PropertyValue
is nif:broaderContext of
nif:broaderContext
is schema:hasPart of
schema:isPartOf
nif:isString
  • This methodological study was part of a larger prospective cohort study completed between 2013 and 2015, the Embarazo Seguro Bebé Sano (Safe Pregnancy Healthy Baby) Study, investigating several potential environmental, nutritional and infectious risk factors for disease during pregnancy and infancy [27, 28]. Study participants were recruited from pregnant women presenting to Hospital Regional de Occidente—San Juan de Dios, the public primary care clinics in San Juan Ostuncalco or Concepción Chiquirichapa, and the clinics of the non-governmental organizations Asociación de Comadronas del Área Mam and Pies del Occidente from May 2013 to February 2014. Medical chart reviews and interviews of the pregnant women were used to determine eligibility. Women were eligible if they were non-smokers between the ages of 18 and 40 years old, had a low-risk, singleton pregnancy less than 20 weeks gestation by LMP or ultrasound and not greater than 24 weeks by the other method, lived and planned to stay in the area for at least six months post-delivery and had access to a mobile phone. Low risk pregnancies were defined as maternal blood pressure less than139/89 mmHg and no protein, glucose or ketones in the urine at the baseline clinical exam. A home visit was made within one week after enrollment to assess baseline household and maternal characteristics. Weekly phone calls and cell phone text messages were conducted after baseline to assess pregnancy status until the mother delivered her infant. Mothers were instructed to contact the study personnel if they went into labor. Study personnel assessed newborns either at the hospital or in the home, depending on place of delivery and timing of the assessment. Determination of gestational age at birth: Three methods were used to calculate the gestational age at birth: fetal ultrasound, New Ballard and LMP. Preterm births were those infants having a gestational age at birth less than 37 weeks. As defined by the WHO, moderate to late preterm births are those between 32 and 37 weeks, very preterm between 28 and 32 weeks, and extremely preterm less than 28 weeks [29]. LMP was self-reported during the baseline clinical exam. If the exact date of the month could not be recalled, the 15th day of the month was used. We determined the estimated delivery date (LMP + 280 days). The difference between the estimated delivery date by LMP and actual delivery date was then used to determine gestational age in weeks: (40weeks-LMPestimateddelivery-actualdelivery7). Project nurses assessed newborn gestational age using the New Ballard (NB) maturational assessment. The New Ballard Score measures 6 physical and 6 neuromuscular criteria; the total score ranges from -10 to 50. Nurses were trained over 2 days; didactic training including videos of the New Ballard assessment was followed by hands-on assessments of newborns in the postpartum ward of the hospital. NB scores were calculated for infants examined within 96 hours of birth in 156 (83%) of the live births. Typically the New Ballard gives gestational age in 2 week increments. To provide gestational ages in smaller increments, gestational age (in weeks) was determined from the NB score by a regression equation from published scores and estimated gestational ages (0.4*NB score + 24) [18]. All mothers had an ultrasound taken as part of the screening for recruitment eligibility. To standardize measures among those performing ultrasounds, study physicians received didactic and hands-on training on ultrasound use before the study began, complemented by follow-up in-person trainings during the course of the study. Consistent quality measurements were required before physicians were permitted to perform ultrasounds for the study. For quality control, 33% of study images were assessed by a board-certified obstetrician and ultrasounds were repeated if they were found to be of low quality. When multiple ultrasounds were performed, the earliest set of high quality images was used. Before 14 weeks gestational age, crown-rump length was used to determine gestational age [30]. After 14 weeks gestation, gestational age was taken as the mean age determinations from biparietal diameter measurement, head circumference, femur length and abdominal circumference. If gestational age computed from any individual measure differed from the mean gestational age by more than the level of accuracy at that gestational age (1 week for ultrasounds obtained before 21 weeks or 2 weeks for ultrasounds obtained at 21 weeks or later) [8], that measure was excluded from the mean estimated gestational age. Additionally, all available images were evaluated to verify poor caliper placement used during fetal measurement and those with improper caliper placements were excluded from analysis. The difference between estimated delivery date by ultrasound and the actual delivery date was used to calculate gestational age at birth in weeks. For the calculation of small for gestational age, newborn weight was measured to the nearest 5 grams using a Seca 334 digital infant scale. Weight was measured in triplicate and clothing weight was subtracted from infant weight. Measurements within 50 g of each other were used to determine the average. If either the birthweight or clothing weight were implausible (<1400 grams or >5000 grams for birthweight and <100 g or >1000 g for clothing) or taken later than two weeks after birth, the recorded birthweight from the hospital medical record was used, if available (n = 25), or excluded from analysis, if unavailable. Only weights taken within 2 weeks of birth (n = 184) were used in the estimation of small-for-gestational age. All statistical analyses were performed in Stata 13. Descriptive statistics (proportion, mean, standard deviation, median, interquartile range) were used to describe the data. Lin’s concordance correlations [31] and Bland-Altman plots [32] were used to determine agreement between different methods of measuring gestational age. Lin’s concordance correlation is the product of the Pearson correlation coefficient and a bias correction factor, a measure of accuracy that indicates how close the best-fit line deviates from the line y = x or perfect agreement between methods. Fixed biases (systematic differences that can be adjusted for in analyses), or mean difference between methods, were assessed using the Bland-Altman plots. In addition, Bland-Altman plots were used to assess the 95% limits of agreement or the range in which 95% of differences would be expected to fall. Spearman rank correlation was used to determine correlation between ultrasound-determined gestational age and individual New Ballard item scores or physical and neuromuscular subscores. Multivariate linear regression was used to assess the association of maternal and infant characteristics on the difference in gestational age estimates compared to ultrasound. Differences in these differences based on sociodemographic characteristics were estimated from a linear regression equation using maternal and infant characteristics as explanatory variables. Cohen’s kappa statistic [33] was used to assess agreement in the determination of preterm birth between the different methods of measuring gestational age. Using ultrasound as the gold-standard method, the sensitivities, specificities, and diagnostic odds ratios [34] for each method were determined. The diagnostic odds ratio (DOR) is the ratio of the odds that the method identifies a preterm infant as preterm to the odds that it identifies a full-term infant as preterm. Higher DORs are indicative of better performance and a DOR of 1 indicates that the test is as likely to predict that an infant is preterm regardless of whether they are. Associations between NB item scores and preterm birth were determined by Fisher’s exact test and differences in physical and neuromuscular subscores between full and preterm birth were tested for statistical significance using the Wilcoxon rank sum test. Small-for-gestational age was defined as an infant being below the 10th weight centile for gestational age as determined by the INTERGROWTH-21st Newborn Size Application Tool using standards derived from births in eight ethnically distinct countries (http://intergrowth21.ndog.ox.ac.uk/). This study was approved by human subjects committees at Universidad del Valle de Guatemala, Emory University, and the Guatemalan Ministry of Public Health and Social Welfare. This study was reviewed in accordance with CDC human subjects protection policy and was determined to be human subjects research, but CDC involvement did not constitute engagement in human subjects research. All participants signed a written consent that was read to them in Spanish and, if they were unable to read or write, a witness also signed to ensure comprehension by the participant.
rdf:type