Treatments for tobacco use, including behavioral counseling and FDA-approved medications, increase the likelihood that smokers interested in quitting will make and succeed in a quit attempt. These medications are safe, including among smokers with psychiatric and medical comorbidities, including cancer, HIV, and cardiovascular disease. Yet, the majority of smokers who try to quit do not use FDA-approved medications or guideline-based counseling in their attempt. Data from Medicaid, Medicare, outpatient medical settings, and primary care show that, at best, only 25% of those interested in quitting use evidence-based tobacco treatments in their attempt. This reality may be worse for populations that traditionally show higher smoking rates and lower access to treatments; 40-75% of people with HIV (PWH) are current smokers, which is >2 times the general population rate. PWH also show very low rates of tobacco treatment utilization, including clinician-based treatments such as the 5As, NRT, and web-based treatments, and varenicline use among PWH is ~4%. The widespread use of antiretroviral therapy (ART) for PWH has substantially improved life expectancy but PWH now lose more life-years to tobacco use than to their HIV infection, primarily from cancer and cardiovascular disease, which account for almost one-quarter of all deaths among PWH.

The investigators have shown that clinician willingness to treat patient smoking is related to clinician bias, including their perceived role in treating tobacco, patient culpability, and effectiveness. In 2 studies, the investigators showed that didactic instruction to counter these biases significantly increased clinician willingness to treat patient tobacco use. In a single-arm study, the researchers showed that a message delivered to oncologists through the EMR that addressed omission bias about tobacco treatment improved actual tobacco treatment rates for cancer patients from 0% to 36%. The study team recently completed a pragmatic RCT testing EMR-delivered patient and clinician nudges directed at omission and status quo biases to promote tobacco treatment in oncology (P50 CA244690). With 2,146 cancer patients accrued, a generalized estimating equation in an ITT analysis showed that the clinician nudge led to about a 3-fold increase in tobacco treatment engagement vs. usual care (35.6% vs. 13.5%; OR = 3.36, 95% CI: 2.12-5.32, p < 0.001). Advanced Practice Practitioners were significantly more likely than physicians to provide tobacco treatment and non-white patients were significantly more likely than white patients to engage in tobacco treatment (p's < 0.001). These findings were presented at the 15th Annual Conference on the Science of Dissemination and Implementation in Health and a manuscript is in preparation.

While these data show a >3-fold increase in tobacco treatment rates from the clinician nudge (vs. usual care), the patient nudge did not increase tobacco treatment rates. Notably, however, the investigators in this study did not conduct a discrete choice experiment (DCE) to identify the patient bias to target with the patient nudge and the team relied on the EMR for patient nudge delivery (~25% of patients are not registered users of the patient portal). Conducting a DCE to identify the bias target and expanding patient nudge delivery to include text messages may affect the impact of the patient nudge in the proposed study.

This study is particularly important because it will generate: 1) implementation strategies informed by behavioral economics and directed at clinicians and patients to increase tobacco use treatment for PWH; 2) optimal EHR-based infrastructure to facilitate tobacco use treatment; and 3) knowledge about which strategies are effective that can be tested more widely. New approaches to increasing tobacco treatment for PWH could lead to benefits at the patient and population levels. Insights from this study could be applied in other clinical settings to increase tobacco treatment, and thus improve patient outcomes in other populations.