Summary less than 27 weeks), birthing center, and

Summary of the StudyIn this large-scale study conducted across 25 centers in 7 countries, researchers investigated whether delaying clamping of the umbilical cord had effects on the mortality and morbidity of preterm infants born prior to 30 weeks gestation. Researchers assigned fetuses either to immediate cord clamping (within 10 seconds of birth) or delayed cord clamping (at least 60 seconds after birth), and the mortality/morbidity outcome was measured at 36 weeks gestation. The study did not find any significant difference in this outcome between the immediate clamping and delayed clamping groups.Research Question and Study Design     Research Question & Current EvidenceThe central question posed in this study concerns the preferred timing of umbilical cord clamping upon the birth of preterm infants. In the past, the normal practice was immediate cord clamping “because of concerns about harm from delayed resuscitation, hypothermia, hyperbilirubinemia, or polycythemia” (Tarnow-Mordi et al., 2017, p. 2446). However, there is some evidence that delayed cord clamping, which increases the amount of placental blood received by a newborn, results in lower mortality and decreased incidence of necrotizing enterocolitus and infection (Tarnow-Mordi et al.). This has led some professional organizations to establish guidelines for delays in cord clamping, ranging from 30 to 180 seconds (Tarnow-Mordi et al.). However, based on current evidence, the authors of this study conclude that “whether delayed clamping alone has benefits with respect to mortality or the incidence of neurodevelopmental disability remains unknown” (Tarnow-Mordi et al., p. 2446).     Study DesignIn this study, fetuses from women expected to deliver before 30 weeks of gestation were randomly assigned to either immediate cord clamping (10 seconds or less after delivery) or delayed clamping (60 seconds or more after) (Tarnow-Mordi et al., 2017). This study was designed to capture a large sample across 25 birthing centers in 7 countries. Randomization was done shortly before the time of the birth and stratified based on gestational age (more or less than 27 weeks), birthing center, and singleton vs. multiple births (Tarnow-Mordi et al.). The study used, as its primary outcome, a composite measure of death or major morbidity, which is discussed in more detail below (see “Research Methods and Measurement Tools”). This outcome was measured at 36 weeks of gestation.The authors describe their study design as one that “minimized data collection in order to maximize enrollment” (Tarnow-Mordi et al., 2017, p. 2453). For example, data was not collected on administration of antenatal glucocorticoids, time to onset of breathing, intubation, or fetal heart rate or oxygen saturation (Tarnow-Mordi et al.) The absence of this data is a severe limitation of the study design. Although the authors do not say that scarcity of data is a study limitation, they do suggest antenatal glucocorticoid administration data would have aided an analysis of why their “finding contrasts with those of earlier systematic reviews involving smaller populations” (Tarnow-Mordi et al., p. 2451). Also, more comprehensive data collection would have supported an analysis of the underlying causes and implications of the poor adherence to treatment in the delay arm, as discussed below (see “Study Findings and Limitations”).Sample and SettingA sample size of 1,634 fetuses were enrolled and randomized across 25 birthing centers in seven different countries, with study eligibility based on a medical determination that a fetus was considered to have an expected delivery date before 30 weeks gestation (Tarnow-Mordi et al., 2017). Of the 1,634 fetuses enrolled and randomized, 1,566 infants were born alive before 30 weeks and included in the study results, of which 782 were assigned to immediate cord clamping and 784 to delayed cord clamping (Tarnow-Mordi et al.). The number of fetuses enrolled met the goal of the study, which was an enrollment of 1,600, and analysis was performed on an intent to treat basis, under which a fetus that had been assigned to a treatment was included in the analysis even if the actual time between birth and cord clamping was nonadherent with the assigned treatment (Tarnow-Mordi et al.).Research Methods & Measurement ToolsIn this study, the primary outcome was a composite measure of death or major morbidity, defined as occurrence or diagnosis, before 36 completed weeks of post-menstrual age, of any of the following: death, severe brain injury, severe retinopathy, necrotizing enterocolitus, or late-onset sepsis. The authors chose a composite measure, rather than a measure based simply on mortality, because their study, even with its large sample, was not “powered for mortality,” that is, lacked the statistical power to detect anticipated differences in mortality (Tarnow-Mordi et al., 2017 p. 2451).A significant problem with this study is that the measure of primary outcome was changed during the course of the study. The original primary outcome measure had also included, as one of the morbidity components of the composite measure, chronic lung disease. (Tarnow-Mordi et al., 2017). During the study, the researchers found that their definition of chronic lung disease was sweeping in 64% of subjects, rather than the expected 26%, so they dropped chronic lung disease from the primary outcome measure (Tarnow-Mordi et al.). The implications of this change are discussed under the “Study Limitations” heading below.Study Findings and Limitations     Study FindingsThe median time to cord clamping was 5 seconds in the immediate-clamping group and 60 seconds in the delayed-clamping group. The study found no statistically significant difference in the composite primary outcome (death or major morbidity) at 36 completed weeks between infants assigned to immediate cord clamping and those assigned to delayed cord clamping. While fewer infants in the delayed-clamping group died at 36 completed weeks, the researchers explain that this may be a chance finding. There was no significant difference in median APGAR scores at 1 minute and 5 minutes. Of note, the rate of adherence to the assigned cord clamping time was 94.9% in the immediate-clamping group but only 73.2% in the delayed-clamping group.     Study LimitationsThere are several limitations to this study. As noted above, the study was intentionally designed to collect only minimal data in hopes of increasing enrollment (Tarnow-Mordi et al., 2017, p. 2453). It would be beneficial to have more data though, especially when exploring why the adherence rate was just 73.2% in the delayed-clamping group. Was it due to miscommunication, logistical issues, need for immediate neonatal resuscitation interventions, or another reason entirely? More rigorous data collection could help shape future studies on the subject.The researchers acknowledged that neither their own study nor any similar study to date had been powered for a mortality, as noted above (Tarnow-Mordi et al., 2017, p. 2451). They calculated a required sample size of 11,000 patients to yield 90% power to detect a 20% difference in the risk of death between the immediate-clamping and delayed-clamping groups. It would be incredibly challenging to design and implement such a large study.With regard to dropping lung disease from the primary outcome measure (as described above), it is a shortcoming of the study report that there is no discussion of the implications this may have had on the validity of the study design, and the authors fail to describe as study limitations either the change in the primary outcome measurement or the absence of an appropriate measure of lung disease.Relevance to Clinical PracticeTraditionally, placental transfusions are a relatively easy way to provide preterm babies with the extra blood they may need at birth, however, delayed cord clamping has the potential to deliver larger amounts of blood, oxygen and nutrients to infants delivered preterm. Briefly delaying the clamping of the umbilical cord requires no extra procedures or equipment, thus making it a very low-cost intervention. Because of this, delaying the clamping of the cord is an intervention that can be implemented anywhere, from hospitals to deliveries in the field, and developing countries at no cost. While the results of this particular study did not yield results that were determined to be statistically relevant, the study raises intriguing questions about the potential effects of delayed cord clamping that warrant further analysis. For example, what might have been the outcome for the 26.8% of infants that were selected to receive delayed cord clamping yet the physicians felt clamping was medically necessary to perform other life saving measures? What might have been found if the researches prolonged the amount of time delayed before clamping? While this study does not provide us with the data to answer such questions, it does warrant conduction of further studies altering the variables and controls to further investigate delayed cord clamping effects.