by Johanna Charlotte Neubauer, Anna Kaiser, Leon Lettermann, Tobias Volkert, Alexander Häge
ObjectiveThis study evaluates the performance of four large language models—ChatGPT 4o, ChatGPT o1-mini, Gemini 2.0 Flash, and Gemini 1.5 Flash—in answering multiple-choice questions in child and adolescent psychiatry to assess their level of factual knowledge in the field.
MethodsA total of 150 standardized multiple-choice questions from a specialty board review study guide were selected, ensuring a representative distribution across different topics. Each question had five possible answers, with only one correct option. To account for the stochastic nature of large language models, each question was asked 10 times with randomized answer orders to minimize known biases. Accuracy for each question was assessed as the percentage of correct answers across 10 requests. We calculated the mean accuracy for each model and performed statistical comparisons using paired t-tests to evaluate differences between Gemini 2.0 Flash and Gemini 1.5 Flash, as well as between Gemini 2.0 Flash and both ChatGPT 4o and ChatGPT o1-mini. As a post-hoc exploration, we identified questions with an accuracy below 10% across all models to highlight areas of particularly low performance.
ResultsThe accuracy of the tested models ranged from 68.3% to 78.9%. Both ChatGPT and Gemini demonstrated generally solid performance in the assessment of in child and adolescent psychiatry knowledge, with variations between models and topics. The superior performance of Gemini 2.0 Flash compared with its predecessor, Gemini 1.5 Flash, may reflect advancements in artificial intelligence capabilities. Certain topics, such as psychopharmacology, posed greater challenges compared to disorders with well-defined diagnostic criteria, such as schizophrenia or eating disorders.
ConclusionWhile the results indicate that language models can support knowledge acquisition in child and adolescent psychiatry, limitations remain. Variability in accuracy across different topics, potential biases, and risks of misinterpretation must be carefully considered before implementing these models in clinical decision-making.
by Charlotte J. Whiffin, Kathleen Joy O. Khu, Brandon G. Smith, Isla Kuhn, Santhani M. Selveindran, Laura Hobbs, Samin Davoody, Yusuf Docrat, Orla Mantle, Upamanyu Nath, Lara Onbaşı, Stasa Tumpa, Ignatius N. Esene, Harry Mee, Fergus Gracey, Shobhana Nagraj, Tom Bashford, Angelos G. Kolias, Peter J. Hutchinson
Following calls for more qualitative research in neurosurgery, this scoping review aimed to describe the range and reach of qualitative studies relevant to the field of neurosurgery and the patients and families affected by neurosurgical conditions. A systematic search was conducted in September 2024 across six databases: Medline via Ebsco; Embase via OVID; PsycINFO via Ebsco; Scopus; Web of Science Core Collection; and Global Health via Ebsco. Eligibility criteria were based on Population, Concept, and Context. The search identified 18,809 hits for screening with 812 included in the final analysis. Seven themes were identified from a content analysis of study aims: 1 Perspectives of living with a neurosurgical condition; 2 Family perspectives; 3 Perceptions of neurosurgery; 4 Perceptions of general healthcare care; 5 Decision making; 6 Advancing neurosurgery; and, 7 Understanding neurosurgical conditions. Traumatology was identified as the most researched sub-specialty (43.2%) yet few studies were led explicitly by a neurosurgeon (1.6%) or those with a neurosurgical affiliation (10.5%). Lead authors were predominantly from high income countries (93.7%), as were most multi-author teams (86.6%). There was a trend towards increasing publication over time; however, only 8.4% of papers were published in neurosurgical specific journals. The data set had an average Field Weighted Citation Impact of 0.96 and Field Weighted Views Impact of 1.11, 18.9% were cited in policy documents in 15 countries. This scoping review provides a comprehensive picture of the current qualitative research base in neurosurgery and suggests ways to improve the conduct and reporting of such studies in the future. Addressing these challenges is crucial if qualitative research is to advance the neurosurgical evidence base in a rigorous way.by Alessandro Roman, Charlotte Linthout, Ben Raymond, Constantianus J. M. Koenraadt
Various vector control strategies are in place to reduce the spread of arthropod-borne viruses. Some of these, such as application of insecticides, are encountering operational challenges and a reduced overall effectiveness due to evolution of resistance. Alternative approaches for mosquito population control, such as the sterile insect technique, depend on efficient mass-rearing of healthy mosquitoes prior to mass-release in the field. Therefore, improving efficiency and quality of mass-rearing techniques is crucial to obtain fit mosquitoes. Previous studies have shown that Acetic Acid Bacteria of the genus Asaia can have a mutualistic effect on larval development in different mosquito species and can thus contribute to improved rearing output. However, whether improved performance in the larval stages may have knock-on effects in the adult stage, for example by increasing their capability to transmit arbovirus, remains unclear. Such effects may jeopardize future control efforts. We tested the effects of two Asaia species, Asaia krungthepensis and Asaia bogorensis, on development time and adult size under two rearing conditions: individual rearing and group rearing of Culex pipiens larvae. Besides investigating development and size, we also investigated whether Asaia spp. exposure during the larval stage can influence the vector competence of Culex pipiens pipiens for West Nile virus (WNV). Our work shows the potential of improving mass-rearing efficiency by employing Asaia krungthepensis as a mutualist for Culex pipiens pipiens. Importantly, this study reveals no significant increase in dissemination and transmission rate of WNV by Culex pipiens pipiens when inoculated with Asaia spp., although an increase in viral titer in the legs and the saliva was observed when the mosquitoes were inoculated with the two Asaia species. Interestingly, we confirmed that Asaia spp. bacteria did not establish as a permanent member of the microbiota of Culex pipiens pipiens. As Asaia spp. did not establish in adult mosquitoes, the observed change in WNV titers can be a result of indirect interactions of Asaia with the native Culex pipiens pipiens microbiome. Our results stress the importance of carefully evaluating host-symbiont interactions to avoid the potential of releasing mosquitoes with enhanced vector competence.by Gabriel Dumitrescu, Jovan Antovic, Nida Soutari, Charlotte Gran, Aleksandra Antovic, Kais Al-Abani, Jonathan Grip, Olav Rooyackers, Apostolos Taxiarchis
Complement and extracellular vesicles (EVs) association with thrombogenic tendencies is acknowledged, but limited evidence exists for their link to COVID-19 venous thromboembolism. This study aims to examine the relationship between pulmonary embolism and the expression of complement and other proteins related to thrombogenesis in severe Covid-19 patients. We included prospectively 207 severe COVID-19 patients and retrospectively screened for pulmonary embolism (PE). This analysis comprises 20 confirmed PE cases and 20 matched patients without PE. Blood samples taken at the admission in the intensive care unit were analyzed for complement using ELISA. EVs derived from neutrophils, endothelium, or platelets, as well carrying complement or tissue factor were analyzed using flow cytometry. Complement levels were markedly elevated, with a notable increase in C3a and Terminal Complement Complex. The most prevalent EV population was identified as tissue factor (TF)-carrying EVs which peaked in patients with PE during ICU days 4–9. However, for both the complement and analyzed EV populations, no statistically significant differences were found between the patients who developed pulmonary embolism and those who did not. In conclusion, complement factors and EVs expressing tissue factor, along with EVs derived from endothelial cells and platelets, are elevated in severe COVID-19 patients, regardless of the presence of pulmonary embolism. However, the involvement of complement and procoagulant EVs in peripheral plasma in the development of pulmonary embolism is still unclear and requires further investigation.