P3.Q321: Population-level impact of an automated tobacco treatment delivery system on household tobacco smoking rates

Background: Helping parents quit smoking is a public health priority. However, parents are rarely, if ever, offered tobacco treatment through pediatric settings. Our objective was to evaluate the population impact on self-reported household tobacco smoking rates of an automated tobacco treatment delivery system within the pediatric EHR that screens, motivates, and automates evidence-based treatment for parents who smoke.

Methods: As part of an NIH-funded cluster-randomized controlled trial, we deployed a pre-visit household tobacco use screener completed at preventive visits by parents in 6 intervention and 6 control pediatric primary care practices from July 2021-February 2023. This screener assessed parent reports of household tobacco use in both arms. Intervention arm parents were automatically enrolled in the quitline, smokefreeTXT and assigned receipt of NRT unless they opted out of these treatments. Our main outcome was both positive and negative state changes from baseline to follow-up in household smoking status in intervention and control arms as obtained from the screener. Positive changes were defined as going from parents’ report of “Only Me” as the smoker to “Nobody” in the Household; “Me and Someone Else” going to “Only Me” or “Only Someone Else”; and “Only Someone Else” going to “Nobody.” Negative changes were defined as “Nobody” going to “Only Me,” “Nobody” going to “Only Someone Else” or “Me and Someone Else”, “Only Me” going to “Me and Someone Else” and “Only Someone Else” going to “Me and Someone Else.” We analyzed the data at the child level. To determine the isolated effect of the intervention on households reporting smoking at baseline, we examined only positive changes as well. To assess comparability of data collection between conditions, we assessed mean days between baseline and follow-up in each arm.

Results: A total of 34,027 intervention and 32,129 control arm parents completed the household screener at baseline and had at least one follow-up visit. Of these, 3,478 (10.2%) reported having at least one household smoker at baseline in the intervention vs. 3,363 (10.5%) in the control arm. Overall, the intervention arm had 1.6% net positive changes vs. 0.3% in the control arm (p=0.022, Table 1). Among households with at least one smoker, there were 4.81% more positive changes in the intervention arm than control (45.3% vs 40.5% P<.001) see Table 2. The mean days of follow up between arms was 489 and 492 in the intervention vs. control arms.

Conclusion: An automated, systematic tobacco use screening and treatment delivery intervention led to a small but significant reduction in population-level household smoking rates in the intervention arm compared to control arm. Our results suggest that deploying automated parental tobacco treatment throughout pediatric settings would likely decrease family tobacco use over time.