Pregnant smokers find online ads more effective than some direct medical interventions – and it’s cheaper too

Alan Martin April 20, 2018

It’s fair to say that online ads – and Facbeook ads in particular – are at a low ebb right now. At a time when the Cambridge Analytica scandal is making people widely question the Facebook business model, a new study from Cambridge University and the University of East Anglia provides a timely reminder that online ads can be an objective force for good – both morally and financially.

Pregnant smokers find online ads more effective than some direct medical interventions – and it's cheaper too

In short, the study found that online ads were both more effective at getting pregnant women to give up cigarettes, and a far more cost-effective way of doing so when compared to our current methods.

Before I get into the “how”, here’s a little bit on the “why”: smoking while pregnant increases the likelihood of miscarriage, stillbirth and premature delivery, as well as being linked to a wide selection of infant health problems. Despite these risks, some 11% of women smoke during pregnancy. This isn’t intended to be a judgemental point – the addictive qualities of cigarettes are hardly up for debate, and giving up is both chemically and psychologically taxing. The health risks though, make bringing that 11% figure down a high priority, not just for the health of future citizens, but also to ensure that money being spent on preventable health problems could be used elsewhere.

Most pregnant smokers want to quit – especially when the health impacts are known and understood, but one system seems particularly promising at getting results. It’s called MiQuit, and is a free text message-based intervention service that pregnant women can sign up to at any time. Previously, MiQuit had been advertised with a leaflet slipped into pregnant women’s maternity booking packs without any extra introduction from the health professional. “As MiQuit is a fully automated and user-initiated digital intervention, women can start using it without the need for any health professional involvement – and this keeps potential implementation costs to a minimum,” Dr Joane Emery from UEA’s school of health sciences tells me via email.

“The ads give very concise information designed to make someone click on them for further information out of interest – that MiQuit is for pregnant smokers, is free and is NHS-supported,” Emery explains. “The intervention itself uses a range of theory-based behaviour change techniques, tailored to the individual user, including: increasing motivation to quit, information about health risks, advice about quit attempt preparation, boosting self-confidence to quit, boosting non-smoker identity, advice about managing cravings and trigger situations, dealing with lapses, etc.”

Meaningful intervention

pregnant_smokers_find_online_ads_more_helpful_than_a_medical_intervention_-_2

This leaflet-based approach resulted in an estimated uptake of around 4% of pregnant smokers, which is a decent start, but clearly had room for improvement. To this end, the researchers began to push MiQuit online in three ways: alongside Google search results for related terms; with targeted Facebook ads and on the National Childbirth Trust and NHS Choices websites (these had the advantage of being free to place.)

The Google and Facebook ads cost, on average, £24.73 for each sign-up, which scaled up to an estimated £736 for each pregnant woman confirmed to have completed the programme. That may sounds like a lot, but it actually compares pretty well to our current best methods, such as offering financial inducement to quit (£1,127 per quitter) or passing confirmed smokers on to specialist NHS services (£952 per quitter.) Suffice it to say that each of these is deemed cost-effective for the NHS compared to dealing with the likely repercussions of not giving up while pregnant – but the MiQuit model is a not insignificant improvement.

In all, 3.4% of people who clicked on the adverts took on the MiQuit program and two thirds of those continued to the end of the 12-week programme – but bear in mind that not everyone who clicked will have actually been a pregnant smoker: even in the shadow of Cambridge Analytica, there’s some things that can’t be known for sure. Emery tells me that for Facebook, the researchers were targeting women of likely child-bearing age. “Our advert was also targeted to women who specified ‘pregnancy’ or ‘childbirth’ as interests,” Emery explains. “We were not able to specifically target pregnant smokers, so we cannot know for sure how many of the people who clicked on our Facebook advert were pregnant smokers.”

Google ads were more targeted, looking at specific search terms such as “stop smoking in pregnancy”, which led to some interesting differences between the two: most intriguingly in the time of intervention. While Facebook ads attracted interest throughout the woman’s pregnancy, around half of all Google interventions occurred during the first five weeks’ gestation. That’s particularly interesting, because currently the earliest offline cessation interventions are aimed during the 8-12 week mark.

Could MiQuit be more impactful if it wasn’t just tied to SMS messages? Facebook Messenger, for example, could follow women from device to device. There’s no reason why it couldn’t be ported, Emery says, but for now SMS is best purely because it remains a (virtually) universal way of talking to women – even those without smartphones. “One advantage currently of SMS is that all phones work on it (so not excluding those who don’t own a smartphone) and no data/wifi/internet connection is required. We currently feel that SMS works best but MiQuit could be moved to another platform if usage was to decline.”

This kind of intervention could be expanded to all kinds of public health areas from diet to sexual health, but it’s a particularly good fit for smoking and pregnancy because of the stigma attached. As Emery says: “online advertising might be particularly helpful for pregnant women who smoke given its anonymous nature,” as they may not be comfortable disclosing to a potentially judgemental third party – even if it is a medical professional.

What the research doesn’t reveal is the best way to approach people who aren’t actively looking for help, but a life saved is a life saved. It’s a timely reminder that not all ads are created equal – and while using personal data to change voting behaviour will leave many feeling queasy, Google and Facebook can be extremely helpful on these type of issues – areas which are quite literally matters of life and death.