Skip to content
BY-NC-ND 4.0 license Open Access Published by De Gruyter October 1, 2014

Response

  • Robert T. Hasty , Ryan C. Garbalosa and Gabriel Suciu

We appreciate the comments by Fraser and Temple,1 Leo and Lacasse,2 Chen and Xiong,3 and Gurzell4 regarding our May 2014 study 5 on Wikipedia articles.

As we identified in our article,5 one of the study's limitations was that reviewers could use any peer-reviewed reference that was published or updated (within the past 5 years) as a standard. A future study could be conducted that would limit the type of references that could be used. Fraser and Temple1 also question whether there might have been a selection bias to the articles that were representative of the 10 most costly medical conditions. The representative articles were chosen before any analysis of the content to reduce the chance of a selection bias. It is important to note that there are more than 30,000 articles in Wikipedia pertaining to medical conditions6 and that our study only examined 10 of those articles. In addition, our study was designed to look for errors rather than omissions of information. We made no comment or criticism on the breadth of information covered on each subject or its completeness. A future study could be conducted to examine a larger number of representative articles. However, for the purposes of our study, examining more articles would not have lessened the importance of the errors we found in the articles we studied.

Leo and Lacasse2 present an analysis of a sampling of raw data on major depressive disorder from our study. One Wikipedia assertion they mention in their analysis was “antidepressants have not been found to be beneficial in children.”7 They challenge the finding the reviewer cited from UpToDate, in which the reviewer reported, “No, fluoxetine is approved for age 8 and older” and discussed the US Food and Drug Administration approval process.5 Unfortunately, Leo and Lacasse did not mention that the reviewer in our study went on to cite a meta-analysis that examined 19 studies and found “an overall benefit of SSRI [selective serotonin reuptake inhibitor] medications compared to placebo for children and adolescents,” although this meta-analysis did go on to say that these results “should be interpreted with caution.”8

Another Wikipedia assertion mentioned in the Leo and Lacasse analysis was, “Psychotherapy is the treatment of choice for people younger than 18 years.” The researcher in our study cited a guideline for adolescent depression, which stated, “however, the results of a recent RCT [randomized controlled trial] demonstrated superior efficacy of combination therapy (medication and CBT [cognitive behavioral therapy]) versus CBT alone…. When indicated by clinical presentation (clear diagnosis of MDD [major depressive disorder] with no comorbid conditions) and patient/family preference, an SSRI should be used.”9 This finding does not support the assertion made in Wikipedia.

Although their analysis highlights controversial topics in major depressive disorder, it is important to note that the Wikipedia article Leo and Lacasse referenced did not include the controversy in their 2 assertions.7 If Wikipedia articles are considered review articles, then it would be expected that major controversial points would be discussed rather than presented from one perspective. If such a discussion occurred in the Wikipedia article (as would likely have occurred in a peer-reviewed review article), Wikipedia would not have been found as discordant with the literature on this topic. Overall, the analysis of the subset of data given to Leo and reported by Leo and Lacasse is not complete, and the 2 examples they cited support the study as designed. We would encourage them to reproduce our study to see if they find similar results.

Chen and Xiong3 as well as Gurzell4 question the use of the McNemar test and the statistical analysis of the data in our study. The study's null hypothesis was that “there would be concordance between the Wikipedia article and peer-reviewed sources.” Our study dealt with correlated proportions; therefore, we matched the peer-reviewed results with the Wikipedia results (by keywords).5 For greater clarity, our table 3 should have indicated “Peer-Reviewed” as a row header instead of “Reviewer 1” and “Reviewer 2,” and then under this header we should have had 2 rows: “Concordance” and “Discordance.” The column header should have been “Wikipedia.” For example, we calculated the following proportions for the lung cancer assertions: PPeerR=(73+27)/201=0.49 and PWiki=(73+83)/201=0.77. The McNemar test results for 2 tails shows P<.0001 (using the exact binomial probability calculation). Thus, we rejected the null hypothesis for the similar assertions. Because of the large number of possible errors, we did not measure the accuracy but estimated the significance of the difference between the peer-reviewed and Wikipedia statements, which is why we did not report CIs.

References

1 Fraser JH Temple N . Wikipedia vs peer-reviewed medical literature for information about the 10 most costly medical conditions [letter]. J Am Osteopath Assoc.2014;114(10):761. doi:10.7556/jaoa.2014.146.10.7556/jaoa.2014.146Search in Google Scholar PubMed

2 Leo J Lacasse JR . Wikipedia vs peer-reviewed medical literature for information about the 10 most costly medical conditions [letter]. J Am Osteopath Assoc.2014;114(10):761-764. doi:10.7556/jaoa.2014.147.10.7556/jaoa.2014.147Search in Google Scholar PubMed

3 Chen GS Xiong Y . Wikipedia vs peer-reviewed medical literature for information about the 10 most costly medical conditions [letter]. J Am Osteopath Assoc.2014;114(10):764-765. doi:10.7556/jaoa.2014.148.10.7556/jaoa.2014.148Search in Google Scholar PubMed

4 Gurzell E . Wikipedia vs peer-reviewed medical literature for information about the 10 most costly medical conditions [letter]. J Am Osteopath Assoc.2014;114(10):765-766. doi:10.7556/jaoa.2014.149.10.7556/jaoa.2014.149Search in Google Scholar PubMed

5 Hasty RT Garbalosa RC Barbato VA et al. . Wikipedia vs peer-reviewed medical literature for information about the 10 most costly medical conditions. J Am Osteopath Assoc.2014;114(5):368-373. doi:10.7556/jaoa.2014.035.10.7556/jaoa.2014.035Search in Google Scholar PubMed

6 Wikipedia : WikiProject Medicine. Wikipedia website. http://en.wikipedia.org/wiki/Wikipedia:WikiProject_Medicine. Accessed August 25, 2014.Search in Google Scholar

7 Major depressive disorder. Wikipedia website. http://en.wikipedia.org/wiki/Major_depressive_disorder. Accessed April 25, 2012.Search in Google Scholar

8 Drews AA Antonuccio DO Kirsch I . A meta-analysis of randomized placebo controlled trials of antidepressant medications in depressed children: do the benefits justify the risks?J Mind-Body Regul.2011;1(2):85-95.Search in Google Scholar

9 Cheung AH Zuckerbrot RA Jensen PS Ghalib K Laraque D Stein RE GLAD-PC Steering Group . Guidelines for Adolescent Depression in Primary Care (GLAD-PC): II—treatment and ongoing management. Pediatrics. 2007;120(5):e1313-e1326.10.1542/peds.2006-1395Search in Google Scholar PubMed

Published Online: 2014-10-01
Published in Print: 2014-10-01

© 2014 The American Osteopathic Association

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Downloaded on 29.3.2024 from https://www.degruyter.com/document/doi/10.7556/jaoa.2014.150/html
Scroll to top button