Skip to main content

Evidence > CHIPS (Control of Hypertension In Pregnancy Study)

Control of Hypertension In Pregnancy Study

Update on 28 January 2015: The CHIPS Trial results are now published in the NEJMRead more.

Goal

In the CHIPS Trial, the investigators sought to determine whether ‘less tight’ control (aiming for a diastolic blood pressure [dBP] of 100 mmHg), compared with ‘tight’ control (aiming for a diastolic blood pressure [dBP] of 85 mmHg) would decrease the risks of adverse baby outcomes without increasing the risk of problems for the mother.

Background

Women with non-severe non-proteinuric pre-existing or gestational hypertension remote from term are deemed a high-risk group. It is unclear how best to manage this type of hypertension to do more good than harm. Allowing blood pressure to be higher could improve uteroplacental perfusion, fetal growth, and ultimately, neonatal wellbeing. Based on meta-analyses of randomised trials, less tight control could decrease the risk of small-for-gestational-age infants, but may increase the risk of (transient) severe maternal hypertension, antenatal hospitalisation, and proteinuria at delivery. However, there is insufficient evidence on which to base clinical decisions because of reporting bias and between-trial heterogeneity in outcome. Guidelines are founded mainly on expert opinion. We established a definitive randomised trial to inform clinical decision-making.

Control of Hypertension In Pregnancy Study (CHIPS) was an open, multicentre, international, randomised controlled trial with an intention-to-treat analysis. The study was pragmatic in nature, and was undertaken to reflect real clinical practice rather than the very tightly controlled circumstances of explanatory trials. The pragmatic approach was important in this trial, where the interventions were policies of ‘less tight’ and ‘tight’ control, and it was not possible to blind clinicians or patients to the dBP goals of each treatment group.

Aim

We aimed to determine whether, for pregnant women with non-severe non-proteinuric maternal hypertension at 14-33 weeks' gestation, less tight control (target diastolic blood pressure 100 mm Hg) compared with tight control (target diastolic blood pressure 85 mm Hg), would increase or decrease the likelihood of pregnancy loss, high-level neonatal care for more than 48 h, or serious maternal complications.

Methods

Design

CHIPS was an international multicentre randomised trial of 1028 women (514 per group) from 50 tertiary and community centres (2009-13). Eligible women were randomised centrally to either less tight or tight control of their hypertension. Randomisation was stratified by centre and type of hypertension. In the less tight control group, if diastolic blood pressure was 105 mm Hg or higher, antihypertensive drugs were started or increased in dose. In the tight control group, if diastolic blood pressure was 80 mm Hg or lower, antihypertensives were decreased in dose or discontinued. For both groups, centres provided their usual care. Data was also collected on potential co-interventions (eg, bed-rest).

Population

Inclusion criteria were: pre-existing/gestational hypertension; office diastolic blood pressure 90-105 mm Hg (or 85-105 mm Hg if on antihypertensives); live fetus; and 14-33 weeks' gestation. Exclusion criteria were: severe systolic hypertension (≥160 mm Hg); proteinuria; contraindication to either group of the trial or to prolongation of pregnancy; use of an angiotensin-converting-enzyme inhibitor at or after 14 weeks' gestation; known multiple gestation; lethal/major fetal anomaly; plan to terminate pregnancy; or previous participation in CHIPS.

Outcomes

The primary outcome was pregnancy loss (miscarriage, ectopic pregnancy, pregnancy termination, stillbirth, or neonatal death) or high-level neonatal care for 48 hr or longer in the first 28 days of life or before primary hospital discharge, whichever was later. The secondary outcome was one or more serious maternal complication(s) until 6 weeks' postpartum. Other outcomes included compliance with interventions within 4 weeks after randomisation. Women were contacted 6-12 weeks after delivery (or loss) and, for preterm babies, when the baby was at 36 weeks' corrected postgestational-age to inquire about satisfaction with care and any major maternal/neonatal morbidity after hospital discharge.

Analysis plan

The primary outcome was compared between groups with multivariate logistic regression (α = 0·046, two-sided). We based our sample size on a clinically important reduction in pregnancy loss or high-level neonatal care for more than 48 h from 33% to 25%. Two interim analyses took place after primary outcome data were available for one-third and then two-thirds of enrolled women. An independent data and safety monitoring board considered early termination of the trial at the first and second interim analyses, with a two-tailed test of significance at the p<0·0002 and p<0·012 levels, respectively.

Status

The CHIPS study is now published in the NEJM. Read more here

Goals

  • Further CHIPS analyses: The impact on outcome of both BP level and variability will be explored, in an effort to fully understand the predictive power of BP, as well as an overview paper is planned.
  • CHIPS-Child: In this follow-up study of CHIPS, the developmental programming hypothesis was being tested by determining if children born of mothers in ‘less tight’ (vs. ‘tight’) BP control arms of CHIPS showed differences in postnatal growth and health. The results were presented in October 2016 at the International Society for the Study of Hypertension in Pregnancy (ISSHP) meeting, Brazil.
  • CHIPS implementation: The next steps will be to further our understanding of the views about ‘tight’ and ‘less tight’ control of BP, and to undertake implementation planning in low- and middle-income countries (LMICs). The views of women were explored with regards to the choices that they would make for BP control and why. The views of clinicians were also explored in a formal stakeholder analysis in 2017-8. Furthermore, we will work towards implementation of a ‘tight’ BP control target in LMICs, using methyldopa as the antihypertensive. This is particularly important in under-resourced settings where maternal surveillance is less available and less frequent, whether that surveillance is delivered by the women through home BP (and other) monitoring, by community health workers at home, in primary health centres (PHCs), or by clinicians at PHCs or facilities.

Publications