Predicting CPAP failure after less invasive surfactant administration (LISA) in preterm infants by machine learning model on vital parameter data: a pilot study

Objective. Less invasive surfactant administration (LISA) has been introduced to preterm infants with respiratory distress syndrome on continuous positive airway pressure (CPAP) support in order to avoid intubation and mechanical ventilation. However, after this LISA procedure, a significant part of infants fails CPAP treatment (CPAP-F) and requires intubation in the first 72 h of life, which is associated with worse complication free survival chances. The aim of this study was to predict CPAP-F after LISA, based on machine learning (ML) analysis of high resolution vital parameter monitoring data surrounding the LISA procedure.Approach. Patients with a gestational age (GA) <32 weeks receiving LISA were included. Vital parameter data was obtained from a data warehouse. Physiological features (HR, RR, peripheral oxygen saturation (SpO2) and body temperature) were calculated in eight 0.5 h windows throughout a period 1.5 h before to 2.5 h after LISA. First, physiological data was analyzed to investigate differences between the CPAP-F and CPAP-Success (CPAP-S) groups. Next, the performance of two types of ML models (logistic regression: LR, support vector machine: SVM) for the prediction of CPAP-F were evaluated.Main results. Of 51 included patients, 18 (35%) had CPAP-F. Univariate analysis showed lower SpO2, temperature and heart rate variability (HRV) before and after the LISA procedure. The best performing ML model showed an area under the curve of 0.90 and 0.93 for LR and SVM respectively in the 0.5 h window directly after LISA, with GA, HRV, respiration rate and SpO2as most important features. Excluding GA decreased performance in both models.Significance. In this pilot study we were able to predict CPAP-F with a ML model of patient monitor signals, with best performance in the first 0.5 h after LISA. Using ML to predict CPAP-F based on vital signals gains insight in (possibly modifiable) factors that are associated with LISA failure and can help to guide personalized clinical decisions in early respiratory management.

Prognostic model for predicting survival in very preterm infants: an external validation study

OBJECTIVE

To perform a temporal and geographical validation of a prognostic model, considered of highest methodological quality in a recently published systematic review, for predicting survival in very preterm infants admitted to the neonatal intensive care unit. The original model was developed in the UK and included gestational age, birthweight and gender.

DESIGN

External validation study in a population-based cohort.

SETTING

Dutch neonatal wards.

POPULATION OR SAMPLE

All admitted white, singleton infants born between 23⁺⁰ and 32⁺⁶  weeks of gestation between 1 January 2015 and 31 December 2019. Additionally, the model's performance was assessed in four populations of admitted infants born between 24⁺⁰ and 31⁺⁶  weeks of gestation: white singletons, non-white singletons, all singletons and all multiples.

METHODS

The original model was applied in all five validation sets. Model performance was assessed in terms of calibration and discrimination and, if indicated, it was updated.

MAIN OUTCOME MEASURES

Calibration (calibration-in-the-large and calibration slope) and discrimination (c statistic).

RESULTS

Out of 6092 infants, 5659 (92.9%) survived. The model showed good external validity as indicated by good discrimination (c statistic 0.82, 95% CI 0.79-0.84) and calibration (calibration-in-the-large 0.003, calibration slope 0.92, 95% CI 0.84-1.00). The model also showed good external validity in the other singleton populations, but required a small intercept update in the multiples population.

CONCLUSIONS

A high-quality prognostic model predicting survival in very preterm infants had good external validity in an independent, nationwide cohort. The accurate performance of the model indicates that after impact assessment, implementation of the model in clinical practice in the neonatal intensive care unit could be considered.

TWEETABLE ABSTRACT

A high-quality model predicting survival in very preterm infants is externally valid in an independent cohort.

DOSE EVALUATION FOR DIGITAL X-RAY IMAGING OF PREMATURE NEONATES

X-ray radiography is a commonly used diagnostic method for premature neonates. However, because of higher radiosensitivity and young age, premature neonates are more sensitive to the detrimental effects of ionising radiation. Therefore, it is important to monitor and optimise radiation doses at the neonatal intensive care unit (NICU). The number of x-ray examinations, dose-area product (DAP) and effective doses are evaluated for three Dutch NICUs using digital flat panel detectors. Thorax, thorax-abdomen and abdomen protocols are included in this study. Median number of examinations is equal to 1 for all three hospitals. Median DAP ranges between 0.05 and 1.02 μGy m2 for different examination types and different weight categories. These examinations result in mean effective doses between 4 ± 4 and 30 ± 10 μSv per examination. Substantial differences in protocols and doses can be observed between hospitals. This emphasises the need for up-to-date reference levels formulated specifically for premature neonates.

Pilot study demonstrates that salivary oxytocin can be measured unobtrusively in preterm infants

Aim

This study assessed the feasibility and obtrusiveness of measuring salivary oxytocin in preterm infants receiving Kangaroo care, because this is a period of maximal bonding or co‐regulation. We also analysed possible influential determinants, including maternal oxytocin.

Methods

The saliva of preterm infants and their mothers was collected prior to, and during, Kangaroo care using cotton swabs and pooled into vials until sufficient volumes were obtained to measure oxytocin levels using a radioimmunoassay. The obtrusiveness of the infants’ collections was measured with a Likert scale.

Results

Saliva was collected unobtrusively prior to, and during, 30 Kangaroo care sessions in 21 preterm infants. This resulted in three vials with sufficient volumes of before‐Kangaroo care saliva and three with during‐Kangaroo care saliva. Oxytocin was detectable in all six vials. The Kangaroo care duration and the intensity of the mother–infant interaction before and during Kangaroo care seemed to be the most important determinants, and these should preferably be standardised in any future trials.

Conclusion

Oxytocin was measured unobtrusively in the pooled saliva of preterm infants both before and during Kangaroo care and could therefore be investigated as a biomarker in future studies.

[Two-year follow-up of infants born at 24 weeks gestation; first outcomes following implementation of the new 'Guideline for perinatal policy in cases of extreme prematurity']

OBJECTIVE

Since 2010 the guideline 'Guideline for perinatal policy in cases of extreme prematurity' has advised an active policy in infants born at 24 weeks gestation. We investigated how infants born at 24 and 25 weeks gestation in the first year following the implementation of the guideline had developed by the age of 2 years.

DESIGN

Retrospective national cohort study.

METHOD

The study population consisted of all surviving infants born in the Netherlands at 24 or 25 weeks gestation in the period from 1 October 2010 to 1 October 2011. At a corrected age of 2 years the children underwent a general physical and neurological examination, and their cognitive scores were determined on the 'Bayley scales of infant and toddler development' (Bayley III). Examinations took place in the 10 neonatal intensive care units (NICU's) in the Netherlands.

RESULTS

Of 185 extremely premature infants, 166 were admitted to a NICU. A total of 95 survived to a corrected age of 2 years; 78 (82%) children were examined. Their average cognitive score on the Bayley III scale was 88 (SD: 16). Among the children born at 24 weeks gestation, 20% had mild disabilities and 20% had moderate to severe disabilities. Among the children born at 25 weeks gestation, 17% had mild disabilities and 12% had moderate to severe disabilities.

CONCLUSION

Of the children born at 24 weeks gestation in the first year after the introduction of active policy in the Netherlands and surviving to 2 years of age (46%), more than half had developed without disabilities. This was comparable to children born at 25 weeks gestation. Of all children born at 24 weeks gestation, 25% survived to 2 years of age without disabilities.

Safe patient monitoring is challenging but still feasible in a neonatal intensive care unit with single family rooms

AIM

Patient monitoring generates a large number of alarms. The aim of this study was to evaluate the rate, type and management of alarms and to determine the risks of a distributed alarm system in a neonatal intensive care unit (NICU) with single family rooms.

METHODS

A risk analysis was performed before implementation. Alarms from patient monitors recorded for a year were identified, classified and counted. The first alarms, which went to the nurse responsible for the patient, were distinguished from the repeat alarms that were generated if the nurse failed to respond within 45 sec.

RESULTS

The alarm handling protocol was changed as staff felt they needed a greater overview of the NICU alarms to avoid risks. In 1 year, 222 751 critical alarms including 12 309 repeat alarms were generated by patient monitors, equivalent to two alarms per patient per hour. Most of the alarms were oxygen desaturation alarms, followed by bradycardia alarms. About 3% of the desaturation alarms and 0.2% of the bradycardia alarms were repeated.

CONCLUSION

Safe patient monitoring was challenging in a NICU with single family rooms, but possible by employing a distributed alarm system. The low number of repeat alarms indicated quick response times.

Multiple intravenous doses of paracetamol result in a predictable pharmacokinetic profile in very preterm infants

AIM

The therapeutic options available to treat neonatal pain are limited, and one alternative for nonopioid systemic treatment is paracetamol. However, pharmacokinetic data from prolonged administration of intravenous paracetamol in neonates are limited. The aim of this study was to present pharmacokinetics after multiple dose of intravenous paracetamol in very preterm infants of <32 weeks' gestation.

METHODS

Fifteen very preterm infants received five, six-hourly doses of intravenous paracetamol (7.5 mg/kg). Blood samples were taken to measure paracetamol, glutathione and hepatic function, together with urine samples for paracetamol metabolites.

RESULTS

A two-compartment pharmacokinetic model gave the best fit for all individual patients and resulted in a predictable pharmacokinetic profile. The estimated pharmacokinetic population parameters were volume of distribution 0.764 ± 0.225 L/kg, elimination rate constant (ke ) 0.117 ± 0.091/h and intercompartment rate constants k12 0.607 ± 0.734/h and k21 1.105 ± 0.762/h.

CONCLUSION

Our study found that multiple doses of intravenous paracetamol resulted in a predictable pharmacokinetic profile in very preterm infants. Increases in postmenstrual age and weight were associated with increased clearance. No evidence of hepatotoxicity was found.

QRS classification and spatial combination for robust heart rate detection in low-quality fetal ECG recordings

Non-invasive fetal electrocardiography (ECG) can be used for prolonged monitoring of the fetal heart rate (FHR). However, the signal-to-noise-ratio (SNR) of non-invasive ECG recordings is often insufficient for reliable detection of the FHR. To overcome this problem, source separation techniques can be used to enhance the fetal ECG. This study uses a physiology-based source separation (PBSS) technique that has already been demonstrated to outperform widely used blind source separation techniques. Despite the relatively good performance of PBSS in enhancing the fetal ECG, PBSS is still susceptible to artifacts. In this study an augmented PBSS technique is developed to reduce the influence of artifacts. The performance of the developed method is compared to PBSS on multi-channel non-invasive fetal ECG recordings. Based on this comparison, the developed method is shown to outperform PBSS for the enhancement of the fetal ECG.

Acquired progressive hypotonia in infancy: consider compressive cervical myelopathy

This case report presents a rare cause of progressive hypotonia due to a congenital bony defect of the atlas in a 2-month-old girl. The patient was initially referred to the paediatric department with feeding problems. Within days after admission she developed progressive hypotonia and showed decreased reflexes of the upper extremities. Magnetic resonance imaging showed compressive myelopathy at the level of vertebra C1. After laminectomy of the C1-vertebra the neurological symptoms resolved.

CONCLUSION

Most anomalies of the vertebras are asymptomatic. Only symptomatic anomalies with compression and neurological symptoms need surgery.

Quantitative analysis of maturational changes in EEG background activity in very preterm infants with a normal neurodevelopment at 1 year of age

BACKGROUND

The electroencephalographic (EEG) background pattern of preterm infants changes with postmenstrual age (PMA) from discontinuous activity to continuous activity. However, changes in discontinuity have been investigated by visual analysis only.

AIM

To investigate the maturational changes in EEG discontinuity in healthy preterm infants using an automated EEG detection algorithm.

STUDY DESIGN

Weekly 4h EEG recordings were performed in preterm infants with a gestational age (GA)<32weeks and normal neurological follow-up at 1year. The channel C3-C4 was analyzed using an algorithm which automatically detects periods of EEG inactivity (interburst intervals). The interburst-burst ratio (IBR, percentage of EEG inactivity during a moving time window of 600s) and mean length of the interburst intervals were calculated. Using the IBR, discontinuous background activity (periods with high IBR) and continuous background activity (periods with low IBR) were automatically detected and their mean length during each recording was calculated. Data were analyzed with regression and multivariate analysis.

RESULTS

79 recordings were performed in 18 infants. All recordings showed a cyclical pattern in EEG discontinuity. With advancing PMA, IBR (R(2)=0.64; p<0.001), interburst interval length (R(2)=0.43; p<0.001) and length of discontinuous activity (R(2)=0.38; p<0.001) decreased, while continuous activity increased (R(2)=0.50; p<0.001). Multivariate analysis showed that all EEG discontinuity parameters were equally influenced by GA and postnatal age.

CONCLUSION

Analyzing EEG background activity in preterm infants is feasible with an automated algorithm and shows maturational changes of several EEG derived parameters. The cyclical pattern in IBR suggests brain organisation in preterm infant.

Quantitative analysis of amplitude-integrated electroencephalogram patterns in stable preterm infants, with normal neurological development at one year

BACKGROUND

The amplitude-integrated EEG (aEEG) is feasible for monitoring cerebral activity in preterm infants. However, quantitative data on normal patterns in these infants are limited.

OBJECTIVE

To study maturational aEEG changes in a cohort of stable preterm infants by automated quantification.

METHODS

In a cohort of stable preterm infants with gestational age (GA) <32 weeks and normal neurological follow-up at 1 year, weekly 4 h EEG recordings were performed. aEEG traces were obtained from channel C(3)-C(4). The upper margin amplitude (UMA), lower margin amplitude (LMA) and bandwidth (BW) were quantitatively calculated using an expert software system. In addition, the relative duration of discontinuous background pattern (discontinuous background defined as activity with LMA <5 microV, expressed as DC-%) was calculated.

RESULTS

79 aEEG recordings (4-6 recordings/infant) were obtained in 18 infants. Analysis of the first week recordings demonstrated a strong positive correlation between GA and LMA, while DC-% decreased significantly. Longitudinally, all infants showed increase of LMA. Multivariate analysis showed that GA and postnatal age (PA) both contributed independently and equally to LMA and DC-%. We found a strong correlation between postmenstrual age (GA + PA) and LMA and DC-%, respectively.

CONCLUSION

To our knowledge, this is the first study where aEEG development was studied by automated quantification of aEEG characteristics in a cohort of stable preterm infants with a normal neurological development at 1 year of age. LMA and DC-% are simple quantitative measures of neurophysiologic development and may be used to evaluate neurodevelopment in infants.

Neonatal airway obstruction caused by rapidly growing nasopharyngeal teratoma

A case report is presented of a rapidly growing congenital nasopharyngeal teratoma (epignathus) in a preterm infant, leading to severe upper airway obstruction. Prenatal diagnosis by ultrasonography did not reveal the condition because the tumour masses were initially small and there was no polyhydramnios. Epignathus is a rare cause of upper airway obstruction of the newborn that can grow rapidly in the neonatal period and should be treated surgically.

CONCLUSION

Epignathus is a rare cause of upper airway obstruction of the newborn that can grow rapidly in the neonatal period.

Systemic air embolism after cardiopulmonary resuscitation in a preterm infant

We report a preterm infant with extensive systemic air embolism after cardiopulmonary resuscitation for cardiac arrest due to an occluding thrombus in the inferior vena cava. After excluding other potential causes (air infusion, necrotizing enterocolitis or pulmonary leakage syndrome), we postulate that the pressure gradient needed for air embolism to occur is related to the resuscitation procedure. An important clue of air embolism was noted on the chest X-ray taken before death showing intracardial air.

CONCLUSION

Systemic air embolism may occur as a very rare complication after cardiopulmonary resuscitation.

Amplitude-integrated electroencephalographic changes in a newborn induced by overdose of morphine and corrected with naloxone

The amplitude-integrated electroencephalogram (aEEG) is a useful tool to assess brain function after perinatal asphyxia in term infants. We report a full-term newborn with moderate perinatal asphyxia, who accidentally received an overdose of morphine (5000 microg/kg). The overdose of morphine resulted in a clear and immediate change of aEEG background activity from a continuous (C) to discontinuous (DC) background pattern. After administration of naloxone, the background activity restored immediately to continuous background pattern. The aEEG was used to monitor the stepwise reduction in continuous naloxone infusion.

CONCLUSION

An overdose of morphine leads to clear and immediate changes in aEEG which restore after naloxone treatment. The aEEG can be used to monitor naloxone infusion.

An algorithm for the automatic detection of seizures in neonatal amplitude-integrated EEG

AIM

To develop and evaluate an algorithm for the automatic screening of electrographic neonatal seizures (ENS) in amplitude-integrated electroencephalography (aEEG) signals.

METHODS

CFM recordings were recorded in asphyxiated (near)term newborns. ENS of at least 60 sec were detected based on their characteristic pattern in the aEEG signal, an increase of its lower boundary. The algorithm was trained using five CFM recordings (training set) annotated by a neurophysiologist, observer1. The evaluation of the algorithm was based on eight different CFM recordings annotated by observer1 (test set observer 1) and an independent neurophysiologist, observer2 (test set observer 2).

RESULTS

The interobserver agreement between observer1 and 2 in interpreting ENS from the CFM recordings was high (G coefficient: 0.82). After dividing the eight CFM recordings into 1-min segments and classification in ENS or non-ENS, the intraclass correlation coefficient showed high correlations of the algorithm with both test sets (respectively, 0.95 and 0.85 with observer1 and 2). The algorithm showed in five recordings a sensitivity > or = 90% and approximately 1 false positive ENS per hour. However, the algorithm showed in three recordings much lower sensitivities: one recording showed ENSs of extremely high amplitude that were incorrectly classified by the algorithm as artefacts and two recordings suffered from low interobserver agreement.

CONCLUSION

This study shows the feasibility of automatic ENS screening based on aEEG signals and may facilitate in the bed-side interpretation of aEEG signals in clinical practice.

Epidemiology of pyridoxine dependent seizures in the Netherlands

BACKGROUND

Pyridoxine dependent epilepsy is a rare cause of seizures in childhood. The diagnosis is made on clinical criteria, that in many cases are never met. Therefore, epidemiological data on pyridoxine dependency are scarce.

AIMS

To study the epidemiology of pyridoxine dependent epilepsy in the Netherlands, and to determine whether the diagnosis is based on the appropriate criteria.

METHODS

Nationwide all departments of paediatrics (n = 113) and of paediatric or neonatal neurology (n = 17) were asked to report cases of pyridoxine dependent seizures. Birth incidences were calculated using national data on live births from 1991 to 2003.

RESULTS

Response was received from 67% of paediatric departments, including all university hospitals and 94% of child neurology departments. Thirteen patients were reported. Four definite (31%), three probable (23%), and four possible cases (31%) were identified. Two cases (15%) did not meet criteria for either of these groups. The birth incidence was 1:396,000 for definite and probable cases and 1:252,000 when possible cases are included.

CONCLUSIONS

Thus far, epidemiological data on pyridoxine dependent seizures were only available from the UK and Ireland. A higher incidence was found in the Netherlands, in accordance with earlier suggestions of a regional difference. The study shows that the diagnosis is often made without performance of a formal trial of withdrawal. The importance of confirming the diagnosis, concerning the consequences as for individual prognosis, the potential side effects of prolonged pyridoxine substitution, and the possibility of treating the mother in case of future pregnancies are emphasised.

Customized spectral band analysis compared with conventional Fourier analysis of heart rate variability in neonates

A customized filtering technique is introduced and compared with fast Fourier transformation (FFT) for analyzing heart rate variability (HRV) in neonates from short-term recordings. FFT is classically the most commonly used spectral technique to investigate cardiovascular fluctuations. FFT requires stability of the physiological signal within a 300 s time window that is usually analyzed in adults. Preterm infants, however, show characteristics of rapidly fluctuating heart rate and blood pressure due to an immature autonomic regulation, resulting in non-stationarity of these signals. Therefore neonatal studies use (half-overlapping or moving) windows of 64 s length within a recording time of 2-5 min. The proposed filtering technique performs a filtering operation in the frequency range of interest before calculating the spectrum, which allows it to perform an analysis of shorter periods of only 42 s. The frequency bands of interest are 0.04-0.15 Hz (low frequency, LF) and 0.4-1.5 Hz (high frequency, HF). Although conventional FFT analysis as well as the proposed alternative technique result in errors in the estimation of LF power, due to spectral leakage from the very low frequencies, FFT analysis is more sensitive to this effect. The response times show comparable behavior for both the techniques. Applying both the methods to heart rate data obtained from a neonate before and after atropine administration (inducing a wide range of HRV), shows a very significant correlation between the two methods in estimating LF and HF power. We conclude that a customized filtering technique might be beneficial for analyzing HRV in neonates because it reduces the necessary time window for signal stability.

Supraventricular escape rhythms during transient episodes of bradycardia in preterm infants

OBJECTIVE

To evaluate the origin of transient episodes of sinus bradycardia, atrial escape rhythm, and atrioventricular nodal escape rhythm in preterm infants.

MATERIAL AND METHODS

The study was observational, and was carried out in a third level neonatal intensive care unit. We studied 19 spontaneously breathing infants born healthy but prior to term, the examinations being carried out between the ages of 3 and 28 days. The mean gestational age was 29.2 +/- 1.9 weeks, and the mean birth weight was 1154 +/- 264 g. Transient episodes of bradycardia were defined as a decline in heart rate equal to or greater than 25% from baseline, lasting for at least 3 successive RR-intervals. To discriminate between different types of escape mechanisms, we used the P wave and the P axis of the electrocardiogram. Sinus bradycardia was diagnosed when the P axis was from +0 to +90 degrees; atrial escape rhythm when it was from +91 to +359 degrees, and atrioventricular nodal escape rhythm when the P wave was absent, hidden, or followed the QRS complex.

RESULTS

The mean P axis was +50 +/- 11 degrees. We observed 60 transient episodes of bradycardia in 11 of the 19 infants. Of these, 34 (57%) were classified as sinus bradycardia, and 13 (22%) as atrial escapes. Atrioventricular nodal escapes developed during 6 episodes (10%), while 7 episodes (11%) could not be classified.

CONCLUSIONS

Atrial excitation as evidenced by the P axis during sinus rhythm is similar in very preterm infants to that seen in children and adults. Escape rhythms with different origins occur during transient episodes of bradycardia in healthy preterm infants. In at least one third, the episodes are due to atrial or atrioventricular nodal escape.

A severe case of myasthenia gravis during pregnancy

Myasthenia gravis is characterised by muscle weakness and fatigability, particularly of the facial and extremity muscles, deteriorating during the day. During pregnancy, myasthenia gravis is rare and the course of illness is unpredictable. The present case illustrates that first diagnosed during pregnancy, recurrent exacerbations can appear and lead to life-threatening situations.

[Congenital complete atrioventricular block in neonatal lupus erythematosus]

Fetal bradycardia of 60 beats per minute was diagnosed in the 22nd week of a normal pregnancy. Echocardiography showed a complete fetal atrioventricular heart block. The fetal heart was without structural abnormalities. Auto-antibodies (Ro/SS-A and antinuclear antibodies) could be demonstrated in both mother and child. The mother, however, showed no signs or symptoms of a connective tissue disease. We describe neonatal lupus erythematosus as an important cause of congenital heart block.

[Effect of age of erythrocyte concentration administered to premature infants: a retrospective study]

Traditionally fresh red blood cells (RBCs, age less than 10 days) are used for neonatal transfusion. Retrospectively we studied the influence of the age (range 2-34 days, divided into age groups 1-7, 8-14, 15-21, 22-28, 29-35 days) of RBCs on hemoglobin, pH, bicarbonate, and potassium after administration of a small amount of packed RBC (10-15 ml/kg). We reviewed the hospital records of 58 preterm infants (mean birth weight 1316 +/- 543 g, mean gestational age 30 6/7 +/- 3 weeks) who received 201 transfusions (mean of 3.5 RBC transfusions per infant; range 1-22). Following transfusion there was a significant increase in hemoglobin in all age groups. No significant change occurred in pH, bicarbonate or potassium. We conclude that RBCs used for neonatal transfusions do not need to be fresh. Multiple donor exposure of neonates can be limited by splitting blood of a single donor into small portions and using them up to 35 days.