Journal Club Papers
Paper 1: Perinatal morbidity among women with a previous caesarean delivery (PRISMA trial): a cluster-randomised trial.
Chaillet et al. for the PRISMA Trial Research Group. The Lancet. 11 December 2023. DOI: 10.1016/S0140-6736(23)01855-X
Planning a vaginal birth after caesarean section (VBAC) requires women and their care providers to weigh up the potential risk of complications of a repeat caesarean section with the risks associated with a trial of labour. Chaillet et al assess a multifaceted approach to help navigate this decision.
This was a cluster-randomised controlled trial, with forty Quebec hospitals randomised to either control (routine care), or a multifaceted intervention (hospitals randomised, not individual participants). The intervention consisted of three components:
- a patient decision aid;
- the calculated likelihood of achieving a vaginal birth;
- ultrasound assessment of lower uterine segment thickness at 36 weeks’ gestation, triaging women into having a low (≥2.5mm thick), medium (≥2mm and <2.5mm) or high risk (<2mm) of rupture.
Healthcare providers at the intervention sites were additionally invited to attend a one-day workshop in intrapartum VBAC management.
- Women were eligible for the study if they received care at a participating hospital, had a singleton pregnancy with birth occurring from 24 weeks’ gestation, and a history of one caesarean section.
- The primary outcome was a composite of neonatal mortality and major morbidity, including: intracerebral or severe intraventricular haemorrhage, hypoxic-ischaemic encephalopathy, seizure, invasive mechanical ventilation, hypotension requiring vasopressors, proven sepsis or infection, necrotising enterocolitis, Apgar score <4 at 5 minutes, metabolic acidosis, major trauma, periventricular leukomalacia.
- Secondary outcomes included minor perinatal complications and major and minor maternal complications, mode of birth, induction of labour, epidural analgesia, and episiotomy.
- From April 2016 – Dec 2019, 21,281 women were included, with 10,514 in the intervention group and 10,767 in the control group. Participants randomised to the control group were more likely to be ≥35 years of age (36.6%, compared with 28.8% in the intervention group).
The intervention significantly reduced the likelihood of neonatal mortality and major morbidity.
- Compared with routine care, the intervention was associated with a 31% reduction in neonatal mortality and major morbidity (aOR 0.72 (95%CI 0.52, 0.99); 4.3% control vs. 3.8% intervention), and a 46% reduction in major maternal morbidity (aOR 0.54 (95%CI 0.33, 0.89); 2.5% control vs. 1.8% intervention).
- Most women in the study had an elective repeat caesarean section (67.5% in the intervention; 61.9% in the control).
- Among those who underwent a trial of labour, there was no difference between the groups in secondary outcomes, including odds of vaginal birth, need for emergency caesarean section, and uterine rupture.
This low-risk multifaceted intervention reduced major maternal and neonatal morbidity in a Canadian cohort – without reducing emergency caesarean section or uterine rupture rates. The authors concluded that the outcome is most likely explained by more effective pre-delivery risk assessment, where those at highest risk of morbidity were encouraged have an elective caesarean section (and avoid labour).
This is an important study as it shows the use of decision aids and VBAC calculators to support decision making may improve outcomes. It is notable that such aids have yet to see widespread use in the clinic
Paper 2: Calcium supplementation for the prevention of pre-eclampsia: Challenging the evidence from meta-analyses.
Wright et al. BJOG. 9 January 2024. DOI: 10.1111/1471-0528.17769
In 2011, the World Health Organisation (WHO) issued a guideline recommending calcium supplementation in pregnancy to prevent preeclampsia, in areas with low dietary calcium intake. This recommendation followed a 2010 Cochrane review (updated in 2014 and 2018) of 13 randomised trials. The Cochrane review showed a 55% decreased risk of preeclampsia associated with calcium supplementation of ≥1g per day, and a 74% reduction when baseline calcium intake was low.
The authors of this new study propose that, by reporting only the mean risk ratio (RR) generated through random-effects meta-analysis (a method of meta-analysis which assumes that the underlying effect follows a normal distribution), too much weight may have been given to smaller trials in those Cochrane analyses. This means smaller trials, less powered to detect the true effect of calcium supplementation, may have been given too much ‘weight’ and have overly swayed the results of this meta-analysis that suggested benefit.
To support their viewpoint, the authors point to the I² statistics generated in the Cochrane review. The I² statistic is a way to measure the difference (heterogeneity) between studies in a meta-analysis. If I² = 0%, then the studies are virtually identical in their methodology, and easier to compare (it’s a cleaner meta-analysis). If I² = 100%, then the studies are very different, and not easy to compare (and less clean to combine into a meta-analysis).
- The primary analysis from the Cochrane review consisted of 13 trials of 15,750 women and found a 55% reduction in the risk of preeclampsia with calcium supplementation (RR 0.45; 95% CI 0.31 – 0.65). However, there was significant heterogeneity among studies, with I² = 70%.
- The authors then performed a new meta-analysis examining calcium to prevent preeclampsia. In this new analysis, when only the three largest trials were included and small studies were excluded (90% of participants, N=13,815), the heterogeneity dropped to I² = 0%. In this analysis, the association between calcium supplementation and preeclampsia disappeared (RR 0.92; 95% CI 0.80 – 1.06).
- The Cochrane review also examined a subgroup of trials conducted in settings where there is low baseline calcium levels (8 trials, N=10,678). In this analysis, calcium supplementation was again associated with a significant reduction in the risk of preeclampsia (RR 0.36, 95% CI 0.20, 0.65), but heterogeneity was substantial (I² = 76%).
- The authors of the current review performed a new meta-analysis of trials done in areas with low calcium intake but only included the two largest trials (almost 90% of participants, N=9479). The heterogeneity again reduced to I² = 0%. The association between calcium and preeclampsia risk again disappeared (RR 0.92; 95% CI 0.75 – 1.13).
The authors suggest that, because of the significant heterogeneity between studies included in the Cochrane review, the results (mean effect) are misleading. The smaller trials, which made up only 10% of the meta-analyses’ populations and report large effect sizes across small populations, were likely weighted too heavily under the assumptions made by a random-effects meta-analysis. This may be a symptom of publication bias – where small trials are only published if their results are considered significant, while small trials that are nonsignificant trials are never published.
Overall, the results do not support the use of calcium to prevent preeclampsia.
The authors concluded that guiding bodies need to carefully consider heterogeneity and other potential biases in meta-analyses before establishing guidelines, and that alternative methods of meta-analysis may be required to account for significant differences between small and large trials.
Paper 3: Two Randomized Trials of Low-Dose Calcium Supplementation in Pregnancy.
Dwarkanath et al. NEJM. 11 January 2024. DOI: 10.1056/NEJMoa2307212
Within days of publication of the above re-analysis of the Cochrane review, these dual randomised trials were published in NEJM (calcium supplementation has suddenly become news again!). These trials aimed to examine the efficacy of smaller doses of calcium in reducing preeclampsia and preterm birth in low-income settings. Two separate double-blind RCTs were conducted, in India and Tanzania, and published in the one paper. The noninferiority of 500mg oral calcium was compared with 1500mg oral calcium in reducing the primary outcomes of preeclampsia and preterm birth (The noninferiority cut-offs were determined by the findings of risk reduction from the Cochrane review….). 11,000 nulliparous women were recruited to each trial before 20 weeks gestation. In India:
- Cumulative incidence of preeclampsia was 3.0% in the 500mg group and 3.6% in the 1500mg group (RR 0.84; 95%CI 0.68-1.03) – noninferior!
- Preterm birth occurred in 11.4% of the 500mg group and 12.8% in the 1500mg group (RR 0.89; 95%CI 0.80-0.98) – noninferior!
In Tanzania:
- Cumulative incidence of preeclampsia was 3.0% in the 500mg group and 2.7% in the 1500mg group (RR 1.10; 95%CI 0.88-1.36) – noninferior!
- Preterm birth occurred in 10.4% of the 500mg group and 9.7% in the 1500mg group (RR 1.07; 95%CI 0.95-1.21) – NOT noninferior! (as the upper confidence interval of 1.21 exceeded their pre-specified non-inferior margin of 1.16)! The interpretation of this finding is that the smaller dose was inferior in reducing preterm birth in this trial group.
The application of these findings to clinical practice in high resource settings is unclear. Also, the findings of the prior publication we discussed also makes it difficult to interpret these trials: if calcium does not truly alter the risk of preeclampsia as previously thought, then varying the dose of calcium given may be inconsequential.
Overall, these recent contributions call into question the efficacy of calcium to prevent preeclampsia, even in areas where calcium intake is low
Paper 4: Repeat placental growth factor-based testing in women with suspected preterm pre-eclampsia (PARROT-2): a multicentre, parallel-group, superiority, randomised controlled trial.
Hurrell et al. The Lancet. 8 February 2024. DOI: 10.1016/S0140-6736(23)02357-7
Outside of Australia, Placental Growth Factor (PlGF)-based testing is routinely used in the management of preeclampsia. PlGF levels >5th percentile have been shown to accurately rule out preeclampsia needing delivery within the next 14 days. And, among women with suspected preeclampsia PlGF testing reduces time to diagnosis and severe maternal outcomes. Based on these findings PlGF-based testing is now recommended by the National Institute for Health and Care Excellence (NICE) in the UK and the International Society for the Study of Hypertension in Pregnancy (ISSHP).
Given the utility of PlGF-based testing at diagnosis and during early management it has been pondered whether repeat testing may confer further benefits. To address this, Hurrell and team launched the PARROT-2 trial.
- PARROT-2 was an individual-level, multicentre, parallel-group, superiority, randomised controlled trial, aimed to determine whether repeat PlGF-based testing would have any impact on neonatal or maternal outcomes. It was conducted across 22 units in the United Kingdom, and it recruited women aged 18+, with a viable singleton pregnancy, who were between 22-36 weeks at the time of their first PlGF-based test for suspected preeclampsia.
- Participants were randomised (1:1 ratio) to unblinded repeat PlGF-based testing, or usual care with blinded repeat PlGF-based testing. Repeat tests were conducted up to four times throughout pregnancy.
- The primary outcome was a composite of stillbirth, early neonatal death (within 7 days of delivery), or neonatal unit admission. Neonatal outcomes were selected for the primary outcome, due to the results of PARROT-1 suggesting clear maternal benefit of PlGF-based testing. Secondary outcomes included severe adverse maternal outcomes, preterm birth, caesarean delivery, and time from initial PlGF-based testing to diagnosis of preeclampsia.
There was no difference in the rate of the primary outcome. The composite of stillbirth, early neonatal death, or neonatal unit admission occurred in 195 of 625 (31.2%) of infants randomised to the unblinded group and in 174 of 626 (27.8%) of infants randomised to the usual care/blinded group (RR 1.21; 95% CI 0.95 – 1.13).
- There was also no difference in any of the individual components: stillbirth (RR 0.67; 95%CI 0.11 – 3.99); early neonatal death (RR 1.00; 95%CI 0.06 – 15.98); or neonatal unit admission (RR 1.13; 95%CI 0.95 – 1.35).
- There was a significant reduction in gestational age at delivery among women in the unblinded repeat-PlGF group, compared with the usual care/blinded group (36.7 weeks vs 37.1 weeks; mean difference [MD] –0.40 weeks; 95% CI –0.68, –0.12; p=0.005).
- There was a significant increase in the number of deliveries prior to 34 weeks’ gestation among women in the unblinded repeat-PlGF group (90, 14.4%) compared with the blinded group (55, 8.8%); (RR 1.63; 95% CI 1.19 – 2.24).
- There was no difference between groups for the composite severe maternal outcome. However, the unblinded repeat-PlGF group had a significant (14%) increase in the rate of caesarean section (aRR 1.14; 95% CI 1.05 – 1.23; p=0.002), and a reduced time from initial test to diagnosis of preeclampsia (MD –3.79 days; 95% CI –7.10, –0.47; p=0.025).
While repeat PlGF-based testing was associated with a faster diagnosis of preeclampsia, there were no positive downstream benefits and rather, the potential for harm with a reduction in gestational age at birth and an increase in preterm birth <34 weeks’. The authors note that for women with initially abnormal PlGF, their results rarely normalised over time, and only 3.3% of participants with normal initial results had very low PlGF by their final test. Whilst this study supports the findings of PARROT-1 and emphasises the importance of initial PlGF testing in the clinical setting, it does not provide substantial evidence to warrant repeat testing in clinical practice.
[We note PIGF is not widely available in Australia, but if we ever get it, we know how to use it – initially, but not do repeat testing].