. Here we assessed ChatGPT's performance in adhering to the American Academy of Otolaryngology-Head and Neck Surgery guidelines. METHODS: We presented ChatGPT with 24 clinical otolaryngology questions based on the guidelines of the American Academy of Otolaryngology. This was done three times (N = 72) to test the model's consistency. Two otolaryngologists evaluated the responses for accuracy and relevance to the guidelines. Cohen's Kappa was used to measure evaluator agreement, and Cronbach's alpha assessed the consistency of ChatGPT's responses. RESULTS: The study revealed mixed results; 59.7% (43/72) of ChatGPT's responses were highly accurate, while only 2.8% (2/72) directly contradicted the guidelines. The model showed 100% accuracy in Head and Neck, but lower accuracy in Rhinology and Otology/Neurotology (66%), Laryngology (50%), and Pediatrics (8%). The model's responses were consistent in 17/24 (70.8%), with a Cronbach's alpha value of 0.87, indicating a reasonable consistency across tests. CONCLUSIONS: Using a guideline-based set of structured questions, ChatGPT demonstrates consistency but variable accuracy in otolaryngology. Its lower performance in some areas, especially Pediatrics, suggests that further rigorous evaluation is needed before considering real-world clinical use.

Virtual Reality Helps Describe the Progression of Thyroid Cartilage Calcification.

OBJECTIVES: Thyroid cartilage (TC) calcifications may impact surgical planning and clinical management. However, few studies to date have implemented virtual reality (VR) to evaluate these calcifications. This study assessed the feasibility of evaluating TC calcifications in various regions and measuring their volumes through VR models generated from computed tomography scans. We also investigated age and gender-related differences in calcification patterns. METHODS: Ninety-two participants were categorized into younger, middle-aged, and older age groups. Calcification patterns (degree in Hounsfield units and volume of calcification in cm3) in different TC regions were identified by VR analysis, which enabled comparisons between age groups and genders. RESULTS: Significant differences in calcification patterns were observed between males and females, particularly in the middle right, middle left, bottom left, and vertex regions. Age-related differences in the vertex region showed increased calcification in the older age group. CONCLUSION: This study points to the contribution of VR in the evaluation of complex anatomical structures. The findings revealed significant gender and age patterns in TC calcification. These insights can inform surgical planning and highlight the potential of using VR to gain a better understanding of TC calcification clinically.

Zoom in: factors affecting vocal habits during online meetings, a prospective trial on 40 subjects.

PURPOSE: To identify factors that influence vocal habits during online meetings (OMs). METHODS: A prospective trial of forty participants without any known hearing or vocal cord disorders. Subjects participated in an OM divided into six randomly ordered sections, with alterations in audio/speaking equipment and language: the computer's speaker-microphone, a single earbud, two-earbuds or headphones; with/without video, native-language-speaking (Hebrew) versus second language-speaking (English). Each section included free speech, sustained phonation, and a standardized passage. Participants ranked their vocal-effort for each section. Three blinded raters independently scored the voice using the GRBAS scale, and acoustic analyses were performed. RESULTS: No significant difference in self-reported vocal effort was demonstrated between sections. Second-language speaking resulted in significantly increased intensity (p < 0.0001), frequency (p = 0.015), GRBAS (p = 0.008), and strain (p < 0.0001) scores. Using the computer's speaker/microphone resulted in significantly higher strain (p < 0.0001). Using headphones, single or two earbuds resulted in lower intensity and a lower strain score. No differences were detected between OMs with or without video. CONCLUSIONS: Using the computer's microphone/speaker or speaking in a second language during OMs, may result in vocal habits associated with vocal trauma.

Advancing Medical Practice with Artificial Intelligence: ChatGPT in Healthcare.

BACKGROUND: Advancements in artificial intelligence (AI) and natural language processing (NLP) have led to the development of language models such as ChatGPT. These models have the potential to transform healthcare and medical research. However, understanding their applications and limitations is essential. OBJECTIVES: To present a view of ChatGPT research and to critically assess ChatGPT's role in medical writing and clinical environments. METHODS: We performed a literature review via the PubMed search engine from 20 November 2022, to 23 April 2023. The search terms included ChatGPT, OpenAI, and large language models. We included studies that focused on ChatGPT, explored its use or implications in medicine, and were original research articles. The selected studies were analyzed considering study design, NLP tasks, main findings, and limitations. RESULTS: Our study included 27 articles that examined ChatGPT's performance in various tasks and medical fields. These studies covered knowledge assessment, writing, and analysis tasks. While ChatGPT was found to be useful in tasks such as generating research ideas, aiding clinical reasoning, and streamlining workflows, limitations were also identified. These limitations included inaccuracies, inconsistencies, fictitious information, and limited knowledge, highlighting the need for further improvements. CONCLUSIONS: The review underscores ChatGPT's potential in various medical applications. Yet, it also points to limitations that require careful human oversight and responsible use to improve patient care, education, and decision-making.

Molecular Profiling for Bethesda III to VI Nodules: Results of a Multicenter International Retrospective Study.

OBJECTIVE: Molecular testing is a well-established tool that assists in the management of thyroid nodules. We describe our experience using molecular testing of thyroid nodules with Bethesda III to VI cytology. METHODS: This is a retrospective multicenter, multinational study of thyroid nodules that underwent preoperative molecular profiling with ThyGenX/ThyGeNEXT or ThyroSeq V3 between 2015 and 2022. The clinical characteristics and mutational profiles of tumors were compared. Collected data included demographics, cytology results, surgical pathology, and molecular alterations. Molecular alterations were categorized into 3 main phenotypes: BRAF-like, RAS-like, and non-BRAF-non-RAS (NBNR). RESULTS: Overall, 784 patients who had surgery were included, of which 603 (76.2%) were females. The most common histologic type was papillary thyroid cancer (PTC) with 727 (91.9%) cases. In total, 205 (28.2%) cases showed an aggressive subtype of PTC (eg, tall cell and hobnail). BRAF-like alterations were most likely to be found in Bethesda V and VI nodules and show extrathyroidal extension (ETE), nodal disease, and/or aggressive subtypes of PTC (P < .001 for all). RAS-like alterations were more commonly found in Bethesda III and IV nodules and were less likely to show ETE, nodal disease, and/or aggressive histology (P < .001 for all). NBNR alterations were more commonly found in Bethesda III and IV nodules and were less likely to show ETE, nodal disease, and/or aggressive subtypes of PTC. However, they were rarely but significantly associated with poorly differentiated thyroid cancer (P < .005). CONCLUSION: Molecular testing of thyroid nodules can help determine the likelihood of malignancy and classify nodules into several tumor phenotypes, predicting their behaviors and potentially allowing for a more tailored treatment. NBNR alterations should be managed with caution.

Deep learning in voice analysis for diagnosing vocal cord pathologies: a systematic review.

OBJECTIVES: With smartphones and wearable devices becoming ubiquitous, they offer an opportunity for large-scale voice sampling. This systematic review explores the application of deep learning models for the automated analysis of voice samples to detect vocal cord pathologies. METHODS: We conducted a systematic literature review following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) reporting guidelines. We searched MEDLINE and Embase databases for original publications on deep learning applications for diagnosing vocal cord pathologies between 2002 and 2022. Risk of bias was assessed using Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2). RESULTS: Out of the 14 studies that met the inclusion criteria, data from a total of 3037 patients were analyzed. All studies were retrospective. Deep learning applications targeted Reinke's edema, nodules, polyps, cysts, unilateral cord paralysis, and vocal fold cancer detection. Most pathologies had detection accuracy above 90%. Thirteen studies (93%) exhibited a high risk of bias and concerns about applicability. CONCLUSIONS: Technology holds promise for enhancing the screening and diagnosis of vocal cord pathologies. While current research is limited, the presented studies offer proof of concept for developing larger-scale solutions.

Paradigm Change for Intraoperative Surgical Margin Assessment for Oral Squamous Cell Carcinoma.

OBJECTIVE: Achieving clear surgical margins is one of the primary surgical goals in treating oral squamous cell carcinoma (OSCC) and thus aiming to improve overall and disease-specific survival. Therefore, we developed the Goal-Oriented Assessment for Intraoperative Margin ('GAIM') protocol, a novel intraoperative approach for margin assessment, and present here our 5-year experience and outcomes. METHODS: 'GAIM' is a 7-step procedure comprising systematic ruler-aided resection of labeled tumor-bed margins, frozen section (FS) co-produced by both pathologists and operating surgeons, and immediate extension of resection according to FS findings. Data from all patients operated using the 'GAIM' protocol at a single tertiary center between 2018 to 2022 were analyzed, including margin status on FS and final pathology (FP) records, recurrence, and mortality. RESULTS: A total of 196 patients were included, 56.6% (n = 111) stages I-II, and 43.4% (n = 85) stages III-IV. Using the 'GAIM' protocol, we achieved an overall 94.4% of clean and revised clean surgical margins. Patients with a 2-year and longer follow-up (n = 141) had local recurrence in 3.5% when both FS and final margins were clean, 8.1% when FP margins were clean, and 16.7% with close/positive final margins. CONCLUSIONS: The proposed 'GAIM' protocol is a novel, effective, reproducible, and safe approach for margin evaluation that can be systematically applied. It can increase the rate of final clean surgical margins and potentially improve patients' outcomes. LEVEL OF EVIDENCE: 3 Laryngoscope, 134:1725-1732, 2024.

A Five-Decade Text Mining Analysis of Cochlear Implant Research: Where We Started and Where We Are Heading.

Background and Objectives: Since its invention in the 1970s, the cochlear implant (CI) has been substantially developed. We aimed to assess the trends in the published literature to characterize CI. Materials and Methods: We queried PubMed for all CI-related entries published during 1970-2022. The following data were extracted: year of publication, publishing journal, title, keywords, and abstract text. Search terms belonged to the patient's age group, etiology for hearing loss, indications for CI, and surgical methodological advancement. Annual trends of publications were plotted. The slopes of publication trends were calculated by fitting regression lines to the yearly number of publications. Results: Overall, 19,428 CIs articles were identified. Pediatric-related CI was the most dominant sub-population among the age groups, with the highest rate and slope during the years (slope 5.2 ± 0.3, p < 0.001), while elderly-related CIs had significantly fewer publications. Entries concerning hearing preservation showed the sharpest rise among the methods, from no entries in 1980 to 46 entries in 2021 (slope 1.7 ± 0.2, p < 0.001). Entries concerning robotic surgery emerged in 2000, with a sharp increase in recent years (slope 0.5 ± 0.1, p < 0.001). Drug-eluting electrodes and CI under local-anesthesia have been reported only in the past five years, with a gradual rise. Conclusions: Publications regarding CI among pediatrics outnumbered all other indications, supporting the rising, pivotal role of CI in the rehabilitation of children with sensorineural hearing loss. Hearing-preservation publications have recently rapidly risen, identified as the primary trend of the current era, followed by a sharp rise of robotic surgery that is evolving and could define the next revolution.

Novel Association of the NOTCH Pathway Regulator MIB1 Gene With the Development of Bicuspid Aortic Valve.

Importance: Nonsyndromic bicuspid aortic valve (nsBAV) is the most common congenital heart valve malformation. BAV has a heritable component, yet only a few causative genes have been identified; understanding BAV genetics is a key point in developing personalized medicine. Objective: To identify a new gene for nsBAV. Design, Setting, and Participants: This was a comprehensive, multicenter, genetic association study based on candidate gene prioritization in a familial cohort followed by rare and common association studies in replication cohorts. Further validation was done using in vivo mice models. Study data were analyzed from October 2019 to October 2022. Three cohorts of patients with BAV were included in the study: (1) the discovery cohort was a large cohort of inherited cases from 29 pedigrees of French and Israeli origin; (2) the replication cohort 1 for rare variants included unrelated sporadic cases from various European ancestries; and (3) replication cohort 2 was a second validation cohort for common variants in unrelated sporadic cases from Europe and the US. Main Outcomes and Measures: To identify a candidate gene for nsBAV through analysis of familial cases exome sequencing and gene prioritization tools. Replication cohort 1 was searched for rare and predicted deleterious variants and genetic association. Replication cohort 2 was used to investigate the association of common variants with BAV. Results: A total of 938 patients with BAV were included in this study: 69 (7.4%) in the discovery cohort, 417 (44.5%) in replication cohort 1, and 452 (48.2%) in replication cohort 2. A novel human nsBAV gene, MINDBOMB1 homologue MIB1, was identified. MINDBOMB1 homologue (MIB1) is an E3-ubiquitin ligase essential for NOTCH-signal activation during heart development. In approximately 2% of nsBAV index cases from the discovery and replication 1 cohorts, rare MIB1 variants were detected, predicted to be damaging, and were significantly enriched compared with population-based controls (2% cases vs 0.9% controls; P = .03). In replication cohort 2, MIB1 risk haplotypes significantly associated with nsBAV were identified (permutation test, 1000 repeats; P = .02). Two genetically modified mice models carrying Mib1 variants identified in our cohort showed BAV on a NOTCH1-sensitized genetic background. Conclusions and Relevance: This genetic association study identified the MIB1 gene as associated with nsBAV. This underscores the crucial role of the NOTCH pathway in the pathophysiology of BAV and its potential as a target for future diagnostic and therapeutic intervention.

Investigation of genetic sex-specific molecular profile in well-differentiated thyroid cancer: Is there a difference between females and males?

BACKGROUND: Although more common in females, thyroid cancer is deemed to be more aggressive in males. The reasons for sex disparities in thyroid cancer are not well understood. We hypothesised that differences in molecular mutations between females and males contribute to this phenomenon. METHODS: Retrospective multicentre multinational study of thyroid nodules that underwent preoperative molecular profiling between 2015 and 2022. The clinical characteristics and mutational profiles of tumours in female and male patients were compared. Collected data included demographics, cytology results, surgical pathology, and molecular alterations. RESULTS: A total of 738 patients were included of which 571 (77.4%) were females. The extrathyroidal extension was more common in malignancies in males (chi-squared, p = 0.028). The rate of point mutations and gene fusions were similar in both sex groups (p > 0.05 for all mutations). Patients with nodules with BRAFV600E mutations were significantly younger than BRAF wild-type nodule patients (t-test, p = 0.0001). Conversely, patients with TERT promoter mutations were significantly older than patients with wild-type TERT (t-test, p < 0.0001). For patients harbouring both BRAFV600E and TERT mutations, the difference in age at presentation was significantly different in females (t-test, p = 0.009) but not in males (t-test, p = 0.433). Among females, patients with BRAFV600E and TERT mutations were significantly older than their wild-type or single-mutation counterpart (t-test, p = 0.003). CONCLUSION: The absolute rate of molecular mutations was similar in females and males. We found that extrathyroidal extension was more common in males. Moreover, BRAFV600E and TERT mutations occur at a younger age in males than in females. These two findings are factors that may explain the tendency of more aggressive disease in males.

In-office KTP laser treatment for laryngeal pathologies: 5-years outcomes.

INTRODUCTION: In the past two decades, laser systems were introduced into the office setting for laryngeal pathologies, offering the advantages of a shorter procedure and recovery. To date, long-term data on outcomes is limited. This study aims to evaluate the efficacy and safety of the office-based potassium-titanyl-phosphate (KTP) laser procedure for laryngeal pathologies. METHODS: A retrospective cohort of in-office KTP laser procedures for two main vocal folds lesions groups: (i) benign and pre-malignant; and (ii) intraepithelial lesions in a prior invasive cancer field between 2010 and 2020. Data were collected from electronic medical records, telephone interviews, and video documentation of the procedure, including treatment completion, disease control, and whether additional interventions were required. RESULTS: A total of 81 patients underwent 153 in-office KTP laser procedures for benign (36, 44.4%), pre-malignant (15, 18.5%), and lesions in a prior malignancy field (30, 37.1%) with a mean of 1.89 ± 1.81 procedures per patient. One hundred and thirty-eight (90.2%) procedures were well tolerated and completed successfully. During the 5-years of follow-up, 63% of the patients with previous malignancy were managed exclusively in the office. In the pre-malignant group, 76% required no additional type of intervention. Patients with papilloma required significantly more procedures per patient compared with other pathologies (3.6 ± 4 vs. 1.61 ± 1, p-value = 0.02). Surgery was required only in 18.2% of the papilloma patients. Three (1.9%) patients had short-term complications, all resolved within 6 months. Failure to complete the procedure was significantly associated with active smoking (p-value < 0.001) and, in most cases (90%), was related to patient intolerance. CONCLUSION: Office-based KTP laser laryngeal procedures have shown promising results for both benign and selected cases of lesions in a prior malignancy field with a high compliance and a very low complication rate, suggesting its use as an effective and safe treatment modality for selected patients.

Is There a Role for Molecular Testing for Low-Risk Differentiated Thyroid Cancer? A Cost-Effectiveness Analysis.

Molecular testing for thyroid nodules has been rapidly developed in recent years, aiming to predict the presence of malignancy and aggressive features. While commonly utilized to predict malignancy, its role in guiding the management approach is still developing. The high cost of genetic tests and long-term sequences of thyroid cancer is limiting to real-life studies. Objective: To evaluate the cost effectiveness of molecular testing for low-risk differentiated thyroid cancer (lrDTC). Methods: We developed a Markovian decision tree model of a simulated lrDTC cohort, comparing two management strategies: (I) Conducting genetic tests (GT)-patients are stratified into three risk groups for distant metastasis by the identified molecular markers: low-, intermediate- and high-risk molecular profile; followed by management accordingly: patients with low-risk will undergo hemithyroidectomy (HT), patients with intermediate-risk will undergo total thyroidectomy (TT), and high-risk patients will undergo TT with central neck dissection; (II) Without genetic tests (wGT)-all patients will undergo HT according to the ATA recommendations for lrDTC. Outcomes were measured as quality-adjusted life years (QALYs) and costs of each strategy. Results: GT was found as cost effective, leading to a gain of 1.7 QALYs with an additional cost of $327 per patient compared to wGT strategy. This yielded an incremental cost-effectiveness ratio of $190 per QALY. Sensitivity analysis demonstrated robust results across the variables' ranges. The most impactful variable was the benefit from performing TT rather than HT for intermediate to high-risk patients. Conclusions: Our model found that molecular testing for lrDTC is cost-effective, allowing tailored management according to the patient's personal risk level reflected in the genetic profile, hence improving outcomes.

Mitral valve prolapse: From new mechanisms to diagnostic challenges.

Mitral valve prolapse (MVP) is the most common primary valvular abnormality, associated with various degrees of incompetent function and sequelae, including heart failure and sudden cardiac death. Recent improvements in echocardiographic techniques and new insights into mitral valve anatomy and physiology have rendered the diagnosis of this condition more accurate and reliable. Here we review the genetic etiology, clinical significance, diagnosis, and treatment options for MVP patients.

Sino-nasal hemangiopericytoma: a case series and systematic literature review.

BACKGROUND: Hemangiopericytoma is a rare tumor of the sino-nasal tract. Its clinical behavior is controversial. Whereas some describe an indolent course, others consider it to be an aggressive lesion with a tendency toward rapid local recurrence. Here, we describe our experience in the management of sino-nasal hemangiopericytoma (SN-HPC), comparing our experience with the current literature, and evaluating signs and tools to improve diagnosis and treatment. METHODS: All cases of SN-HPC between 2010 and 2020 were extracted and reviewed from our institutional electronic medical records. SN-HPC cases from PubMed and EMBASE between 2010 and 2020 were analyzed in a systematic literature review using the preferred reporting items for systematic review and meta-analysis (PRISMA) guidelines. Data regarding demographics, presentation, diagnosis, treatment, and outcome were collected. RESULTS: We identified four cases of SN-HPC in the nasal cavity in our institution and an additional 53 cases in previous reports. The mean age at the time of diagnosis was 59 years, with a 1.2:1 male to female ratio. SN-HPC mostly appears unilaterally, arising in the ethmoid sinus (42.1%). The most common presenting symptoms were epistaxis (47.3) and nasal obstruction (47.3%). Both computed tomography (CT) and magnetic resonance imaging (MRI) were required for diagnosis and for tailoring the treatment plan. Endoscopic surgical excision was used in 85.9% of the patients, and in 15.7%, an additional preoperative embolization was performed, which was associated with septal necrosis in one patient (2.6%). The recurrence rate was 7%. CONCLUSION: Although previous reports attribute an aggressive tumoral behavior to SN-HPC, our experience and the literature review support a more indolent course with low recurrence rates following complete endoscopic resection. Preoperative embolization can be useful in certain cases, but due to potential complications, it should not be routinely indicated.

Outcomes of the Bethesda system for reporting thyroid cytopathology in community- vs. institution-performed cytology.

OBJECTIVES: Despite the important role of the community in the assessment and risk stratification of patients with thyroid nodules, evidence-based data on the Bethesda System for Reporting Thyroid Cytopathology (BSRTC) performance in community clinics is lacking. We aim to evaluate BSRTC performance of aspirations taken in community clinics compared with primary referral center. METHODS: Patients who underwent thyroid surgery between 2013 and 2018 at our institution were divided according to the fine needle aspirations (FNA) settings: community FNA (cFNA) vs. institutional FNA (iFNA). Demographics, BSRTC results and final pathology were collected. Diagnostic values were calculated for BSRTC categories (sensitivity, specificity, positive predictive value [PPV], and negative predictive value [NPV]), and were compared between the groups. RESULTS: A total of 268 nodules were included in the study; 77% (207) cFNA and 23% (61) iFNA. Patients in the community were younger (51.7 ± 15.1 vs. 56.6 years±14.8, p = 0.03) and with less epidemiology risk factors for thyroid cancer (1.9% vs 13.1%, p < 0.001). cFNA malignancy rate for BSRTC I-VI was 0%, 6.4%, 11.8%, 32.1%, 91.6% and 93.8% respectively. Best sensitivity was found for BSRTC III-VI in both groups (88% and 83%, cFNAs and iFNAs, respectively). Overall best performance was obtained for BSRTC V-VI for both groups (cfNA: 85%, 97%, 93%, 94% and 93%; iFNAs: 81%, 100%, 100%, 87% and 91%, for sensitivity, specificity, PPV, NPV and accuracy, respectively). CONCLUSIONS: Community-performed FNAs demonstrate acceptable BSRTC distribution and malignancy rates, comparable with a primary referral academic hospital. This supports the universality of the BSRTC 2017 and its recommendations also in the community.

Is Bicuspid Aortic Valve Morphology Genetically Determined? A Family-Based Study.

Bicuspid aortic valve (BAV) is a common congenital heart disease, with a 10-fold higher prevalence in first-degree relatives. BAV has different phenotypes based on the morphology of cusp fusion. These phenotypes are associated with different clinical courses and prognoses. Currently, the determinants of the valve phenotype are unknown. In this study we evaluated the role of genetics using familial cohorts. Patients with BAV and their first-degree relatives were evaluated by echocardiography. The concordance in BAV phenotype between pairs of family members was calculated and compared with the concordance expected by chance. We then performed a systematic literature review to identify additional reports and calculated the overall concordance rate. During the study period, 70 cases from 31 families and 327 sporadic cases were identified. BAV was diagnosed in 14% of the screened relatives. The proportions of the morphologies identified was: 12.3% for type 0, 66.2% for type 1-LR, 15.4% for type 1-RN, 4.6% for type 1-NL, and 1.5% for type 2. For the assessment of morphologic concordance, we included 120 pairs of first-degree relatives with BAV from our original cohort and the literature review. Concordance was found only in 62% of the pairs which was not significantly higher than expected by chance. In conclusion, our finding demonstrates intrafamilial variability in BAV morphology, suggesting that morphology is determined by factors other than Mendelian genetics. As prognosis differs by morphology, our findings may suggest that clinical outcomes may vary even between first-degree relatives.

Validating the 'CUT score' risk stratification tool for indeterminate thyroid nodules using the Bethesda system for reporting thyroid cytopathology.

PURPOSE: Managing intermediate thyroid nodules remains challenging. The CUT score is an Italian metanalysis-based cytologic (SIAPEC-IAP) scoring system, designed to assist clinicians. However, it was never evaluated against the Bethesda system for reporting thyroid cytopathology (BSRTC). This study aims to validate its utility for BSRTC III and IV nodules in a non-Italian population. METHODS: We collected all BSRTC III and IV thyroid nodules with a documented final pathology between 2010 and 2020. We calculated the C + U components of the CUT score using retrospective clinical (C) data collection and reevaluation of preoperative sonography (U) examination. The cytology (T) component which originally referred to the five-tiered SIAPEC-IAP cytologic classification was replaced by the corresponding BSRTC categories. Optimal test performances were calculated using receiver operating characteristic (ROC) curve analysis. Data were analyzed twice with considering of NIFTP as benign and as malignant. RESULTS: After exclusions, 62 nodules from 61 patients were included (50% BSRTC III, 50% BSRTC IV). Malignant nodules demonstrated a significantly higher C + U score compared with benign in both categories. The C + U cutoff value for BSRTC III was 5.25 (sensitivity and specificity of 69.23% and 66.67%, respectively, AUC = 0.72, p-value = 0.016), and 5.75 for BSRTC IV (sensitivity and specificity of 85.7% and 76.5%, respectively, AUC = 0.84, p-value < 0.001). CONCLUSION: Our study suggests that the CUT score is applicable for both BSRTC III and IV nodules, and highlights the need for internal validations, since the cutoffs found were higher than previously reported.