Background: Unobserved, or "home" buprenorphine induction is common in some clinical practices. Patients take the initial and subsequent doses of buprenorphine after, rather than during, an office visit. This review summarizes the literature on the feasibility and acceptability, safety, effectiveness, and prevalence of unobserved induction.
Methods: We searched the English language literature for studies describing unobserved buprenorphine induction and associated outcomes. Clinical studies were assessed by strength of design, bias, and internal and external validity. Surveys of provider practices and unobserved induction adoption were reviewed for prevalence data and key findings. We also examined previous review papers and international buprenorphine treatment guidelines.
Results: N = 10 clinical studies describing unobserved induction were identified: 1 randomized controlled trial, 3 prospective cohort studies, and 6 retrospective cohort studies. The evidence supports the feasibility of unobserved induction, particularly in office-based primary care practices. Evidence is weak to moderate in support of no differences in adverse event rates between unobserved and observed inductions. There is insufficient or weak evidence in terms of any or no differences in overall effectiveness (treatment retention, medication adherence, illicit opioid abstinence, other drug use). N = 9 provider surveys assessed unobserved induction: observed induction logistics are seen as barriers to buprenorphine prescribing; unobserved induction appears widespread in specific locations. International guidelines reviewed emphasize clinician or pharmacist observed induction (the United States, the United Kingdom, France, Australia); only one (Denmark) explicitly endorses unobserved induction.
Conclusions: There is insufficient evidence supporting unobserved induction as more, less, or as effective as observed induction. However, the predominantly observational and naturalistic studies of unobserved induction reviewed, all of which have significant sources of bias and limited external validity, document feasibility and low rates of adverse events. Unobserved induction seems to be widely adopted in US and French regional provider surveys. Prescribers, policy makers, and patients should balance the benefits of observed induction such as maximum clinical supervision with the ease-of-use and comparable safety profile of unobserved induction.