Control of Hypertension in Pregnancy Study

High blood pressure (BP) in pregnancy carries risks for both mother and baby, but for years doctors have been treating it without knowing what target BP to aim for. Everyone agrees with lower than 160/110 to prevent maternal stroke, but they worry about getting near normal (120/80) for fear of reducing placental blood flow and harming the baby.

Such concerns were justified by both theoretical arguments and some rather poor quality trials, but in this week’s New Engl J Med (click here or CHIPS Trial – NEJM 2015), Laura Magee and her colleagues from Canada, show they were misplaced.

The CHIPS trial compared less tight control (target diastolic 100) with tight control (target 85). It was registered here in 2008, with a planned sample size of 1028.  They recruited 1030 women, but on the advice of the independent steering committee, excluded 43 women from one site due to concerns about data integrity and consent, leaving 987 for analysis. The odds ratio of the predefined primary composite outcome (fetal loss or >48 hours neonatal intensive care) with less tight control, was 1.02; 95% CI 0.77-1.35. In other words there were no fetal risks from aiming for a target diastolic of 85.

Since lower is clearly better for the mother*, we should now aim for tight control.  I’m changing my policy today, and the National Institute for Clinical Excellence (NICE) will surely soon amend its advice.

Robin Hood’s contribution

I’m also proud. Nottingham University Hospitals NHS Trust sponsored the UK arm. Dozens of UK doctors and midwives (listed here), supported by the UK National Institutes of Health Research, patiently explained the uncertainty to pregnant women and their partners, recruited those who wished to participate, and followed everyone up. The UK was the highest recruiting country worldwide.

Next steps

There is a rumour that the US National Institutes of Health (NIH) have just funded pretty much a repeat of the CHIPS trial. Some colleagues blame the “not discovered here” syndrome, but I think that’s unfair. Even after randomising nearly 1000 women the confidence intervals around the fetal effect in CHIPS is compatible with a 20% reduction and a 30% increase in bad outcomes with less tight control. Millions of women have hypertension treated in pregnancy every year. A repeat trial to narrow these confidence intervals is worthwhile.

I’m following a “tight control” policy, but I’ll be looking out for the US trial results.

Jim Thornton

*CHIPS didn’t prove that. Although the primary adverse maternal outcome was higher in the “less tight” group, the difference (3.7 v 2.0%) was not statistically significant. I doubt any feasible pregnancy trial could prove maternal harm, because bad maternal outcomes are rare. But many observational studies have shown that very high BP is associated with maternal strokes, and the rate of dangerously high BP (>160/110) was greater in the “less tight” group.