ChatGPT vs. web search for patient questions: what does ChatGPT do better?

PURPOSE: Chat generative pretrained transformer (ChatGPT) has the potential to significantly impact how patients acquire medical information online. Here, we characterize the readability and appropriateness of ChatGPT responses to a range of patient questions compared to results from traditional web searches. METHODS: Patient questions related to the published Clinical Practice Guidelines by the American Academy of Otolaryngology-Head and Neck Surgery were sourced from existing online posts. Questions were categorized using a modified Rothwell classification system into (1) fact, (2) policy, and (3) diagnosis and recommendations. These were queried using ChatGPT and traditional web search. All results were evaluated on readability (Flesch Reading Ease and Flesch-Kinkaid Grade Level) and understandability (Patient Education Materials Assessment Tool). Accuracy was assessed by two blinded clinical evaluators using a three-point ordinal scale. RESULTS: 54 questions were organized into fact (37.0%), policy (37.0%), and diagnosis (25.8%). The average readability for ChatGPT responses was lower than traditional web search (FRE: 42.3 ± 13.1 vs. 55.6 ± 10.5, p < 0.001), while the PEMAT understandability was equivalent (93.8% vs. 93.5%, p = 0.17). ChatGPT scored higher than web search for questions the 'Diagnosis' category (p < 0.01); there was no difference in questions categorized as 'Fact' (p = 0.15) or 'Policy' (p = 0.22). Additional prompting improved ChatGPT response readability (FRE 55.6 ± 13.6, p < 0.01). CONCLUSIONS: ChatGPT outperforms web search in answering patient questions related to symptom-based diagnoses and is equivalent in providing medical facts and established policy. Appropriate prompting can further improve readability while maintaining accuracy. Further patient education is needed to relay the benefits and limitations of this technology as a source of medial information.

Public perceptions of radiofrequency ablation versus standard surgery for benign thyroid nodules.

BACKGROUND: Ultrasound-guided radiofrequency ablation has shown promising results for nonsurgical treatment of benign thyroid nodules. The purpose of this study is to investigate public perceptions of radiofrequency ablation and identify salient decision factors for benign thyroid nodules treatment. METHODS: An internet-based survey was distributed via an online platform. Survey participants were prompted to envision having a benign thyroid nodules and were assessed on risk acceptance, willingness to pay, and importance of decision factors (eg, cost, risk, scarring) regarding treatment with either radiofrequency ablation or standard surgery. RESULTS: A total of 830 respondents (male 46.3%, median age 35 years) were included. Respondents ranked the most important factors for benign thyroid nodules treatment as risk of missing cancer and risk of permanent hoarseness (mean score 5.23 and 4.50 out of 7, respectively). Female respondents ranked missing cancer higher in importance compared with other decision factors (coefficient = 0.251, P = .0002). Younger respondents (coefficient = 0.009, P = .014) or those with higher education levels (coefficient = 0.092, P = .010) ranked having a scar as a more important decision factor. Prior scars were associated with choosing a nonsurgical procedure over surgery (coefficient = 0.478, P = .00), even if scars were well-healed (coefficient = 0.781, P < .0001). On average, respondents are willing to pay less for radiofrequency ablation than for standard surgery (radiofrequency ablation: $7,612.44 vs surgery: $8,298.50; P < .0001). CONCLUSION: Respondents identified risk of missing cancer and risk of permanent hoarseness as the most important decision factors for benign thyroid nodules treatment. Education level and history of previous scars are both associated with a proclivity toward nonsurgical treatment and radiofrequency ablation.

Understanding Public Perceptions Regarding Cochlear Implant Surgery in Adults.

BACKGROUND AND OBJECTIVE: Approximately 6% of adults eligible for cochlear implantation (CI) undergo surgery. This study aims to understand how general perceptions about CI contribute to barriers causing this low utilization rate. METHODS: Participants completed an online survey regarding their perceptions about cochlear implantation. They were asked to rank CIQOL-10 Global priorities and corresponding risk tolerance for minor complications (changes in taste, vertigo) and major complications (infections requiring hospitalization, meningitis, reimplantation, facial paralysis, and cerebrospinal fluid [CSF] leak). RESULTS: A total of 615 responses (female 55%, mean age 39 years [range 20-78]) were included. Respondents identified issues with insurance (46%) and fear of undergoing surgery (21%) as barriers preventing eligible adults from receiving CI. Regarding surgical risk, respondents significantly underestimated rates of minor complications (p < 0.0001) and almost all major complications (all p < 0.0001) while overestimating rates of serious infections and need for reimplantation (both p < 0.005). The ability to hear strangers in noisy environments was identified as the highest priority for CI (26%). Individuals who knew someone with a cochlear implantation were willing to tolerate higher rates of all minor and major complications (all p < 0.05). Healthcare workers were similarly willing to tolerate significantly higher rates of meningitis, facial paralysis, and CSF leak to achieve their highest priority (all p < 0.05). CONCLUSIONS: Respondents identify insurance coverage and fear of surgery as primary reasons for low utilization of adult CI in the United States. However, healthcare workers or those who knew someone with a cochlear implantation indicated significantly higher rates of risk tolerance to achieve their highest quality of life priorities, indicating the potential that the benefits of cochlear implantation are underappreciated in the general population.

Psychophysical chemosensory dysfunction in eating disorders: a qualitative systematic review.

PURPOSE: Patients with chemosensory dysfunction experience significant quality of life disruptions, including reduced enjoyment of eating. While chemosensory dysfunction has been associated with eating disorders, the relationship is poorly understood. This systematic review aims to characterize psychophysical gustation and olfaction in patients with eating disorders. METHODS: Systematic review of investigations assessing psychophysical chemosensory function in patients with organic eating disorders. RESULTS: 26 studies were included. Five studies assessed both chemosenses, while 12 and 9 assessed exclusively gustation or olfaction, respectively. In total, 779 patients were included [72.4% anorexia nervosa (AN), 26.7% bulimia nervosa (BN), 0.8% combined AN/BN]. Patients with eating disorders experienced rates of hypogeusia up to 87% in AN and 84.6% in BN. There was evidence for alterations in psychophysical olfaction, but orientation of trends were less clear. Chemosensory dysfunction was more evident in AN patients. Treatment correlated with improved chemosensory function. CONCLUSIONS: Despite heterogeneity in study methodology and results, this review demonstrates that patients with eating disorders experience some degree of chemosensory dysfunction, particularly in gustation. This symptomatology overlaps with those experienced by patients with other causes of chemosensory impairment. These findings suggest potential broad psychosocial, dietary, and mental health implications in patient populations experiencing chemosensory dysfunction. LEVEL OF EVIDENCE: Level II.

Audience of Academic Otolaryngology on Twitter: Cross-sectional Study.

BACKGROUND: Despite the ubiquity of social media, the utilization and audience reach of this communication method by otolaryngology-head and neck surgery (OHNS) residency programs has not been investigated. OBJECTIVE: The purpose of this study was to evaluate the content posted to a popular social media platform (Twitter) by OHNS residency programs. METHODS: In this cross-sectional study, we identified Twitter accounts for accredited academic OHNS residency programs. Tweets published over a 6-month period (March to August 2019) were extracted. Tweets were categorized and analyzed for source (original versus retweet) and target audience (medical versus layman). A random sample of 100 tweets was used to identify patterns of content, which were then used to categorize additional tweets. We quantified the total number of likes or retweets by health care professionals. RESULTS: Of the 121 accredited programs, 35 (28.9%) had Twitter accounts. Of the 2526 tweets in the 6-month period, 1695 (67.10%) were original-content tweets. The majority of tweets (1283/1695, 75.69%) were targeted toward health care workers, most of which did not directly contain medical information (954/1283, 74.36%). These tweets contained information about the department's trainees and education (349/954, 36.6%), participation at conferences (263/954, 27.6%), and research publications (112/954, 11.7%). Two-thirds of all tweets did not contain medical information. Medical professionals accounted for 1249/1362 (91.70%) of retweets and 5616/6372 (88.14%) of likes on original-content tweets. CONCLUSIONS: The majority of Twitter usage by OHNS residency programs is for intra and interprofessional communication, and only a minority of tweets contain information geared toward the public. Communication and information sharing with patients is not the focus of OHNS departments on Twitter.

Melanoma metastatic to the hyoid bone.

Metastatic melanoma may be included in the differential diagnosis of hyoid masses in patients with a history of melanoma. Hyoid resection is well tolerated and of diagnostic and therapeutic benefit in patients with tumors metastatic to the hyoid bone.

OHNS Residency Program and Applicant Social Media Presence During the COVID-19 Pandemic.

OBJECTIVES: In addition to clinical and social disruption, the Coronavirus Disease 2019 (COVID-19) pandemic has affected many aspects of the otolaryngology residency application process. With delays in the 2021 Electronic Residency Applications Service (ERAS) timeline, students and programs have had more time to interact prior to the formal application process. This communication will report recent trends in social media presence by OHNS residency programs, and discuss mechanisms to compensate for decreased applicant-program interactions using social media ahead of the 2021 Match. METHODS: In a cross-sectional study of the accredited otolaryngology residency programs in the United States, the number of social media profiles on Twitter, Instagram, and Facebook from 2009 to 2019 were recorded and compared. RESULTS: Most programs (61%) have at least 1 social media profile. Over the past 10 years, the number of programs on social media has increased. During the COVID-19 pandemic, Twitter and Instagram showed higher rates of growth compared to Facebook. With the reduction of in-person opportunities for interactions, both applicants and programs are utilizing social media to showcase their values and their research. Twitter, in particular, also serves as a platform for professional networking. CONCLUSION: Both Twitter and Instagram are growing in popularity among programs and applicants to enhance networking. Social media is a powerful tool for networking and may help compensate for limitations imposed on the residency match process by the COVID-19 pandemic while maintaining professionalism considerations. The impact of social media on the 2021 otolaryngology residency match is an evolving phenomenon.

A Systematic Review and Meta-Analysis of Taste Dysfunction in Chronic Rhinosinusitis.

OBJECTIVES: Patients with chronic rhinosinusitis (CRS) often describe alterations in sense of taste. These complaints have historically been attributed to olfactory dysfunction; however, there is evidence of direct, objective, gustatory disturbances in the setting of CRS that are not thoroughly characterized. This study sought to investigate and summarize gustatory dysfunction experienced by patients with CRS. METHODS: PubMed, EMBASE, Cochrane Library, Web of Science, and Scopus databases were reviewed following PRISMA guidelines. English language, original studies investigating objective taste in adult patients with CRS were included. A meta-analysis with inverse variance, random-effects model was performed. RESULTS: Of 2750 studies screened, 11 articles with 471 unique patients were included. Patients with CRS exhibit worse gustatory function compared to healthy controls (standardized mean difference 0.94 [95% CI, 0.44-1.45]). Hypogeusia was identified in 32/95 (33.7%) patients from three studies that used methods with a validated definition of hypogeusia. Older age, male gender, and smoking history were associated with taste dysfunction, while objective gustatory and olfactory dysfunction were not correlated. Subjective taste and quality of life measures were also not associated with objective taste. The impact of sinus surgery on objective taste is unclear. CONCLUSION: Approximately 34% of patients with CRS experience hypogeusia. Neither olfactory function nor subjective taste were associated with objective gustatory function. Given the substantial prevalence of taste dysfunction patients with CRS, there is significant potential for growth in understanding of pathogenesis, impact on quality of life, and potential treatment strategies of taste impairment in the CRS patient population. LEVEL OF EVIDENCE: 1 Laryngoscope, 131:482-489, 2021.

Treatment of post-viral olfactory dysfunction: an evidence-based review with recommendations.

BACKGROUND: Post-viral olfactory dysfunction (PVOD) is one of the most common causes of olfactory loss. Despite its prevalence, optimal treatment strategies remain unclear. This article provides a comprehensive review of PVOD treatment options and provides evidence-based recommendations for their use. METHODS: A systematic review of the Medline, Embase, Cochrane, Web of Science, Scopus, and Google Scholar databases was completed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Studies with defined olfactory outcomes of patients treated for PVOD following medical, surgical, acupuncture, or olfactory training interventions were included. The Clinical Practice Guideline Development Manual and Conference on Guideline Standardization (COGS) instrument recommendations were followed in accordance with a previously described, rigorous, iterative process to create an evidence-based review with recommendations. RESULTS: From 552 initial candidate articles, 36 studies with data for 2183 patients with PVOD were ultimately included. The most common method to assess olfactory outcomes was Sniffin' Sticks. Broad treatment categories included: olfactory training, systemic steroids, topical therapies, a variety of heterogeneous non-steroidal oral medications, and acupuncture. CONCLUSION: Based on the available evidence, olfactory training is a recommendation for the treatment of PVOD. The use of short-term systemic and/or topical steroids is an option in select patients after careful consideration of potential risks of oral steroids. Though some pharmacological investigations offer promising preliminary results for systemic and topical medications alike, a paucity of high-quality studies limits the ability to make meaningful evidence-based recommendations for the use of these therapies for the treatment of PVOD.

Laryngotracheal Reconstruction in Adults Aged 60 Years and Older.

OBJECTIVE: The primary aim of this study is to evaluate the safety, efficacy, and execution of major open laryngotracheal operations for patients in the advanced decades. STUDY DESIGN: Case series with chart review. SETTING: Multidisciplinary clinic at a tertiary care academic hospital. SUBJECTS AND METHODS: Patient characteristics, operative course, and postoperative outcomes were retrospectively recorded for all airway reconstruction operations performed between 1999 and 2016 on patients aged ≥60 years Long-term success was defined as prosthesis-free survival at last follow-up. Descriptive statistics were performed. RESULTS: Twenty-nine patients met inclusion criteria, and the median age was 71 years (interquartile range, 63-74). Tracheal resection was the most common procedure (13 patients), followed by laryngotracheal reconstruction (7 patients). Fifteen patients began their operation with a tracheostomy, 6 of whom underwent decannulation prior to leaving the operating room. Three additional patients underwent decannulation at follow-up appointments and were prosthesis-free at most recent follow-up. The mean time to decannulation among these patients was 3 months. Of the 14 patients beginning their procedure without a tracheostomy, only 2 required permanent airway prosthesis. The overall long-term rate of prosthesis-free survival was 72.4% (21 of 29 patients). Factors suggestive of long-term success include lower McCaffrey grade and lack of pulmonary disease, hypertension, or diabetes, as well as decreased red blood cell distribution width on preoperative complete blood count. CONCLUSION: Through careful patient selection, preoperative workup, and meticulous postoperative care, airway reconstruction procedures in patients aged ≥60 years are reasonably successful. Of 29 patients, 21 (72.4%) were successfully breathing long-term without airway prosthesis.

Association Between Red Blood Cell Distribution Width and Outcomes of Open Airway Reconstruction Surgery in Adults.

Importance: Airway reconstruction for adults with laryngotracheal stenosis (LTS) is directed toward improving airway caliber to mitigate the patient's dyspnea and achieve prosthesis-free breathing (ie, without tracheostomy, intraluminal stent, or T-tube). Despite the importance of preoperative risk stratification to minimize postoperative complications, consensus on an objective predictive algorithm for open airway reconstruction is lacking. Objective: To determine whether the ability to achieve a prosthesis-free airway in adults after open airway reconstruction is associated with red blood cell distribution width (RDW) at the time of surgery. Design, Setting, and Participants: Case series study investigating 92 consecutive patients 18 years and older with laryngotracheal stenosis who underwent open airway reconstruction at a US tertiary care hospital from January 1, 2006, to January 1, 2017. Main Outcomes and Measures: The main outcome was a prosthesis-free airway (absence of tracheostomy, intraluminal stent, or T-tubes) at last follow-up. Multivariate logistic regression modeling was used to identify independent factors associated with this outcome. Results: Of the 92 patients who met inclusion criteria, the median (interquartile range) age was 44 (33.0-60.3) years; 50 (53%) were female, and 82 (89%) were white. In all, 74 patients (80%) were prosthesis free at the last follow-up (mean, 833 days; 95% CI, 10-4229 days). In multivariate analyses, airway decannulation was significantly correlated with reduced RDW (odds ratio [OR], 0.40; 95% CI, 0.19-0.84) and the absence of posterior glottic stenosis (OR, 0.12; 95% CI, 0.04-0.37). Conclusions and Relevance: These data suggest that surgical success in open airway reconstruction is significantly associated with RDW and whether the patient had posterior glottic stenosis. The RDW is a routine laboratory parameter that may provide some insight to the preoperative probability of prosthesis removal, facilitate risk stratification, promote informed patient decision making, and optimize health care resource management.

Evaluation of Social Media Presence of Otolaryngology Residency Programs in the United States.

Importance: Over two-thirds of the adult population in the United States use Facebook. Despite the high interest in and use of social media by the general public, the presence and accessibility of health care organizations on social media has not yet been fully evaluated. Objective: To determine the use and popularity of social media among otolaryngology residency programs in the United States. Design, Setting, and Participants: A cross-sectional study of the presence of accredited otolaryngology residency programs in the United States in an internet data repository was conducted. Programs were stratified by Doximity Residency Navigator reputation rankings (dividing programs into quartiles) and US News & World Report (comparing programs affiliated with hospitals ranked in the top 50 vs programs affiliated with unranked hospitals). Social media sites and activity for each program were assessed using internet searches. The study was conducted in April 2017. Results: Among 101 otolaryngology residency programs, 30 were found to have social media sites (29.7%). Facebook and Twitter were the most commonly used services, with 25 (24.8%) and 14 (13.9%) accounts, respectively. Based on Doximity Residency Navigator rankings, programs in the first quartile were more likely to have Facebook profiles than programs in the fourth quartile (42.3% vs 12.0%; absolute difference, 30%; 95% CI, 2.9% to 52.6%). First- and second-quartile programs showed increased Facebook activity. There was greater Twitter presence in first- vs fourth-quartile programs (19.2% vs 8.0%; absolute difference, 11.2%; 95% CI, -11.6% to 33.0%). Higher-quartile programs were more active on Twitter and exhibited increased numbers of likes and followers. Analysis of US News & World Report rankings revealed that ranked programs had higher rates of presence, activity, and popularity on both Facebook and Twitter. However, these were smaller differences than seen when comparing Doximity Residency Navigator rankings. Correlation between the 2 ranking systems was indicated (Spearman ρ = 0.59; 95% CI, 0.34 to 0.76). Conclusions and Relevance: This study suggests that otolaryngology residency programs with higher Doximity Residency Navigator reputation rankings have a stronger presence on social media. Smaller trends were observed for programs in the top 50 US News & World Report rankings. Overall, social media use among otolaryngology programs seems relatively low, and this may present an opportunity to increase communication with the public via these technologies.

Understanding sociodemographic factors related to health outcomes in pediatric obstructive sleep apnea.

OBJECTIVES: (1) To examine relationships between socioeconomic status (SES) and successful treatment of pediatric obstructive sleep apnea (OSA) with adenotonsillectomy (T&A). (2) To explore sociodemographic factors and medical comorbidities that separate OSA and refractory OSA populations in children. METHODS: We retrospectively reviewed pediatric OSA patients (ages 0-18). Patients evaluated for OSA by pediatric otolaryngology between January 2014 and December 2015 were included. OSA was defined as requiring T&A. Refractory OSA (ROSA) was defined as recurring, polysomnography-proven, OSA after T&A, ultimately requiring another intervention, such as a multi-level airway operation. Clinical data were complemented with sociodemographic data. ZIP codes were used to approximate median household income. RESULTS: Our cohort included 105 ROSA and 53 OSA patients. These patients came from similar rates of single parent households and coverage by public insurance. Median household income for OSA patients was $47,086 (IQR $36,395-$60,196), compared to $45,696 (IQR $37,669-$56,203) for ROSA patients. Over 60% of all patients fell below the national household income average. Nearly half of the cohort resided in the three largest metro counties closest to our institution. These patients represented higher rates of single-parent households (p = 0.045) and public insurance (p = 0.002), and trends towards lower rates of ROSA (p = 0.138). CONCLUSION: Our results identified sociodemographic factors that may influence healthcare compliance and subsequently overall health outcomes. We identified no statistically significant difference in measures of SES between patients with refractory vs non-refractory OSA. Patients living closest to our medical center had lowest rates of ROSA, suggesting that access to care may affect outcomes of pediatric OSA.

Endoscopic Ear Surgery Skills Training Improves Medical Student Performance.

OBJECTIVE: Analyze medical student performance using an endoscopic ear surgery (EES) skills trainer over time. DESIGN: Simulation experience. SETTING: Surgical skills lab. PARTICIPANTS: Five medical students and 1 expert surgeon completed 3 training sessions using an EES trainer to practice specific tasks designed to improve instrument control. During each session, participants performed 3 tasks, such as placing beads onto wires, 5 times per session for a total of 15 trials per task. Exercises were scored based on completion time. Results were compared before and after training. Pre- and post-task surveys were administered. RESULTS: Prior to skills training, 0% of medical students reported feeling comfortable holding an endoscope or ear instruments. For students, overall mean completion times decreased significantly for each exercise by the conclusion of the training experience: placing beads on wires improved from 152 to 44 seconds (p < 0.001), placing simulated prostheses from 264 to 93 seconds (p < 0.001), and navigating a pattern from 193 to 66 seconds (p = 0.002). Individual analysis showed varying learning curves among participants and between exercises, with some students exhibiting rapid improvement. At the conclusion of training, 80% of students felt comfortable holding the endoscope and using otologic instruments, and 100% were satisfied with the experience. CONCLUSIONS: Novice surgeons can exhibit significant skill improvement with repetitive practice on an EES trainer. Variable baseline skill and improvement rates highlight the individual characteristics of skill acquisition, which may be important factors during otolaryngology and endoscopic ear surgery training.

Respiratory events after adenotonsillectomy requiring escalated admission status in children with obstructive sleep apnea.

OBJECTIVES: To characterize postoperative respiratory complications following adenotonsillectomy (AT) in children with obstructive sleep apnea (OSA) and to identify variables associated with pediatric intensive care unit (PICU) admission. METHODS: Retrospective analysis of 133 pediatric OSA patients with prior AT. Assessment of the postoperative hospital course informed patient stratification based on respiratory event severity, PICU admission status, and unscheduled escalation of care. RESULTS: Thirty-six (26.8%) patients were admitted to the PICU. Compared to non-PICU admissions, these patients were significantly younger and with greater preoperative apnea-hypopnea indices, comorbidities, and percentage of post-anesthesia care unit (PACU) time requiring supplemental oxygen. Seventy-one respiratory events occurred in 59 patients, with 60.6% affecting PICU patients. Fifteen severe events occurred, affecting 31% of PICU patients. Of 14 unscheduled escalations of care, 7 were PICU admissions who, compared to planned PICU admissions, spent significantly more time in the PACU and exhibited a trend towards greater PACU time on supplemental oxygen. CONCLUSIONS: Pediatric patients requiring post-AT PICU care have more risk factors for respiratory compromise. Total PACU time and total PACU time requiring supplemental oxygen may indicate patient risk for postoperative respiratory complications and need for intensive care. Future work includes prospective determination of appropriate post-AT PICU admission.

Readability of online patient education materials for velopharyngeal insufficiency.

OBJECTIVE: Evaluate the readability of online and mobile application health information about velopharyngeal insufficiency (VPI). METHODS: Top website and mobile application results for search terms "velopharyngeal insufficiency", "velopharyngeal dysfunction", "VPI", and "VPD" were analyzed. Readability was determined using 10 algorithms with Readability Studio Professional Edition (Oleander Software Ltd; Vandalia, OH). Subgroup analysis was performed based on search term and article source - academic hospital, general online resource, peer-reviewed journal, or professional organization. RESULTS: 18 unique articles were identified. Overall mean reading grade level was a 12.89 ± 2.9. The highest reading level among these articles was 15.47-approximately the level of a college senior. Articles from "velopharyngeal dysfunction" had the highest mean reading level (13.73 ± 2.11), above "velopharyngeal insufficiency" (12.30 ± 1.56) and "VPI" (11.66 ± 1.70). Articles from peer-reviewed journals had the highest mean reading level (15.35 ± 2.79), while articles from academic hospitals had the lowest (12.81 ± 1.66). There were statistically significant differences in reading levels between the different search terms (P < 0.01) and article source types (P < 0.05). Only one mobile application was identified with VPI information, with a readability of 10.68. CONCLUSION: Despite published reading level guidelines, online patient education materials for VPI are disseminated with language too complex for most readers. There is also a lack of VPI-related mobile application data available for patients. Patients will benefit if future updates to websites and disseminated patient information are undertaken with health literacy in mind. Future studies will investigate patient comprehension of these materials.