We need better data on smoking in pregnancy

My work as a smoking cessation midwife involved me directly with the initial national target to reduce smoking in pregnancy from 23% in 1995 to 18% by 2005 and now to 15% by 2010. An additional requirement is to reduce the rate of mothers who are smoking at delivery by 1% year on year, specifically focusing on disadvantaged women to tackle inequalities in infant mortality. There is considerable pressure to meet this target, which the Healthcare Commission is monitoring. Reducing smoking in pregnancy is also a health objective for Sure Start Children’s Centres.

The 1% annual target was set centrally but has not been discussed with key stakeholders involved in its implementation. These stakeholders are now concerned by how realistic the target is and whether it is achievable within the required time scale. How was the target derived? Why has good practice guidance not been issued in support? Each primary care trust has a plan specifying how it intends to decrease the percentage of mothers who are smoking at delivery year on year and this is how their performance is managed. However, there are concerns about the quality of current smoking data, making it difficult to set local targets and baselines and to monitor progress.

Reducing inequalities is proving challenging; the latest infant feeding survey shows marked variations in smoking in pregnancy by mother’s socioeconomic classification and age. In 2005 mothers in routine and manual groups were four times more likely than women in managerial and professional occupations to have smoked (29% versus 7%). Teenagers were also five times more likely to have smoked than older mothers (45% versus 9%).

Over the past five years rates have consistently increased in these two groups. Tackling social disadvantage is a wider problem than training staff in cessation skills. Pregnant women who smoke get less support and have fewer financial resources, more family problems, less residential stability, and more psychological and emotional problems than non-smokers. Unsurprisingly, smoking is often used in these circumstances to relieve stress. Besides their multiple disadvantages, many women are typically highly nicotine dependent and their partners or family also smoke, which is a barrier to stopping.

Historically the infant feeding survey monitored the national target, but it is only undertaken every five years and cannot provide local information. To provide a more timely and regional breakdown of the number of mothers smoking at delivery, all National Health Service hospital trusts with maternity services should collect data on smoking. However, since 2003-4 the basic collection of smoking data has been reported to be incomplete (more than 5% unknowns) and inconsistent (doesn’t add up; see www.dh.gov.uk/tobacco).

Although there were improvements last year, and 87% of the 152 primary care trusts submitted data, which met the Department of Health’s minimum standard for data quality, there are still problem areas. Furthermore, in a recent area audit of data on smoking at Good Hope Hospital, part of the Heart of England NHS Foundation Trust, a number of postnatal women were contacted to determine whether their recorded smoking status at delivery was correct. In a considerable number of cases this was not reassessed, resulting in inaccurate recorded data, which is probably a national problem.

Another weakness in the system of data collection is that the data are based on self reporting and should be interpreted with caution. Smoking has become less acceptable as a social habit, and with greater public knowledge of the risks of smoking in pregnancy and pressure to stop, pregnant women may be less inclined to admit to smoking. So the true prevalence is almost certainly unknown. We need to debate the use of measurements of cotinine in saliva or urine in maternity units to validate smoking behaviour objectively. Attention to smoking behaviour in this way should be as routine a part of antenatal care as the blood pressure check with testing during pregnancy and at delivery to identify changes and measure progress. Although cotinine testing is a more reliable measure, increasingly breath testing for carbon monoxide at the booking visit is being introduced into routine midwifery practice.

Though these measures have practical and cost implications it may help to increase the trustworthiness of data and clarify the true scale of the problem. For example, the latest infant feeding survey, which is based on retrospective and self reported data, shows that in 2005 17% of mothers in England smoked throughout their pregnancy. (Interestingly, this is in line with the interim government target.) Whereas in June the Department of Health reported that throughout England smoking rates at delivery last year ranged from 4.4% for Richmond and Twickenham Primary Care Trust to a staggering 38.2% for Blackpool Primary Care Trust.

Clearly the target setting process is flawed because it relies on the effectiveness of current data collection and monitoring mechanisms. Given these inherent problems it is difficult to establish accurate baselines, set local targets, monitor progress, and compare countries. My view is that the resultant targets are unreliable and unrealistic, which can be demotivating and frustrating and cause needless stress for the people involved.

Clearly the target is challenging, according to the Department of Health figures, published in June, only a quarter of Primary Care Trusts achieved the target for a 1% reduction in smoking at delivery in 2005-6 and 2006-7. Collecting quality data isn’t just about meeting targets, it is key to knowing whether our interventions are improving health. Data inadequacies aside, the latest infant feeding survey figures show that smoking is still common, especially among young mothers and mothers in lower socioeconomic groups, who require proactive targeting and tailored resources.

At the coalface we have experienced some successes; supporting women in difficult circumstances—that is, women with material and emotional pressures that work against stopping; preventing postnatal relapse back to smoking; supporting women in the third trimester of their pregnancy; and supporting partners and others. Not all of these may be measurable in the target driven NHS.

I thank the consultant obstetrician and gynaecologist Michael Moloney for his help and support in writing this article and for finding time to read and comment on early drafts.

Competing interests: The author is employed by Good Hope Hospital and Birmingham East and North Primary Care Trust as a smoking cessation midwife; is a member of the programme development group for forthcoming National Institute for Health and Clinical Excellence guidance on smoking cessation; is a member of the International Network of Women Against Tobacco; is the Network Coordinator for the West Midlands Stop Smoking in Pregnancy network and In2focus and Pfizer have sponsored network meetings; and has had conference travel fees reimbursed by In2focus.

Pregnant women who smoke get less support and have fewer financial resources, more family problems, less residential stability, and more psychological and emotional problems than non-smokers