Leveraging Predictive Pharmacometrics-Based Algorithms to Enhance Perinatal Care-Application to Neonatal Jaundice.

The field of medicine is undergoing a fundamental change, transforming towards a modern data-driven patient-oriented approach. This paradigm shift also affects perinatal medicine as predictive algorithms and artificial intelligence are applied to enhance and individualize maternal, neonatal and perinatal care. Here, we introduce a pharmacometrics-based mathematical-statistical computer program (PMX-based algorithm) focusing on hyperbilirubinemia, a medical condition affecting half of all newborns. Independent datasets from two different centers consisting of total serum bilirubin measurements were utilized for model development (342 neonates, 1,478 bilirubin measurements) and validation (1,101 neonates, 3,081 bilirubin measurements), respectively. The mathematical-statistical structure of the PMX-based algorithm is a differential equation in the context of non-linear mixed effects modeling, together with Empirical Bayesian Estimation to predict bilirubin kinetics for a new patient. Several clinically relevant prediction scenarios were validated, i.e., prediction up to 24 h based on one bilirubin measurement, and prediction up to 48 h based on two bilirubin measurements. The PMX-based algorithm can be applied in two different clinical scenarios. First, bilirubin kinetics can be predicted up to 24 h based on one single bilirubin measurement with a median relative (absolute) prediction difference of 8.5% (median absolute prediction difference 17.4 µmol/l), and sensitivity and specificity of 95.7 and 96.3%, respectively. Second, bilirubin kinetics can be predicted up to 48 h based on two bilirubin measurements with a median relative (absolute) prediction difference of 9.2% (median absolute prediction difference 21.5 µmol/l), and sensitivity and specificity of 93.0 and 92.1%, respectively. In contrast to currently available nomogram-based static bilirubin stratification, the PMX-based algorithm presented here is a dynamic approach predicting individual bilirubin kinetics up to 48 h, an intelligent, predictive algorithm that can be incorporated in a clinical decision support tool. Such clinical decision support tools have the potential to benefit perinatal medicine facilitating personalized care of mothers and their born and unborn infants.

Serum neurofilament light chain (sNfL) values in a large cross-sectional population of children with asymptomatic to moderate COVID-19.

BACKGROUND: Serum neurofilament light chain (sNfL) is an established biomarker of neuro-axonal damage in multiple neurological disorders. Raised sNfL levels have been reported in adults infected with pandemic coronavirus disease 2019 (COVID-19). Levels in children infected with COVID-19 have not as yet been reported. OBJECTIVE: To evaluate whether sNfL is elevated in children contracting COVID-19. METHODS: Between May 22 and July 22, 2020, a network of outpatient pediatricians in Bavaria, Germany, the Coronavirus antibody screening in children from Bavaria study network (CoKiBa), recruited healthy children into a cross-sectional study from two sources: an ongoing prevention program for 1-14 years, and referrals of 1-17 years consulting a pediatrician for possible infection with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). We determined sNfL levels by single molecule array immunoassay and SARS-CoV-2 antibody status by two independent quantitative methods. RESULTS: Of the 2652 included children, 148 (5.6%) were SARS-CoV-2 antibody positive with asymptomatic to moderate COVID-19 infection. Neurological symptoms-headache, dizziness, muscle aches, or loss of smell and taste-were present in 47/148 cases (31.8%). Mean sNfL levels were 5.5 pg/ml (SD 2.9) in the total cohort, 5.1 (SD 2.1) pg/ml in the children with SARS-CoV-2 antibodies, and 5.5 (SD 3.0) pg/ml in those without. Multivariate regression analysis revealed age-but neither antibody status, antibody levels, nor clinical severity-as an independent predictor of sNfL. Follow-up of children with pediatric multisystem inflammatory syndrome (n = 14) showed no association with sNfL. CONCLUSIONS: In this population study, children with asymptomatic to moderate COVID-19 showed no neurochemical evidence of neuronal damage.

Serum Neurofilament Levels in Children With Febrile Seizures and in Controls.

OBJECTIVE: Neuroaxonal damage is reflected by serum neurofilament light chain (sNfL) values in a variety of acute and degenerative diseases of the brain. The aim of this study was to investigate the impact of febrile and epileptic seizures on sNfL, serum copeptin, and prolactin levels in children compared with children with febrile infections without convulsions. METHODS: A prospective cross-sectional study was performed in children aging 6 months to 5 years presenting with fever (controls, n = 61), febrile seizures (FS, n = 78), or epileptic seizures (ES, n = 16) at our emergency department. sNfL, copeptin, and prolactin were measured within a few hours after the event in addition to standard clinical, neurophysiological, and laboratory assessment. All children were followed up for at least 1 year after presentation concerning recurrent seizures. RESULTS: Serum copeptin values were on average 4.1-fold higher in FS and 3.2-fold higher in ES compared with controls (both p < 0.01). Serum prolactin values were on average 1.3-fold higher in FS compared with controls ( p < 0.01) and without difference between ES and controls. There was no significant difference of mean sNfL values (95% CI) between all three groups, FS 21.7 pg/ml (19.6-23.9), ES 17.7 pg/ml (13.8-21.6), and controls 23.4 pg/ml (19.2-27.4). In multivariable analysis, age was the most important predictor of sNfL, followed by sex and C reactive protein. Neither the duration of seizures nor the time elapsed from seizure onset to blood sampling had an impact on sNfL. None of the three biomarkers were related to recurrent seizures. SIGNIFICANCE: Serum neurofilament light is not elevated during short recovery time after FS when compared with children presenting febrile infections without seizures. We demonstrate an age-dependent decrease of sNfL from early childhood until school age. In contrast to sNfL levels, copeptin and prolactin serum levels are elevated after FS.

Enhanced early prediction of clinically relevant neonatal hyperbilirubinemia with machine learning.

BACKGROUND: Machine learning models may enhance the early detection of clinically relevant hyperbilirubinemia based on patient information available in every hospital. METHODS: We conducted a longitudinal study on preterm and term born neonates with serial measurements of total serum bilirubin in the first two weeks of life. An ensemble, that combines a logistic regression with a random forest classifier, was trained to discriminate between the two classes phototherapy treatment vs. no treatment. RESULTS: Of 362 neonates included in this study, 98 had a phototherapy treatment, which our model was able to predict up to 48 h in advance with an area under the ROC-curve of 95.20%. From a set of 44 variables, including potential laboratory and clinical confounders, a subset of just four (bilirubin, weight, gestational age, hours since birth) suffices for a strong predictive performance. The resulting early phototherapy prediction tool (EPPT) is provided as an open web application. CONCLUSION: Early detection of clinically relevant hyperbilirubinemia can be enhanced by the application of machine learning. Existing guidelines can be further improved to optimize timing of bilirubin measurements to avoid toxic hyperbilirubinemia in high-risk patients while minimizing unneeded measurements in neonates who are at low risk.

Birth experience in newborn infants is associated with changes in nociceptive sensitivity.

Vaginal birth prepares the fetus for postnatal life. It confers respiratory, cardiovascular and homeostatic advantages to the newborn infant compared with elective cesarean section, and is reported to provide neonatal analgesia. We hypothesize that infants born by vaginal delivery will show lower noxious-evoked brain activity a few hours after birth compared to those born by elective cesarean section. In the first few hours of neonatal life, we record electrophysiological measures of noxious-evoked brain activity following the application of a mildly noxious experimental stimulus in 41 infants born by either vaginal delivery or by elective cesarean section. We demonstrate that noxious-evoked brain activity is related to the mode of delivery and significantly lower in infants born by vaginal delivery compared with those born by elective cesarean section. Furthermore, we found that the magnitude of noxious-evoked brain activity is inversely correlated with fetal copeptin production, a surrogate marker of vasopressin, and dependent on the experience of birth-related distress. This suggests that nociceptive sensitivity in the first few hours of postnatal life is influenced by birth experience and endogenous hormonal production.

Personalized weight change prediction in the first week of life.

BACKGROUND & AIMS: Almost all neonates show physiological weight loss and consecutive weight gain after birth. The resulting weight change profiles are highly variable as they depend on multiple neonatal and maternal factors. This limits the value of weight nomograms for the early identification of neonates at risk for excessive weight loss and related morbidities. The objective of this study was to characterize weight changes and the effect of supplemental feeding in late preterm and term neonates during the first week of life, to identify and quantify neonatal and maternal influencing factors, and to provide an educational online prediction tool. METHODS: Longitudinal weight data from 3638 healthy term and late preterm neonates were prospectively recorded up to 7 days of life. Two-thirds (n = 2425) were randomized to develop a semi-mechanistic model characterizing weight change as a balance between time-dependent rates of weight gain and weight loss. The dose-dependent effect of supplemental feeding on weight gain was characterized. A population analysis applying nonlinear mixed-effects modeling was performed using NONMEM 7.3. The model was evaluated on the remaining third of neonates (n = 1213). RESULTS: Key population characteristics (median [range]) of the whole sample were gestational age 39.9 [34.4-42.4] weeks, birth weight 3400 [1980-5580] g, maternal age 32 [15-51] years, cesarean section 26%, and girls 50%. The model demonstrated good predictive performance (bias 0.01%, precision 0.56%), and is able to accurately predict individual weight change (bias 0.15%, precision 1.43%) and the dose-dependent effects of supplemental feeding up to 1 week after birth based on weight measurements during the first 3 days of life, including birth weight, and the following characteristics: gestational age, gender, delivery mode, type of feeding, maternal age, and parity. CONCLUSIONS: We present the first mathematical model not only to describe weight change in term and late preterm neonates but also to provide an educational online tool for personalized weight prediction in the first week of life.

Electrophysiological Measurement of Noxious-evoked Brain Activity in Neonates Using a Flat-tip Probe Coupled to Electroencephalography.

Pain is an unpleasant sensory and emotional experience. In non-verbal patients, it is very difficult to measure pain, even with pain assessment tools. Those tools are subjective or determine secondary physiological indicators which also have certain limitations particularly when exploring the effectiveness of analgesia. As cortical processing is essential for pain perception, brain activity measures may provide a useful approach to assess pain in infants. Here we present a method to assess nociception with electrophysiological brain activity recordings optimized for the use in newborn infants. To produce highly standardized and reproducible noxious stimuli we applied mechanical stimulation with a flat-tip probe, e.g., PinPrick, which is not skin-breaking and does not cause behavioral distress. The noxious-evoked potential allows the objective measurement of nociception in non-verbal patients. This method can be used in newborn infants as early as 34 weeks of gestational age. Moreover, it could be applied in different situations such as measuring the efficacy of analgesic or anesthetic drugs.

Characterizing and Forecasting Individual Weight Changes in Term Neonates.

OBJECTIVES: To develop a mathematical, semimechanistic model characterizing physiological weight changes in term neonates, identify and quantify key maternal and neonatal factors influencing weight changes, and provide an online tool to forecast individual weight changes during the first week of life. STUDY DESIGN: Longitudinal weight data from 1335 healthy term neonates exclusively breastfed up to 1 week of life were available. A semimechanistic model was developed to characterize weight changes applying nonlinear mixed-effects modeling. Covariate testing was performed by applying a standard stepwise forward selection-backward deletion approach. The developed model was externally evaluated on 300 additional neonates collected in the same center. RESULTS: Weight changes during first week of life were described as a function of a changing net balance between time-dependent rates of weight gain and weight loss. Males had higher birth weights (WT0) than females. Gestational age had a positive effect on WT0 and weight gain rate, whereas mother's age had a positive effect on WT0 and a negative effect on weight gain rate. The developed model showed good predictive performance when externally validated (bias = 0.011%, precision = 0.52%) and was able to accurately forecast individual weight changes up to 1 week with only 3 initial weight measurements (bias = -0.74%, precision = 1.54%). CONCLUSIONS: This semimechanistic model characterizes weight changes in healthy breastfed neonates during first week of life. We provide a user-friendly online tool allowing caregivers to forecast and monitor individual weight changes. We plan to validate this model with data from other centers and expand it with data from preterm neonates.