Tocolytics: do they actually work?

BMJ 2012; 345 doi: http://dx.doi.org/10.1136/bmj.e6531 (Published 09 October 2012) Cite this as: BMJ 2012;345:e6531
  1. Zarko Alfirevic, professor of fetal and maternal medicine
  1. 1Department of Women’s and Children’s Health, University of Liverpool, Liverpool Women’s Hospital, Liverpool L8 7SS, UK
  1. zarko{at}liv.ac.uk

Never mind “which one works?” We still don’t know if delaying delivery is safe

In a linked research paper (doi:10.1136/bmj.e6226), Haas and colleagues assessed the relative effectiveness of various tocolytic drugs aimed at preventing preterm birth.1 Being born too early may have serious detrimental effects on health. It is hardly surprising that extremely preterm children, born before 26 weeks’ gestation, who survive to school age have significantly lower scores for cognitive ability, reading, and mathematics.2 However, even babies born a few weeks before term are more likely to underperform on most measures of general health than their peers born at term.3 With an estimated 14.9 million babies born preterm each year (11.1% of all live births worldwide), the burden of preterm births on families and healthcare systems in high income and low income countries is considerable.4 It is important to remember that some preterm births are iatrogenic. They are a consequence of labour induction or prelabour caesarean sections, which are often life saving interventions for mother and baby, particularly in severe pre-eclampsia, intrauterine growth restriction, or placental abruption. Such iatrogenic preterm births should be clearly distinguished from preterm births that result from spontaneous or “true” preterm labour.

Women in spontaneous preterm labour present with uterine contractions, ruptured membranes, or both, and most obstetricians make every effort to stop such labour. But is it really possible to stop true preterm labour? The basic idea is to restore uterine quiescence by giving drugs that will prevent uterine contractility. These drugs include oxytocin antagonists, prostaglandin inhibitors, calcium channel blockers, β mimetics, magnesium sulphate, and nitric oxide donors. Antibiotics, hydration, ethanol, progesterone, relaxin, electro-inhibition, and alternative therapies have also been tried. The choice of agent to use is important because evidence suggests that an attempt to keep babies in a potentially “hostile” uterine environment may be harmful. A seven year follow-up study of babies born to mothers with intact membranes, who were thought to be in preterm labour and were given erythromycin or co-amoxiclav to prevent preterm birth, found an unexpected increase in cerebral palsy compared with those given placebo.5

Before the availability of studies such as the current network meta-analysis by Haas and colleagues, decisions on which tocolytic to use were often made by weighing up relative risks from randomised trials of various drugs. Network meta-analysis provides a methodologically robust means of combining both direct and indirect comparisons of tocolytic drugs on the basis of a common comparator.6 Like any other meta-analytical approach, its usefulness depends on the quality of the primary data and on the relevance of individual studies’ clinical outcomes. Haas and colleagues pooled results from 95 randomised trials and found that prostaglandin inhibitors and calcium channel blockers were most effective at prolonging pregnancy for a few days. However, they found no evidence that any tocolytic drug improved mortality or neonatal morbidity, or any other clinically important outcome.

Why have randomised trials of various tocolytics focused on gestational age as a surrogate for neonatal morbidity? The answer is that trials with fewer than 100 participants have a reasonably good chance of detecting a statistically significant difference in gestational age between two groups, whereas much larger trials would be needed to determine clinically meaningful effects. The “idolatry of surrogates” has been criticised in the BMJ recently,7 and prolongation of gestation as a surrogate for neonatal morbidity is another example of a surrogate outcome that is not helpful. By focusing on gestational age at birth, the question of whether forcing babies to stay in a potentially hostile uterine environment for another week, or even longer, is clinically beneficial is not answered.

The evidence that tocolytics enable timely transport to specialist neonatal intensive care facilities and exposure to a full course of antenatal corticosteroids to improve lung maturity may not be enough. We need a proof of a sustained improvement in important health outcomes that matter to women, babies, and their families. However, a placebo controlled randomised trial that answers this question would probably be too big, difficult, and expensive to make it onto the agenda of most researchers and funders. Obstetricians who seek clinically useful evidence continue to urge researchers, the drug industry, and public funders to join forces in setting up definitive trials of tocolytic therapy with childhood morbidity as the main outcome. It does not matter which tocolytic is tested, we need only to know that one really works.

Unfortunately, despite Haas and colleagues’ well done meta-analysis, the Royal College of Obstetricians and Gynaecologists’ 2011 recommendation that it is reasonable not to use tocolytic therapy still stands.8 What should we do in the absence of the definitive evidence that tocolytic therapy improves children’s health, both in the short and long term? Most clinicians faced with a woman who is distressed with painful “contractions,” which may or may not be a sign of preterm labour, will continue to prescribe their favourite tocolytic. At least we should be honest and tell women that we are giving them drugs that will hopefully prolong pregnancy, but that these drugs may not make their babies healthier. And let’s hope that babies are not coming to greater harm by our attempts to keep them in utero.


Cite this as: BMJ 2012;345:e6531


  • Research, doi:10.1136/bmj.e6226
  • Competing interests: The author has completed the ICMJE uniform disclosure form at www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declares: no support from any organisation for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work in the previous three years; no other relationships or activities that could appear to have influenced the submitted work.

  • Provenance and peer review: Commissioned; not externally peer reviewed.


View Abstract