Protocol for a multicenter, double-blinded placebo-controlled randomized controlled trial comparing intravenous ferric derisomaltose to oral ferrous sulfate for the treatment of iron deficiency anemia in pregnancy: the IVIDA2 trial

Introduction

Iron deficiency anemia (IDA) affects more than 1 in 3 pregnant women worldwide¹ and nearly 1 in 5 pregnant women in the United States.² IDA is an important risk factor for both maternal and fetal morbidity. IDA is associated with higher rates of cesarean delivery, depression, and blood transfusion,^3-5 the major driver of the Centers for Disease Control and Prevention (CDC)’s severe maternal morbidity composite quality metric.⁶ Maternal anemia is also associated with increased risk of preterm birth, low birth weight, and small-for-gestational-age neonates.^3,4 Moreover, iron-deficient mothers are at risk of delivering iron-deficient neonates who, despite iron repletion, remain at risk for delayed growth and development.⁷ Fetal-neonatal iron deficiency has been linked to neurological impairments in infants⁸ that may persist into adulthood.⁷ Animal studies have demonstrated a clear role for iron in normal brain development, including myelination, dendritic growth, and synapse formation.^9-11 Thus, there appears to be a strong association between IDA in pregnancy and long-term adverse neurodevelopmental outcomes.

Guidelines from the American College of Obstetricians and Gynecologists (ACOG) and the CDC recommend screening all pregnant women for anemia at the first prenatal visit with a complete blood count (CBC).¹² If anemia is diagnosed, additional testing may be performed to investigate the potential cause, usually with a serum ferritin, which is a more sensitive and specific marker for iron deficiency than other markers including serum iron and transferrin saturation.^13,14

Iron supplementation is recommended to treat IDA in pregnancy, but important questions remain about the optimal route of delivery. Oral iron, administered in doses higher than found in prenatal vitamins, is the current standard for treating IDA in pregnancy in the United States. Lower iron doses and alternate day dosing have increasingly been used because higher and more frequent doses do not improve iron uptake but appear to increase adverse medication effects.^15,16 Ferrous sulfate is the most commonly prescribed oral formulation as it is inexpensive, safe, readily available, and, when tolerated, effective. However, a meta-analysis of 43 randomized controlled trials (RCTs) reported that up to 70% of patients prescribed oral iron experienced significant gastrointestinal perturbation, decreasing adherence to therapy.¹⁷

Intravenous (IV) iron is an attractive alternative because it mitigates the adherence and absorption challenges of oral iron. It is usually administered in the second and third trimesters, as there are no safety data for first-trimester use. All IV iron products currently on the market have equivalent safety and efficacy¹⁸; thus, the choice of formulation is based on cost and administration burden. Formulations that allow a complete replacement dose in a single visit are preferred. Ferric derisomaltose (Monoferric, Pharmacosmos Therapeutics Inc., Morristown, NJ) has been used since 2009 in other countries under the drug name ferric isomaltoside and was approved by the United States Food and Drug Administration (FDA) in 2020 based on two randomized, open-label RCTs performed in a total of 3050 non-pregnant patients with IDA.^19,20 In these trials, serious adverse events were reported in 0.3% (6/2008) of the ferric-derisomaltose-treated subjects. Thus, ferric derisomaltose has the advantages of being an FDA-approved single 1000 mg infusion with demonstrated efficacy and safety.²¹

It is plausible that IV iron is superior to oral iron in promoting rapid correction of anemia and iron deficiency. Two meta-analyses of randomized trials found that compared with oral iron, IV iron was associated with significantly higher hemoglobin (Hb) level following therapy among pregnant women with IDA.^22,23 One of the meta-analyses (eight studies) additionally evaluated maternal and neonatal outcomes and found that IV iron was associated with higher neonatal birth weight, higher neonatal ferritin levels, and less frequent adverse effects and therapy discontinuation.²³ However, the primary trials were nearly all conducted in developing countries (none in the United States), included small sample sizes (50 – 252), and did not assess meaningful maternal or neonatal outcomes.²³

In a subsequent large randomized controlled trial conducted in India, pregnant women at 20–28 weeks with a Hb of 5–8 g/dL, or at 29–32 weeks with a Hb of 5–9g/dL, were randomly assigned to receive oral iron (100 mg elemental iron twice daily) or up to five divided doses of IV iron sucrose infusions.²⁴ The results showed no difference in maternal composite morbidity, defined as any one of postpartum hemorrhage, blood transfusion during and after delivery, sepsis, shock, prolonged hospital stay and intensive care unit admission or referral to higher centers.²⁴ There was no significant difference in serious maternal adverse events or serious fetal and neonatal adverse events.²⁴ This study is limited by using iron sucrose, which required multiple infusions, resulting in a wide range of IV iron doses infused (200 – 1600 mg). In addition, the primary outcome was a maternal composite outcome including some components not directly related to anemia.

A recent study in Denmark compared the efficacy of a single-dose (1000 mg) IV iron (ferric derisomaltose) with 100 mg daily oral iron (ferrous fumarate) in preventing anemia among 201 pregnant women 14–21 weeks’ with persistent iron deficiency (ferritin < 30 μg/L).²⁵ Over the 18-week follow-up period, those receiving IV iron had a higher mean Hb increase and were less likely to develop anemia than those receiving oral iron (9% vs. 27%).²⁵ There were no significant differences between the two groups in treatment-related adverse events.²⁵ While results are promising, this study only included a small proportion of patients (11%) with anemia at baseline and thus is not informative for patients information with IDA.

There is, therefore, an urgent need for a rigorous study to test the clinical effectiveness, safety, and cost-effectiveness of IV iron on clinically relevant maternal and neonatal outcomes among those with IDA. In this protocol document, we will describe the study methods, statistical analytic plans, ethical approval, and dissemination plan of a multicenter, placebo controlled double-blinded randomized controlled trial that will provide definitive data on optimal route of iron repletion among pregnant women with IDA.

Methods and Analysis

Trial Objectives:

The primary objective of this trial is to determine whether, among pregnant women in the United States with moderate-to-severe IDA, a single dose of IV ferric derisomaltose is a safe, superior alternative to oral iron repletion in reducing peripartum blood transfusion. Additional secondary outcomes include adverse medication reactions, maternal and neonatal hematologic indices, and perinatal outcomes.

This study also has two secondary objectives: (1) assess the effect of IV iron, compared with oral iron, on offspring brain myelin content and neurodevelopment six months after delivery. (2) compare the cost-effectiveness of IV to oral iron for the treatment of IDA in pregnancy.

Study Design:

The IVIDA2 trial is a double-blind, placebo-controlled, multicenter randomized trial to test the central hypothesis that IV iron in pregnant women with moderate-to-severe IDA (Hemoglobin (Hb)<10 g/dL and serum ferritin<30 ng/mL) at 24-28 weeks will be effective, safe, and cost-effective, compared with oral iron. The trial will recruit pregnant women with IDA at 24-28 weeks and will follow them until six weeks’ postpartum. Those who enroll in IVIDA2 and deliver at Women and Infants Hospital of Rhode Island will be eligible to participate in the neonatal imaging and neurodevelopmental component of the study, which will occur six months after delivery.

Study settings and participants.:

The IVIDA2 trial will be conducted at three medical centers in the United States: (1) Women & Infants Hospital of Rhode Island (WIHRI); (2) University of Michigan; and (3) Washington University in St. Louis. Partipants will be recruited from prenatal clinics affiliated with these three centers. WIHRI—the IVIDA2 coordinating center—is a major teaching affiliate of Alpert Medical School of Brown University and a large regional referral center, performing over 8,500 deliveries annually. The University of Michigan Medical Center is the major perinatal referral facility for Michigan and delivers approximately 4,900 women annually. Washington University in St. Louis will enroll patients who deliver at Barnes-Jewish Hospital, a tertiary care center with 4,500 deliveries annually.

Eligibility Criteria:

Participants will be randomized only if they fulfill all inclusion criteria and none of the exclusion criteria (Table 1). Broad inclusion criteria will increase generalizability of our results.

Table 1:

Study enrollment criteria

Inclusion criteria

• Pregnant women between the ages of 18 – 45 years

• Singleton gestation

• Iron-deficiency anemia (serum ferritin <30 ng/mL and Hb<10 g/dL)

• At 24 – 28 weeks’ gestation

• Plan to deliver at participating hospital

Exclusion criteria

• Non-iron-deficiency anemia e.g. thalassemia, sickle cell disease, B12 or folate deficiency, hypersplenism

• Malabsorptive syndrome, inflammatory bowel disease, gastric bypass, or hypersensitivity to oral or IV iron

• Multiple gestation

• Inability or unwillingness to provide informed consent

• Inability to communicate with members of the study team, despite the presence of an interpreter

• Planned delivery at a non-study affiliated hospital

Trial Interventions:

We will compare IV versus oral iron in pregnant women with moderate-to-severe IDA at 24-28 weeks’ gestation in double-blinded fashion.

1. Oral Iron: Participants assigned to the oral iron group will receive a single 250 mL IV normal saline infusion given over 20 minutes and 325 mg ferrous sulfate (65 mg of elemental iron) to be taken until delivery

2. Intravenous iron: Participants assigned to the IV iron group will receive a single IV infusion of 1000 mg ferric derisomaltose in 250 mL saline given over 20 minutes and daily placebo tablets until delivery.

Oral Tablet Management:

The study pharmacy will prepare and label the active iron pills and matching placebo tablets and ship them to the study sites to be dispensed. The dosing of both the active and placebo tablets is 1 to 3 per day, at the discretion of the patient’s provider. This will mimic clinical practice, where patients are usually started on 1 tablet per day with escalation or reduction as needed at the discretion of the provider. Research staff will monitor adherence and adverse effects related to the tablets through tablet counts and patient interview at 4 weeks after the IV infusion and on admission for delivery.

IV Infusion Management:

The active IV iron will be shipped by Pharmacosmos Therapeutics Inc. to the study sites The IV iron infusion will be prepared in 250 ml of normal saline. The placebo infusion will be 250 mL IV normal saline only. To blind the participants, the IV bag and tubing will be covered with a brown plastic infusion bag and tubing covering to obscure the contents of the IV solution. The infusions will occur on the labor and delivery floor, antepartum unit, clinical trials unit, or infusion center, per standard protocols at each site. Participants will undergo pre- and post-infusion confirmation of fetal viability via bedside Doptone or ultrasound. Participants will be monitored during infusions and for at least 30 minutes and until clinically stable following completion of the infusion. Research staff will also contact patients within 24-48 hours of infusion to ascertain transfusion reactions.

Screening and enrollment of participants:

We will employ efficient recruitment techniques that we have used in recent randomized clinical trials in the same settings.^26,27 Pregnant women at the study sites will be screened for anemia at their first prenatal visit with a CBC (or CBC with serum ferritin, per clinic or provider practice patterns). Patients with anemia may be managed expectantly, tested for IDA with a serum ferritin, or prescribed oral iron by their care provider for presumed IDA per standard of care. Patients without anemia will be managed expectantly. All patients will be re-screened with a CBC at 24-28 weeks’ gestation. Those with Hb<10 g/dL will be tested for IDA using serum ferritin if not previously tested. Moderate-to-severe IDA will be defined as Hb<10 g/dL at 24-28 weeks’ and a serum ferritin <30 ng/mL at any point from their first prenatal visit until 28 weeks’. Patients who initiate prenatal care visit at 24-28 weeks’ and found to have moderate-to-severe IDA will be eligible for enrollment at that time.

Randomization, allocation concealment, and blinding:

Enrolled participants will be randomly assigned in a 1:1 ratio using computer-generated randomization sequence. A web-based randomization sequence will be prepared by using variable block sizes (4 and 6), stratified by study site, maternal age (<35 vs. ≥35 years) and severity of anemia (Hb <8 vs. 8-10 g/dL) at enrollment. Stratification by study site promotes balance of groups in each study site to control for any differences in patient management at the different centers. Taking severity of anemia at randomization into account promotes balance in the baseline hemoglobin level, which has the potential to influence outcomes.

The randomization sequence will be maintained centrally by the study statistician. A subject’s group assignment will be generated only after the patient has agreed to participate in the study,informed consent signed and is scheduled for infusion. The procedure for revealing the blinded group assignment will be implemented in REDCap (Research Electronic DataCapture), a secure internet-based data management system.^28,29 The study staff will click a button in REDCap after confirming a patient’s eligibility to reveal the group assignment.

Participants and providers will be blinded. Both groups will receive identical 250 ml IV infusions containing either 1000 mg of ferric derisomaltose or normal saline. They will also receive similar tablets of either ferrous sulfate containing 65 mg of elemental iron or placebo beginning on day of the infusion. Although adverse events such as constipation and Fishbane reactions are more common with certain treatment types and could result in patients guessing the group to which they were assigned, the use of placebo would allow us to more accurately estimate the differences in these adverse reactions that are due to the active oral or intravenous iron, rather than a placebo effect. Trained research staff at each site will abstract clinical data from the medical record without reference to participant study group. Once all data are collected at all three sites, the database will be unlocked and ready for unblinding and analyses.

Post-Randomization Prenatal Care and Anemia Management:

Participants will receive routine prenatal care. There will be no study-specific tests for anemia between randomization and delivery. If additional hemoglobin determinations are performed for routine clinical purposes after randomization and reveal ongoing anemia, the patient’s care provider will have options of expectant management, increasing the oral tablet dose, or, if the anemia is severe, blood transfusion. In the anticipated rare event that a subject’s care provider makes the decision to offer IV iron to the patient outside the study, the patient’s group assignment will be revealed by the unblinded study pharmacist to avoid iron overload from a second IV iron infusion if the patient previously received IV iron from the study. Figure 1 describes a summary of study visit timeline.

Figure 1:

Summary of Study Procedures

Neonatal imaging and neurodevelopmental assessments:

Infants born at the WIHRI site to IVIDA2 participants will be eligible for a brain MRI and neurodevelopmental assessment at 6 months of age. The infants of consenting participants will receive a brain MRI and neurodevelopmental assessment at centers affiliated with Hasbro Children’s Hospital. Infants will first receive a non-sedating MRI using a novel, noninvasive neuroimaging technique termed mcDESPOT (multicomponent-Driven Equilibrium Single-Pulse Observation of T1 and T2), to quantify myelin water volume fraction (MWF), a surrogate measure for myelin content.^30–32 This has been used previously to characterize normative patterns of myelination in healthy infants^33,34 and to investigate relationships between myelin content and evolving brain function and cognitive skills.^35,36 Infants will undergo MRI during natural, non-sedated sleep at either nap or bedtime on a GE Premier 3 Tesla scanner (GE Medical Systems). Notably, this technique has been used extensively to study myelination patterns in infancy and early childhood. ^33–38

Within 7 days of a successful MRI, each child will be assessed with the Mullen Scales of Early Learning. The Mullen Scales of Early Learning is a standardized and population-normed tool for assessing fine and gross motor control, visual reception, and expressive and receptive language for children up to 5 years, 9 months of age.³⁹ In addition to individual age-normalized domain scores, there are 3 composite Mullen scores that reflect overall cognitive ability (Early Learning Composite) as well as verbal and non-verbal development quotients. Each of these composite scores is expressed as a standard score with a mean of 100 and standard deviation of 15.

Cost Effectiveness Analysis:

The planned cost-effectiveness analysis will adhere closely to the guidelines for economic analysis accompanying clinical trials set forth by the international Task Force on Good Research Practices on RCT Cost-Effectiveness Analysis.^40,41 Please refer to Supplement A: Cost Effectiveness for more information about our planned cost-effectiveness analysis.

Data Safety and Monitoring Board (DSMB):

An independent DSMB has been set up to review on a regular basis the safety data of the upcoming trial. The DSMB is composed of international experts in clinical trials, perinatal pharmacology, obstetrics, neonatology, epidemiology, and statistics. The DSMB has already met and established guidelines to study monitoring in terms of efficacy and harm. Ethical considerations require that if the intervention is overwhelmingly effective, the data should be released for implementation, and it would no longer be ethical to randomize patients. The DSMB will use the Haybittle-Peto stopping rule as a guide for early stopping, which has the advantages that the exact number of interim analyses need not be specified in advance and the overall type I error is preserved at 0.05 (i.e. sample size adjustment for multiple testing is not needed).⁴² Under this rule, the study will only be stopped if there is an extreme difference in the primary outcome (p<0.001 or >2 standard deviations difference between groups for continuous outcomes). In terms of harm, if new information is discovered during the study that indicates that the intervention provides an unreasonable risk to subjects, the DSMB would recommend enrollment be suspended or terminated. Enrollment will only be resumed once the risk has been appropriately mitigated.

The DSMB plans to meet every 3-6 months during the trial in open and closed sessions. Open sessions will focus on reviewing aggregate data, conduct and progress of the study. This includes participant accrual, protocol compliance, and problems encountered; of note, data by treatment group are not presented in the open session. Closed sessions will be attended only by DSMB members, who will review data by study group to ensure participant safety and well-being. The DSMB will recommend to the sponsors and investigators whether to continue, modify, or terminate the trial on ethical grounds.

Outcomes:

The primary outcome is peripartum blood transfusion, defined as receiving at least one blood transfusion during delivery hospitalization until 7 days postpartum (Table 2). Perinatal blood transfusion is the main driver of the CDC’s severe maternal morbidity composite metric⁶; as such, our study will directly examine whether differences in iron repletion method reduce severe maternal morbidity among those with IDA during pregnancy. Refer to Table 2 for secondary outcomes and Table 3 for timing and location of data collection for all outcomes. Of note, Fishbane reactions—a treatment response characterized by transient flushing, chest tightness, and joint pains—will be categorized as mild/moderate reactions for this study, as per clinical consensus.⁴³

Table 2:

Primary and secondary secondary outcomes

PRIMARY OUTCOME
Outcome Measure	Definition/Assessment
Maternal
Peripartum blood transfusion	Any blood transfusion during delivery hospitalization or within 7 days of hospital discharge
SECONDARY OUTCOMES
Outcome Measure	Definition/Assessment
Maternal
Maternal Hb at delivery	On admission to labor & delivery
Maternal anemia at delivery	Hb <11 mg/dL on admission to labor & delivery
Maternal ferritin at delivery	On admission to labor and delivery
Maternal hemoglobin postpartum day 1	On postpartum day 1
Cesarean delivery	Cesarean delivery for any indication in patients without prior cesarean deliveries
Edinburgh Perinatal Depression Scale (EPDS) score at randomization (baseline) and at 4 – 6weeks postpartum	From EPDS routinely collected as part of standard of care at randomization and postpartum visit at 4 – 6 weeks
Maternal infection	Any infection diagnosed from initiation of treatment until 4 – 6 weeks postpartum
Composite maternal complications	Maternal mortality or any one of several maternal morbidities
Adverse reactions
Mild/moderate reactions	Nausea/vomiting, abdominal pain, diarrhea, constipation, chest tightness, itching, urticaria, flushing, headache, dizziness, arthralgia, back or joint aches, mild shortness of breath, tachycardia, hypertension or hypotension, or Fishbane reaction.
Major infusion reaction	Anaphylaxis-like reactions Non-reassuring fetal heart rate tracing within 24 hours of infusionrequiring delivery
Neonatal
Gestational age at delivery	Based on standard obstetric criteria
Preterm birth <37 weeks	Gestational age at delivery <37 0/7 weeks (spontaneous or indicated)
NICU Admission	Admission to the neonatal intensive care unit for any indication
Birth weight	First weight after delivery
Umbilical artery pH, bicarbonate, lactate, and base excess	Based on routine cord gases
Neonatal hemoglobin	Based on cord blood or first neonatal complete blood count
Neonatal Ferritin	Based on cord blood or first neonatal blood draw
APGAR scores at 1 and 5 minutes	Routine APGAR scores at delivery
Composite neonatal complications	Neonatal mortality or any one of several neonatal morbidities

Table 3:

Data Sources

Data	Source	Frequency
Baseline characteristics	Chart abstraction	Once at enrollment
Medications	Chart review of medical reconciliation	Three times at prenatal visit 4 weeks after enrollment, at deliveryand 4 – 6-week postpartum visit
Blood transfusion	Chart abstraction	Once, at delivery
Maternal Hb at delivery	Chart abstraction	Once, at delivery
Maternal ferritin at delivery	Study specific test	Once, at delivery
Maternal hemoglobin postpartum day 1	Chart abstraction	Once, after delivery
Composite maternal complications	Chart abstraction
Maternal Hb at 4 – 6 weeks postpartum	Chart abstraction if done or studyspecific test	Once, at 4 – 6-week postpartum visit
Maternal ferritin at 4 – 6 weekspostpartum	Study specific test	Once, at 4 – 6-weeks postpartumvisit
Mode of delivery	Chart abstraction	Once, after delivery
Edinburgh Perinatal Depression Scale(EPDS) score	Chart abstraction of routinelyadministered EPDS	Twice, at enrollment and at 4 – 6-week postpartum visit
Maternal infection	Chart abstraction	Multiple throughout pregnancy and postpartum
Mild/moderate reactions	Study specific questions and chartabstraction	Twice, 4 weeks after infusion and at delivery
Major infusion reaction	Chart abstraction	Twice, at and 24 – 48 hours after infusion
Gestational age at delivery	Chart abstraction	Once, at delivery
Preterm birth <37 weeks	Chart abstraction	Once, at delivery
NICU Admission	Chart abstraction	Once, after delivery
Birth weight	Chart abstraction	Once, after delivery

Sample size:

From institutional data, the proportion of patients with IDA treated mostly with oral iron who received a blood transfusion ranged from 5.3% to 15%. Using a baseline blood transfusion rate of 5%, 670 patients will provide 80% power to detect a 75% or greater relative difference in proportion of patients receiving peripartum blood transfusion (5.0% versus 1.25%). To accommodate an anticipated 10% loss to follow-up, we will recruit 746 patients (373 IV iron; 373 oral iron). The anticipated 75% relative reduction in peripartum blood transfusion is plausible. In the only study that reported persistent anemia at delivery, IV iron reduced persistent anemia by 88% compared with oral iron (4.4% vs 37.8%, RR 0.12, 95% CI 0.03, 0.48).⁴⁴ In addition, our pilot trial showed a 100% reduction in blood transfusion among women treated with IV iron compared with oral iron (0% versus 15%).⁴⁵

Given the importance of demonstrating the safety of IV iron, a power analysis was calculated to determine whether this study had sufficient power to detect a difference in serious adverse reactions or mild/moderate reactions between study groups. Assuming the risk for serious adverse events is 0% with oral iron and up to 0.21% with IV iron,⁴⁶ a sample size of 746 (373 per group) will afford >90% power to detect a non-inferiority margin of 1% in serious adverse events, using a one-sided Z-test with the significance level of 0.025. Assuming a rate of mild/moderate adverse events of 20% with oral iron and up to 20.4% with IV iron, the sample size of 746 (373 per group) will provide >90% power to detect a non-inferiority margin of 10% in mild/moderate adverse events, using a one-sided Z-test with the significance level of 0.025.

Based on our prior recruitment for similar studies, we anticipate that approximately 30% (n=75) of parents at WIHRI will consent to have their infants receive non-sedated MRI and neurodevelopmental assessments. The standard deviation of MWF estimates in white matter is 5% in healthy children.⁴⁷ To reliably detect a one standard deviation (5%) difference in MWF between the IV iron and oral iron groups, using a 2-sample t test (alpha = 0.05, power = 0.90), a total of 56 observations (28 per group) are required, including 20% inflation for attrition. For the Mullen Scales of Early Learning, each of the composite scores is expressed as a mean of 100 with standard deviation of 15. The sample size of 75 will provide >90% power to detect one standard deviation (15) difference in the composite score.

Statistical Analysis Plan

A detailed statistical analysis plan will be finalized before unblinding the database. Analyses will be performed according to the intention-to-treat principle, where participants are classified according to their randomized intervention group.⁴⁸ To verify the comparability of the randomized groups, baseline characteristics (e.g. age, BMI, nulliparity, gestational age, education levels) will be compared between groups. Continuous variables will be compared using the student’s t-test or Mann-Whitney U test as appropriate. Dichotomous variables will be compared using the Chi-square or Fisher exact tests as appropriate. Although planned models for outcomes will only have the stratification variables, if any baseline characteristics are unbalanced, additional analyses including these covariates in the model will be performed. The distributions of continuous variables will be examined, with a plan to use transformation or nonparametric methods in cases of violation to the normal distribution assumption. The frequency distribution of all categorical variables will be examined, with the plan to use exact inference procedures in cases of zero or small cell size.

Primary Outcome: Peripartum blood transfusion:

The primary efficacy outcome is the proportion of participants who received peripartum blood transfusion. For this outcome and other categorical secondary outcomes, generalized estimating equations (GEE) will be used to account for study site, maternal age (<35 and ≥35 years), and severity of anemia (Hb<8 and 8-10 g/dL) at enrollment and to estimate risk differences (RDs) and relative risks (RRs), associated 95% confidence intervals (CIs), and p-values using the identity and log link function for binary outcomes and normal link function for continuous outcomes. Prespecified subgroup analyses will be performed for the primary outcome by study site, maternal age (<35 and ≥35 years), BMI (<30, ≥30 Kg/m²), severity of anemia (<8 and 8-10 g/dL), and gestational age at enrollment. The Breslow-Day test will be used to test for homogeneity, assessing whether the relative effect of IV iron differs across subgroups.

Safety Analyses:

We will test the non-inferiority hypothesis that the prevalence of serious and mild/moderate adverse events are not significantly greater in those randomly assigned to IV iron. We will report RD and RR and their associated 95% CIs, and likelihood ratio p-values corresponding to the treatment group indicator from a GEE model including as covariates the stratification variables of study site, maternal age and severity of anemia at enrollment. If the upper confidence limit of the difference is less than the non-inferiority margin (i.e., 1% for serious and 10% for mild/moderate adverse events) we will reject the null hypothesis and conclude that the IV iron is non-inferior to oral iron with regard to serious or mild/moderate adverse reactions.

Neonatal brain MRI and neurodevelopmental outcomes:

Voxel-wise MWF differences between the IV iron and oral iron groups will be evaluated at each image voxel using unpaired t-test and a general linear model (GLM) that includes gestational age at delivery and birth weight as additional variables of in interest. The FMRIB Software Library package (FMRIB Analysis Group, Oxford, United Kingdom) will be used to construct the GLM. Both the GLM and group differences will be tested non-parametrically using permutation testing (randomize) and 5,000 permutations. Significance will be defined as P <.05, with correction for the multiple comparisons in MRI data using a cluster-based technique.⁴⁹ Each of the mean composite scores of the Mullen Scales of Early Learning will be compared using the unpaired t-test and GLM that includes variables such as gestational age at delivery and birth weight known to impact neurodevelopmental scores.

Additionally, we will conduct subgroup analyses of all neonatal outcomes by neonatal sex. We will use tests of interaction to assess if the effects of IV iron on neonatal outcomes, compared with oral iron, vary by sex.

Data management

Data Collection:

Detailed antepartum, intrapartum, and postpartum information from study participants will be collected. Please refer to Supplement A: Data Collection for more information.

Data Processing:

The Data Coordinating Center (DCC) will be responsible for data management and analysis. Data will be collected and managed with REDCap, an established, secure, web-based data capture and management tool developed at Vanderbilt University and supported by the bioinformatics team at WIHRI.^28,29 Refer to Supplement A: Data Processing for more information.

Loss to follow-up and missing data:

Strategies used in prior studies will be employed to achieve excellent follow-up and minimal missing data.^26,27 Written authorization will be obtained to access delivery records if participants end up delivering at a facility not participating in the study. Active data management will occur at each site, including monthly data review for rates and patterns of missing data. Remedial measures including data re-abstraction and staff retraining will be used as needed to minimize missing data. Despite these efforts, data may still be missing. If significant deficiency is noted it will be addressed analytically.⁵⁰

Ethics and Dissemination

Ethical approval has been granted by a central institutional review board at Advarra (Pro00060930), and participating centers—WIHRI, University of Michigan Medical Center, Washington University School of Medicine in St. Louis, and Hasbro Children’s Hospital—have ceded ethical approval to Advarra. Witnessed informed written consent will be obtained from each participant before conducting any study-related procedures. The results will be presented to and shared with members of each participating center and presented to the international medical community via publications and scientific presentations.

The IVIDA2 protocol is registered on Clinical Trials Registry (NCT05462704).

Discussion:

Iron deficiency anemia is common condition during pregnancy worldwide and in the United States and is associated with maternal morbidity and adverse neonatal outcomes. Treatment is recommended, but ideal route of iron administration remains unclear. Via a multicenter and double-blinded, placebo-controlled approach, IVIDA2 aims to provide the high-quality data needed for clinicians, medical governing bodies, and policymakers to determine whether intravenous iron should become the preferred treatment modality for IDA in pregnancy. The results of this study will likely be of major significance and will be widely disseminated to local and national medical authorities, obstetric governing bodies, and clinicians via peer-reviewed journal publications and presentations at high-impact scientific meetings.