Predictive Model for Late Stillbirth Among Antenatal Hypertensive Women

Prediction Models for Adverse Pregnancy Outcomes in India: Methodological Considerations for an Emerging Topic

“…A research question for prediction or prognosis [ 3 ] is different to the more commonly investigated goal of explanation (aetiology) [ 7 ] and warrants a targeted approach to study design, model development and bias assessment. Contrary to the implicit assumption by Kumar et al[ 2 ], it is possible for single risk factors to be not statistically significantly associated with stillbirth yet be predictive when considered together with other risk factors. Internal validation via a single hold-out sample or n-fold cross-validation has its limitations but at least provides a degree of confidence in predictive performance because the validation set is somewhat independent of the data used to develop the model.…”

“…Ideally researchers would require a registry of all births to enable estimation of incidence (or perhaps more accurately, prevalence at birth) and to ensure the study sample is both representative of pregnancies that end in stillbirth as well as live birth. Consequently, it is unclear as to whether the performance for the model reported by Kumar et al [2] is impacted by selection of the control population.…”

“…Internal validation via a single hold-out sample or n-fold cross-validation has its limitations but at least provides a degree of confidence in predictive performance because the validation set is somewhat independent of the data used to develop the model. Lack of internal validation, as is the situation in the study by Kumar et al[ 2 ], motivates postponement of external validation until this is achieved, and consequently proscribes application of the model for screening. Model development for stillbirth prediction would typically require observation of large numbers of stillbirths of the order of thousands and perhaps tens of thousands ( cf.…”

“…If stillbirth can be predicted in such countries, a large number of cases might be prevented. In response to this significant global issue, Kumar et al [ 2 ] developed a model for the prediction of stillbirth after 28 weeks of pregnancy for women with hypertensive disorders of pregnancy. Abruption, gestational hypertension, family history of hypertension, maternal education, low number of antenatal visits (< 4 visits) and low foetal weight (< 2000 g) were used as predictors.…”

Effect of implementing acupressure therapy on gastrointestinal tolerance and growth monitoring for premature neonates

“…This finding may be interpreted by the fact that acupressure can trigger certain mechanisms such as, improvement of GIT motility and regularity as well as the release of food absorption hormones which can be explained by the effect of moderate pressure massage. Moderate pressure massage stimulates the release of ornithine decarboxylase (anp index of growth hormone) [10] It also stimulates the vagus nerve leading to increased vagal activity and in turn to increased gastric motility and the release of insulin (food absorption hormone)and Insulin-like Growth Factor1(IGF-1) [11] . Time to reach full feeds has significant effect on long term neurodevelopment outcome in premature neonates.…”