• Stephen F Butler, Synne Wing Venuti, Christine Benoit, Richard L Beaulaurier, Brian Houle, and Nathaniel Katz.
• Inflexxion, Inc, Newton, MA 02464, USA. sfbutler@inflexxion.com
• Clin J Pain. 2007 Sep 1; 23 (7): 619-28.

ObjectivesThis study describes the development of a systematic approach to the analysis of Internet chatter as a means of monitoring potentially abusable opioid analgesics.MethodsMessage boards dedicated to drug abuse were selected using specific inclusion criteria. Threaded discussions containing 48,293 posts were captured. A coding system was created to compare content of posts related to 3 opioid analgesics: Kadian, Vicodin, and OxyContin.ResultsThe number of posts containing mentions of the target drugs were significantly different [OxyContin (1813)>Vicodin (940)>Kadian (27), P<0.001]. Analyses revealed that these differences were not simply a reflection of the availability of each product (ie, number of prescriptions written). Reliability tests indicated that the content coding system achieved good interrater reliability coefficients (average kappa across all categories=0.76, range=0.52 to 1.0). Content analysis of a sample of 234 randomly selected posts indicated that the proportion of Internet posts endorsing abuse of Kadian was statistically significantly less than OxyContin (45.5% vs. 68.4%, P=0.036, not adjusted for multiple comparisons).DiscussionThese results suggest that a systematic approach to postmarketing surveillance of Internet chatter related to pharmaceutical products is feasible and yields reliable information about the quantity of discussion of specific products and qualitative information regarding the nature of the discussions. Kadian was associated with fewer Internet mentions than either OxyContin or Vicodin. This investigation stands as a first attempt to establish systematic methods for conducting Internet surveillance.

15 keywords...

hide…