Harmful air where we should feel most safe: a multi-agency quality improvement project to reduce passive smoking outside a tertiary neonatal and maternity unit

Why did you do this work?
Smoking is the largest driver of health inequalities.1 Smoke free sites, especially where there are children, are widely supported,2 and help break societal norms perpetuating smoking.1 Despite our Smoke Free pledge3 many still smoke on hospital grounds, especially at the maternity/neonatal entrance. This impacts on user and staff experience, litter, complaints, and smoke and pollution4 5 exposure. This project aimed to reduce smoking episodes at our maternity/neonatal entrance by 25% over 2 months, without increasing aggression incidents, with a hypothesised reduction in air pollution.

What did you do?
We formed a multi-agency team, including public health, quality-improvement, communications, estates, air quality and sustainability expertise – alongside community and hospital smoking-support teams.

Following an audit (December 2023), we ran a two-way factorial design to evaluate efficacy of a pre-existing tannoy and/or newly piloted twice-daily tobacco-dependency advisors at the neonatal/maternity hospital entrance (June-July 2024). Smoking episodes over 10 minutes (outcome measure) were counted, nine times a day, 3 days a week for 7 weeks, with analysis via run charts and StudyIt software. Particulate matter (PM)2.5 and PM10 air pollutants were also recorded and compared with smoking data (process measure).

Stakeholder engagement included a large-scale survey (post pilot) of patients, public and staff advertised via Trust communications, posters, and social media. Smoking related Datixes and complaints (2021–2023) were reviewed alongside advisor feedback on aggression (balancing measure).

What did you find?
Baseline smoking was high (0–9/10 minutes, median 4). We demonstrated two shifts, a trend and a 38% reduction in median from baseline (figure 1). Both interventions reduced smoking, with advisors being most effective (figure 2). A significant correlation was found between particulate matter and traffic, but not smoking.

Of 1,811 survey respondents, 7% were parents of an inpatient/outpatient child. Smoke free policy support was high (mean 4.52/5) with many comments regarding children. Awareness of support services was low (47–61%), as was staff confidence in approaching smokers (mean 2.06/5) – 40% would take up further training.

Complaints at the neonatal/maternity entrance were high (7/26). Of 289 Datixes, aggression ranked top, although advisors experienced no aggression during this project.

To sustain change, we are implementing smoking advisors at entrances, and exploring new tannoy and comms messages incorporating patient voice. We have positive feedback on piloted staff training which will be promoted across neonates and paediatrics.