National targets to reduce smoking in pregnancy are unreliable and unrealistic because they are based on incomplete data, argues a midwife in this week’s BMJ.

The initial national target was 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.

Yet, according to Department of Health figures published in June, only a quarter of Primary Care Trusts achieved this target in 2005-7. Carmel O’Gorman, a midwife specialising in smoking cessation at the Good Hope Hospital in Birmingham, is concerned by how realistic this target is and whether it is achievable within the required timescale.

The quality of current smoking data makes it difficult to set local targets and baselines and to monitor progress, she warns. The latest infant feeding survey shows marked variations in smoking in pregnancy by mother’s social class and age. But this survey 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, she believes that all hospital trusts with maternity services should collect data on smoking, though she acknowledges there are problems with data recording. Another weakness is that the data are based on self-reporting and should be interpreted with caution, she warns. She suggests measuring cotinine levels (a by-product of nicotine) during pregnancy and at delivery to help increase the trustworthiness of data and clarify the true scale of the problem.