Dataset Information

Scientific considerations for evaluating cancer bioassays conducted by the Ramazzini Institute.

ABSTRACT:

Background

The Ramazzini Institute (RI) has completed nearly 400 cancer bioassays on > 200 compounds. The European Food Safety Authority (EFSA) and others have suggested that study design and protocol differences between the RI and other laboratories by may contribute to controversy regarding cancer hazard findings, principally findings on lymphoma/leukemia diagnoses.

Objective

We aimed to evaluate RI study design, protocol differences, and accuracy of tumor diagnoses for their impact on carcinogenic hazard characterization.

Methods

We analyzed the findings from a recent Pathology Working Group (PWG) review of RI procedures and tumor diagnoses, evaluated consistency of RI and other laboratory findings for chemicals identified by the RI as positive for lymphoma/leukemia, and examined evidence for a number of other issues raised regarding RI bioassays. The RI cancer bioassay design and protocols were evaluated in the context of relevant risk assessment guidance from international authorities.

Discussion

Although the PWG identified close agreement with RI diagnoses for most tumor types, it did not find close agreement for lymphoma/leukemia of the respiratory tract or for neoplasms of the inner ear and cranium. Here we discuss a) the implications of the PWG findings, particularly lymphoma diagnostic issues; b) differences between RI studies and those from other laboratories that are relevant to evaluating RI cancer bioassays; and c) future work that may help resolve some concerns.

Conclusions

We concluded that a) issues related to respiratory tract infections have complicated diagnoses at that site (i.e., lymphoma/leukemia), as well as for neoplasms of the inner ear and cranium, and b) there is consistency and value in RI studies for identification of other chemical-related neoplasia.

SUBMITTER: Gift JS

PROVIDER: S-EPMC3852791 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:BackgroundWith the rapid evolution of artificial intelligence (AI), particularly large language models (LLMs) such as ChatGPT-4 (OpenAI), there is an increasing interest in their potential to assist in scholarly tasks, including conducting literature reviews. However, the efficacy of AI-generated reviews compared with traditional human-led approaches remains underexplored.ObjectiveThis study aims to compare the quality of literature reviews conducted by the ChatGPT-4 model with those conducted by human researchers, focusing on the relational dynamics between physicians and patients.MethodsWe included 2 literature reviews in the study on the same topic, namely, exploring factors affecting relational dynamics between physicians and patients in medicolegal contexts. One review used GPT-4, last updated in September 2021, and the other was conducted by human researchers. The human review involved a comprehensive literature search using medical subject headings and keywords in Ovid MEDLINE, followed by a thematic analysis of the literature to synthesize information from selected articles. The AI-generated review used a new prompt engineering approach, using iterative and sequential prompts to generate results. Comparative analysis was based on qualitative measures such as accuracy, response time, consistency, breadth and depth of knowledge, contextual understanding, and transparency.ResultsGPT-4 produced an extensive list of relational factors rapidly. The AI model demonstrated an impressive breadth of knowledge but exhibited limitations in in-depth and contextual understanding, occasionally producing irrelevant or incorrect information. In comparison, human researchers provided a more nuanced and contextually relevant review. The comparative analysis assessed the reviews based on criteria including accuracy, response time, consistency, breadth and depth of knowledge, contextual understanding, and transparency. While GPT-4 showed advantages in response time and breadth of knowledge, human-led reviews excelled in accuracy, depth of knowledge, and contextual understanding.ConclusionsThe study suggests that GPT-4, with structured prompt engineering, can be a valuable tool for conducting preliminary literature reviews by providing a broad overview of topics quickly. However, its limitations necessitate careful expert evaluation and refinement, making it an assistant rather than a substitute for human expertise in comprehensive literature reviews. Moreover, this research highlights the potential and limitations of using AI tools like GPT-4 in academic research, particularly in the fields of health services and medical research. It underscores the necessity of combining AI's rapid information retrieval capabilities with human expertise for more accurate and contextually rich scholarly outputs.

Project description:OBJECTIVE:The peer review of completed Patient-Centered Outcomes Research Institute (PCORI) funded research includes reviews from patient reviewers (patients, caregivers, and patient advocates). Very little is known about how best to support these reviewers in writing helpful comments from a patient-centred perspective. This study aimed to evaluate the effect of a new training in peer review for patient reviewers. DESIGN:Observational study. SETTING:Online. PARTICIPANTS:Adults registered in the PCORI Reviewer Database as a patient stakeholder. INTERVENTION:A new online training in peer review. MAIN OUTCOME MEASURES:Changes in reviewers' knowledge and skills; change in self-efficacy and attitudes, satisfaction with the training and perceived benefits and relevance of the training. RESULTS:Before-after training survey data were analysed for 37 (29.4% of 126) patient reviewers invited to participate in an online training as part of a quality improvement effort or as part of a PCORI peer review. The reviewers improved their answers to the knowledge questions (p<0.001, median number of answers improved 4 (95% CI 3 to 5), large effect size (ES) Cohen's w=0.94) after the training, particularly in the questions targeting the specifics of PCORI peer review. Reviewers improved their skills in recognising helpful review comments, but those without peer-review background improved proportionally more (p=0.008, median number of answers improved 2 (95% CI 1 to 3), medium ES w=0.60). The use of training modestly increased reviewers' confidence in completing a high-quality peer review (p=0.005, mean increase in 5-point Likert rating 0.51 (95% CI 0.17 to 0.86), small-to-medium ES Cliff's delta=0.32) and their excitement about providing a review slightly increased (p=0.019, mean increase in 5-point Likert rating 0.35 (95% CI 0.03 to 0.68), small ES delta=0.19). All reviewers were satisfied with the training and would recommend it to other reviewers. CONCLUSIONS:Training improved knowledge, skills and self-efficacy and slightly increased enthusiasm for completing a PCORI peer review.

Project description:Domain analysis by means of scientific collaboration enables evidencing aspects that are involved in the establishment of relationships between researchers and institutions, such as the influence of institutional management models for the development of collaborative networks. This article aims to analyze the domain through the scientific collaboration network of the National Institute of the Atlantic Forest (INMA), a research unit currently affiliated to the Brazilian Ministry of Science, Technology and Innovation (MCTI), formerly known as the Professor Mello Leitão Museum of Biology (MBML), in order to acknowledge the institutional research identity in its historical journey as a public institution. It is thus analyzed how co-authorship constitutes this network and what research profile it reveals. Co-authorship analysis is adopted as a methodology, as well as the analysis of administrative documents with the survey and categorization of employees, regarding their types of ties to the institution, combined with searches in the Scopus database for the corroboration of institutional affiliations. A corpus of 138 articles published by 41 researchers from 1993 to 2019 is consolidated in this base, which represents 44% of the Institute's total research collaborators (93 collaborators). Of these 41, 92.5% have temporary links, such as scholarship holders and/or volunteers, with the remaining being public workers. It is recognized that the citation impact of the scientific production of scholarship holders, consigned to the Institute, is less than the citation impact of the volunteers' and public workers' production. It is evidenced that eight of the ten publications with the greatest impact and thematic prominence correspond to the field of zoology, with emphasis on the fields of herpetology and primatology. Macro-level collaborative relations are more intense with the United States, in both areas mentioned, covering 16% of the total corpus of articles in cooperation with that country. Zoology, besides its greater impact, accounts for more than half of the corpus production (65.9%).On the other hand, botany is responsible for 30.4% of the corpus, with its dispersed international cooperation in a broad variety of countries. Individual authorship articles are 57% consigned to botany. In summary, the accomplished analysis will contribute to the development of institutional domain analysis methodologies that present scientific collaboration as a basic procedure.