The Aortic Compression Trial (ACT) - A randomized controlled trial to reduce blood loss in cesarean delivery
PURPOSE AND AIMS The purpose is to reduce the prevalence of severe postpartum hemorrhage (PPH), causing maternal mortality and morbidity worldwide. The aim is to investigate if routine manual external aortic compression (EAC) in cesarean delivery (CD) is an effective and safe measure to prevent severe PPH, in a multicenter randomized controlled superiority trial.
SURVEY OF THE FIELD Severe PPH is defined as blood loss \>1000 ml at childbirth. Severe PPH is associated with 25% of maternal mortality worldwide and the leading cause of severe morbidity, such as cardiac, respiratory or renal failure and hysterectomy, notably in low-income countries. PPH also entails blood transfusion, anemia, prolonged postpartum recovery, and depression. The global prevalence of severe PPH is 6%. In Sweden, severe PPH occurs in 7% in vaginal birth, 12% in elective CD, and 17% in emergency CD with large variation between hospitals. Routine intravenous oxytocin is recommended to prevent PPH in CD. Several mechanical methods may treat excessive bleeding, but none has been assessed for prevention of PPH. One method for temporizing blood loss is manual external aortic compression (EAC), usually applied until appropriate care is available, for example during transport. The aorta is compressed by placing a fist onto the aorta through the abdominal wall which cuts off the blood supply to the uterus and hence reduces uterine bleeding. A Cochrane review in 2020 (by Kellie et al) stated that high-quality randomized trials into mechanical and surgical methods for the treatment of PPH are urgently needed. Swedish SBU stated in 2022 that prevention of complications in CD is a top prioritized research area. WHO stated in the Global Summit on PPH 2023 that research on PPH should be prioritized to reach global goals on maternal health. The effect or safety of preventive manual EAC has never been evaluated in a clinical trial.
STUDY DESIGN A multicenter randomized controlled superiority trial (RCT) using 1:1 randomization and intention-to-treat analysis. Randomization will be stratified by site using randomized permuted blocks. Core outcomes for prevention of PPH will be reported (8).
Main research questions
Does routine EAC in patients with term/near term pregnancies undergoing elective CD:
1. Reduce the prevalence of severe PPH?
2. Reduce severe morbidity and/or mortality in women?
PICO Population: Patients undergoing elective CD. Intervention: Manual external aortic compression immediately after delivery of the baby until the bleeding from the uterine incision is controlled, usually with the first suture row of uterine incision closure. Cord clamping will be timed and performed regardless of EAC.
Control: No manual external aortic compression, unless deemed vital (standard care).
Outcome: Primary: calculated blood loss \>1000 ml or blood transfusion within 2 days.
Other exploratory and safety questions are if EAC affects average blood loss; postpartum hemoglobin or hematocrit levels; time in surgery; pain, breathing, nausea or overall patient experience during surgery; conversion to general anesthesia; use of extra medication during surgery for pain, uterine tone, blood pressure, or blood loss; maternal kidney function; duration of hospital stay postpartum; neonatal outcomes; maternal well-being after childbirth; healing of the uterine scar; or healthcare costs.
Variables and measures Primary Endpoint The primary endpoint is severe postpartum hemorrhage, defined as a calculated estimated blood loss greater than 1000 ml or a red-cell transfusion within 48 hours after delivery. The calculated estimated blood loss = the estimated blood volume × (preoperative hematocrit - postoperative hematocrit) ÷ preoperative hematocrit. The maternal blood volume in milliliters is calculated as the admission body weight in kilograms × 85.
Secondary Endpoints
In clinical routine, blood loss is measured gravimetrically by collecting the blood in canisters and weighing gauzes and drapes, including vaginal bleeding, up to 2 hours after CD. This outcome is also source for the diagnostic codes O67.8 and O72.1. In accordance with the recommended core outcome set (9), five important measures of severe PPH are designated as secondary endpoints:
1. Gravimetrically estimated blood loss \>1000 ml within 2 hours of childbirth or red-cell transfusion within 48 hours after childbirth (binary)
2. Total blood loss (in g) up to 2 hours after childbirth (continuous)
3. Red-cell transfusion within 48 hours after childbirth (binary)
4. Composite outcome of transfer to higher level care (intensive care or similar), prolonged stay in post anesthesia care \>4 hours, or prolonged postnatal hospital care \>4 days (binary)
5. Composite outcome of hypovolemic shock, hysterectomy, uterine tamponade, uterine artery ligation or embolization treatment or other re-operation for internal or vaginal hemorrhage, surgical trauma to bladder, ureters or intestines, thromboembolism, organ failure, cardiac arrest, or maternal death within 42 days (binary).
Exploratory endpoints
Blood loss:
1. Venous B-hemoglobin 1-3 days after childbirth (continuous)
2. Hematocrit 1-3 days after childbirth (continuous)
3. Anemia diagnosed within 7 days postpartum (B-hemoglobin \<100 g/l) (binary)
Morbidity (including use of resources):
1. Analgetic treatment in addition to regional anesthesia during surgery or conversion to general anesthesia (binary)
2. Pharmacological treatment for low blood pressure during surgery (phenylephrine, ephedrine, norepinephrine, intravenous crystalloid fluids) (binary)
3. Use of additional uterotonic drugs (\>5 IU oxytocin) (binary)
4. Use of hemostatic drugs (e.g., tranexamic acid) (binary)
5. Use of intravenous iron replacement therapy (ferric carboxymaltose) (binary)
6. S-creatinine 1-3 days after childbirth (continuous)
7. Glomerular filtration rate 1-3 days after childbirth (continuous)
8. Start-to-finish surgery duration (min) (continuous)
9. Days in postnatal care (mother) from operation finish time to discharge (continuous)
10. B-hemoglobin at postnatal visit 8-12 weeks after delivery (continuous)
Patient well-being, satisfaction, and acceptability:
1. Perioperative pain (numeric rating scale 0-10)
2. Perioperative breathing discomfort (numeric rating scale 0-10)
3. Perioperative nausea (numeric rating scale 0-10)
4. Perioperative overall experience (numeric rating scale 0-10)
5. Edinburgh Postnatal Depression Scale at 6-8 weeks postpartum
6. Self-rated health 6-12 weeks postpartum (Likert scale 0-5, retrieved from the Swedish Pregnancy Register \[SPR\])
Neonatal outcomes:
1. Apgar score at 5 minutes \<7 (binary)
2. Apgar score at 10 minutes \<6 (binary)
3. Neonatal acidosis (cord pH \<7.0) (binary)
4. Admission to neonatal care (binary)
5. Neonatal hyperbilirubinemia (binary)
6. Neonatal anemia (binary)
7. Time to cord clamping (min, continuous)
8. Severe neonatal adverse event composite outcome (hypothermic treatment, hypoxic ischemic encephalopathy grade 2¬-3, or neonatal death within 28 days) (binary)
9. Breastfeeding Self-Efficacy Scale short form at 6-8 weeks postpartum
10. Self-reported breastfeeding 6-12 weeks postpartum (retrieved from SPR)
Long-term health:
1. Uterine scar healing measured by sonographic evaluation of the hysterotomy scar at 6-12 months after delivery (mean anterior wall thickness, prevalence of a niche)
2. Uterine rupture in the subsequent childbirth, scar dehiscence, or isthmocele surgery (5-year register-based follow-up)
Cost-effectiveness and healthcare costs: This will be detailed in a separate analysis plan and calculated including the first 42 days postpartum based on an estimated cost for the utilization of medical resources preliminary including (but not limited to):
* Duration of surgery
* Blood transfusion and transfusion of other blood products
* Intensive care
* Maternal and neonatal length of hospital stay
* Number of maternal visits postpartum
Material: Patient selection - population, sample The trial will include patients undergoing elective CD to facilitate timely informed consent. Around 10 000 elective CD are performed annually in Sweden. Patients will be recruited and screened at scheduling or admission for CD. Results will likely be transferrable to emergency CD with higher risk of PPH in which obtaining informed consent is more challenging. Using EAC as preventive method makes safety a high priority (severe adverse effects should be avoided in health individuals). Patients with planned cesarean hysterectomy or preoperative anemia will be excluded since their risk of PPH or blood transfusion postpartum may not be related to blood loss at CD.
Inclusion criteria: Patients with elective CD, live fetus (fetuses if multiple pregnancy), gestational week 34+0 or more.
Exclusion criteria: Preoperative B-hemoglobin \<100 g/l, planned cesarean hysterectomy, other condition as deemed by attending surgeon.
Treatment will be allocated in the operating theatre using opaque sealed envelopes professionally prepared by KTA. There is no masking due to the nature of the intervention. Analyses will be blinded for allocated treatment.
Estimated sample size and power Around 12.4% of elective CD have PPH \>1000 ml. To demonstrate a clinically relevant relative risk reduction of 30% at 80% power, n=1069 women in each treatment arm is needed, assuming significance level α=0.05, two-sided test. The sample size will be inflated by 5% to account for drop-out and heterogeneity between centers, resulting in total sample size n=2246 women. This sample size is also sufficient to detect relatively rare outcomes, such as intensive care or hysterectomy. Sample size calculations were performed by using the POWER procedure in SAS/STAT Software, version 9.4 of the SAS System for Windows (SAS Institute Inc., Cary, NC).
Statistical methods The primary efficacy analysis will be based on the intention-to-treat (ITT) population using a generalized linear mixed-effects model with log-link, Poisson distribution, treatment as fixed effect and site as random effect to estimate the relative risk of event while accounting for center effects. Using this model, the 80% power for the primary endpoint will be retained provided that recruitment is terminated with complete blocks, i.e., with balanced treatment allocation at all sites (12). Similar methods will be employed for secondary and exploratory endpoints using appropriate models and distributions depending on the type and distribution of the outcome, i.e., Poisson model to estimate the relative risk of binary events (e.g., maternal composite outcome), log-normal distribution for continuous and positively skewed outcomes (e.g., duration of surgery), negative binomial distribution for counts (e.g., days in hospital), and normal distribution for other numeric outcomes. Patient-reported outcomes and questionnaire scales will be treated as numeric in this regard. All estimates will be provided with 95% confidence intervals. If the primary endpoint is significant, the secondary endpoint maternal composite outcome will be confirmatory tested at the 5% significance level. All safety outcomes will be presented for both the ITT population and the safety population (all randomized women according to actual treatment). The statistical analysis plan will be established in detail and documented before data lock. Responsible statistician is Henrik Imberg, PhD, at Statistiska Konsultgruppen Sweden AB.
FEASIBILITY The trial has received a planning grant (VR 2021-06579), ethical approval Sept 20, 2022 from the Swedish Ethical Review Authority (2022-04327-01) with amendment approved Dec 20, 2022 (2022-06377-02), and formal support from www.snaks.se. The trial has started at one site (Danderyd) Dec 1, 2022, and had recruited 317 patients by Jan 19, 2025 (30% of elective CD) and proven acceptable both to patients and physicians. Nine hospitals in five regions have agreed to participate in the trial: Danderyd, Karolinska Huddinge and Solna, Södersjukhuset, Södertälje, Norrköping, Falun, Uppsala, and Göteborg. These sites perform approximately 3500 elective CD per year (www.graviditetsregistret.se).
We have assessed safety outcomes after 156 randomized patients. EAC was applied during mean 8.7 minutes (median 9, range 1-22). There were no clinically or statistically significant differences between the groups regarding safety outcomes (P-Creatinine postoperative (µmol/l), P-GFR postoperative (ml/min), overall maternal experience, umbilical vein pH, umbilical vein pO2, Apgar 5 min \<7, or admission to neonatal care.
