bims-librar 2018-06-17 papers

Issue of 2018–06–17
four papers selected by
Thomas Krichel, Open Library Society

Spine Deform. 2018 Jul - Aug;6(4):pii: S2212-134X(17)30442-2. [Epub ahead of print]6(4): 373-383

Predominance of Studies With Poor Level of Evidence Among the Top 100 Most Cited Studies on Idiopathic Scoliosis: A Bibliometric and Content Analysis.

Joel Gambín-Botella, María Ayala, Joaquín Alfonso-Beltrán, Carlos Barrios.

STUDY DESIGN: Bibliometric review of current literature.
OBJECTIVE: To identify and analyze the characteristics of the 100 most cited articles on idiopathic scoliosis focusing on the level of evidence.
SUMMARY OF BACKGROUND DATA: The scientific literature on idiopathic scoliosis has been constantly evolving, but many aspects of its true etiology, natural history, and response to treatment continue to be discussed. To date, no study has used bibliometric analysis to review the most influential articles about idiopathic scoliosis.
METHODS: The Thompson Reuters Web of Science was accessed to find the 100 most cited articles on idiopathic scoliosis. The number and citation density, authorship, institutions, country of origin, year of publication, source journals, type of study, topic, study design, and level of evidence were analyzed.
RESULTS: The 100 most frequently cited articles accumulated 13,749 citations. The number of citations ranged between 616 and 80. The 10 most cited articles represent 24.6% of all citations. The treatment of idiopathic scoliosis is the most commonly studied issue (n = 46), and specifically surgical correction (n = 36). Most studies originated in the United States (n = 62) and were published in Spine (n = 56). Almost half of the papers (n = 49) were published between 2000 and 2008. The majority of studies have a case series design (n = 35). Most of the cited articles have low levels of evidence (Level III = 36; Level IV = 35).
CONCLUSION: This bibliometric analysis includes the 100 most cited articles on idiopathic scoliosis, recognizing its importance as a basic milestone in today's spine knowledge. The results indicate that the evolution of the knowledge on idiopathic scoliosis has been through case reports and case series, which analyzed retrospectively today are considered to have a poor level of evidence. This observation seems to be paradoxical because they are the most influential articles on IS and had been published in the top, leading journals of spine surgery.

Keywords: Bibliometric study; Citation analysis; Classic papers; Idiopathic scoliosis; Level of evidence

DOI: https://doi.org/10.1016/j.jspd.2017.12.003

Acad Radiol. 2018 Jul;pii: S1076-6332(18)30093-X. [Epub ahead of print]25(7): 951-954

Collaboration Metrics Among Female and Male Researchers: A 5-Year Review of Publications in Major Radiology Journals.

James C Campbell, Sora C Yoon, Lars J Grimm.

RATIONALE AND OBJECTIVES: Women are under-represented in radiology, but the implications of this under-representation are poorly understood. Therefore, the purpose of this study was to determine if articles published by women in major radiology journals were more collaborative.
MATERIALS AND METHODS: Following an institutional review board exemption, we reviewed all original research articles in Radiology, in the American Journal of Roentgenology, and in Academic Radiology from 2011 to 2015. For each article, the gender of the first and the last authors and proxy measures of collaboration were recorded, including the total number of authors, female authors, departments, and institutions. Nominal logistic regression analysis was used to test for associations while controlling for confounders.
RESULTS: There were 1934 articles analyzed. Female first and last authors represented 30.2% (585 of 1934) and 24.4% (473 of 1934) of the articles, respectively. A female first author was associated with more female last authors (36% vs 20%, P < .001), total female authors (2.9 vs 1.2, P < .001), and departments (3.3 vs 3.0, P < .001). Similarly, a female last author was associated with more female first authors (44% vs 26%, P = .001), total female authors (3.1 vs 1.2, P < .001), departments (3.5 vs 3.0 P < .001), and institutions (2.3 vs 1.9, P = .006). Each additional female author increased the mean number of institutions by 0.33 and departments by 0.46 on linear regression. First- or last-author gender was not associated with total authors (P = .17).
CONCLUSIONS: Original research articles published with a female first or last author were associated with more departments and institutions, but not with the total number of authors, suggesting that women engage in some metrics of more collaborative research.

Keywords: Gender; collaboration; journals; publications; women

DOI: https://doi.org/10.1016/j.acra.2017.12.034

Artif Intell Med. 2018 Jun 07. pii: S0933-3657(17)30591-2. [Epub ahead of print]

Automatic classification of radiological reports for clinical care.

Alfonso Emilio Gerevini, Alberto Lavelli, Alessandro Maffi, Roberto Maroldi, Anne-Lyse Minard, Ivan Serina, Guido Squassina.

Radiological reporting generates a large amount of free-text clinical narratives, a potentially valuable source of information for improving clinical care and supporting research. The use of automatic techniques to analyze such reports is necessary to make their content effectively available to radiologists in an aggregated form. In this paper we focus on the classification of chest computed tomography reports according to a classification schema proposed for this task by radiologists of the Italian hospital ASST Spedali Civili di Brescia. The proposed system is built exploiting a training data set containing reports annotated by radiologists. Each report is classified according to the schema developed by radiologists and textual evidences are marked in the report. The annotations are then used to train different machine learning based classifiers. We present in this paper a method based on a cascade of classifiers which make use of a set of syntactic and semantic features. The resulting system is a novel hierarchical classification system for the given task, that we have experimentally evaluated.

Keywords: Analysis of radiological reports; Machine learning for information extraction; NLP for biomedical texts; Text classification; Text mining

DOI: https://doi.org/10.1016/j.artmed.2018.05.006

Neurosci Biobehav Rev. 2018 Jun 07. pii: S0149-7634(17)30950-8. [Epub ahead of print]

How would publication bias distort the estimated effect size of prototypic antidepressants in the forced swim test?

Ana B Ramos-Hryb, Cari Harris, Omorose Aighewi, Cilene Lino-de-Oliveira.

This commentary aims to discuss the impact of publication bias on the estimated effect of prototypic antidepressants in the forced swim test (FST). A systematic review and meta-analysis (SRMA) recently reported by Kara et al. (2018) showed that selected prototypic antidepressants reduced immobility time of mice in the FST across a variety of experimental designs. Despite differences in the procedures for SRMA, these results resemble the interim data collected by our research group according to a protocol deposited in the Systematic Review Facility and Open Science Framework (osf.io/9kxm4). Both studies detected a high amount of publications reporting statistically significant results and agreement with the primary hypothesis raising the possibility of publication bias in the field of FST. In our preliminary analysis, no evidence for publication bias was observed. However, the present work was limited to the effects of imipramine (doses ranging from 4 to 64 mg/kg) in different strains of mice. Therefore, more comprehensive studies are required to evaluate the risk of publication bias in the field of basic antidepressant research. We see the need to expand the current preliminary studies to evaluate the risk of publication bias within the preclinical research using the FST. Appraisal of the risk of publication bias may avoid misestimated effects of drugs in the FST providing better bases for the discovery of new antidepressants.

Keywords: Animal models; Antidepressants; Meta-analysis; Reproducibility; Systematic Review; Validity

DOI: https://doi.org/10.1016/j.neubiorev.2018.05.025