Parametric analysis of an integrated cardio-respiratory model in preterm newborns during apnea.

The analysis of the complex interactions involved in the acute physiological response to apnea-bradycardia events in preterm newborns remains a challenging task. This paper presents a novel integrated model of cardio-respiratory interactions, adapted to preterm newborns. A sensitivity analysis, based Morris' screening method, was applied to study the effects of physiological parameters on heart rate and desaturation, during the simulation of a 15-seconds apnea-bradycardia episode. The most sensitive parameters are associated with fundamental, integrative physiological mechanisms involving: (i) respiratory mechanics (intermediate airways and lung compliance), (ii) fraction of inspired oxygen, (iii) metabolic rates (oxygen consumption rate), (iv) heart rate regulation and (v) chemoreflex (gain). Results highlight the relevant influence of physiological variables, involved in preterm apnea-bradycardia events.

Early detection of late-onset neonatal sepsis from noninvasive biosignals using deep learning: A multicenter prospective development and validation study.

BACKGROUND: Neonatal sepsis is responsible for significant morbidity and mortality worldwide. Its accurate and timely diagnosis is hindered by vague symptoms and the urgent necessity for early antibiotic intervention. The gold standard for diagnosing the condition is the identification of a pathogenic organism from normally sterile sites via laboratory testing. However, this method is resource-intensive and cannot be conducted continuously. OBJECTIVE: This study aimed to predict the onset of late-onset sepsis (LOS) with good diagnostic value as early as possible using non-invasive biosignal measurements from neonatal intensive care unit (NICU) monitors. METHODS: In this prospective multicenter study, we developed a multimodal machine learning algorithm based on a convolutional neural network (CNN) structure that uses the power spectral density (PSD) of recorded biosignals to predict the onset of LOS. This approach aimed to discern LOS-related pathogenic spectral signatures without labor-intensive manual artifact removal. RESULTS: The model achieved an area under the receiver operating characteristic score of 0.810 (95 % CI 0.698-0.922) on the validation dataset. With an optimal operating point, LOS detection had 83 % sensitivity and 73 % specificity. The median early detection was 44 h before clinical suspicion. The results highlighted the additive importance of electrocardiogram and respiratory impedance (RESP) signals in improving predictive accuracy. According to a more detailed analysis, the predictive power arose from the morphology of the electrocardiogram's R-wave and sudden changes in the RESP signal. CONCLUSION: Raw biosignals from NICU monitors, in conjunction with PSD transformation, as input to the CNN, can provide state-of-the-art prediction performance for LOS without the need for artifact removal. To the knowledge of the authors, this is the first study to highlight the independent and additive predictive potential of electrocardiogram R-wave morphology and concurrent, sudden changes in the RESP waveform in predicting the onset of LOS using non-invasive biosignals.

Excessive workload and insufficient night-shift remuneration are key elements of dissatisfaction at work for French neonatologists.

AIM: Neonatologists are exposed to ethical issues and unplanned emergencies that require 24-h in-house coverage. These elements may affect quality of life at work, which we surveyed. METHODS: This was a self-administered, voluntary and anonymous cross-sectional survey of French neonatologists. An online questionnaire was sent to members of the French Society of Neonatology from June to October 2022. RESULTS: Of approximately 1500 possible responses, 721 were analysed, with a response rate of 48%. Respondents were mostly women (77%), aged 35-50 years (50%), and hospital practitioners (63%). Reported weekly working time was over 50 h for 80%. Among the 650 neonatologists with on-call duty, 47% worked ≥5 shifts per month. For 80% of practitioners, on-call duty was perceived to have a negative impact on personal life; 49% indicated having sleep disorders. The mean satisfaction score at work was 5.7 ± 1.7 on a scale of 0-10. The main reasons for dissatisfaction were excessive working hours and insufficient remuneration for on-call duty. CONCLUSION: This first evaluation of the quality of life at work of French neonatologists showed high workload. The working conditions and specificities of NICU activity may have significant consequences for their mental health.

Trial of labor versus elective cesarean delivery for patients with two prior cesarean-sections: A retrospective propensity score analysis.

INTRODUCTION: Despite awareness of obstetricians to the constant increase in the number of cesarean sections in recent years, the fear of a uterine scar rupture is still present and influences the choice of the mode of delivery in patients with two previous cesarean sections. However, several clinical studies have suggested that, under certain conditions, vaginal birth after two cesarean sections is usually successful and safe. OBJECTIVE: The objective of this study was to compare maternal and neonatal issues according to the planned mode of delivery in patients with two previous cesarean sections. METHODS: It was a retrospective observational comparative study at Rennes University Hospital between January 1, 2013, and December 31, 2020. We performed a propensity score for the comparison of neonatal outcomes: cord pH, cord lactates, Apgar scores, transfer to neonatal unit and deaths, according to the planned delivery mode. Secondary outcomes were maternal issues: uterine rupture, post-partum hemorrhage, deaths. RESULTS: A total of 410 patients with two previous cesarean section were eligible for our study. Prophylactic cesarean was performed in 358 cases (87.3%). Trial of labor was attempted in the 52 remaining patients (12.7%), 67.3 % of whom were successful. Neonatal weight, APGAR score at 1-5-10 min, and pH on cord blood were comparable in both groups. One case of uterine rupture occurred in the trial of labor group. CONCLUSION: Trial of labor seems to be a reasonable option for women with two previous cesarean sections in a selected population.

Assessment of compliance with a color code protocol for non-elective cesarean section and its impact on time to delivery interval and neonatal outcomes.

INTRODUCTION: We set out to assess the compliance with a cesarean section color code protocol and its impact on maternal and neonatal outcomes since its implementation in our maternity ward. METHODS: This was a retrospective study including a sample of 200 patients per year who underwent a non-elective cesarean section delivery in Rennes University Hospital from January 1, 2015 to December 31, 2018. Patients were grouped by year and by color code (red, orange or green). The main outcome was compliance with the protocol (color code in accordance with indication for cesarean section) and compliance with the corresponding decision-delivery interval. Secondary outcomes were maternal and neonatal outcomes. RESULTS: Eight hundred patients were included during the study period. There was no significant difference in patient characteristics over the years. There was a significant improvement in protocol compliance: full compliance increased from 22.4% in 2015 to 76.5% in 2018 (p < 0.0001). The respect of the 15 min decision-delivery interval in red code protocol increased between 2015 and 2018 (p = 0.0020). CONCLUSION: We observed a significant improvement in compliance with the color code protocol between 2015 and 2018 and in the 15 min decision-delivery deadline for the red code.

Evaluation of Maturation in Preterm Infants Through an Ensemble Machine Learning Algorithm Using Physiological Signals.

This study was designed to test if heart rate variability (HRV) data from preterm and full-term infants could be used to estimate their functional maturational age (FMA), using a machine learning model. We propose that the FMA, and its deviation from the postmenstrual age (PMA) of the infants could inform physicians about the progress of the maturation of the infants. The HRV data was acquired from 50 healthy infants, born between 25 and 41 weeks of gestational age, who did not present any signs of abnormal maturation relative to their age group during the period of observation. The HRV features were used as input for a machine learning model that uses filtering and genetic algorithms for feature selection, and an ensemble machine learning (EML) algorithm, which combines linear and random forest regressions, to produce as output a FMA. Using HRV data, the FMA had a mean absolute error of 0.93 weeks, 95% CI [0.78, 1.08], compared to the PMA. These results demonstrate that HRV features of newborn infants can be used by an EML model to estimate their FMA. This method was also generalized using respiration rate variability (RRV) and bradycardia data, obtaining similar results. The FMA, predicted either by HRV, RRV or bradycardia, and its deviation from the true PMA of the infants, could be used as a surrogate measure of the maturational age of the infants, which could potentially be monitored non-invasively and in real-time in the setting of neonatal intensive care units.

Patent ductus arteriosus, tracheal ventilation, and the risk of bronchopulmonary dysplasia.

BACKGROUND: An increased risk for bronchopulmonary dysplasia (BPD) exists when moderate-to-large patent ductus arteriosus shunts (hsPDA) persist beyond 14 days. GOAL: To examine the interaction between prolonged exposures to tracheal ventilation (≥10 days) and hsPDA on the incidence of BPD in infants <28 weeks gestation. STUDY DESIGN: Predefined definitions of prolonged ventilation (≥10 days), hsPDA (≥14 days), and BPD (room air challenge test at 36 weeks) were used to analyze deidentified data from the multicenter TRIOCAPI RCT in a secondary analysis of the trial. RESULTS: Among 307 infants who survived >14 days, 41 died before 36 weeks. Among survivors, 93/266 had BPD. The association between BPD and hsPDA depended on the length of intubation. In multivariable analyses, prolonged hsPDA shunts were associated with increased BPD (odds ratio (OR) (95% confidence interval (CI)) = 3.00 (1.58-5.71)) when infants required intubation for ≥10 days. In contrast, there was no significant association between hsPDA exposure and BPD when infants were intubated <10 days (OR (95% CI) = 1.49 (0.98-2.26)). A similar relationship between prolonged hsPDA and length of intubation was found for BPD/death (n = 307): infants intubated ≥10 days: OR (95% CI) = 2.41 (1.47-3.95)); infants intubated <10 days: OR (95% CI) = 1.37 (0.86-2.19)). CONCLUSIONS: Moderate-to-large PDAs were associated with increased risks of BPD and BPD/death-but only when infants required intubation ≥10 days. IMPACT: Infants with a moderate-to-large hsPDA that persist beyond 14 days are only at risk for developing BPD if they also receive prolonged tracheal ventilation for ≥10 days. Infants who receive less ventilatory support (intubation for <10 days) have the same incidence of BPD whether the ductus closes shortly after birth or whether it persists as a moderate-to-large shunt for several weeks. Early PDA closure may be unnecessary in infants who require short durations of intubation since the PDA does not seem to alter the incidence of BPD in infants who require intubation for <10 days.

Early arterial pressure monitoring and term-equivalent age MRI findings in very preterm infants.

INTRODUCTION: Variability of arterial blood pressure (ABP) has been associated with intraventricular hemorrhage in very preterm neonates (VPT) and may predict other brain lesions assessed at term-equivalent of age (TEA). METHODS: This was a prospective single-center study including VPT with early invasive continuous ABP monitoring and assessed at TEA using brain magnetic resonance imaging (TEA-MRI). The association between early mean ABP (MABP) and TEA-MRI findings was modeled by multivariate logistic regression analysis using covariates selected by the LASSO method. RESULTS: Among 99 VPT, the LASSO procedure selected consecutive periods of lowest MABP of 30 min on day 1 (d1) and 10 min on day 2 (d2) as the most relevant durations to predict TEA-MRI findings (OR [95% CI], 1.11 [1.02-1.23], p = 0.03 and 1.13 [1.01-1.27], p = 0.03, respectively). ROC curve analysis showed optimal thresholds at 30.25 mmHg on d1 and 33.25 mmHg on d2. This significant association persisted after adjustment with covariates including birthweight, gestational age, sex, and inotrope exposure. Final models selected by LASSO included the decile of the birthweight and lowest MABP for 30 min on d1 and 10 min on d2, for which the areas under the ROC curve were 74% and 75%, respectively. CONCLUSION: Early continuous ABP monitoring may predict brain TEA-MRI findings in VPT. IMPACT: Early arterial blood pressure monitoring may contribute to predicting brain damage upon MRI at term-equivalent of age for infants born very preterm. Careful blood pressure continuous monitoring in very preterm infants may identify infants at risk of long-term brain damage. Umbilical artery catheterization provides the best option for continuously monitoring arterial blood pressure in very preterm infants.

Sensitivity Analysis of a Cardio-respiratory Model in Preterm Newborns for the Study of Patent Ductus Arteriosus.

This paper proposes an integrated model of cardio-respiratory interactions in preterm newborns, focused on the study of the patent ductus arteriosus (PDA). A formal model parameter sensitivity analysis on blood flow through the PDA is performed. Results show that the proposed model is capable of simulating hemodynamics in right-to-left and left-to-right shunts. For both configurations, the most significant parameters are associated with mechanical ventricular properties and circulatory parameters related to left ventricle loading conditions. These results highlight important physiological mechanisms involved in PDA and provide key information towards the definition of patient-specific parameters.

Association Between Early Amino Acid Intake and Full-Scale IQ at Age 5 Years Among Infants Born at Less Than 30 Weeks' Gestation.

Importance: An international expert committee recently revised its recommendations on amino acid intake for very preterm infants, suggesting that more than 3.50 g/kg/d should be administered only to preterm infants in clinical trials. However, the optimal amino acid intake during the first week after birth in these infants is unknown. Objective: To evaluate the association between early amino acid intake and cognitive outcomes at age 5 years. Design, Setting, and Participants: Using the EPIPAGE-2 (Epidemiologic Study on Small-for-Gestational-Age Children-Follow-up at Five and a Half Years) cohort, a nationwide prospective population-based cohort study conducted at 63 neonatal intensive care units in France, a propensity score-matched analysis was performed comparing infants born at less than 30 weeks' gestation who had high amino acid intake (3.51-4.50 g/kg/d) at 7 days after birth with infants who did not. Participants were recruited between April 1 and December 31, 2011, and followed up from September 1, 2016, to December 31, 2017. Full-scale IQ (FSIQ) was assessed at age 5 years. A confirmatory analysis used neonatal intensive care unit preference for high early amino acid intake as an instrumental variable to account for unmeasured confounding. Statistical analysis was performed from January 15 to May 15, 2021. Exposures: Amino acid intake at 7 days after birth. Main Outcomes and Measures: The primary outcome was an FSIQ score greater than -1 SD (ie, ≥93 points) at age 5 years. A complementary analysis was performed to explore the association between amino acid intake at day 7 as a continuous variable and FSIQ score at age 5 years. Data from cerebral magnetic resonance imaging at term were available for a subgroup of preterm infants who participated in the EPIRMEX (Cerebral Abnormalities Detected by MRI, Realized at the Age of Term and the Emergence of Executive Functions) ancillary study. Results: Among 1789 preterm infants (929 boys [51.9%]; mean [SD] gestational age, 27.17 [1.50] weeks) with data available to determine exposure to amino acid intake of 3.51 to 4.50 g/kg/d at 7 days after birth, 938 infants were exposed, and 851 infants were not; 717 infants from each group could be paired. The primary outcome was known in 396 of 646 exposed infants and 379 of 644 nonexposed infants who were alive at age 5 years and was observed more frequently among exposed vs nonexposed infants (243 infants [61.4%] vs 206 infants [54.4%], respectively; odds ratio [OR], 1.33 [95% CI, 1.00-1.71]; absolute risk increase in events [ie, the likelihood of having an FSIQ score >-1 SD at age 5 years] per 100 infants, 7.01 [95% CI, 0.06-13.87]; P = .048). In the matched cohort, correlation was found between amino acid intake per 1.00 g/kg/d at day 7 and FSIQ score at age 5 years (n = 775; ß = 2.43 per 1-point increase in FSIQ; 95% CI, 0.27-4.59; P = .03), white matter area (n = 134; ß = 144 per mm2; 95% CI, 3-285 per mm2; P = .045), anisotropy of the corpus callosum (n = 50; ß = 0.018; 95% CI, 0.016-0.021; P < .001), left superior longitudinal fasciculus (n = 42; ß = 0.018; 95% CI, 0.010-0.025; P < .001), and right superior longitudinal fasciculus (n = 42; ß = 0.014 [95% CI, 0.005-0.024; P = .003) based on magnetic resonance imaging at term. Confirmatory and sensitivity analyses confirmed these results. For example, the adjusted OR for the association between the exposure and the primary outcome was 1.30 (95% CI, 1.16-1.46) using the instrumental variable approach among 978 participants in the overall cohort, and the adjusted OR was 1.35 (95% CI, 1.05-1.75) using multiple imputations among 1290 participants in the matched cohort. Conclusions and Relevance: In this cohort study, high amino acid intake at 7 days after birth was associated with an increased likelihood of an FSIQ score greater than -1 SD at age 5 years. Well-designed randomized studies with long-term follow-up are needed to confirm the benefit of this nutritional approach.

Criteria for assessing the quality of clinical practice guidelines in paediatrics and neonatology: a mixed-method study.

BACKGROUND: Evidenced-based practice is a key component of quality care. This study aims to explore users' expectations concerning paediatric local clinical practice guidelines. METHODS: A mixed method approach was applied, including material from quantitative questionnaire and semi-structured interviews. Data were analysed using descriptive statistics and qualitative content analysis. Data were analysed with constant comparative method. Qualitative data were parsed and categorized to identify themes related to decision-making. RESULTS: A total of 83 physicians answered the survey (response rate 83%). 98% of the participants wanted protocols based on international guidelines, 80% expected a therapeutic content. 24 semi-structured interviews were conducted to understand implementation processes, barriers and facilitators. Qualitative analysis revealed 5 emerging themes: improvement of local clinical practice guidelines, patterns of usage, reasons for non-implementation, alternative sources and perspectives. CONCLUSION: Some criteria should be considered for the redaction of local clinical practice guidelines: focus on therapeutic, ease of access, establish local clinical practice guidelines based on international guidelines adapted to the local setting, document references and include trainees such as residents in the redaction.

New Insights Into Microbiota Modulation-Based Nutritional Interventions for Neurodevelopmental Outcomes in Preterm Infants.

Gut microbiota and the central nervous system have parallel developmental windows during pre and post-natal life. Increasing evidences suggest that intestinal dysbiosis in preterm infants predisposes the neonate to adverse neurological outcomes later in life. Understanding the link between gut microbiota colonization and brain development to tailor therapies aimed at optimizing initial colonization and microbiota development are promising strategies to warrant adequate brain development and enhance neurological outcomes in preterm infants. Breast-feeding has been associated with both adequate cognitive development and healthy microbiota in preterms. Infant formula are industrially produced substitutes for infant nutrition that do not completely recapitulate breast-feeding benefices and could be largely improved by the understanding of the role of breast milk components upon gut microbiota. In this review, we will first discuss the nutritional and bioactive component information on breast milk composition and its contribution to the assembly of the neonatal gut microbiota in preterms. We will then discuss the emerging pathways connecting the gut microbiota and brain development. Finally, we will discuss the promising microbiota modulation-based nutritional interventions (including probiotic and prebiotic supplementation of infant formula and maternal nutrition) for improving neurodevelopmental outcomes.

Frequency of Extracorporeal Membrane Oxygenation Support and Outcomes After Implementation of a Structured PICU Network in Neonates and Children: A Prospective Population-Based Study in the West of France.

OBJECTIVES: To describe the frequency and outcomes on the use of extracorporeal membrane oxygenation (ECMO) among critically ill neonates and children within a structured pediatric critical care network in the West of France. To assess the optimality of decision-making process for patients primarily admitted in non-ECMO centers. DESIGN: Observational prospective population-based study from January 2015 to December 2019. PATIENTS: Neonates over 34 weeks of gestational age, weighing more than 2,000 g and children under 15 years and 3 months old admitted in one of the 10 units belonging to a Regional Pediatric Critical Care Network. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Eight-thousand one hundred eighty-nine children and 3,947 newborns were admitted within one of the 10 units of the network over the study period. Sixty-five children (8.1% [95% CI, 6.2-10‰]) and 35 newborns (9.4% [95% CI, 6.4-12%]) required ECMO support. Of these patients, 31 were first admitted to a non-ECMO center, where 20 were cannulated in situ (outside the regional ECMO center) and 11 after transfer to the ECMO regional center. Cardiogenic shock, highest serum lactate level, and cardiac arrest prior to first phone call with the regional ECMO center were associated with higher rate of in situ cannulation. During the study period, most of the patients were cannulated for underlying cardiac issue (42/100), postoperative cardiac surgery instability (38/100), and pediatric (10/100) and neonatal (10/100) respiratory distress syndrome. Patients primarily admitted in non-ECMO centers or not had similar 28-day post-ICU survival rates compared with those admitted in the referral ECMO center (58% vs 51%; p = 0.332). Pre-ECMO cardiac arrest, ECMO, and lower pH at ECMO onset were associated with lower 28-day post-ICU survival. CONCLUSIONS: Our local results suggest that a structured referral network for neonatal and pediatric ECMO in the region of Western France facilitated escalation of care with noninferior (or similar) early mortality outcome. Our data support establishing referral networks in other equivalent regions.

Early bradycardia detection and therapeutic interventions in preterm infant monitoring.

In very preterm infants, cardio-respiratory events and associated hypoxemia occurring during early postnatal life have been associated with risks of retinopathy, growth alteration and neurodevelopment impairment. These events are commonly detected by continuous cardio-respiratory monitoring in neonatal intensive care units (NICU), through the associated bradycardia. NICU nurse interventions are mainly triggered by these alarms. In this work, we acquired data from 52 preterm infants during NICU monitoring, in order to propose an early bradycardia detector which is based on a decentralized fusion of three detectors. The main objective is to improve automatic detection under real-life conditions without altering performance with respect to that of a monitor commonly used in NICU. We used heart rate lower than 80 bpm during at least 10 sec to define bradycardia. With this definition we observed a high rate of false alarms (64%) in real-life and that 29% of the relevant alarms were not followed by manual interventions. Concerning the proposed detection method, when compared to current monitors, it provided a significant decrease of the detection delay of 2.9 seconds, without alteration of the sensitivity (97.6% vs 95.2%) and false alarm rate (63.7% vs 64.1%). We expect that such an early detection will improve the response of the newborn to the intervention and allow for the development of new automatic therapeutic strategies which could complement manual intervention and decrease the sepsis risk.

Augmenting Critical Care Patient Monitoring Using Wearable Technology: Review of Usability and Human Factors.

BACKGROUND: Continuous monitoring of the vital signs of critical care patients is an essential component of critical care medicine. For this task, clinicians use a patient monitor (PM), which conveys patient vital sign data through a screen and an auditory alarm system. Some limitations with PMs have been identified in the literature, such as the need for visual contact with the PM screen, which could result in reduced focus on the patient in specific scenarios, and the amount of noise generated by the PM alarm system. With the advancement of material science and electronic technology, wearable devices have emerged as a potential solution for these problems. This review presents the findings of several studies that focused on the usability and human factors of wearable devices designed for use in critical care patient monitoring. OBJECTIVE: The aim of this study is to review the current state of the art in wearable devices intended for use by clinicians to monitor vital signs of critical care patients in hospital settings, with a focus on the usability and human factors of the devices. METHODS: A comprehensive literature search of relevant databases was conducted, and 20 studies were identified and critically reviewed by the authors. RESULTS: We identified 3 types of wearable devices: tactile, head-mounted, and smartwatch displays. In most cases, these devices were intended for use by anesthesiologists, but nurses and surgeons were also identified as potentially important users of wearable technology in critical care medicine. Although the studies investigating tactile displays revealed their potential to improve clinical monitoring, usability problems related to comfort need to be overcome before they can be considered suitable for use in clinical practice. Only a few studies investigated the usability and human factors of tactile displays by conducting user testing involving critical care professionals. The studies of head-mounted displays (HMDs) revealed that these devices could be useful in critical care medicine, particularly from an ergonomics point of view. By reducing the amount of time the user spends averting their gaze from the patient to a separate screen, HMDs enable clinicians to improve their patient focus and reduce the potential of repetitive strain injury. CONCLUSIONS: Researchers and designers of new wearable devices for use in critical care medicine should strive to achieve not only enhanced performance but also enhanced user experience for their users, especially in terms of comfort and ease of use. These aspects of wearable displays must be extensively tested with the intended end users in a setting that properly reflects the intended context of use before their adoption can be considered in clinical settings.

Detection of Apnea Bradycardia from ECG Signals of Preterm Infants Using Layered Hidden Markov Model.

Apnea-bradycardia (AB) is a common complication in prematurely born infants, which is associated with reduced survival and neurodevelopmental outcomes. Thus, early detection or predication of AB episodes is critical for initiating preventive interventions. To develop automatic real-time operating systems for early detection of AB, recent advances in signal processing can be employed. Hidden Markov Models (HMM) are probabilistic models with the ability of learning different dynamics of the real time-series such as clinical recordings. In this study, a hierarchy of HMMs named as layered HMM was presented to detect AB episodes from pre-processed single-channel Electrocardiography (ECG). For training the hierarchical structure, RR interval, and width of QRS complex were extracted from ECG as observations. The recordings of 32 premature infants with median 31.2 (29.7, 31.9) weeks of gestation were used for this study. The performance of the proposed layered HMM was evaluated in detecting AB. The best average accuracy of 97.14 ± 0.31% with detection delay of - 5.05 ± 0.41 s was achieved. The results show that layered structure can improve the performance of the detection system in early detecting of AB episodes. Such system can be incorporated for more robust long-term monitoring of preterm infants.

Diagnosis of Neonatal Late-Onset Infection in Very Preterm Infant: Inter-Observer Agreement and International Classifications.

BACKGROUND: The definition of late-onset bacterial sepsis (LOS) in very preterm infants is not unified. The objective was to assess the concordance of LOS diagnosis between experts in neonatal infection and international classifications and to evaluate the potential impact on heart rate variability and rate of "bronchopulmonary dysplasia or death". METHODS: A retrospective (2017-2020) multicenter study including hospitalized infants born before 31 weeks of gestation with intention to treat at least 5-days with antibiotics was performed. LOS was classified as "certain or probable" or "doubtful" independently by five experts and according to four international classifications with concordance assessed by Fleiss's kappa test. RESULTS: LOS was suspected at seven days (IQR: 5-11) of life in 48 infants. Following expert classification, 36 of them (75%) were considered as "certain or probable" (kappa = 0.41). Following international classification, this number varied from 13 to 46 (kappa = -0.08). Using the expert classification, "bronchopulmonary dysplasia or death" occurred less frequently in the doubtful group (25% vs. 78%, p < 0.001). Differences existed in HRV changes between the two groups. CONCLUSION: The definition of LOS is not consensual with a low international and moderate inter-observer agreement. This affects the evaluation of associated organ dysfunction and prognosis.

Early Detection of Late Onset Sepsis in Premature Infants Using Visibility Graph Analysis of Heart Rate Variability.

OBJECTIVE: This study was designed to test the diagnostic value of visibility graph features derived from the heart rate time series to predict late onset sepsis (LOS) in preterm infants using machine learning. METHODS: The heart rate variability (HRV) data was acquired from 49 premature newborns hospitalized in neonatal intensive care units (NICU). The LOS group consisted of patients who received more than five days of antibiotics, at least 72 hours after birth. The control group consisted of infants who did not receive antibiotics. HRV features in the days prior to the start of antibiotics (LOS group) or in a randomly selected period (control group) were compared against a baseline value calculated during a calibration period. After automatic feature selection, four machine learning algorithms were trained. All the tests were done using two variants of the feature set: one only included traditional HRV features, and the other additionally included visibility graph features. Performance was studied using area under the receiver operating characteristics curve (AUROC). RESULTS: The best performance for detecting LOS was obtained with logistic regression, using the feature set including visibility graph features, with AUROC of 87.7% during the six hours preceding the start of antibiotics, and with predictive potential (AUROC above 70%) as early as 42 h before start of antibiotics. CONCLUSION: These results demonstrate the usefulness of introducing visibility graph indexes in HRV analysis for sepsis prediction in newborns. SIGNIFICANCE: The method proposed the possibility of non-invasive, real-time monitoring of risk of LOS in a NICU setting.

Apnea bradycardia detection based on new coupled hidden semi Markov model.

In this paper, a method for apnea bradycardia detection in preterm infants is presented based on coupled hidden semi Markov model (CHSMM). CHSMM is a generalization of hidden Markov models (HMM) used for modeling mutual interactions among different observations of a stochastic process through using finite number of hidden states with corresponding resting time. We introduce a new set of equations for CHSMM to be integrated in a detection algorithm. The detection algorithm was evaluated on a simulated data to detect a specific dynamic and on a clinical dataset of electrocardiogram signals collected from preterm infants for early detection of apnea bradycardia episodes. For simulated data, the proposed algorithm was able to detect the desired dynamic with sensitivity of 96.67% and specificity of 98.98%. Furthermore, the method detected the apnea bradycardia episodes with 94.87% sensitivity and 96.52% specificity with mean time delay of 0.73 s. The results show that the algorithm based on CHSMM is a robust tool for monitoring of preterm infants in detecting apnea bradycardia episodes. Graphical Abstract Apnea Bradycardia detection using Coupled hidden semi Markov Model from electrocardiography. In this model, a sequence of hidden states is assigned to each observation based on the effects of previous states of all observations.