BACKGROUND: Large language models (LLMs) have the potential to support promising new applications in health informatics. However, practical data on sample size considerations for fine-tuning LLMs to perform specific tasks in biomedical and health policy contexts are lacking.
OBJECTIVE: This study aims to evaluate sample size and sample selection techniques for fine-tuning LLMs to support improved named entity recognition (NER) for a custom data set of conflicts of interest disclosure statements.
METHODS: A random sample of 200 disclosure statements was prepared for annotation. All \"PERSON\" and \"ORG\" entities were identified by each of the 2 raters, and once appropriate agreement was established, the annotators independently annotated an additional 290 disclosure statements. From the 490 annotated documents, 2500 stratified random samples in different size ranges were drawn. The 2500 training set subsamples were used to fine-tune a selection of language models across 2 model architectures (Bidirectional Encoder Representations from Transformers [BERT] and Generative Pre-trained Transformer [GPT]) for improved NER, and multiple regression was used to assess the relationship between sample size (sentences), entity density (entities per sentence [EPS]), and trained model performance (F1-score). Additionally, single-predictor threshold regression models were used to evaluate the possibility of diminishing marginal returns from increased sample size or entity density.
RESULTS: Fine-tuned models ranged in topline NER performance from F1-score=0.79 to F1-score=0.96 across architectures. Two-predictor multiple linear regression models were statistically significant with multiple R2 ranging from 0.6057 to 0.7896 (all P<.001). EPS and the number of sentences were significant predictors of F1-scores in all cases ( P<.001), except for the GPT-2_large model, where EPS was not a significant predictor (P=.184). Model thresholds indicate points of diminishing marginal return from increased training data set sample size measured by the number of sentences, with point estimates ranging from 439 sentences for RoBERTa_large to 527 sentences for GPT-2_large. Likewise, the threshold regression models indicate a diminishing marginal return for EPS with point estimates between 1.36 and 1.38.
CONCLUSIONS: Relatively modest sample sizes can be used to fine-tune LLMs for NER tasks applied to biomedical text, and training data entity density should representatively approximate entity density in production data. Training data quality and a model architecture\'s intended use (text generation vs text processing or classification) may be as, or more, important as training data volume and model parameter size.

UNASSIGNED: Although the use of the Web has brought major advances in every step of the research process, this also comes with several methodological challenges.
UNASSIGNED: The article presents the European Society for Sexual Medicine\'s position statements on key methodological concerns relative to Web-based research in sexual medicine.
UNASSIGNED: The authors conducted a systematic scoping review of articles using Web-based research methods in sexual medicine. For the creation of the statements, the authors processed the data from the methodology of the studies and formulated the final statements reaching 100% agreement in the group.
UNASSIGNED: European Society for Sexual Medicine statements were provided on the following domains: definition of the population of interest, selection of the population of interest, data collection quality, response rate, self-reported questionnaire, consent, and legal obligations.
UNASSIGNED: Researchers should justify the relevance of the Internet population to the population of interest; should clearly describe how they identified study participants; should select and employ specific measures to avoid hoax; should explicitly describe the process of calculation of response and completion rates as well as the relative implications; should validate traditional sexual health questionnaires for online and, if possible, multilingual use; should not ignore consent in Web-based research; and need to be knowledgeable of the technical measures and legal obligations to protect anonymity.
UNASSIGNED: Researchers are advised to include trained computer scientists in their group, have a good understanding of their legal obligations as to collecting, storing and disseminating personal data, and design their studies by taking into account the challenges of Web-based research.
UNASSIGNED: The heterogeneity of the included studies and methodological low quality of most of them was a limitation, which also shows the importance of this study and the need for guidelines regarding Web-based research.
UNASSIGNED: Large uncontrolled samples could be a threat to the quality of the studies and increase bias if researchers are not mindful of the methodological challenges they would need to account for.

Cross-linguistically, statements and questions broadly differ in syntactic organization. To learn the syntactic properties of each sentence type, learners might first rely on non-syntactic information. This paper analyzed prosodic differences between infant-directed wh-questions and statements to determine what kinds of cues might be available. We predicted there would be a significant difference depending on the first words that appear in wh-questions (e.g., two closed-class words; meaning words from a category that rarely changes) compared to the variety of first words found in statements. We measured F0, duration, and intensity of the first two words in statements and wh-questions in naturalistic speech from 13 mother-child dyads in the Brent corpus of the CHILDES database. Results found larger differences between sentence-types when the second word was an open-class not a closed-class word, suggesting a relationship between prosodic and syntactic information in an utterance-initial position that infants may use to make sentence-type distinctions.

UNASSIGNED: Assessment tools with the ability to capture WHO/UNICEF/UNFPA standard quality-of-care measures are needed. This study aimed to assess the ability of Every Mother Every Newborn (EMEN) tools to capture WHO/UNICEF/UNFPA maternal and newborn quality improvement standard indicators.
UNASSIGNED: A quantitative study using the EMEN quality assessment framework was applied. The six EMEN tools were compared with the WHO/UNICEF/UNFPA maternal and newborn quality improvement standards. Descriptive statistics analysis was carried out with summaries using tables and figures.
UNASSIGNED: Overall, across all EMEN tools, 100% (164 of 164) input, 94% (103 of 110) output, and 97% (76 of 78) outcome measures were assessed. Standard 2 measures, i.e., actionable information systems, were 100% (17 of 17) completely assessed by the management interview, with 72% to 96% of standard 4-6 measures, i.e., client experiences of care, fulfilled by an exit interview tool.
UNASSIGNED: The EMEN tools can reasonably measure WHO/UNICEF/UNFPA quality standards. There was a high capacity of the tools to capture enabling policy environment and experiences of care measures not covered in other available tools which are used to measure the quality of care.

Most countries compile evidence from witnesses and victims manually, whereby the interviewer assimilates what the interviewee says during the course of an interview to produce an evidential statement. This exploratory research examined the quality of evidential statements generated in real world investigations. Transcribed witness/victim interviews (N = 15) were compared to the resultant written statements produced by the interviewing officer and signed as an accurate record by the interviewee. A coding protocol was devised to assess the consistency of information between what was said by the interviewee in the verbal interview and what was reported in the written statement. Statements contained numerous errors including omissions, distortions, and the inclusion of information not mentioned in the verbal interview. This exploratory work highlights an important area for future research focus.

OBJECTIVE: The utility of institutional statements is said to provide clarity and reinforcement of an institution\'s goal. Unfortunately, it can also be argued that these statements are in clear misalignment between the words described and the environments that faculty, students, trainees, and staff of color face. The purpose of this study was to analyze academic medicine institutional statements that responded to 2020 racial tensions following the murders of George Floyd, among others, and the subsequent nationwide protests against police brutality.
METHODS: We conducted a manifest content analysis of institutional statements generated by academic medical centers after George Floyd\'s murder. We used manifest content analysis to gain insights into how institutional statements connect structural racism to the mission of academic medicine. We collected and examined institutional statements from 26 academic medicine centers. Selection parameters included statements that were publicly available and published by the institutions during a 2-week period. We conducted a four-stage analysis: decontextualization, recontextualization, categorization, and compilation. To better understand the collection of statements, we plotted these institutional statements according to the most salient discussed themes.
RESULTS: Overall, institutional statements discussed racism through three subthemes: identifying the racial implications of health disparities, issuing a call to action to address racism, and decentering race. Absent language is also noted. Second, institutional statements evoked institutional values by expressing an explicit connection with the academic medicine mission, naming the value of social justice, and emphasizing the concept of community. Finally, institutional statements largely discussed public health in connection with racism or with institutional values.
CONCLUSIONS: Our study determines a much-needed reconnection to the mission of academic medicine. Reclaiming the social mission will be a major step toward recentering the foundation of institutional actions. This call is what ultimately will improve the health and well-being of marginalized populations.

Despite the evidence provided by clinical trials, there are some uncertainties and controversies regarding the use of triple inhaled therapy. With the aim of evaluating clinical practice in specialized respiratory units, a Delphi consensus document was implemented on the use of single-inhaler fixed-dose triple therapies after 1 year of use in Spain.
A scientific committee of COPD experts defined a thematic index, guided a systematic literature review and helped design the Delphi questionnaire. This was sent to the other 45 COPD experts between April and June 2019. Agreement/disagreement on 58 statements was tested in two rounds using a Likert scale. Replies were classified as a consensus when ≥80% of the panelists agreed; a majority when a degree of agreement of ≥66% was reached; and divergence if agreement was <66%.
After two rounds, 44.44% of the statements reached consensus, 14.81% reached majority and 40.74% were divergent. Panelists agreed that escalating from double bronchodilation should be phenotype-based and aim to prevent exacerbations but not for improving symptoms. The addition of an antimuscarinic to inhaled corticosteroids combinations achieves improvement in lung function, symptoms and exacerbation prevention. Main safety concerns included the increased risk of pneumonia as compared to bronchodilator therapies, with similar cardiovascular effects. There was no consensus agreement on patient type response based on blood eosinophil counts or obstruction severity.
The low degree of consensus among panelists may reflect the complexity of severe COPD management. The information provided here may be useful to clinicians implementing personalized medicine for COPD patients.

An implementation plan was developed in conjunction with the publication of the European Statements of Hospital Pharmacy. Subsequently a baseline survey on the status of the Statements was conducted with specific questions on awareness, capability and capacity seen as crucial to informing future implementation plans. The baseline survey showed that, 18 months after agreement at the European Summit of Hospital Pharmacy, hospital pharmacists across Europe have a growing awareness of the Statements. There is a significant variation across Europe with awareness levels ranging from 100% to less than 30%. Capacity within European Healthcare systems to achieve implementation is challenging with less than 10% of respondents in most countries indicating sufficient capacity to implement the services described in the Statements. Staff skills and experience to implement the changes was identified by respondents at less than 20% in many countries. The implementation plan will need to be constantly updated and modified to sustain progress.

Background/Objective: Criteria-Based Content Analysis (CBCA) is the tool most extensively used worldwide for evaluating the veracity of a testimony. CBCA, initially designed for evaluating the testimonies of victims of child sexual abuse, has been empirically validated. Moreover, CBCA has been generalized to adult populations and other contexts though this generalization has not been endorsed by the scientific literature. Method: Thus, a meta-analysis was performed to assess the Undeutsch Hypothesis and the CBCA checklist of criteria in discerning in adults between memories of self-experienced real-life events and fabricated or fictitious memories. Results: Though the results corroborated the Undeutsch Hypothesis, and CBCA as a valid technique, the results were not generalizable, and the self-deprecation and pardoning the perpetrator criteria failed to discriminate between both memories. The technique can be complemented with additional reality criteria. The study of moderators revealed discriminating efficacy was significantly higher in filed studies on sexual offences and intimate partner violence. Conclusions: The findings are discussed in terms of their implications as well as the limitations and conditions for applying these results to forensic settings.
Antecedentes/Objetivo: El Criteria-Based Content Analysis (CBCA) constituye la herramienta mundialmente más utilizada para la evaluación de la credibilidad del testimonio. Originalmente fue creado para testimonios de menores víctimas de abuso sexual, gozando de amparo científico. Sin embargo, se ha generalizado su práctica a poblaciones de adultos y otros contextos sin un aval de la literatura para tal generalización. Método: Por ello, nos planteamos una revisión meta-analítica con el objetivo de contrastar la Hipótesis Undeutsch y los criterios de realidad del CBCA para conocer su potencial capacidad discriminativa entre memorias de eventos auto-experimentados y fabricados en adultos. Resultados: Los resultados confirman la hipótesis Undeutsch y validan el CBCA como técnica. No obstante, los resultados no son generalizables y los criterios auto-desaprobación y perdón al autor del delito no discriminan entre ambas memorias. Además, se encontró que la técnica puede ser complementada con criterios adicionales de realidad. El estudio de moderadores mostró que la eficacia discriminativa era significativamente superior en estudios de campo en casos de violencia sexual y de género. Conclusiones: Se discute la utilidad, así como las limitaciones y condiciones para la transferencia de estos resultados a la práctica forense.

UNASSIGNED: The multidisciplinary management of oncologic patients is identified as the bottom line element of quality in tumor care.
UNASSIGNED: In 2015, 7 Italian scientific societies representing the specialists involved in the diagnosis and treatment of genitourinary tumors joined efforts in the Italian uro-oncologic multidisciplinary teams (MDTs) project. The aims were to promote the reorganization of genitourinary cancer care, switching to a multidisciplinary approach, reach a consensus on the core elements for the setup of MDTs in genitourinary oncology, and support health policy makers and managers in remodeling of the assistance and care of uro-oncologic patients on a national level.
UNASSIGNED: The first activity was the setup of 5 working groups, given the task of exploring selected topics: general principles, organization of MDTs, minimal requirements, economic evaluation, and relations with authorities. The groups participated in the writing of a document that was approved by the scientific societies and published on their web sites. Moreover, a few items summarizing the extensive document were approved in the first MDT Consensus Conference held in Milan in December 2015.
UNASSIGNED: The experience of this initial phase led to the opening of the team to other professionals and societies, in line with a correct management of patients with genitourinary tumors, which need a multidisciplinary as well as a multiprofessional approach with emerging techniques and procedures, and with a new project work package on genitourinary paths of care and indicators.