0923 Progesterone to prevent preterm birth in twin pregnancies

Several summaries in this series comment on the ethics and/or value of repeating studies and/or making small changes to previous study designs where there is evidence of limited effect or adverse effects of an intervention. This issue is dealt with to some extent in a recent Lancet article on the lack of effectiveness of progesterone in the prevention of preterm birth in a randomized controlled trial of 500 twin pregnancies. Odds ratio for the intervention was 1.36 with a 95% CI of 0.89-2.09. The authors also performed a meta-analysis of similar trials and found a combined odds ratio of 0.095 – 10.6, i.e. again showing no effect of the intervention, and they further comment that review of the WHO clinical trials registry lists several additional ongoing trials, some of which are so small that they will not yield data that will contribute to the debate on the effect in multiple pregnancies (which is apparently not the same as in singleton pregnancies). An editorial in the same issue makes several important points, focusing in particular on the need to go beyond whether the drug prolongs pregnancy and to also include follow-up of the offspring, ideally into adulthood, to ensure that there are actually gains from the intervention. Examples of long-term negative consequences for offspring have emerged on follow-up in other areas (e.g. following maternal diethylstilbestrol during pregnancy or antenatal steroids to induce lung maturity). Another aspect of this debate is the need to measure subsequent quality of life in terms of QALYs (quality adjusted life years) gained after an intervention. This conventional health economic metric is used to measure, balance or compare outcomes across age ranges so that decisions can be made about the value and affordability of various health interventions. Paediatricians are notoriously poor at contributing to this debate.

Read more:
Lancet 2009; 373: 2000-2 and 2034-40
Pediatrics 2005; 115: e600-14
Int J Tech Ass Hlth Care 2006; 22: 235-41