Evaluating the Implementation and Dissemination of a Tobacco-free Workplace Program in Homeless-serving Agencies
Introduction
Tobacco use has been causally linked to at least 12 different types of cancer and 30% of cancer deaths in the U.S. Although the national prevalence of smoking has declined from 42.6% to 11.5% over the past 50 years, adults experiencing homelessness continue to smoke at an elevated rate (~80%) and, consequently, suffer from greater smoking-related cancer morbidity. Our prior work has shown that a comprehensive tobacco-free workplace program can increase clinicians’ tobacco knowledge, screening rates, and intervention practices in homeless-serving agencies. Herein, we evaluate program implementation, including penetration, fidelity, and acceptance, as well as program dissemination objectives, within three homeless-serving agencies in Texas. This was funded through CPRIT PP210026 (2021-2023).
Methods
The three participating agencies had 34 employees (intra-agency range=2-17; ~50% clinicians) who served 1,355 adults annually in Zavala, Cameron, and Travis counties. Program objectives for penetration, fidelity, and acceptance were assessed using pre- and post-implementation surveys (compared via Wilcoxon test). Dissemination objectives included webpage reach, program promotion, and technical assistance provision as documented in program records across agencies located in Travis, Harris, and Fort Bend counties.
Results
Penetration: The program provided tobacco dependence and treatment education to 55.8% of employees (vs >/=80% targeted), yielding significant knowledge gains (59.3%; p<0.0001, relative to >/=20% gain targeted). Only one agency attended specialized training (vs 100% agency participation targeted). Fidelity: 1 agency (with 2 employees) withdrew, citing capacity concerns (100% agency retention not met). Tobacco screenings increased over time, from 0% of clinicians participating (pre) to 50% (post) (vs >/=85% compliance targeted; objective not met). Intervention delivery increased over time from 0% of smokers receiving them to 100% (objective of >/=85% of patients receiving an intervention by post met). Acceptability: Among completing agencies, 24.2% (pre) and 23.1% (post) of employees cited concerns about maintaining/sustaining the tobacco-free workplace policy (decrease in concerns over time and </=15% of employees with concerns by post; objectives not met). Of employees familiar with the program at post-implementation, 75% reported satisfaction with the program (objective of >/=80% satisfied not met). Reach: Webpage reach objectives of >/=10/month were achieved (avg=11.8). Promotion: The step-by-step implementation guide was shared with 140 homeless-serving agencies (165 recipients) across Texas; 11 meetings were held (objectives of >/=60 agencies contacted and >/=11 meetings met). Technical assistance: Technical assistance and resources were provided to three additional agencies (objective of engaging >/=2 agencies met).
Conclusion
Dissemination objectives were met but implementation objectives were only partially achieved. Contributing factors included high employee turnover (making it difficult to maintain center contacts who could monitor and help implement programming components), limited agency capacity (as few as 2 employees in 1 case), competing priorities (e.g., food distribution) that were evident from implementation observations that hampered the ability to fully engage in program implementation, and contextual circumstances (a global pandemic and the research team’s institutional change causing disruptions that adversely affected the smooth rollout of programming components). Additionally, small sample sizes and low survey completion rates influenced objective achievement. Future work should focus more on capacity-building of engaged agencies and tailoring programming during implementation (e.g., to accommodate competing priorities) given the challenges encountered in PP210026.