Automatic retrieval of current evidence to support update of bibliography in clinical guidelines

Iruetaguena, Ander; Adeva, Juan José García; Pikatza, Juan Manuel; Segundo, Unai; Buenestado, David; Barrena, Raúl

doi:10.1016/j.eswa.2012.10.015

Cited by 10 publications

(11 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the cases where there are a lot of results, the goal articles have a very high chance to get lost outside of the top of the ranking. This reinforces the findings of (Iruetaguena, et al, 2013), who noted similar results. This indicates that the combination of the Rosenfeld-Shiffman filter combined with tf-idf is perhaps not a suitable way to process large numbers of articles, as the resulting ratings are very close to each other for many articles.…”

Section: Results For Rankingsupporting

confidence: 82%

“…Cohen et al (2012) show some success in identifying goal articles with a machine learning approach, but this approach requires a large manually annotated dataset, which is very labor-intensive. Iruetaguena et al(2013) were able to find these articles, but their result set was too large, and their rating and filtering proved insufficient to filter out goal articles to the top of their ranking.…”

Section: Resultsmentioning

confidence: 99%

“…An example of a recommendation, including its evidence is shown below. This example corresponds to recommendation 1 in table 5.…”

Section: Methodsmentioning

confidence: 99%

“…The library is available from its website 4 . The source code is available on the author's GitHub account 5 .…”

Section: Methodsmentioning

confidence: 99%

“…Even though the system has decent results, it is difficult to determine how it got these results and whether the method is generalizable. The second approach is a system developed by (Iruetaguena, et al, 2013). This system takes the referenced articles from a guideline, and then constructs a new set of articles by using the PubMed related articles search.…”

Section: Related Workmentioning

confidence: 99%

See 4 more Smart Citations

Finding Evidence for Updates in Medical Guidelines

Reinders

Teije

Huang

2015

Proceedings of the International Conference on Health Informatics

View full text Add to dashboard Cite

Abstract:Medical guidelines are documents that describe optimal treatment for patients by medical practitioners based on current medical research (evidence), in the form of step-by-step recommendations. Because the field of medical research is very large and always evolving, keeping these guidelines up-to-date with the current state of the art is a difficult task. In this paper, we propose a method for finding relevant evidence for supporting the medical guideline updating process. Our method that takes from the evidence-based medical guideline the recommendations and their corresponding evidence as its input, and that queries PubMed, the world's largest search engine for medical citations, for potential new or improved evidence. We built a prototype and performed a feasibility study on a set of old recommendations, and compared the output to evidence for the newer version. The system succeeded in finding goal articles for 11 out of 16 recommendations, but in total, only 20 out of 71 articles were retrieved. Our ranking method for most relevant articles worked well for small result sets, but for large result sets it failed to rank the goal articles in the top 25 results.

show abstract

Section: Results For Rankingsupporting

confidence: 82%

Section: Resultsmentioning

confidence: 99%

“…An example of a recommendation, including its evidence is shown below. This example corresponds to recommendation 1 in table 5.…”

Section: Methodsmentioning

confidence: 99%

“…The library is available from its website 4 . The source code is available on the author's GitHub account 5 .…”

Section: Methodsmentioning

confidence: 99%

Section: Related Workmentioning