Validation of a Survey Instrument for Screening of Pre-Pregnant Women - The PREPSA SCALE

Background

Over the past 40 years rates of anemia among reproductive age women,especially in low and middle-income countries (reported range 20-60%) have notdecreased despite a multitude of governmental campaigns and fundedprograms. While many women suffer from multiple micronutrientdeficiencies, iron is considered most important and is associated withincreases in mortality and morbidity in both non-pregnant as well as pregnantwomen and their offspring. Additionally, iron deficiencies that present atvarious stages of a woman’s life have been associated with poorerlife-quality. The degree of anemia has been shown in many studies to beassociated with the worst health outcomes. There are a number of timesduring a woman’s reproductive period that effective interventions mightpreferentially be employed -- pre-pregnancy, during gestation orpost-partum. Iron deficiency anemia is an intergenerationalcondition and the initiation of menses adds further burden to alreadyreduced iron stores. A number of initiatives employing a single infusionof IV Iron, as compared to the usual standard of care oral iron, have shownpromise. Large definitive studies are currently ongoing in pregnant women, anda Global Network trial is soon to begin in women with moderate anemia diagnosedimmediately post-delivery. It would be ideal if one could identify womenalready suffering from anemia and determine its severity pre-pregnancy, thusallowing focused public health or personalized interventions to be implemented.Such programs may have more impact and be more cost-effective than interveningduring pregnancy or in the post-partum period. Health agencies often recommend a course of oral iron for those who aremildly anemic, a single dose intravenous iron infusion for those suffering frommoderate anemia, and referral for a complete medical and hematologic work-upprior to treatment for those diagnosed with severeanemia.  Additionally, hormonal therapy and the use of othernutritional supplements could be considered. Regardless of the proposedintervention, an instrument that exhibits good correlation with grade of anemiais likely to be widely utilized. During a woman’s pre-pregnancyreproductive years, heavy menstrual bleeding (HUB) is likely a most importantcontributor to high rates of anemia. The majority of women suffering from HUBhave no underlying pathology, thus allowing for focused interventions toreplete iron stores. Religious beliefs and customs often prevent womenfrom speaking about issues related to menstruation, even if their flow isabnormal, often viewing heavy menstrual bleeding as “normal†or a necessaryburden.

To our knowledge, there does not exist a validated instrument that couldbe easily utilized in LMICs that would convey meaningful associations between awoman’s response to a series of focused questions and her hematologic findingsreferable to anemia. A small trial conducted in Spain utilized a questionnairethat focused almost exclusively on descriptive menstrual blood loss and islikely not applicable to use in LMICs.

Study Aim

To develop and validate the PREPSA scoring instrument in women ofreproductive age prior to pregnancy.

Study Design

The proposed pilot study incorporates 23 individual questions (AppendixA) which will be provided to village/rural women by female community healthworkers. We wish to sample 500 pre-pregnancy women (ages 18-25). Subsequent tothe completion of responses to the questionnaire (and within four weeks) allpreviously consented participants will have blood drawn to obtain hemoglobin,ferritin, and transferrin saturation values. The responses thatmost closely align with low hemoglobin indices will be identified and becomethe basis of a scored, validated instrument which may be weighted based uponthe strength of the associations found. Results will inform futurehealth-outcome trials; interest among funders has already been confirmed.

The timeline includes a 2–3-monthapproval window and a four-month intervention, with 1-2 months to review studydata and generate a primary publication.

Data Analyses

We plan to analyze the data in threesteps. First, we will use a two-parameter logistic item response model tocalibrate the scale and obtain information about the best items to include in arevised measure. Second, we will confirm the one-dimensionality of the scaleusing confirmatory factor analytic methods appropriate for binary andcategorical outcomes. Third, the predictive validity of the revised scale willbe tested via linear regression using the metrics gleamed from the blood samplesas outcomes. While calibration, refinement, and validation of the PREPSAinstrument will be an important outcome of this trial, the results will alsofacilitate future health-outcome trials in LMICs.