PROM, prelabor rupture of membranes.
Changes in risk factors for preterm birth in Western Australia 1984–2006
Article first published online: 11 NOV 2013
© 2013 Royal College of Obstetricians and Gynaecologists
BJOG: An International Journal of Obstetrics & Gynaecology
Volume 120, Issue 13, page 1698, December 2013
How to Cite
Hermans, F., Kamphuis, E. and Mol, B. (2013), Changes in risk factors for preterm birth in Western Australia 1984–2006. BJOG: An International Journal of Obstetrics & Gynaecology, 120: 1698. doi: 10.1111/1471-0528.12410
- Issue published online: 11 NOV 2013
- Article first published online: 11 NOV 2013
- Manuscript Accepted: 3 JUN 2013
We read with interest the article by Hammond et al., that reported that antecedent profiles for all types of preterm birth have changed markedly over the two decades studied, with increasing maternal age and a four-fold increase in the rates of pre-existing medical conditions as main drivers. The authors estimated population attributable fractions to quantify the impact of single and groups of antecedents on preterm birth rates. Multivariate logistic regression was used to calculate odds ratios of risk factors for different subtypes of delivery compared with a normal term delivery.
We calculate from the article that spontaneous preterm birth and preterm prelabour rupture of membranes (PPROM) rates were stable over the two decades (2.7 versus 2.5% and 1.4 versus 1.5%), but medically indicated preterm delivery rates increased from 1.8 to 2.2% (Table 1). The authors presented odds ratios for particular risk factors calculated over the entire period, implying that the rate of a specific risk factor remained constant from 1984 till 2006. Based on this criterion an increase of preterm birth rates was to be expected as the prevalence of such a risk factor increases. This increase was only seen among the medically indicated preterm deliveries. In our opinion, it is very likely that the increase in pre-existing medical diseases is the main cause for the increase of medically indicated births, but the data as presented do not provide insight into this mechanism. It is also important to realise that the authors do not provide information on whether medically indicated preterm deliveries resulted in better maternal or perinatal outcomes.
|Period 1 (1984–1991)||Period 3 (2000–2006)|
|Spontaneous term||106 128||59.3||73 993||45.6|
|Medically indicated term||59 276||33.1||73 151||45.1|
|Medically indicated preterm||3140||1.8||3579||2.2|
|Total||178 894||100||162 283||100|
At issue is also that the authors mention that ‘preterm deliveries could be eliminated after removing six antecedents from the population; up to 50% for medically indicated preterm deliveries, and between 10 and 20% PROM and spontaneous preterm labour'. In our opinion, this statement is challengeable, as many of the pregnancy complications, such as threatened preterm labour, pre-eclampsia and antepartum haemorrhage, are in fact a first sign of the disease process leading to preterm birth rather than true antecedents, i.e. they are risk indicators rather than risk factors.
In this respect, it is important to realise that the authors include all variables in a one-step model. Alternatively, one could imagine a more ‘stepwise’ approach that mimics the occurrence of the subsequent steps leading to preterm birth over time. In such an approach, baseline factors and obstetric history could be used to estimate a baseline risk, while in subsequent steps, the contribution of pre-existing medical conditions and pregnancy complications that precede preterm birth could be modelled. The wonderful data set that the authors have collected would allow the assessment of changes over time in this process, as well as comparison of similar women in their first and second pregnancy.