In a health center that serves an economically disadvantaged community that is strongly affected by the opioid epidemic, people treated for opioid dependence with home-based buprenorphine inductions were as likely to be retained at 30 days as those with traditional office-based inductions. We hypothesized that home-based inductions would be associated with greater treatment retention based on research on chronic disease management that highlights the benefits of greater patient involvement and development of patient self-management skills (Bodenheimer, Lorig, Holman, & Grumbach, 2002
; Hibbard & Cunningham, 2008
; Lorig et al., 2001
; Lorig, Sobel, & Stewart, 1999
; Mosen et al., 2007
). However, our data did not support this hypothesis and did not indicate superiority of either induction type on this short-term outcome in this observational study. There was also no indication of superiority of either induction type on experiencing difficulties with the induction process.
Because this was an observational study, in which patients and providers selected the treatment (office- vs. home-based induction) that they decided would be best, we believe the findings of this study reflects the success of the patients and providers in our study in selecting the buprenorphine induction type that was best suited to patients’ needs, abilities, and beliefs about health care. Our results indicate that, in most cases, the decision about which induction type to use was made appropriately because 78.3% of the sample was retained at 30 days and approximately equal proportions were retained in each induction group. Given the importance of the induction process for longer term opioid addiction management, these data should be encouraging for providers who are considering treating opioid-dependent patients in the primary care setting. Most patients receiving office- or home-based inductions successfully completed the induction process.
We also did not detect differences in 30-day retention by induction type in multivariable analyses that adjusted for patient characteristics by induction type (AOR = 1.10, 95% CI = 0.43–2.78). We identified only two patient characteristics that were associated with 30-day retention: having an insurance type other than Medicaid and having used street methadone in the past 30 days. Patients who did not have Medicaid probably had fairly stable income sources (and therefore were privately insured or self-paid) and were more likely to be retained. Patients who used street methadone in the 30 days before induction were also more likely to be retained. We reported a similar result in an earlier evaluation of our buprenorphine treatment program (Cunningham et al., 2008
) and believe that this finding is worthy of future exploration. Our clinical experiences with this population indicate that patients presenting for buprenorphine treatment who used street methadone rather than heroin or prescription opioids tended to describe their use as self-medication for opioid withdrawal symptoms rather than a desire to “get high.” These patients tended to be highly motivated to find a better treatment approach for their opioid dependence. It will be important to examine whether these factors are also associated with short-term retention in a buprenorphine treatment program in an experimental research design.
Unfortunately our study, based on post hoc medical record review data, was unable to assess fully the social and clinical factors that may have contributed to patient and provider decisions about the appropriate induction type. Factors such as level of support, stability, and severity of substance abuse should play a role in determining which patients will be comfortable and successful with a treatment strategy that draws on self-management skills (Street, Gordon, Ward, Krupat, & Kravitz, 2005
; Tobias, Cunningham, Cunningham, & Pounds 2007
) and therefore should influence decisions about the appropriate induction type for particular patients. Our study identified ethnicity, employment status, type of insurance, history of crack/cocaine or sedative/benzodiazepine use, and prior buprenorphine use as being associated with type of induction. These characteristics may be markers for the underlying factors mentioned above.
There were a number of limitations of this study, the primary one being the limited sample size and power to detect differences. Second, because our data were based on reviews of medical records of routine clinical care rather than data collected for research, we were limited by the type and number of factors that we could examine. For example, urine toxicology tests were conducted based on clinical judgment rather than a standardized research protocol; therefore, we could not compare the two groups of patients on other important outcomes like drug use behaviors. Third, as we follow patients over a longer period, it is possible that differences in outcomes between the two groups might emerge. For example, as patients face ongoing drug craving, interact with others who use opioids, and have challenging life experiences that were previously associated with drug use, the benefits of patient self-management may increase.
Finally, although an advantage of our study is that it compares induction strategies among patients recruited from the same health center population, it is possible that patients at this health center differ from the general population of opioid-dependent patients, and therefore, our study may not be generalizeable to a larger target population. Therefore, the specific characteristics that are markers for level of support, stability, and severity of substance abuse might differ in different populations.
We conclude that both office- and home-based buprenorphine inductions are feasible in the primary care setting. In our study, treatment retention of 78.3% at 30 days is similar or greater than that reported in previous studies (e.g., Alford et al., 2007
; Cunningham et al., 2008
; Fiellin et al., 2002
; Lee et al., 2008
; Mintzer et al., 2007
; O’Connor et al., 1998
; Stein, Cioe, & Friedmann, 2005
; Sullivan et al., 2006
) and adds to the growing literature that demonstrates that patients can successfully undergo buprenorphine induction in office- or home-based settings. In contrast to our hypothesis that a strategy that encourages patient self-management from the earliest stage of treatment for opioid dependence would result in better retention, our data do not support either office- or home-based inductions as being superior with regard to short-term patient retention. Our evaluation of differences between patients who are retained at 30 days and who initially chose office- vs. home-based inductions indicate that, in most cases, patients and providers were able to appropriately assess patients’ needs and self-management abilities and made appropriate induction strategy decisions.
It is crucial to continue to develop and evaluate innovative treatment models to successfully address the growing opioid epidemic. Providing buprenorphine treatment in the primary care setting offers opportunities to apply lessons from chronic disease management described in the CCM for treating opioid dependence.