Friday, 3 March 2017

The challenges (and joys) of evaluating babyClear©: a package of support to help pregnant women to stop smoking

Guest post by Sue Jones, Research Associate, Teesside University

A team of Fuse researchers from Newcastle and Teesside Universities published findings from the babyClear© study a few weeks ago and I thought that I’d put finger to keyboard to share with you the challenges and joys of evaluating the roll out of this innovative intervention.

In 2012, I became involved with evaluating babyClear©, a package of support for maternity and stop smoking services, designed to help them to deliver the stop smoking message more effectively to pregnant women. BabyClear© was due to be rolled out regionally across North East England and evaluated throughout, which presented a number of challenges:
  • Challenge 1: different research questions – we wanted to know if this new approach worked and would it help women quit but we knew that this would not be enough; we wanted to understand what influenced those figures, and what healthcare staff need to do to be most effective.
  • Challenge 2: ethical dilemma – ethically we could not deny pregnant women a test like carbon monoxide monitoring that was known to improve outcomes to some degree, so the regional rollout of babyClear© offered a prime opportunity to evaluate the intervention using a natural experiment1.
  • Challenge 3: wide variety of stop smoking delivery models – the extent of austerity measures experienced by the public sector has been far greater than anticipated when the research was envisaged in 2011. At the same time responsibility for delivery of stop smoking services has been moved to local authorities who themselves are under extreme pressure to reduce spending. This has created a wide variety of stop smoking delivery models, all trying to provide a low cost service but with implications for the implementation. For example: babyClear© was designed to be a package that could easily slot into existing services, however it assumed a number of systems were standard when they were not, such as a midwife available at dating scan appointments and a local stop smoking specialist in pregnancy. All those Heinz 57 varieties of stop smoking service delivery models and systems within maternity services, each one different from every other, made it logistically challenging to implement the new pathway, leading to delays of varying lengths in each Trust area.
  • Challenge 4: researching within a changing system – due to ongoing changes largely in the delivery of stop smoking services, but also in maternity, and their impact on the implementation of babyClear©, data collection plans had to be re-thought again ... and again ... and again to reflect what was happening out in the real world! 
We were greatly helped in approaching some of these challenges by the publication in 2014 of the Medical Research Council (MRC) Guidance on process evaluation of complex interventions. Using this guidance, we were able to start re-shaping our thinking in terms of how the qualitative data could be used synergistically with the numerical data. We set about strengthening the methodology with a retrospective logic model, weaving contextual data into the mix and with an eye on the mechanisms of impact.

After overcoming these challenges, along came the joys: the findings of our study proved that babyClear© was not only effective but also cost-effective, which was a great achievement in such a short timescale. This new approach, which supported midwives to offer universal carbon monoxide screening and refer pregnant smokers quickly to expert help, nearly doubled quit rates.

The findings highlighted that we could systematically help women to stop smoking in pregnancy which will result in already well-evidenced outcomes such as:
  • Help mothers have babies who are heavier and healthier than if they continued smoking
  • Help more mothers lead healthier lives
  • Help mothers live longer and see their children grow up
  • Help the children to live and run and grow up surrounded by smoke free air; and 
  • Enable them to not be held back by smoking-related poor health
So have a read of our paper, this has the nitty-gritty of the statistical outcomes.

Importantly, soon we hope to be publishing the details about the how, what, when, where, why questions that were the focus of the qualitative process evaluation. Without this it is difficult to know how to implement it elsewhere to best effect and why it works well in one place and not another.

Celebrate our findings with us; if the maternity and stop smoking services are able to use the babyClear© approach to implement best practice/national guidance it can offer the support that is needed so that more women stop smoking during their pregnancy than did before. So keep your eyes peeled for my next blog – which will focus on the findings from the process evaluation.

  1. “A natural experiment is an empirical study in which individuals (or clusters of individuals) exposed to the experimental and control conditions are determined by nature or by other factors outside the control of the investigators, yet the process governing the exposures arguably resembles random assignment”. (Reference:     More info: Craig P, Cooper C, Gunnell D, Haw S, Lawson K, Macintyre S, Ogilvie D, Petticrew M, Reeves B, Sutton M, Thompson S. Using natural experiments to evaluate population health interventions: new Medical Research Council guidance. J epidemiol commun h. 2012 May 10:jech-2011.
Related content:

No comments:

Post a Comment