Obstetric landmark trials

Question 1

Discuss the ACTORDS (Neonatal resp distress after repeat steroid exposer) RCT Crowther Lancet 2006

Answer 1

RCT double blinded
23 hospitals (NZ, Aus)
Mothers that remained at risk of PTB up until 32 weeks gestation after course of steroids, repeat dose/placebo weekly until delivery or over 32 weeks. (900ish women)
Outcomes: Reduced RDS 33 vs41%, RR0.8, reduced severe lung disease, O2 therapy, no difference in mean weight, no difference in secondary maternal outcomes (infection, maternal fever)

RANZCOG recommends repeat doses to 32+6, Cochrane review - 2015 agrees

Limitation: no long term FU

Question 2

Outcomes at 2 years post repeat steroid dose Crowther NEMJ 2007

Answer 2

FU at 2 years for neuro disability, intact survival, body side, BP, resp morbidity
No chance or difference other than referral for attention problems, animal studies also suggest hyperactivity
Limitation: still not long term enough to review behvaioural problems

Question 3

ASTEC: Steroids prior CS Stutchfield BMJ 2005

Answer 3

Multi-centre pragmatic RCT (unblinded)
2x doses of steroid given 48hours prior to CS
Primary outcome: admission to NICU, resp distress, TTN
Outcomes: RR 0.46 of admission, RR 0.2 for resp distress, RR 0.54 of TTN
Interpretation: increasing gestation and steroids both help
Limitation: no long term outcomes or blinding
SCN admission 37/40 11 vs 5
38/40 6 v 2.8
39 1.5 v 0.6

Question 4

ACHOIS Crowther NEJM 2005 GDM treatment for improving perinatal outcomes

Answer 4

Multicentre RCT
Included women taking OGTT that was high for glucose intolerant but not overt diabetes (>7 fasting and >7.7-11.1 2hour) to either treatment or routine care (blinded to result…..!)
Results: Composite neonatal outcome 1 v 4% RR 0.33
Secondary: increased IOL, increased SCBU, reduced macrosomia, CS =, reduced PET
Interpretation: Improves neonatal outcomes without increasing CS by increasing IOL
Limitation: results may be due to earlier IOL rather than. treatment

Question 5

Antenatal Betamethasone for Women at Risk of Late Preterm Delivery Gyamfi-Bannerman NEMJ 2016

Answer 5

Multi-centre RCT, double blinded
(note did not tocolyse)
Included women with singletons at 34 to 36+5 weeks gestation at high risk of delivery eg cx >3cm/75% effaced or PROM
Given course of steroids
Outcomes: Composite neonatal outcome RR 0.8 (death, O2 therapy)
Secondary outcomes: (all neonatal things like RDS, hypo, ICVH, pneumonia, sepsis, TTN) - steroids reduced TTN, surfactant use, bronchopulmonary dysplasia, resus, less time in nicu and BF
Adverse outcomes: more transient neonatal hypoglycaemia, no change in chorio, or neonatal sepsis, more drug reactions
NNT 35

Question 6

Arrive Trial Grobman NEJM 2018

Answer 6

Unblinded RCT
Low risk nulliparous women at 34-38+6 weeks randomised to expectant management until at least 40+5 (before 42+2) or IOL at 39 -+4weeks.
Outcomes: No different in perinatal outcome (CI hit 1) death, SB, neonatal complications but CS 18 vs 22%, shorter NICU stay, less uterotomy extensions, longer stay in hospital.
NNT 28

Question 7

ARRIVE: Grobman NEJM 2018

Answer 7

RCT non blinded
41 hospitals across USA
Low risk primps with singleton pregnancies
IOL 39-39+4 cf 40+5-42+2
Primary outcome: perinatal outcome, severe neonatal complications, c-section
4.3 vs 5.3% not statistically significant but 18 vs 22% CS rate was significant.
Limitations: obstetric lead, low chance outcomes

Question 8

CLASP RCT of LDA for prevention of PET
Lancet 1994

Answer 8

Multicentre RCT 9364 women
Pregnant 12/40+ women at risk of PET, IUGR (prophylactic or therapeutic)
60mg aspirin or placebo
Primary outcome: development of proteinuric PET. Reduction of 12% but non significant.
Aspirin not associated with a significant increase in bleeding.
Limitations: Definition of PET described as a rise in baseline and included women with PET already
ASPRE trial showed 62% reduction in pre-term PET
Cochrane review 18% reduction in PET

Question 9

IOL vs expectant management at term for IUGR. The DIGITAT Group 2010 BMJ

Answer 9

Multi-centre RCT (Holland) from 2004-2008
Singleton pregnancies 36-41 weeks with suspected IUGR based on EFW or AC <10th or drop off in growth
Excluded PROM, GDMI, DM, prev CS
Randomised to IOL within 48hours (ARM, PG, Foleys) or expectant management with twice weekly CTG, USS, BP IOL 41-42 weeks
Primary outcome was composite adverse neonatal outcome (mortality, APGAR <7 at 5mins, pH7.05) not significant RR 0.8
Secondary outcomes were CS, mat morbidity with no difference
Interpretation: No increased risk of intervention and study not powered for SB so safe to IOL
Limitations: expectant group monitored very closely, not powered for SB

Question 10

HYPITAT IOL vs expectant management for gHTN/PET after 36 weeks. Koopmans Lancet 2009

Answer 10

RCT multi-centre (parallel ??to what)
Singleton 36-41 weeks with gHTN or mild PET (ie diastolic <110, not needing IV hypertensives, etc)
IOL with ARM/PG/Foleys or expectant management with Bloods/BP/CTG/USS
Primary outcome composite maternal outcome (mortality, eclampsia, HELLP, VTE, abruption etc), progression to severe PET was 31% and 44% (NNT8 = 8), progression of disease RR 0.64
Secondary outcomes was MOD, neonatal mortality and morbidity - no differences
Interpretation: IOL at 37 weeks for women with mild PET improved maternal outcomes
Limitations: progression to severe disease = one off >170/110, ?some weren’t randomised appropriately

Question 11

HAPO trial NEJM 2008

Answer 11

To determine the level of glucose intolerance associated with increased adverse obstetric outcomes
Prospective cohort trial
Recruited women 24-32 weeks after OGTT
Excluded overt GDM, DM etc
Separated and FU’d the women in seven different categories
Primary outcomes of LGA and neonatal hyperinsulinaemia were associated with all levels of glucose intolerance - BSL >1 SD, no obvious level at which they increased
Weaker association with CS and hypoglycaemia
Positive association with all secondary outcomes - PTB, SD, birth injury, NICU, PET
Interpretation: When reviewed with ACHOIS trial - suggests GDM should definitely be treatment.
The NZ/Aus ADIPS guidelines recommend the cut off of 5.1, 10 and 8.5 (OR of 1.75 for risks)
Limitations: Not an RCT, no data on gestational weight gain

Question 12

MAGPIE Lancet 2002: MgSO2 for eclampsia

FU for babies and mothers

Answer 12

Multicentre RCT between 1998-2001
Women who were pregnant or within 24hours of birth with BP based on BP and proteinuria >30 with clinical uncertainty re MgSO4
Given MgSO2 IV or IM or placebo
Primary outcome was eclampsia or death of baby - Eclampsia RR 0.41, NNT 91, mortality for mother lower RR 0.55 (NS), no difference in perinatal mortality
Secondary outcomes: Mat morbidity - reduced abruption, increased side effects
Interpretation: improves maternal outcomes (85% in low income countries)
Limitations: Difference in perception ?MgSO4, some women had had anti-convulsants prior to transfer to facilities

Magpie FU 1.5 year study for babies (7% FU’d up)
- Everyone was blinded until the end
- Composite outcome of death or neurosensory disability at age 18months
- Results: Nothing statistically significant

Magpie FU 2 year study for women (7%)
- Using MgSO4 for PET is associated with a 16% reduction in death and morbidity potentially related to PET
- NOTE: did not contact if left hospital without a surviving child

Question 13

MgSO4 for neuroprotection Doyle Cochrane 2009

Answer 13

Systematic review
- RCTS of MgSO2 therapy for PTB (x5)
Primary outcomes: mortality, neurological disability and impairments
Results; No overall effect on mortality, RR 0.68 of CP
No effect on maternal morbidity, increased side effects
NNT for CP was 63
Limitations: ?gestations that benefit, findings only present in groups given MgSO4 for ‘neuroprotection’

Question 14

Oracle I trial (PROM) Kenyon, Lancet 2001
Oracle I Lancet 2008 FU of childhood outcomes

Answer 14

Multi-centre RCT
pPROM <37 weeks with uncertainty regarding abx
Randomised to erythromycin 250mg QID 10/7, Co-amoxiclav 250/125mg QID 10/7 or both or placebo
Primary outcome: Composite neonatal (death, lung disease, major cerebral abnormalities) - Reduction in erythromycin group = 11 cf 14.4, also significantly prolonged pregnancy, reduced surfactant and O2 use and reduced positive BC
None of the other groups had benefit over placebo, any group with co-amoxiclav had higher rates of NEC

Cochrane review 2013: abx reduces chorio, PTB, less neonatal infection etc
Also agreed with NEC for augmentin

FU:
No change in functional impairments in any four groups. No evidence of benefit or harm (75% FU)

Question 15

ORACLE II Trial (PTL) Kenyon Lancet 2001
FU trial

Answer 15

RCT 1994-2000, women with spontaneous PTL and intact membranes
Randomised to erythromycin, Augmentin, both, placebo
Primary = composite NND, chronic lung disease, major cerebral abnormality on USS before DC from hospital
Secondary = delivery within 48hours, del <7 days, MOD, maternal abx, gestational age, RDS, NEC
None of the abx improved outcomes. Double NEC rates for Augmentin groups.
Cochrane review 2013: 14 trials agreed with findings
FU data - more babies had functional impairment with erythromycin and CP (both abx)
Limitation: patient reported

Question 16

PPROMT - Immediate delivery vs expectant mx in PPROM close to term Morris Lancet 2015

Answer 16

To determine WTF to do with PROM at 34-36+6
Previously IOL at 34+
RCT, unblinded, intention to treat
Immediate delivery or expectant management with delivery at 37 weeks
Primary outcome was neonatal sepsis
Secondary outcomes: morbidity and mortality, RDS, SGA etc
Maternal outcomes APH, PPH, abx, fever, MOD, VTE
Outcomes: Neonatal sepsis RR 0.8 for immediate IOL, but increased RDS 1.6, LBW, NICU
Expectant management had higher APH, fever, use of abx, but better spont labour and lower CS
No difference between sepsis in GBS status
Interpretation: in the absence of overt signs of infection go for expectant management

Question 17

Progesterone to reduce incidence of spont PTB in women at increased risk Fonseca AMJOG 2003

Answer 17

RCT placebo, double blind
In Brazil with a high PTB
Women with high risk singleton pregnancies (previous PTB, excluded PROM, iatrogenic, multiples) (70/70)
100mg PV progesterone or placebo 24-34 weeks
All screened and treated vaginal infections + test of cure
MOA - progesterone blocks the oxytocin effect of prostaglandins increasing alpha adrenergic tocolytic response
Outcomes:
- One hour per week of monitoring for UA
Incidence and PTB
Results:
- UA 23 vs 54 RR 0.4
- PTL 19 vs 31% admitted with PTL
PTB 13 vs 28% RR 0.48
<34 2.7 v 18.5% RR 0.15
Limitations: excluded PROM and iatrogenic
RANZCOG suggest using progesterone for previous PTB but there isn’t good evidence (OPPTIMUM RCT no difference, Cochrane review with 4 trials 2xdid and 2x didn’t)

Question 18

Progesterone and risk PTB amongst women with short cervix Fonseca NEJM 2007

Answer 18

Double blind placebo RCT
Performed across London
Women with cervical length <15mm K22 weeks
PV progesterone 200mg 24-33+6
Primary: spont delivery before K34 (19 cf 34 RR 0,56)
Secondary: BW, fetal death, major adverse outcomes, NICU (no differences)
250 (125/125)
Interpretation: use progesterone for a short cervix
Meta-analysis - 5 trials showed reduction in PTB in women with cx <2.5

Question 19

IOL compared with expectant management for PROM at term Hannah NEML1996

Answer 19

RCT unblinded
Term prom, singleton, cephalic (MEC contraindicated)
Randomised to
- IOL with oxy
- IOL with prostaglandin gel
- Expectant management for up to 4 days then oxy or PG gel
Primary: infection (baby)
Secondary: CS, maternal evaluation of treatment
Results: no difference in neonatal infection, no difference in CS rate, reduced mat chorio, women preferred IOL
Interpretation: 4 babies in the expectant group and 0 babies in the IOL groups died but this may be due to chance
Cochrane review agrees with mat infection without increasing CS

Question 20

Term Breech Trial, Lancet 2000

Answer 20

Multicentre RCT
2000 women
Singleton frank or complete at term (excluded LGA, hyperextension of head, previa, feto/pelvic disproportion)
Planned CS or vaginal breech (by experience clinician)
Primary outcome: perinatal mortality or birth trauma (1.6 vs 5%)
Secondary: mat mortality or morbidity - no difference
Interpretation: CS safer, esp in low neonatal mortality countries
Limitations: exclusion criteria wasn’t very strict, O&G experience wasn’t always there, deaths accounted for the vaginal delivery weren’t always related to the actual delivery
Evidence since suggests 2/1000 death from breech vagina, 1/1000 cephalic and 0.5/1000 CS, hence strict criteria met then breech delivery acceptable

Question 21

Truffle trial, Lancet 2015

Answer 21

Unblinded RCT
Tertiary centres in Europe
Singleton pregnancies 26-32 weeks with very preterm IUGR, >500mg, normal DV and normal CTG
Randomly assigned to
- Delivery once reduced machine calculated STV, early ductus venous changes, late changes - AV wave at or below baseline
All had safety criteria of STV <3, unprovoked recurrent decels
After 32 weeks delivery according to local protocols
Protocol also recommended delivery at 30 weeks with REDF, at 32 with AEDF
Primary end point: Survival without CP/neuro-disability impairment at 2years
Secondary: Mortality, neonatal complications
Results: Late DV changes didn’t improve survival but did improve neuro-disability, not CP
Paediatrician doing the FU assessment was blinded

Question 22

Twin Birth Study Barrett NEJM 2013

Answer 22

RCT in 25 countries, similar trial to Breech birth
Twin pregnancies 32-38+6 weeks (T1 cephalic, excluded MCMA, IUFD, EFW <1500 >4000)
Randomised to planned CS or IOL 37+5-38+6
Primary outcomes: composite PND/morbidity
Secondary: birth trauma, NEC, low APGAR, sepsis, RDS, IVD, maternal morbidity
Results: no difference in primary outcome or maternal outcomes, 40% required unplanned CS
Interpretation: no benefit to CS if first twin cephalic, able to achieve emergency in 30 mins, staff experienced in ECV/IPV
No difference at two year FU either

Question 23

The Woman Trial Lancet 2017

Answer 23

Indication: PPH primary cause of direct mat mortality, trauma studies suggest 1/3 reduction in EBL. TXA reduces fibrinogen breakdown
International RCT blinded, placebo controlled
Anyone having a PPH with clinician uncertainty of use of TXA
Primary: mortality or hysterectomy (changed to death only)
Secondary - VTE, surgical intervention, complications, QoL
20,000 women - results no difference in primary outcome but did have RR 0.8 is just using death, stronger effect within 3 hours, reduced laparotomy, no difference in VTE or blood transfusion
Interpretation: Early administration of TXA improves outcomes (reduced death and laparotomy)