Fetal nondiabetic-macrosomia: risk factors for pregnancy adverse outcome and comparison of two growth curves in the prediction of cesarean section.

BACKGROUND: Randomized trials reported no difference whether induction or expectant management is performed in non-diabetic women with large for gestational age babies but no tool has been validated for the prediction of high risk cases. AIM: Assessing the performance of different growth curves in the prediction of complications. METHODS: Data from 1066 consecutive non-diabetic women who delivered babies ≥4000 g were collected. Logistic regression analysis was used to analyze the impact of the maternal variables on: instrumental delivery, shoulder dystocia (SD), perineal tears, cesarean section (CS), and postpartum hemorrhage. Intergrowth21 curves and customized Gardosi's curves were compared in terms of prediction of adverse outcomes. FINDINGS: Induction of labor was performed in 23.1% cases. The rate of CS was 17%. Hemorrhage, fetal distress, and SD occurred in 2%, 1.3%, and 2.7% of cases, respectively. Induction was significantly associated with instrumental delivery (p < .001), CS (p = .001), third and fourth degree perineal tears (p = .031), and post-partum hemorrhage (p = .02). The cutoff of 90th percentile according to Intergrowth21 did not show significant performance in predicting CS, while the same cutoff according to the Gardosi curves showed an OR 1.92 (CI 1.30-2.84) (p = .0009). DISCUSSION: Gardosi curves showed a better performance in predicting the risk of CS versus Intergrowth curves. Induction is significantly associated with adverse outcome in non-diabetic women with LGA babies.

Contaminations in (meta)genome data: An open issue for the scientific community.

In recent years, the high throughput and the low cost of next-generation sequencing (NGS) technologies have led to an increase of the amount of (meta)genomic data, revolutionizing genomic research studies. However, the quality of sequencing data could be affected by experimental errors derived from defective methods and protocols. This represents a serious problem for the scientific community with a negative impact on the correctness of studies that involve genomic sequence analysis. As a countermeasure, several alignment and taxonomic classification tools have been developed to uncover and correct errors. In this critical review some of these integrated software tools and pipelines used to detect contaminations in reference genome databases and sequenced samples are reported. In particular, case studies of bacterial contaminations, contaminations of human origin, mitochondrial contaminations of ancient DNA, and cross contaminations are examined.