Anemia affects 50–80% of postpartum mothers in low- and middle-income countries (LMICs).1 Postpartum anemia can be caused by untreated antenatal iron-deficiency anemia or excessive blood loss during or after childbirth.
Anemia is a lack of enough red blood cells or hemoglobin needed to transport oxygen through the body. Common symptoms of anemia are fatigue, weakness, shortness of breath, and psychological symptoms such as anxiety or depression. Iron-deficiency anemia is a lack of red blood cells due to insufficient iron, leading to low hemoglobin production. This type of anemia can be treated with iron supplementation, but other types, such as hemolytic anemia, may not respond to treatment with iron supplements. Untreated anemia can lead to other complications such as increased susceptibility to infections, increased or irregular heartbeat, or impaired lung function.
Postpartum depression (PPD) has been reported to be associated with postpartum anemia. PPD is a serious mental health problem, often occurring within four to six weeks after childbirth. Without treatment, PPD has negative consequences for both mothers and their infants.2 The prevalence of PPD varies based on the definition of the disease, country-specific factors, diagnostic tools, threshold of diagnosis for the screening measure, and period over which the prevalence is determined.2
Ki collaborated with a primary investigator in Nigeria, her research team, and a BMGF program strategy team to better understand IV iron as a postpartum intervention. Our goals were to understand IV iron as an intervention for alleviating postpartum anemia as well as consider IV iron as an intervention for the postpartum mental health of mothers with anemia in the region.
While depression has been reported among women with iron-deficiency anemia, little data was available to understand whether treating anemia with IV iron could also resolve or prevent mental health symptoms, especially in the trial region. Therefore, the impact of designing a study for a depression outcome or an anemia outcome was unclear. Through a literature review, real-world evidence (RWE), and clinical trial simulations, the team collaborated to understand if a single study was practical for addressing depression as well as anemia.
We assembled a knowledge base in women’s mental health research, epidemiology, and clinical mental health assessments to facilitate a gap analysis for real-world data (RWD) evaluation and provide quantitative clinical trial parameter estimates.
We found a broad range of thresholds for a depression diagnosis, which contributed to wide parameter estimates for depression. First, we highlighted key characteristics of existing epidemiological surveys and how they correlated to or differed from a subsequent clinical trial. From these literature sources, we evaluated the clinical instruments utilized within included studies, exploring which aspects of depression were covered and how they pertained to the proposed study. We explored how the Edinburgh Postnatal Depression Scale (EPDS) was administered in previous epidemiological studies of mothers in Nigeria.3 Specifically, we looked at the frequency, the threshold for postpartum depression diagnosis, and anemia thresholds in the target populations. We also assessed alternative proposed clinical instruments, such as the Fatigue Severity Scale (FSS) and the mother-to-infant bonding scale (MIBS).
As with many historical evaluations of mental health conditions, it is often difficult to find studies with consistent definitions, tools, and thus, estimates that can be used for trial planning simulations. While our mental health knowledge base did not identify a data source to accurately inform clinical trial design parameters from existing literature, it informed our real-world evidence (RWE) studies.
In addition to the literature review, we sought RWE to understand the relationship between postpartum anemia and depression and establish parameter estimates for clinical trial simulations. Since both outcomes were under consideration for a clinical trial, possibly as a co-primary endpoint, we wanted to evaluate whether we could expect both to have enough statistical power. If the two proposed outcomes were not highly correlated, the proposed trial might not adequately describe the treatment effects of one of the outcomes without a much larger sample size.
We extracted relevant RWD from electronic medical records (EMR) in Optum Integrated Data and the Cerner Health Datalab. RWD is an assembly and curation of data from global health studies and other sources, including retrospective and prospective observational studies, EMRs, and individual-level patient experience data.
Real-world datasets are selected based on attributes of interest, such as diagnosis codes, outpatient or hospital visits, cost information, laboratory values, exposure (or lack of exposure) to selected medicines, pregnancy, age, gender, or clinical history. We chose a population that met the entry criteria of the proposed trial to mimic that population as much as possible. The EMRs we collected for the selected population were diagnoses (e.g., birth outcomes, EPDS score, etc.), labs (e.g., hemoglobin, iron panel, etc.), prescriptions (e.g., iron), and patient demographics. The records from the selected population were analyzed for descriptive statistics.
Descriptive statistics did not yield convincing evidence of an association between postpartum anemia and postpartum depression, or that treatment with iron affects postpartum depression scores at six weeks. From these findings, we expect weak-to-modest effects of IV iron treatment for postpartum depression, which would require a large sample size to confirm as a co-primary endpoint in a clinical trial.
Our results helped prioritize the clinical outcomes for a trial on IV iron as an intervention, with anemia as a primary outcome and postpartum depression as a secondary outcome. An exploratory trial with postpartum depression as a secondary outcome may support the decision to address this outcome in further research. The choice of anemia as a primary outcome was supported by literature review and RWE. Our literature review indicated that depression levels had a broad range of thresholds for diagnosis, which resulted in varied sample size estimates. RWE suggested that there was not a strong relationship between anemia and postpartum depression.
While there are potential limitations due to a lack of available data, our analysis uncovered a gap in the current understanding that can be addressed in the planned trial and other future research. The differences in the literature among regions, time points, and trial criteria made it difficult to determine trial parameter estimates. In addition, the differences in RWD across regions mean that RWE derived from high-income countries (HIC) may only partially apply to the trial region. For example, the association between anemia and postpartum depression may be stronger in the target geography. Likewise, when using RWD to evaluate safety (e,g, increased infection rates could be more likely with the bolus dose of IV Iron compared to oral iron), we would not want to ignore a safety signal in high-resource settings since it would likely be worse or there would be less access to treatment in low-resource settings.
We supported the wider team in understanding the options available in trial design and the impact of those choices. By collaborating and sharing our experience in mental health research, epidemiologic research, RWE, and clinical understanding of mental health assessments, we helped the primary investigator make well-informed clinical trial design choices for her locale.
Datasets Utilized
Literature review of women’s mental health research, epidemiology, and clinical mental health assessments and real-world data (RWD) from electronic medical records
Keywords
Intravenous Iron, postpartum anemia, postpartum depression, epidemiology, real-world data (RWD), real-world evidence (RWE), Nigeria