PARIS Collaboration: antiplatelets to prevent pre-eclampsia - a review using individual patient data

Article type
Askie L, Duley L, Henderson-Smart D, Stewart L, Showell M, Farrell B
Background: Over 38,000 women have participated in randomised trials evaluating antiplatelet agents, principally low dose aspirin, for the prevention of pre-eclampsia (high blood pressure and proteinuria during pregnancy) and its complications. The Cochrane Systematic Review of aggregate data from these trials demonstrates small to moderate reductions in the risk of pre-eclampsia, baby death and preterm delivery. To understand more about which women are most likely to benefit we are conducting a review and meta-analysis using data from individual women in each trial. Whilst the advantages of individual patient data (IPD) reviews are well known, this methodology has not been previously used in the perinatal field.

Objectives: The project aims to confirm that antiplatelet agents, when given to at-risk women, will prevent or reduce the severity of pre-eclampsia without adverse effects, and to investigate whether outcomes are affected by commencing antiplatelets earlier in pregnancy; and/or when a higher dose (greater than 75 mg) of aspirin is used. As data on individual women are being collected, we will also explore whether there are women with particular high risk factors who benefit more (or less) from the intervention.

Methods: Trials comparing antiplatelet/s with placebo or no antiplatelet during pregnancy are potentially eligible. These trialists have been invited to join the PARIS (Perinatal Antiplatelet Review of International Studies) Collaboration. This group has developed and agreed the protocol, which specifies the analysis plan and dataset to be collected. Data for each woman in each study will be provided and used for the meta-analyses.

Results: 55 potentially eligible trials have been identified and invited to provide their data. To date, more than 90% of data have been pledged. At April 2005 over 75% of the data have been supplied, with the remaining expected by October 2005. Only 1% of data are known to be lost. Data analysis will begin in early 2006, with final results being available in 2007.

Conclusions: Progress to date will be presented along with issues relating to the use of IPD methodology in the perinatal field, such as the potentially competing effects of treatment on both mother and infant(s).