When Less is More: A Web-Based Study of User Beliefs about Buprenorphine Dosing in Self-Treatment of Opioid Withdrawal Symptoms

Document Type

Conference Proceeding

Publication Date



There is growing evidence of an alarming increase in the illicit use of buprenorphine in the U.S., but our understanding of its use is limited because current epidemiologic systems do not systematically monitor buprenorphine. This study aims to explore Web-based data on illicit buprenorphine use, focusing on user beliefs about the appropriate dosing in self-treatment of opioid withdrawal. A web forum that allows free discussion of illicit drugs and is accessible for public viewing was selected for analysis. Posts that contained discussions of buprenorphine and opioid withdrawal symptoms were retrieved using PREDOSE, a novel semantic web platform developed for the information extraction and analysis of social web data on illicit drugs. All unique user names were anonymized. A total of 1,140 posts were retrieved, covering a time period between 2005 and May, 2013. These posts were uploaded to an NVivo database. A random sample of 378 (33%) posts was selected for content analysis. The number of buprenorphine-related posts increased from 46 in 2005 to 1,012 in 2009 and 4,376 in 2011. Over 65% of coded posts that contained information about buprenorphine dose in the self-treatment of withdrawal symptoms, endorsed and/or advocated, use of significantly lower amounts of buprenorphine (2 mg and lower) than typical doses of 16-24 mg per day recommended for standard treatment. Such posts expressed a belief that lower doses of buprenorphine are more effective in the self-treatment of opioid dependence, while the physician-prescribed dosage is too high. Thus, prescribed doses can be 'conserved' or shared with others. Social Web data suggest that the 'less is more' approach to buprenorphine dosing may be fairly prevalent among illicit opioid users and may be one of the contributing factors to the increasing availability of diverted buprenorphine. Our findings highlight the importance of Web-based data in drug abuse epidemiology research.


Presented at the College of Problems of Drug Dependence, San Juan, Puerto Rico, June 14-19, 2014.