In response to a patient vignette, few HIV providers attending HIV conferences in six US cities recommended buprenorphine treatment in primary care. One quarter of providers recommended buprenorphine in primary care for the heroin-only using vignette patient; only 9% recommended it for a more complex vignette patient with heroin and cocaine use. The paucity of provider endorsement for buprenorphine treatment in primary care reflects the significant work that remains in the area of provider training and systems change for buprenorphine to become a truly accessible treatment option for opioid-dependent HIV-infected patients nationwide. Several factors associated with endorsement of buprenorphine in primary care may be used to target training efforts to particular groups of providers and to modify current training strategies to further enhance availability of buprenorphine.
As expected, physicians experienced in prescribing buprenorphine were more likely to recommend buprenorphine in the patient vignettes compared with nonprescribers. Thus, creating additional opportunities and incentives for providers to gain experience, through mentorship (such as the national Physician Clinical Support System, www.pcssmentor.org
) or via the creation of quality benchmarks in screening, referring, and/or treating substance abuse may improve access. Residency training may be an optimal time to introduce and model buprenorphine treatment to trainees in primary care; a number of curricula and other resources are available to assist faculty members to teach about buprenorphine.22,23
Generalist physicians were more than twice as likely to endorse treating the vignette patient with buprenorphine in primary care as were infectious disease physicians. This finding was similar to that of a prior study in which faculty and resident physicians from primary care training programs (family medicine and internal medicine) were more likely to have positive attitudes toward buprenorphine prescribing than non-primary care-trained faculty and residents.24
Further, integrating buprenorphine treatment into primary care has been found feasible and efficacious in a number of observational and experimental studies.25–27
To promote appropriate treatment of HIV-infected and opioid-dependent patients in primary care, generalist physicians may be an ideal target group, receptive to integrating such treatment into routine care.
We also found that providers who identified as African American or black were three times as likely as white providers to endorse buprenorphine in HIV primary care. Minority physicians are more likely to work in underserved areas than white physicians and to provide care to vulnerable and underserved patients.28,29
Because of their experience, they may be more sensitive to the need to treat opioid dependence among HIV-infected patients in the communities they serve. Further, physicians who practice in underserved areas may have less readily accessible specialty substance abuse services, and therefore primary care treatment may the more realistic or even the only treatment option. One strategy to promote expanded access to buprenorphine in HIV primary care might be to target providers who work in such underserved areas, who may be more receptive to offering treatment for opioid dependence to their patients.
We found no association between the vignette patients’ race or gender and provider recommendation for buprenorphine in primary care. Our study may have had insufficient power to detect such differences, or the vignette study may have failed to elicit provider biases. National data demonstrates that, compared with patients receiving methadone, patients receiving buprenorphine for opioid dependence are more likely to be white than non-white. Whether this disparity represents differences in access to care or frank racial biases is uncertain.30
The health disparities literature, however, provides ample evidence that racial and ethnic minorities, and women, are less likely to be offered and/or accept recommended treatments for such conditions as coronary artery disease, end-stage renal disease, and arthritis.31–35
Further, racial and ethnic minorities are less likely to be offered opioids for treatment of acute pain syndromes.36,37
Despite our findings, we believe that ongoing assessment of clinical decision making in buprenorphine treatment is warranted to avert disparities in access to buprenorphine for persons of color.
Our study had several limitations. We acknowledge that the use of patient vignettes may not represent actual clinician decision making or treatment recommendations. In addition, the form of our question, in which participants were asked to choose only the single option that “best served” the hypothetical patient, may have underrepresented the degree to which participants may have considered buprenorphine in primary care an acceptable option. Such hypotheticals, however, increasingly have been used in the medical literature to study physician behavior and decision making. In a number of quality of care studies, physician responses to vignettes have been found more accurate than chart review in predicting physician behaviors.38,39
The authors of these studies argue for the utility of vignettes in assessing quality of care to control the variability in illness severity and patient characteristics.
Our low response rate also limits the generalizability of our findings. The HIV providers who completed the survey may have been more interested in the issue of substance use disorders and buprenorphine and therefore more likely to recommend buprenorphine treatment. Therefore, we believe that our findings may over-estimate the degree to which HIV providers endorse buprenorphine treatment in primary care, although without confirmatory studies this is speculative.
In summary, our study of provider and patient factors associated with endorsement of buprenorphine treatment in primary care found that HIV providers rarely endorsed treatment of opioid dependence with buprenorphine in primary care for vignette patients. To promote integration of buprenorphine into routine medical care requires additional work. Our finding that African American providers are more likely to endorse buprenorphine in primary care may indicate their greater experience in working with underserved communities where opioid dependence is identified as a common challenge. Generalist physicians who provide HIV care may be a willing target audience to promote uptake of buprenorphine in primary care. Because integrating HIV and opioid addiction treatment has important individual and public health implications, further study of and interventions with physicians who are receptive to buprenorphine treatment in primary care settings may be one strategy to improve the integration of HIV and opioid addiction treatment.