“Early screening of pregnant women could save 'more than 1,000 premature births a year',” is the headline in the Daily Mail . This is based on comments from British obstetrics and gynaecology consultant Dr Ronnie Lamont, who reportedly suggested that “the links between infections and premature birth are so strong that women should be routinely screened around the 15th week of pregnancy – and given antibiotics if needed”. His comments follow a US study in over 100 women, which found that 15% of women who go on to give birth prematurely have amniotic fluid which is infected with bacteria or fungi.
Although this study provides some evidence about how common infections of the amniotic fluid are among women who experience preterm labour, it did not look at women who did not experience premature labour or look at the effect any antimicrobial treatment may have had. Therefore, by itself, this study is not sufficient evidence to determine whether a microbial screening programme could help to prevent premature births.
Where did the story come from?
Dr Daniel DiGiulio and colleagues from Stanford University School of Medicine and other universities and medical centres in the US carried out this research. The study was funded by the National Institute of Child Health and Human Development and the National Institutes of Health. It was published in the peer-reviewed open-access medical journal: PLoS One .
What kind of scientific study was this?
This was a cohort study looking at whether there were any differences in the microbes (mainly bacteria and fungi) found in the amniotic fluid of women who gave birth prematurely and those who carried their babies to term.
The researchers searched the database of one hospital (Hutzel Women’s Hospital) in Detroit to identify women who had been admitted with spontaneous preterm labour (at least two regular contractions every 10 minutes along with cervical changes, before 37 weeks into the pregnancy) and intact membranes (that is, their waters had not broken) between October 1998 and December 2002. The researchers only included women who had undergone sampling of their amniotic fluid (sampled using a needle though the abdomen, not via the vagina) for testing for the presence of microbes, and for whom there was sufficient extra fluid available for the tests the researchers wanted to conduct. Women carrying more than one baby (for example twins) were excluded, as were women who had not delivered at the hospital, and those whose babies had been found to have major abnormalities.
The researchers identified 166 women who met their inclusion criteria. They then grouped these women into those who had delivered prematurely and those who had gone on to deliver at term. As part of the amniocentesis carried out on women early in their pregnancy, amniotic fluid taken from the women had been tested in various ways, including testing for microbes by culturing the fluid in the laboratory and seeing what organisms grew. Any amniotic fluid left over from this process had been stored in the freezer after these procedures. The researchers took this stored fluid and used a different set of tests to look for microbes. These tests involved using a technique called PCR to find any microbial DNA in the amniotic fluid. This technique identifies specific pieces of DNA, and is very sensitive to even very small amounts of DNA. If any DNA was identified, the researchers then looked at its sequence (the order of the four building blocks called nucleotides that make up the DNA) to find out what type of microbe the DNA belonged to.
The researchers compared the outcomes between the groups of women. These outcomes included the presence of signs of inflammation in the amniotic fluid (such as white blood cells), signs of inflammation of the membranes surrounding the foetus or umbilical cord (chorioamnionitis or funisitis), pregnancy outcome (such as gestational age at birth and length of time from amniocentesis to birth) and outcomes in the newborn (such as complications at birth, including newborn death). They also looked at whether the evidence supported the possibility that the microbes caused preterm delivery, by looking at the timing of infection, level of infection and location of infection. Analyses were adjusted for other factors that might affect results, such as maternal age, gestational age at amniocentesis and cervical dilation at admission.
What were the results of the study?
Among the 166 women that experienced preterm labour, about two thirds (113 women) delivered prematurely and one third (53 women) carried their babies to term. Ten of the women had clinical signs of inflammation of the membranes surrounding the foetus (chorioamnionitis). The researchers found that 25 women (15%) had evidence of microbes in the amniotic fluid either on standard culture testing or by using the PCR technique. Most of these microbes were bacteria, which were identified as belonging to 17 different species. There were a few cases of fungal infection, with only one species found. PCR picked up 19 infections, nine of which had not been picked up by standard culture testing. Standard culture testing picked up 16 infections, six of which were not picked up by PCR.
Women who tested positive for microbes on PCR were more likely to have chorioamnionitis or funisitis than those who tested negative. All women who tested positive by PCR or culture went on to deliver prematurely. Although there was a trend towards more newborn complications among women who tested positive, this association did not reach statistical significance. Women who tested positive had shorter intervals between amniocentesis and delivery than those who tested negative. Women who had higher concentrations of bacterial DNA in their amniotic fluid were associated with earlier delivery.
What interpretations did the researchers draw from these results?
The researchers concluded that the amniotic fluid of women who undergo premature labour contains a greater variety of microbes than previously thought. They also conclude that their findings support the hypothesis that these microbes may be causing premature delivery.
What does the NHS Knowledge Service make of this study?
There are a number of points to consider when interpreting this study:
- The study was relatively small (in particular, there were few women who experienced premature labour but then went on to deliver at term) and carried out retrospectively. The authors themselves acknowledge that their study could not prove that microbial infection causes premature delivery and that larger prospective studies would be needed. There are a large number of potential causes of premature delivery, including cervical and uterine structural features, excess amniotic fluid, under-nutrition and chronic disease in the mother and previous premature births.
- Because the PCR technique is very sensitive, it is especially susceptible to contamination. Although steps were taken to minimise the risk of contamination, this may still be a problem.
- The samples that were tested by PCR had been stored in the freezer for between two and six years and some of the DNA could have broken down in that time.
- This study only included women who experienced preterm labour and therefore the results may not apply to women who do not.
- Amniocentesis carries a low level of risk to the foetus; this means it would be unlikely to be used as part of a general screening programme for all pregnant women.
- Although some women in preterm labour did show evidence of microbial infection, the majority (85%) did not. Therefore, in this study at least, most of the women who went on to deliver prematurely would not have been picked up by these tests.
This study did not look at the effect of antimicrobial treatment on pregnancy or newborn outcomes. Therefore, it is not possible to say for certain from this study that treating these infections would be able to reduce preterm births or complications in the newborn, or what the risks of these treatments might be.
Analysis by Bazian
Edited by NHS Website
Links to the headlines
Daily Mail, 26 August 2008
Links to the science
Cochrane Database Syst Rev 2002, Issue 4
Cochrane Database Syst Rev 2002, Issue 4
Cochrane Database Syst Rev 2008, Issue 2
Cochrane Database Syst Rev 2007, Issue 1
PLoS ONE 2008; 3(8):e3056