Prediction of severe adverse neonatal outcomes at the second stage of labour using machine learning: a retrospective cohort study.

OBJECTIVE: To create a personalised machine learning model for prediction of severe adverse neonatal outcomes (SANO) during the second stage of labour. DESIGN: Retrospective Electronic-Medical-Record (EMR) -based study. POPULATION: A cohort of 73 868 singleton, term deliveries that reached the second stage of labour, including 1346 (1.8%) deliveries with SANO. METHODS: A gradient boosting model was created, analysing 21 million data points from antepartum features (e.g. gravidity and parity) gathered at admission to the delivery unit, and intrapartum data (e.g. cervical dilatation and effacement) gathered during the first stage of labour. Deliveries were allocated to high-risk and low-risk groups based on the Youden index to maximise sensitivity and specificity. MAIN OUTCOME MEASURES: SANO was defined as either umbilical cord pH levels ≤7.1 or 1-minute or 5-minute Apgar score ≤7. RESULTS: The model for prediction of SANO yielded an area under the receiver operating curve (AUC) of 0.761 (95% CI 0.748-0.774). A third of the cohort (33.5%, n = 24 721) were allocated to a high-risk group for SANO, which captured up to 72.1% of these cases (odds ratio 5.3, 95% CI 4.7-6.0; high-risk versus low-risk groups). CONCLUSIONS: Data acquired throughout the first stage of labour can be used to predict SANO during the second stage of labour using a machine learning model. Stratifying parturients at the beginning of the second stage of labour in a 'time out' session, can direct a personalised approach to management of this challenging aspect of labour, as well as improve allocation of staff and resources. TWEETABLE ABSTRACT: Personalised prediction score for severe adverse neonatal outcomes in labour using machine learning model.

Prediction of adverse perinatal outcome by fetal biometry: comparison of customized and population-based standards.

OBJECTIVE: To compare the predictive performance of estimated fetal weight (EFW) percentiles, according to eight growth standards, to detect fetuses at risk for adverse perinatal outcome. METHODS: This was a retrospective cohort study of 3437 African-American women. Population-based (Hadlock, INTERGROWTH-21st , World Health Organization (WHO), Fetal Medicine Foundation (FMF)), ethnicity-specific (Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD)), customized (Gestation-Related Optimal Weight (GROW)) and African-American customized (Perinatology Research Branch (PRB)/NICHD) growth standards were used to calculate EFW percentiles from the last available scan prior to delivery. Prediction performance indices and relative risk (RR) were calculated for EFW < 10th and > 90th percentiles, according to each standard, for individual and composite adverse perinatal outcomes. Sensitivity at a fixed (10%) false-positive rate (FPR) and partial (FPR < 10%) and full areas under the receiver-operating-characteristics curves (AUC) were compared between the standards. RESULTS: Ten percent (341/3437) of neonates were classified as small-for-gestational age (SGA) at birth, and of these 16.4% (56/341) had at least one adverse perinatal outcome. SGA neonates had a 1.5-fold increased risk of any adverse perinatal outcome (P < 0.05). The screen-positive rate of EFW < 10th percentile varied from 6.8% (NICHD) to 24.4% (FMF). EFW < 10th percentile, according to all standards, was associated with an increased risk for each of the adverse perinatal outcomes considered (P < 0.05 for all). The highest RRs associated with EFW < 10th percentile for each adverse outcome were 5.1 (95% CI, 2.1-12.3) for perinatal mortality (WHO); 5.0 (95% CI, 3.2-7.8) for perinatal hypoglycemia (NICHD); 3.4 (95% CI, 2.4-4.7) for mechanical ventilation (NICHD); 2.9 (95% CI, 1.8-4.6) for 5-min Apgar score < 7 (GROW); 2.7 (95% CI, 2.0-3.6) for neonatal intensive care unit (NICU) admission (NICHD); and 2.5 (95% CI, 1.9-3.1) for composite adverse perinatal outcome (NICHD). Although the RR CIs overlapped among all standards for each individual outcome, the RR of composite adverse perinatal outcome in pregnancies with EFW < 10th percentile was higher according to the NICHD (2.46; 95% CI, 1.9-3.1) than the FMF (1.47; 95% CI, 1.2-1.8) standard. The sensitivity for composite adverse perinatal outcome varied substantially between standards, ranging from 15% for NICHD to 32% for FMF, due mostly to differences in FPR; this variation subsided when the FPR was set to the same value (10%). Analysis of AUC revealed significantly better performance for the prediction of perinatal mortality by the PRB/NICHD standard (AUC = 0.70) compared with the Hadlock (AUC = 0.66) and FMF (AUC = 0.64) standards. Evaluation of partial AUC (FPR < 10%) demonstrated that the INTERGROWTH-21st standard performed better than the Hadlock standard for the prediction of NICU admission and mechanical ventilation (P < 0.05 for both). Although fetuses with EFW > 90th percentile were also at risk for any adverse perinatal outcome according to the INTERGROWTH-21st (RR = 1.4; 95% CI, 1.0-1.9) and Hadlock (RR = 1.7; 95% CI, 1.1-2.6) standards, many times fewer cases (2-5-fold lower sensitivity) were detected by using EFW > 90th percentile, rather than EFW < 10th percentile, in screening by these standards. CONCLUSIONS: Fetuses with EFW < 10th percentile or EFW > 90th percentile were at increased risk of adverse perinatal outcomes according to all or some of the eight growth standards, respectively. The RR of a composite adverse perinatal outcome in pregnancies with EFW < 10th percentile was higher for the most-stringent (NICHD) compared with the least-stringent (FMF) standard. The results of the complementary analysis of AUC suggest slightly improved detection of adverse perinatal outcome by more recent population-based (INTERGROWTH-21st ) and customized (PRB/NICHD) standards compared with the Hadlock and FMF standards. Published 2019. This article is a U.S. Government work and is in the public domain in the USA.

FK778 does not impair intestinal allograft absorption in a preclinical model of total small bowel transplantation.

Malononitrilamide 715 (FK778), a new low-molecular weight immunosuppressant, inhibits both T-cell and B-cell function by acting on the pathway for de novo pyrimidine biosynthesis. Pyrimidines are important for intestinal trophism; their inhibition may predispose to metabolic and functional impairments, such as diarrhea and malabsorption. In this study we assessed the absorptive capacity of intestinal allografts in a large-animal model of small bowel transplantation (SBTx) in pigs chronically treated with FK778. Ten outbred pigs underwent total orthotopic SBTx. Immunosuppression consisted of oral tacrolimus (trough levels 5-15 ng/mL) and oral FK778 (4 mg/kg per day) administered for 60 days. The D-xylose absorption test was performed at day 60 to evaluate carbohydrate absorption. Results were compared to normal controls. Eight of the 10 animals were alive and in good condition at day 60. All of their allografts were free of rejection. The animals had a mean maximal weight loss of 6.4% during the study period; the final weight was comparable to the initial weight (P > .05). Diarrhea was present in all animals (mean 16% of postoperative days). The D-xylose curves showed that absorption in the transplanted animals at day 60 was similar to that in the untreated controls (P > .05). The absence of differences was confirmed by the statistical analysis. In conclusion, our preclinical study in pigs showed that chronic treatment with FK778 in combination with tacrolimus did not impair carbohydrate absorption by the allograft after SBTx.

FK778 and tacrolimus combination therapy to control acute rejection after pig intestinal transplantation.

Malononitrilamide 715 (FK778) is a new class of low molecular weight immunosuppressant. Experimental studies in heart, liver, and kidney transplantation have shown a strong synergism when FK778 is used in combination with tacrolimus and when its administration is delayed by 7 days after the transplant. Following this indication, in a swine model of orthotopic small bowel transplantation (SBT), we assessed the efficacy of combined low dose tacrolimus and FK778 administered from day 0 or day 7. The entire small bowel was replaced in 16 piglets: group 1 (n = 5), no immunosuppression; group 2 (n = 6) oral tacrolimus to maintain whole blood trough levels between 5 and 15 ng/mL plus FK778 4 mg/kg per day; group 3 (n = 6) oral tacrolimus as in group 2 plus FK778 4 mg/kg per day administered after a 7-day delay posttransplant. The median survival was 8 days in group 1, 60 days in group 2, and 13 days in group 3. The differences between group 2 and 1 and between group 2 and 3 are statistically significant. Three episodes of major bacterial infection were detected in both group 2 and 3 (0.5 episode/animal). The infectious-related mortality was 0% in group 2 and 50% in group 3 (P < .05). Acute cellular rejection was absent or mild in all group 2 and 3 stomal biopsies. In conclusion, combining tacrolimus and FK778 allowed prolonged survival after SBT in swine when FK778 was started at the time of SBT. The delayed administration of FK778 resulted in a high incidence of lethal infectious complications.