- The main randomised comparison did not establish a statistically significant difference at 12 months.
- A secondary analysis among people who received interpersonal support gave a more favourable result.
- Workforce pressures, missed appointments and administrative problems affected delivery.
Inside this story 2 perspectives
One of the most revealing figures in a UK study of support after childbirth is the number of people the programme struggled to reach. About one in five participants assigned to BabyBreathe missed the interpersonal element. Health visitor shortages, appointment difficulties and administrative errors all played a part. That makes a difference to the answer a family, a health visitor or the manager of a health service can take from the research.
The trial, published in The BMJ on 18 August, recruited 887 participants who had already stopped smoking before or during pregnancy. The researchers tested whether a package of personal and digital support could help them remain smoke-free after giving birth. They randomly allocated people to receive BabyBreathe or usual care. The study therefore addresses the maintenance of a change participants had already made.
The main comparison includes the missed appointments
At twelve months, 242 of 441 people assigned to BabyBreathe and included in the main comparison met the study’s measure of continuing abstinence. In usual care, 222 of 445 did. That is about 54.9% against 49.9%. The uncertainty around the estimated difference included the possibility of no benefit. The primary analysis did not establish a statistically significant improvement.
The study combined participants’ accounts with a measurement of carbon monoxide in their breath. Its main analysis kept people in their originally allocated groups even when part of the support did not arrive. That preserves a central advantage of random assignment. It also answers a practical question for a service deciding whether to introduce a programme: what happened when a real system, with its existing workforce and competing demands, tried to offer it?
A second look at the people reached
An additional analysis considered the people in the programme group who actually received the interpersonal support as intended. Among those 347 participants, 57.6% met the twelve-month measure, compared with 49.9% in the usual-care group. That comparison was statistically significant, but the researchers added it after the trial plan. The people reached may also differ from those who missed out. Restricting the analysis in that way makes it harder to separate the effect of the contact itself from the circumstances that made the contact possible.
A service thinking about testing the approach again could learn from both findings. It could record how soon the first appointment is offered, whether the family can keep it and how often a missed visit can be rearranged. Such records would help distinguish an invitation being sent from support actually reaching a household. That distinction is especially important when the intervention is an ordinary conversation with someone trained to help, at a time of considerable change for a family.



