Some VIVO Things Blog

A Story of VIVO Data

February 16, 2014 by Mike

This is a true story of using VIVO to answer a question that led to a great benefit for the University of Florida. VIVO wasn't the only way to answer the question. Well, it was the only way to answer the question at UF, we didn't have other data or systems that could do what VIVO could do. But there are other such systems. Regardless, we created the value using VIVO. Here's the story.

The Clinical Research Forum is a leading advocacy group for clinical research. Each year they solicit submissions from universities across the country of "top" papers in clinical research for the year. They recognize the authors, issue a prize, and invite the authors to testify in front of Congress regarding the value of clinical research to the health of the nation. Having a paper selected as one of the top 10 papers in clinical research for the year is a bit like winning an oscar. It's the top recognition a clinical researcher can receive, and a top honor for the researcher's university.

Our Senior Vice President for Health Affairs asked our Clinical and Translational Science Institute Director if we could identify top papers in clinical research written by UF authors for nomination to the Clinical Research Forum. The director of the CTSI asked me if it would be possible to identify top papers using VIVO. We discussed for a few minutes how we might define top papers and determined that the appearance of an academic article in 2013 in a top journal as determined by impact factor would qualify the paper.

I determined top journals by using publicly available web sites listing journals by impact factor and by correlating those results with sites listing journals by other criteria. I settled on 15 journals. I then wrote a SPARQL query to find the papers with UF authors appearing in these journals in 2013. The query is available here. 38 papers were identified by VIVO. I reviewed the list and removed papers that were not in clinical research. As you might expect, the Clinical Research Forum is looking for papers in clinical research. One of the top journals on our list was Science. UF authors had several papers in Science over the course of 2013, but all were in planetary discovery, none in clinical research. I removed articles that had been identified as academic articles, but were actually letters to the editor or comments on papers by others. Following this review I had identified 23 clinical research papers that had been written by UF authors and appeared in 2013 in the top 15 journals.

I looked up each article in PubMed and prepared a Word document with one page per article cut and pasting the article's information from PubMed into the Word document. I gathered the title, keywords, abstract for each. I added the SPARQL query as an appendix.

From the time of the first discussion to the final report, I had spent 2 hours, with most of the time spent identifying the top journals. This list of top journals is in the query and can be reused.

The report went back to our senior vice president and a UF selection committee he convened nominated several of the papers from the report to the Clinical Research Forum. One of the papers was chosen by the Clinical Research Form as the number 5 paper in clinical research in the United States for 2013. The author will receive a prize, and will testify before Congress. This is an unprecedented accomplishment for UF clinical research. We would never have found the paper had it not been for VIVO.

Some takeaways

First, the communication path led quickly from our senior vice president to our CTSI to our VIVO. Our senior vice president believed the CTSI could be helpful. At the CTSI VIVO was well known and immediately put to use. Communication channels to senior leadership, such as these, are critical if VIVO is to have impact. Second, our UF VIVO is disambiguated. This means that we not only know that someone named J. Johnson, a UF author, has written a paper, we have identified which J. Johnson (UF has six) actually wrote the paper. We do this for all papers written by UF authors each week, about 150 per week. Commercial indexing services can't do this. They only indicate that "someone" with that name has written the paper. UF VIVO has 45,000 disambiguated papers. These are all the papers written by UF authors since 2008. As a result, the request led directly to specific authors. Third, alternate methods for identifying papers to be nominated were proposed but would not have identified the paper in question. The proposed methods all involved asking people what papers they thought should be nominated. These methods are both incomplete (UF College of Medicine authors published over 2,500 papers in 2013. Very few of these papers are known to the people who might be asked their opinion) and heavily influenced by relationships of the authors and the people nominating the papers. The "VIVO Method" described here was unbiased. All papers were identified and put before the selection committee.

Winning an unprecedented recognition for UF clinical research is not justification for creating a VIVO. But it serves as an example of the power of VIVO data, and the good that can come from being ready when a request for data comes along.