neonatal care thesis

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• My Account Login
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Review Article
• Open access
• Published: 27 November 2023

The past, current, and future of neonatal intensive care units with artificial intelligence: a systematic review

• Elif Keles   ORCID: orcid.org/0000-0001-8103-797X 1 &
• Ulas Bagci   ORCID: orcid.org/0000-0001-7379-6829 1 , 2 , 3  

npj Digital Medicine volume  6 , Article number:  220 ( 2023 ) Cite this article

5489 Accesses

8 Altmetric

Metrics details

• Paediatric research
• Translational research

Machine learning and deep learning are two subsets of artificial intelligence that involve teaching computers to learn and make decisions from any sort of data. Most recent developments in artificial intelligence are coming from deep learning, which has proven revolutionary in almost all fields, from computer vision to health sciences. The effects of deep learning in medicine have changed the conventional ways of clinical application significantly. Although some sub-fields of medicine, such as pediatrics, have been relatively slow in receiving the critical benefits of deep learning, related research in pediatrics has started to accumulate to a significant level, too. Hence, in this paper, we review recently developed machine learning and deep learning-based solutions for neonatology applications. We systematically evaluate the roles of both classical machine learning and deep learning in neonatology applications, define the methodologies, including algorithmic developments, and describe the remaining challenges in the assessment of neonatal diseases by using PRISMA 2020 guidelines. To date, the primary areas of focus in neonatology regarding AI applications have included survival analysis, neuroimaging, analysis of vital parameters and biosignals, and retinopathy of prematurity diagnosis. We have categorically summarized 106 research articles from 1996 to 2022 and discussed their pros and cons, respectively. In this systematic review, we aimed to further enhance the comprehensiveness of the study. We also discuss possible directions for new AI models and the future of neonatology with the rising power of AI, suggesting roadmaps for the integration of AI into neonatal intensive care units.

Similar content being viewed by others

AI in health and medicine

Pranav Rajpurkar, Emma Chen, … Eric J. Topol

The next generation of evidence-based medicine

Vivek Subbiah

A deep learning algorithm to predict risk of pancreatic cancer from disease trajectories

Davide Placido, Bo Yuan, … Chris Sander

Introduction

The AI tsunami fueled by advances in artificial intelligence (AI) is constantly changing almost all fields, including healthcare; it is challenging to track the changes originated by AI as there is not a single day that AI is not applied to anything new. While AI affects daily life enormously, many clinicians may not be aware of how much of the work done with AI technologies may be put into effect in today’s healthcare system. In this review, we fill this gap, particularly for physicians in a relatively underexplored area of AI: neonatology. The origins of AI, specifically machine learning (ML), can be tracked all the way back to the 1950s, when Alan Turing invented the so-called “learning machine” as well as military applications of basic AI 1 . During his time, computers were huge, and the cost of increased storage space was astronomical. As a result, their capabilities, although substantial for their day, were restricted. Over the decades, incremental advancements in theory and technological advances steadily increased the power and versatility of ML 2 .

How do machine learning (ML) and deep learning (DL) work? ML falls under the category of AI 2 . ML’s capacity to deal with data brought it to the attention of computer scientists. ML algorithms and models can learn from data, analyze, evaluate, and make predictions or decisions based on learning and data characteristics. DL is a subset of ML. Different from this larger class of ML definitions, the underlying concept of DL is inspired by the functioning of the human brain, particularly the neural networks responsible for processing and interpreting information. DL mimics this operation by utilizing artificial neurons in a computer neural network. In simple terms, DL finds weights for each artificial neuron that connects to each other from one layer to another layer. Once the number of layers is high (i.e., deep), more complex relationships between input and output can be modeled 3 , 4 , 5 . This enables the network to acquire more intricate representations of the data as it learns. The utilization of a hierarchical approach enables DL models to autonomously extract features from the data, as opposed to depending on human-engineered features as is customary in conventional ML 3 . DL is a highly specialized form of ML that is ideally modified for tasks involving unstructured data, where the features in the data may be learnable, and exploration of non-linear associations in the data can be possible 6 , 7 , 8 .

The main difference between ML and DL lies in the complexity of the models and the size of the datasets they can handle. ML algorithms can be effective for a wide range of tasks and can be relatively simple to train and deploy 6 , 7 , 9 , 10 , 11 . DL algorithms, on the other hand, require much larger datasets and more complex models but can achieve exceptional performance on tasks that involve high-dimensional, complex data 7 . DL can automatically identify which aspects are significant, unlike classical ML, which requires pre-defined elements of interest to analyze the data and infer a decision 10 . Each neuron in DL architectures (i.e., artificial neural networks (ANN)) has non-linear activation function(s) that help it learn complex features representative of the provided data samples 9 .

ML algorithms, hence, DL, can be categorized as either supervised, unsupervised, or reinforcement learning based on the input-output relationship. For example, if output labels (outcome) are fully available, the algorithm is called “supervised,” while unsupervised algorithms explore the data without their reference standards/outcomes/labels in the output 3 , 12 . In terms of applications, both DL and ML are typically used for tasks such as classification, regression, and clustering 6 , 9 , 10 , 13 , 14 , 15 . DL methods’ success depends on the availability of large-scale data, new optimization algorithms, and the availability of GPUs 6 , 10 . These algorithms are designed to autonomously learn and develop as they gain experience, like humans 3 . As a result of DL’s powerful representation of the data, it is considered today’s most improved ML method, providing drastic changes in all fields of medicine and technology, and it is the driving force behind virtually all progress in AI today 5 (Fig. 1 ).

a Hierarchical diagram of AI. How do machine learning (ML) and deep learning (DL) work? ML falls under the category of AI. DL is a subset of ML. b Ongoing hurdles of AI when applied to healthcare applications. Key concerns related to AI and each concern affects the outcome of AI in Neonatology including; (1) challenges with clinical interpretability; (2) knowledge gaps in decision-making mechanisms, with the latter requiring human-in-the-loop systems (3) ethical considerations; (4) the lack of data and annotations, and (5) the absence of Cloud systems allowing for secure data sharing and data privacy.

There are three major problem types in DL in medical imaging: image segmentation, object detection (i.e., an object can be an organ or any other anatomical or pathological entity), and image classification (e.g., diagnosis, prognosis, therapy response assessment) 3 . Several DL algorithms are frequently employed in medical research; briefly, those approaches belong to the following family of algorithms:

Convolutional Neural Networks (CNNs) are predominantly employed for tasks related to computer vision and signal processing. CNNs can handle tasks requiring spatial relationships where the columns and rows are fixed, such as imaging data. CNN architecture encompasses a sequence of phases (layers) that facilitate the acquisition of hierarchical features. Initial phases (layers) extract more local features such as corners, edges, and lines, later phases (layers) extract more global features. Features are propagated from one layer to another layer, and feature representation becomes richer this way. During feature propagation from one layer to another layer, the features are added certain nonlinearities and regularizations to make the functional modeling of input-output more generalizable. Once features become extremely large, there are operations within the network architecture to reduce the feature size without losing much information, called pooling operations. The auto-generated and propagated features are then utilized at the end of the network architecture for prediction purposes (segmentation, detection, or classification) 3 , 16 .

Recurrent Neural Networks (RNNs) are designed to facilitate the retention of sequential data, namely text, speech, and time-series data such as clinical data or electronic health records (EHRs). They can capture temporal relationships between data components, which can be helpful for predicting disease progression or treatment outcomes 11 , 17 , 18 . RNNs use similar architecture components that CNNs have. Long Short-Term Memory (LSTM) models are types of RNNs and are commonly used to overcome their shortcomings because they can learn long-term dependencies in data better than conventional RNN architectures. They are utilized in some classification tasks, including audio 17 , 19 . LSTM utilizes a gated memory cell in the network architecture to store information from the past; hence, the memory cell can store information for a long period of time, even if the information is not immediately relevant to the current task. This allows LSTMs to learn patterns in data that would be difficult for other types of neural networks to learn.

Generative adversarial networks (GANs) are a class of DL models that can be used to generate new data that is like existing data. In healthcare, GANs have been used to generate synthetic medical images. There are two CNNs (generator and discriminator); the first CNN is called the generator, and its primary goal is to make synthetic images that mimic actual images. The second CNN is called the discriminator, and its main objective is to identify between artificially generated images and real images 20 . The generator and discriminator are trained jointly in a process called adversarial training, where the generator tries to create data that is so realistic that the discriminator cannot distinguish it from real data. GANs are used to generate a variety of different types of data, including images, videos, and text. GANs are used to enhance image quality, signal reconstruction, and other tasks such as classification and segmentation too 20 , 21 , 22 .

Transfer learning (TL) is a concept derived from cognitive science that states that information is transferred across related activities to improve performance on a new task. It is generally known that people can accomplish similar tasks by building on prior knowledge 23 . TL has been implemented to minimize the need for annotation by transferring DL models with knowledge from a previous task and then fine-tuning them in the current task 24 . The majority of medical image classification techniques employ TL from pretrained models, such as ImageNet , which has been demonstrated to be inefficient due to the ImageNet consisting of natural images 25 . The approaches that utilized ImageNet pre-trained images in CNNs revealed that fine-tuning more layers provided increased accuracy 26 . The initial layers of ImageNet-pretrained networks, which detect low-level image characteristics, including corners and borders, may not be efficient for medical images 25 , 26 .

New and more advanced DL algorithms are developed almost daily. Such methods could be employed for the analysis of imaging and non-imaging data in order to enhance performance and reliability. These methods include Capsule Networks, Attention Mechanisms, and Graph Neural Networks (GNNs) 27 , 28 , 29 , 30 . Briefly, these are:

Capsule Networks are a relatively new form of DL architecture that aim to address some of the shortcomings of CNNs: pooling operations (reducing the data size) and a lack of hierarchical relations between objects and their parts in the data. Capsules can capture spatial relationships between features and are more capable of handling rotations and deformations of image objects thanks to their vectorial representations in neuronal space. Capsule Networks have shown potential in image classification tasks and could have applications in medical imaging analysis 27 . However, its implementation and computational time are two hurdles that restrict its widespread use.

Attention Mechanisms , represented by Transformers, have contributed to the development of computer vision and language processing. Unlike CNNs or RNNs, transformers allow direct interaction between every pair of components within a sequence, making them particularly effective at capturing long-term relationships 29 , 30 . More specifically, a self-attention mechanism in Transformers is an important piece of the DL model as it can dynamically focus on different parts of the input data sequence when producing an output, providing better context understanding than CNN based systems.

Graph Neural Networks (GNNs) are a form of data structure that describes a collection of objects (nodes) and their relationships (edges). There are three forms of tasks, including node-level, edge-level, and graph level 31 . Graphs may be used to denote a wide range of systems, including molecular interaction networks, and bioinformatics 31 , 32 , 33 . GNNs have demonstrated potential in both imaging and non-imaging data analysis 28 , 34 .

Physics-driven systems are needed in imaging field. Several studies have demonstrated the effectiveness of DL methods in the medical imaging field 35 , 36 , 37 , 38 , 39 . As the field of DL continues to evolve, it is likely that new methods and architectures will emerge to address the unique challenges and constraints of various types of data. One of the most common problems faced with DL-based MRI construction 35 . Specific algorithms for this problem can be essentially categorized into two groups: data driven and physics driven algorithms. In purely data-driven approaches, a mapping is learned between the aliased image and the image without artifacts 39 . Acquiring fully sampled (artifact-free) datasets is impractical in many clinical imaging studies when organs are in motion, such as the heart, and lung. Recently developed models can employ these under sampled MRI acquisitions as input and generate output images consistent with fully-sampled (artifact free) acquisitions 37 , 38 , 39 .

What is the Hybrid Intelligence? A highly desirable way of incorporating advances in AI is to let AI and human intellect work together to solve issues, and this is referred to as “hybrid intelligence“ 40 (e.g., one may call this “mixed intelligence” or “human-in-the-loop AI systems”). This phenomenon involves the development of AI systems that serve to supplement and amplify human decision-making processes, as opposed to completely replacing them 3 . The concept involves integrating the respective competencies of artificial intelligence and human beings in order to attain superior outcomes that would otherwise be unachievable 41 . AI algorithms possess the ability to process extensive amounts of data, recognize patterns, and generate predictions rapidly and precisely. Meanwhile, humans can contribute their expertise, understanding, and intuition to the discussion to offer context, analyze outcomes, and render decisions 42 . The hybrid intelligence strategy can help decision-makers in a variety of fields make decisions that are more precise, effective, and efficient by combining these qualities 3 , 4 , 43 , 44 . Human in the loop and hybrid intelligence systems are promising for time-consuming tasks in healthcare and neonatology.

Where do we stand currently? AI in medicine has been employed for over a decade, and it has often been considered that clinical implementation is not completely adapted to daily practice in most of the clinical field 5 , 45 , 46 . In recent years, increasingly complex computer algorithms and updated hardware technologies for processing and storing enormous datasets have contributed to this achievement 6 , 7 , 46 , 47 . It has only been within the last decade that these systems have begun to display their full potential 6 , 9 . The field of AI research appears to have been taken up with differing degrees of enthusiasm across disciplines. When analyzing the thirty years of research into AI, DL, and ML conducted by several medical subfields between the years 1988 and 2018, one-third of publications in DL yielded to radiology, and most of them are within the imaging sciences (radiology, pathology, and cell imaging) 48 . Software systems work by utilizing biomedical images with predictive/diagnostic/prognostic features and integrating clinical or pre-clinical data. These systems are designed with ML algorithms 46 . Such breakthrough methods in DL are nowadays extensively applied in pathology, dermatology, ophthalmology, neurology, and psychiatry 6 , 47 , 49 . AI has its own difficulties with the increasing utilization of healthcare (Fig. 1b ).

What are the needs in clinics? Clinicians are concerned about the healthcare system’s integration with AI: there is an exponential need for diagnostic testing, early detection, and alarm tools to provide diagnosis and novel treatments without invasive tests and procedures 50 . Clinicians have higher expectations of AI in their daily practices than before. AI is expected to decrease the need for multiple diagnostic invasive tests and increase diagnostic accuracy with less invasive (or non-invasive) tests. Such AI systems can easily recognize imaging patterns on test images (i.e., unseen or not utilized efficiently in daily routines), allowing them to detect and diagnose various diseases. These methods could improve detection and diagnosis in different fields of medicine.

The overall goal of this systematic review is to explain AI’s potential use and benefits in the field of neonatology. We intend to enlighten the potential role of AI in the future in neonatal care. We postulate that AI would be best used as a hybrid intelligence (i.e., human-in-the-loop or mixed intelligence) to make neonatal care more feasible, increase the accuracy of diagnosis, and predict the outcome and diseases in advance. The rest of the paper is organized as follows: In results, we explain the published AI applications in neonatology along with AI evaluation metrics to fully understand their efficacy in neonatology and provide a comprehensive overview of DL applications in neonatology. In discussion, we examine the difficulties of AI utilization in neonatology and future research discussions. In the methods section, we outline the systematic review procedures, including the examination of existing literature and the development of our search strategy.

We review the past, current, and future of AI-based diagnostic and monitoring tools that might aid neonatologists’ patient management and follow-up. We discuss several AI designs for electronic health records, image, and signal processing, analyze the merits and limits of newly created decision support systems, and illuminate future views clinicians and neonatologists might use in their normal diagnostic activities. AI has made significant breakthroughs to solve issues with conventional imaging approaches by identifying clinical variables and imaging aspects not easily visible to human eyes. Improved diagnostic skills could prevent missed diagnoses and aid in diagnostic decision-making. The overview of our study is structured as illustrated in Fig. 2 . Briefly, our objectives in this systematic review are:

to explain the various AI models and evaluation metrics thoroughly explained and describe the principal features of the AI models,

to categorize neonatology-related AI applications into macro-domains, to explain their sub-domains and the important elements of the applicable AI models,

to examine the state-of-the-art in studies, particularly from the past several years, with an emphasis on the use of ML in encompassing all neonatology,

to present a comprehensive overview and classification of DL applications utilized and in neonatology,

to analyze and debate the current and open difficulties associated with AI in neonatology, as well as future research directions, to offer the clinician a comprehensive perspective of the actual situation.

It is provided an overview of our paper’s structure and objectives: 1. Explaining AI Models and Evaluation Metrics: 2. Evaluating ML applied studies in Neonatology 3. Evaluating DL applied studies in Neonatology 4. Analyzing Challenges and Future Directions.

AI covers a broad concept for the application of computing algorithms that can categorize, predict, or generate valuable conclusions from enormous datasets 46 . Algorithms such as Naive Bayes, Genetic Algorithms, Fuzzy Logic, Clustering, Neural Networks (NN), Support Vector Machines (SVM), Decision Trees, and Random Forests (RF) have been used for more than three decades for detection, diagnosis, classification, and risk assessment in medicine as ML methods 9 , 10 . Conventional ML approaches for image classification involve using hand-engineered features, which are visual descriptions and annotations learned from radiologists, that are encoded into algorithms.

Images, signals, genetic expressions, EHR, and vital signs are examples of the various unstructured data sources that comprise medical data (Fig. 3 ). Due to the complexity of their structures, DL frameworks may take advantage of this heterogeneity by attaining high abstraction levels in data analysis.

Unstructured data such as medical images, vital signals, genetic expressions, EHRs, and signal data contribute to the wide variety of medical information. Analyzing and interpreting different data streams in neonatology requires a comprehensive strategy because each has unique characteristics and complications.

While ML requires manual/hand-crafted selection of information from incoming data and related transformation procedures, DL performs these tasks more efficiently and with higher efficacy 9 , 10 , 46 . DL is able to discover these components by analyzing a large number of samples with a high degree of automation 7 . The literature on these ML approaches is extensive before the development of DL 5 , 7 , 45 .

It is essential for clinicians to understand how the suggested ML model should enhance patient care. Since it is impossible for a single metric to capture all the desirable attributes of a model, it is customarily necessary to describe the performance of a model using several different metrics. Unfortunately, many end-users do not have an easy time comprehending these measurements. In addition, it might be difficult to objectively compare models from different research models, and there is currently no method or tool available that can compare models based on the same performance measures 51 . In this part, the common ML and DL evaluation metrics are explained so neonatologists could adapt them into their research and understand of upcoming articles and research design 51 , 52 .

AI is commonly utilized everywhere, from daily life to high-risk applications in medicine. Although slower compared to other fields, numerous studies began to appear in the literature investigating the use of AI in neonatology. These studies have used various imaging modalities, electronic health records, and ML algorithms, some of which have barely gone through the clinical workflow. Though there is no systematic review and future discussions in particular in this field 53 , 54 , 55 . Many studies were dedicated to introducing these systems into neonatology. However, the success of these studies has been limited. Lately, research in this field has been moving in a more favorable direction due to exciting new advances in DL. Metrics for evaluations in those studies were the standard metrics such as sensitivity (true-positive rate), specificity (true-negative rate), false-positive rate, false-negative rate, receiver operating characteristics (ROC), area under the ROC curves (AUC), and accuracy (Table 1 ).

This systematic review was guided by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) protocol 56 . The search was completed on 11st of July 2022. The initial search yielded many articles (approximately 9000), and we utilized a systematic approach to identify and select relevant articles based on their alignment with the research focus, study design, and relevance to the topic. We checked the article abstracts, and we identified 987 studies. Our search yielded 106 research articles between 1996 and 2022 (Fig. 4 ). Risk of bias summary analysis was done by the QUADAS-2 tool (Figs. 5 and 6 ) 57 , 58 , 59 .

Initial research conducted on 11th of July 2022, yielded 9000 articles, of which 987 article abstracts were screened. Of those, 106 research articles published between 1996 and 2022 were eligible for inclusion in this systematic review. The PRISMA flow diagram illustrates the study selection process in more detail.

Risk of bias summary analysis was done by the QUADAS-2 tool.

Our findings are summarized in two groups of tables: Tables 2 – 5 summarize the AI methods from the pre-deep learning era (“Pre-DL Era”) in neonatal intensive care units according to the type of data and applications. Tables 6 , 7 , on the other hand, include studies from the DL Era. Applications include classification (i.e., prediction and diagnosis), detection (i.e., localization), and segmentation (i.e., pixel level classification in medical images).

ML applications in neonatal mortality

Neonatal mortality is a major factor in child mortality. Neonatal fatalities account for 47 percent of all mortality in children under the age of five, according to the World Health Organization 60 . It is, therefore, a priority to minimize worldwide infant mortality by 2030 61 .

ML investigated infant mortality, its reasons, and its mortality prediction 62 , 63 , 64 , 65 , 66 , 67 , 68 . In a recent review, 1.26 million infants born from 22 weeks to 40 weeks of gestational age were enrolled 67 . Predictions were made as early as 5 min of life and as late as 7 days. An average of four models per investigation were neural networks, random forests, and logistic regression (58.3%) 67 . Two studies (18.2%) completed external validation, although five (45.5%) published calibration plots 67 . Eight studies reported AUC, and five supplied sensitivity and specificity 67 . The AUC was 58.3–97.0% 67 . Sensitivities averaged 63 to 80%, and specificities 78 to 98% 67 . Linear regression analysis was the best overall model despite having 17 features 67 . This analysis highlighted the most prevalent AI neonatal mortality measures and predictions. Despite the advancement in neonatal care, it is crucial that preterm infants remain highly susceptible to mortality due to immaturity of organ systems and increased susceptibility to early and late sepsis 69 . Addressing these permanent risks necessitates the utilization of ML to predict mortality 63 , 64 , 65 , 66 , 68 , 70 . Early studies employed ANN and fuzzy linguistic models and achieved an AUC of 85–95% and accuracy of 90% 62 , 68 . New studies in a large preterm populations and extremely low birthweight infants found an AUC of 68.9–93.3% 65 , 71 . There are some shortcomings in these studies; for example, none of them used vital parameters to represent dynamic changes, and hence, there was no improvement in clinical practice in neonatology. Unsurprisingly, gestational age, birthweight, and APGAR scores were shown as the most important variables in the models 64 , 72 . Future research is suggested to focus on external evaluation, calibration, and implementation of healthcare applications 67 .

Neonatal sepsis, which includes both early onset sepsis and late onset sepsis, is a significant factor contributing to neonatal mortality and morbidity 73 . Neonatal sepsis diagnosis and antibiotic initiation present considerable obstacles in the field of neonatal care, underscoring the importance of implementing comprehensive interventions to alleviate their profound negative consequences. The studies have predicted early sepsis from heart rate variability with an accuracy of 64–94% 74 . Another secondary analysis of multicenter data revealed that clinical biomarkers weighed the ML decision by integrating all clinical and lab variables and achieved an AUC of 73–83% 75 .

ML applications in neurodevelopmental outcome

Recent advancements in neonatal healthcare have resulted in a decrease in the incidence of severe prenatal brain injury and an increase in the survival rates of preterm babies 76 . However, even though routine radiological imaging does not reveal any signs of brain damage, this population is nonetheless at significant risk of having a negative outcome in terms of neurodevelopment 77 , 78 , 79 , 80 . It is essential to discover early indicators of abnormalities in brain development that might serve as a guide for the treatment of preterm children at a greater risk of having negative neurodevelopmental consequences 81 , 82 .

The most common reason for neurodevelopmental impairment is intraventricular hemorrhage (IVH) in preterm infants 83 . Two studies predicted IVH in preterm infants. Both studies have not deployed the ultrasound images in their analysis, they only predicted IVH according to the clinical variables 84 , 85 .

Morphological studies have demonstrated that preterm birth is linked to smaller brain volume, cortical folding, axonal integrity, and microstructural connectivity 86 , 87 . Studies concentrating on functional markers of brain maturation, such as those derived from resting-state functional connectivity (rsFC) analyses of blood-oxygen-level dependent (BOLD) fluctuations, have revealed further impacts of prematurity on the developing connectome, ranging from decreased network-specific connectivity 82 , 88 , 89 . Many studies investigated brain connectivity in preterm infants 88 , 90 , 91 , 92 and brain structural analysis in neonates 93 and neonatal brain segmentation 94 with the help of ML methods. Similarly, one of the most important outcomes of neurodevelopment at 2-year-old-age is neurocognitive evaluations. The studies evaluated the morphological changes in the brain in relation to neurocognitive outcome 95 , 96 , 97 and brain age prediction 98 , 99 . It has been found that near-term regional white matter (WM) microstructure on diffusion tensor imaging (DTI) predicted neurodevelopment in preterm infants using exhaustive feature selection with cross-validation 96 and multivariate models of near-term structural MRI and WM microstructure on DTI might help identify preterm infants at risk for language impairment and guide early intervention 95 , 97 (Table 4 ). One of the studies that evaluated the effects of PPAR gene activity on brain development with ML methods 100 revealed a strong association between abnormal brain connectivity and implicating PPAR gene signaling in abnormal white matter development. Inhibited brain growth in individuals exposed to early extrauterine stress is controlled by genetic variables, and PPARG signaling has a formerly unknown role in cerebral development 100 (Table 2 ).

Alternative to morphological studies, neuromonitorization is shown to be an important tool for which ML methods have been frequently employed, for example, in automatic seizure detection from video EEG 101 , 102 , 103 and EEG biosignals in infants and neonates with HIE 104 , 105 , 106 , 107 , 108 . The detection of artifacts 109 , 110 , sleep states 102 , rhythmic patterns 111 , burst suppression in extremely preterm infants 112 , 113 from EEG records were studied with ML methods. EEG records are often used for HIE grading 114 too. It has been shown in those studies that EEG recordings of different neonate datasets found an AUC of 89% to 96% 104 , 105 , 115 , accuracy 78–87% 114 , 116 regarding seizure detection with different ML methods (Table 3 ).

ML applications in predictions of prematurity complications (BPD, PDA, and ROP)

Another important cause of mortality and morbidity in the NICU is PDA (Patent Ductus Arteriosus). The ductus arteriosus is typically present during the fetal stage, when the circulation in the lungs and body is regularly supplied by the mother; in newborns, the ductus arteriosus closes functionally by 72 h of age 117 . 20–50% of infants with a gestational age (GA) 32 weeks have the ductus arteriosus on day 3 of life 118 , while up to 60% of neonates with a GA 29 weeks have the ductus arteriosus. The presence of PDA in preterm neonates is associated with higher mortality and morbidity, and physicians should evaluate if PDA closure might enhance the likelihood of survival vs. the burden of adverse effects 119 , 120 , 121 , 122 .

ML methods were utilized on PDA detection from EHR 123 and auscultation records 124 such that 47 perinatal factors were analyzed with 5 different ML methods in 10390 very low birth weight infants’ predicted PDA with an accuracy of 76% 123 and 250 auscultation records were analyzed with XGBoost and found to have an accuracy of 74% 124 (Table 3 ).

Bronchopulmonary dysplasia (BPD) is a leading cause of infant death and morbidity in preterm births. While various biomarkers have been linked to the development of respiratory distress syndrome (RDS), no clinically relevant prognostic tests are available for BPD at birth 125 . There are ML studies aiming to predict BPD from birth 70 , 126 , gastric aspirate content 125 and genetic data 127 and it has been shown that BPD could be predicted with an accuracy of up to 86% in the best-case scenario 70 (Table 5 ), analysis of responsible genes with ML could predict BPD development with an AUC of 90% 127 (Table 3 ) and combination of gastric aspirate after birth and clinical information analysis with SVM predicted BPD development with a sensitivity of 88% 125 (Table 5 ).

In relation to published studies in BPD with ML-based predictions, long-term invasive ventilation is considered one of the most important risk factors for BPD, nosocomial infections, and increased hospital stay. There are ML-based studies aiming to predict extubation failure 128 , 129 , 130 and optimum weaning time 131 using long-term invasive ventilation information. It has been shown in those studies that predicted extubation failure with an accuracy of 83.2% to 87% 128 , 129 , 130 (Tables 2 and 3 ).

Retinopathy of prematurity (ROP) is another area of interest in the application of machine learning in neonatology 132 . ROP is a serious complication of prematurity that affects the blood vessels in the retina and is a leading cause of childhood blindness in high and middle-income countries, including the United States, among very low-birthweight (1500 g), very preterm (28–32 weeks), and extremely preterm infants (less than 28 weeks) 132 . Due to a shortage of ophthalmologists available to treat ROP patients, there has been increased interest in the use of telemedicine and artificial intelligence as solutions for diagnosing ROP 132 . Some ML methods, such as Gaussian mixture models, were employed to diagnose and classify ROP from retinal fundus images in studies 132 , 133 , 134 , and it has been reported that the i-ROP 134 system classified pre-plus and plus disease with 95% accuracy. This was close to the performance of the three individual experts (96%, 94%, and 92%, respectively), and much higher than the mean performance of 31 nonexperts (81%) 134 (Table 2 ).

Other ML applications in neonatal diseases

EHR and medical records were featured in ML algorithms for the diagnosis of congenital heart defects 135 , HIE (Hypoxic Ischemic Encephalopathy) 136 , IVH (Intraventricular Hemorrhage) 84 , 85 , neonatal jaundice 137 , 138 , prediction of NEC (Necrotizing Enterocolitis) 139 , prediction of neurodevelopmental outcome in ELBW (extremely low birth weight) infants 65 , 140 , 141 , prediction of neonatal surgical site infections 142 , and prediction of rehospitalization 143 (Table 5 ).

Electronically captured physiologic data are evaluated as signal data, and they were analyzed with ML to detect artifact patterns 144 , late onset sepsis 145 , and predict infant morbidity 146 . Electronically captured vital parameters (respiratory rate, heart rate) of 138 infants (≤34 weeks’ gestation, birth weight ≤2000 gram) in the first 3 h of life predicted an accuracy of overall morbidity and an AUC of 91% 146 (Table 5 ).

In addition to physiologic data, clinical data up to 12 h after cardiac surgery in HLHS (hypoplastic left heart syndrome) and TGA (transposition of great arteries) infants were analyzed to predict PVL (periventricular leukomalacia) occurrence after surgery 147 . The F-score results for infants with HLHS and those without HLHS were 88% and 100%, respectively 147 (Table 5 ). Voice records were used to diagnose respiratory phases in infant cry 148 , to classify neonatal diseases in infant cry 149 , and to evaluate asphyxia from infant cry voice records 150 . Voice records of 35 infants were analyzed with ANN, and accuracy was found 85% 149 . Cry records of 14 infants in their 1st year of life were analyzed with SVM and GMM, and phases of respiration and crying rate were quantified with an accuracy of 86% 148 (Table 3 ).

SVM was the most commonly used method in the diagnosis of metabolic disorders of newborns, including MMA (methylmalonic acidemia) 151 , PKU (phenylketonuria) 152 , 153 , MCADD (medium-chain acyl CoA dehydrogenase deficiency) 152 . During the Bavarian newborn screening program, dried blood samples were analyzed with ML and increased the positive predictive value for PKU (71.9% versus 16,2) and for MCADD (88.4% versus 54.6%) 152 (Table 3 ).

Neonatology with deep learning

The main uses of DL in clinical image analysis are categorized into three categories: classification, detection, and segmentation. Classification involves identifying a specific feature in an image, detection involves locating multiple features within an image; and segmentation involves dividing an image into multiple parts 7 , 9 , 154 , 155 , 156 , 157 , 158 , 159 , 160 .

Neuroradiological evaluation with AI in neonatology

Neonatal neuroimaging can establish early indicators of neurodevelopmental abnormality to provide early intervention during a time of maximal neuroplasticity and fast cognitive and motor development 79 , 96 . DL methods can assist in an earlier diagnosis than clinical signs would indicate.

The imaging of an infant’s brain using MRI can be challenging due to lower tissue contrast, substantial tissue inhomogeneities, regionally heterogeneous image appearance, immense age-related intensity variations, and severe partial volume impact due to the smaller brain size. Since most of the existing tools were created for adult brain MRI data, infant-specific computational neuroanatomy tools are recently being developed. A typical pipeline for early prediction of neurodevelopmental disorders from infant structural MRI (sMRI) is made up of three basic phases. (1) Image preprocessing, tissue segmentation, regional labeling, and extraction of image-based characteristics (2) Surface reconstruction, surface correspondence, surface parcellation, and extraction of surface-based features (3) Feature preprocessing, feature extraction, AI model training, and prediction of unseen subjects 161 . The segmentation of a newborn brain is difficult due to the decreased SNR (signal to noise ratio) resulting from the shorter scanning duration enforced by predicted motion restrictions and the diminutive size of the neonatal brain. In addition, the cerebrospinal fluid (CSF)-gray matter border has an intensity profile comparable to that of the mostly unmyelinated white matter (WM), resulting in significant partial volume effects. In addition, the high variability resulting from the fast growth of the brain and the continuing myelination of WM imposes additional constraints on the creation of effective segmentation techniques. Several non-DL-based approaches for properly segmenting newborn brains have been presented over the years. These methods may be broadly classified as parametric 162 , 163 , 164 , classification 165 , multi-atlas fusion 166 , 167 , and deformable models 168 , 169 . The Dice Similarity Coefficient metric is used for image segmentation evaluation; the higher the dice, the higher the segmentation accuracy 10 (Table 1 ).

In the NeoBrainS12 2012 MICCAI Grand-Challenge ( https://neobrains12.isi.uu.nl ), T1W and T2W images were presented with manually segmented structures to assess strategies for segmenting neonatal tissue 162 . Most methods were found to be accurate, but classification-based approaches were particularly precise and sensitive. However, segmentation of myelinated vs. unmyelinated WM remains a difficulty since the majority of approaches 162 failed to consistently obtain reliable results.

Future research in neonatal brain segmentation will involve a more thorough neural segmentation network. Current studies are intended to highlight efficient networks capable of producing accurate and dependable segmentations while comparing them to existing conventional computer vision techniques. In the perspective of comparing previous efforts on newborn brain segmentation, the small sample size of high-quality labeled data must also be recognized as a significant restriction 169 . The field of artificial intelligence in neonatology has progressed slowly due to a shortage of open-source algorithms and the availability of datasets.

Future research should also focus on improving the accuracy of DL for diagnosing germinal matrix hemorrhage and figuring out how DL can help a radiologist’s workflow by comparing how well sonographers identify studies that look suspicious. More studies could also look at how well DL works for accurately grading germinal matrix hemorrhages and maybe even small hemorrhages that a radiologist can see on an MRI but not on a head ultrasound. This could be useful in improving the diagnostic capabilities of head ultrasound in various clinical scenarios 157 .

Evaluation of prematurity complications with DL in neonatology

In the above discussion, we have addressed the primary applications of DL in relation to disease prediction. These include DL for analyzing conditions such as PDA (patent ductus arteriosus) 158 , IVH (intraventricular ventricular hemorrhage) 155 , 157 , BPD (bronchopulmonary dysplasia) 170 , ROP (retinopathy of prematurity) 171 , 172 , 173 , retinal hemorrhage 174 diagnosis. This also includes DL applications for analyzing MR images 159 , 175 and combined with EHR data 176 , 177 for predicting neurocognitive outcome and mortality. Additionally, DL has potential applications in treatment planning and discharge from the NICU 178 , including customized medicine and follow-up 6 , 67 , 125 (Tables 6 and 7 ).

Digital imaging and analysis with AI are promising and cost-effective tools for detecting infants with severe ROP who may need therapy 132 , 171 , 172 , 179 . Despite limitations such as image quality, interpretation variability, equipment costs, and compatibility issues with EHR systems, AI has been shown to be effective in detecting ROP 180 . Studies comparing BIO (Binocular Indirect Ophthalmoscope) to telemedicine have shown that both methods have equivalent sensitivity for identifying zone disease, plus disease, and ROP. However, BIO was found to be slightly better at identifying zone III and stage 3 ROP 181 , 182 . DL algorithms were applied to 5511 retinal images, achieving an AUC of 94% (diagnosis of normal) and 98% (diagnosis of plus disease), outperforming 6 out of 8 ROP experts 171 . In another study, DL was used to quantify the clinical progression of ROP by assigning ROP vascular severity scores 172 . A consecutive study with a large dataset showed in 4175 retinal images from 32 NICUs, resulting in an AUC of 98% for detecting therapy required ROP with DL 173 . The use of AI in ROP screening programs may increase access to care for secondary prevention of ROP and enable the evaluation of disease epidemiology 173 (Table 6 ).

Signal detection for sleep protection in the NICU is another ongoing discussion. DL has been used to analyze infant EEGs and identify sleep states. Interruptions of sleep states have been linked to problems in neuronal development 183 . Automated sleep state detection from EEG records 184 , 185 and from ECG monitoring parameters 186 were demonstrated with DL. The underperformance of the all-state classification (kappa score 0.33 to 0.44) was likely owing to the difficulties in differentiating small changes between states and a lack of enough training data for minority classes 186 (Table 6 ).

DL has been found to be effective in real-time evaluation of cardiac MRI for congenital heart disease 187 . Studies have shown that DL can accurately calculate ventricular volumes from images rebuilt using residual UNet, which are not statistically different from the gold standard, cardiac MRI. This technology has the potential to be particularly beneficial for infants and critically ill individuals who are unable to hold their breath during the imaging process 187 (Table 6 ).

DL-based 3D CNN algorithms have been used to demonstrate the automated classification of brain dysmaturation from neonatal brain MRI 188 . In a study, brain MRIs of 90 term neonates with congenital heart diseases and 40 term healthy controls were analyzed using this method, which achieved an accuracy of 98%. This technique could be useful in detecting brain dysmaturation in neonates with congenital heart diseases 188 (Table 6 ).

DL algorithms have been used to classify neonatal diseases from thermal images 189 , 190 , 191 , 192 . These studies analyzed neonatal thermograms to determine the health status of infants and achieved good AUC scores 189 , 190 , 191 , 192 . However, these studies didn’t include any clinical information (Table 6 ).

Two large scale studies showed breakthrough results regarding the effect of nutrition practices in NICU 170 and wireless sensors in NICU 193 . A nutrition study revealed that nutrition practices were associated with discharge weight and BPD 170 . This exemplifies how unbiased ML techniques may be used to effectively bring about clinical practice changes 170 . Novel, wireless sensors can improve monitoring, prevent iatrogenic injuries, and encourage family-centered care 193 . Early validation results show performance equal to standard-of-care monitoring systems in high-income nations. Furthermore, the use of reusable sensors and compatibility with low-cost mobile phones may reduce monitoring.

The studies in neonatology with AI were categorized according to the following criteria.

The studies were performed with ML or DL,

imaging data or non-imaging data were used,

according to the aim of the study: diagnosis or other predictions.

Most of the studies in neonatology were performed with ML methods in the pre-DL era. We have listed 12 studies with ML and imaging data for diagnosis. There are 33 studies that used non-imaging data for diagnosis purposes. Imaging data studies cover BA diagnosis from stool color 194 , postoperative enteral nutrition of neonatal high intestinal obstruction 195 , functional brain connectivity in preterm infants 82 , 90 , 91 , 94 , 100 , ROP diagnosis 133 , 134 , neonatal seizure detection from video records 101 , newborn jaundice screening 137 . Non-imaging studies for diagnosis include the diagnosis of congenital heart defects 135 , baby cry analysis 148 , 149 , 150 , inborn metabolic disorder diagnosis and screening 151 , 152 , 153 , HIE grading 104 , 106 , 114 , 136 , 196 , EEG analysis 102 , 104 , 106 , 107 , 110 , 111 , 112 , 113 , 115 , 184 , 197 , 198 , PDA diagnosis 123 , 124 , vital sign analysis and artifact detection 144 , extubation and weaning analysis 129 , 130 , 131 , 144 , BPD diagnosis 127 . ML studies with imaging data for prediction are focused on neurodevelopmental outcome prognosis from brain MRIs 95 , 96 , 97 , 127 , 164 , 199 . ML-based non-imaging data for prediction encompassed mortality risk 63 , 64 , 65 , 68 , NEC prognosis 139 , morbidity 66 , 146 , BPD 125 , 126 .

When it comes to DL applications, there has been less research conducted compared to ML applications. The focus of DL with imaging and non-imaging data focused on brain segmentation 159 , 169 , 175 , 177 , 188 , IVH diagnosis 157 , EEG analysis 184 , 185 , neurocognitive outcome 176 , PDA and ROP diagnosis 171 , 172 , 173 . Upcoming articles and research will surely be from the DL field, though.

It is worth noting that there have also been several articles and studies published on the topic of the application of AI in neonatology. However, the majority of these studies do not contain enough details, are difficult to evaluate side-by-side, and do not give the clinician a thorough picture of the applications of AI in the general healthcare system 66 , 67 , 93 , 95 , 96 , 97 , 99 , 125 , 126 , 127 , 140 , 142 , 147 , 169 , 174 , 177 , 185 , 188 , 200 , 201 , 202 , 203 , 204 , 205 .

There are several limitations in the application of AI in neonatology, including a lack of prospective design, a lack of clinical integration, a small sample size, and single center evaluations. DL has shown promise in bioscience and biosignals, extracting information from clinical images, and combining unstructured and structured data in EHR. However, there are some issues that limit the success of DL in medicine, which can be grouped into six categories. In the following paragraphs, we’ll examine the key concerns related to DL, which have been divided into six components:

Difficulties in clinical integration, including the selection and validation of models;

the need for expertise in decision mechanisms, including the requirement for human involvement in the process;

lack of data and annotations, including the quality and nature of medical data; distribution of data in the input database; and lack of open-source algorithms and reproducibility;

lack of explanations and reasoning, including the lack of explainable AI to address the “black-box” problem;

lack of collaboration efforts across multi-institutions; and

ethical concerns 4 , 5 , 6 , 9 , 10 , 206 .

Difficulties in clinical integration

Despite the accuracy that AI has reached in healthcare in recent years, there are several restrictions that make it difficult to translate into treatment pathways. First, physicians’ suspicion of AI-based systems stems from the lack of qualified randomized clinical trials, particularly in the field of pediatrics, showing the reliability and/or improved effectiveness of AI systems compared to traditional systems in diagnosing neonatal diseases and suggesting appropriate therapies. The studies’ pros and cons are discussed in tables and relevant sections. Studies are mainly focused on imaging-based or signal-based studies in terms of one variable or disease. Neonatologists and pediatricians need evidence-based proven algorithm studies. There are only six prospective clinical trials in neonatology with AI 197 , 207 , 208 , 209 , 210 , 211 . The one is detecting neonatal seizures with conventional EEG in the NICU which is supported by the European Union Cost Program in 8 European NICU 197 . Neonates with a corrected gestational age between 36 and 44 weeks who had seizures or were at high risk of having seizures and needed EEG monitoring were given conventional EEG with ANSeR (Algorithm for Neonatal Seizure Recognition) coupled with an EEG monitor that displayed a seizure probability trend in real time (algorithm group) or continuous EEG monitoring alone (non-algorithm group) 197 . The algorithm is not available, and the code is not shared. Another one is a study showing the physiologic effects of music in premature infants 208 . Even so, it could not be founded on any AI analysis in this study. The third study, “Rebooting Infant Pain Assessment: Using Machine Learning to Exponentially Improve Neonatal Intensive Care Unit Practice (BabyAI),” is newly posted and recruiting 209 . The fourth study, “Using sensor-fusion and machine learning algorithms to assess acute pain in non-verbal infants: a study protocol,” aims to collect data from 15 subjects: preterm infants, term infants within the first month of age in NICU admission and their follow-up data at 3rd and 6th months of age. They record pain signals using facial electromyography(EMG), ECG, electrodermal activity, oxygen saturation, and EEG in real time, and they will analyze the data with ML methods to evaluate pain in neonates. The data is in iPAS (NCT03330496) and is updated as recruitment completed 210 . However, no result has been submitted. The fifth study, “Prediction of Extubation Readiness in Extreme Preterm Infants by the Automated Analysis of Cardiorespiratory Behavior: APEX study” 211 records revealed that the recruitment was completed in 266 infants. Still, no results have been released yet (NCT01909947). To sum up, there is only one prospective multicenter randomized AI study that has been published with its results.

There is an unmet need to plan clinically integrated prospective and real-time data collection studies in neonatology. The clinical situation of infants changed rapidly, and real-time designed studies would be significant by analyzing multimodal data and including imaging and non-imaging components.

The need for expertise in the decision mechanisms

In terms of neonatologists determining whether to implement a system’s recommendation, it may be required for that system to present supporting evidence 95 , 96 , 125 , 202 . Many suggested AI solutions in the medical field are not expected to be an alternative to the doctor’s decision or expertise but rather to serve as helpful assistance. When it comes to struggling neonatal survival without sequela, AI may be a game changer in neonatology. The broad range of neonatal diseases and different clinical presentations of neonates according to gestational age and postnatal age make accurate diagnosis even harder for neonatologists. AI would be effective for early disease detection and would assist clinicians in responding promptly and fostering therapy outcomes.

Neonatology has multidisciplinary collaborations in the management of patients, and AI has the potential to achieve levels of efficacy that were previously unimaginable in neonatology if more resources and support from physicians were allocated to it. Neonatology collaborates and closely works with other specialties of pediatrics, including perinatology, pediatric surgery, radiology, pediatric cardiology, pediatric neurology, pediatric infectious disease, neurosurgery, cardiovascular surgery, and other subspecialties of pediatrics. Those multidisciplinary workflows require patient follow-up and family involvement. AI-based predictive analysis tools might address potential risks and neurologic problems in the future. AI supported monitoring systems could analyze real time data from monitors and detect changes simultaneously. These tools could be helpful not only for routine NICU care but also for “family centered care” 212 , 213 implications. Although neonatologists could be at the center of decision making and giving information to parents, AI could be actively used in NICUs. Hybrid intelligence would provide a follow-up platform for abrupt and subtle clinical changes in infants’ clinical situations.

Given that many medical professionals have a limited understanding of DL, it may be difficult to establish contact and communication between data scientists and medical specialists. Many medical professionals, including pediatricians and neonatologists in our instance, are unfamiliar with AI and its applications due to a lack of exposure to the field as an end user. However, the authors also acknowledge the increasing efforts in building bridges among many scientists and institutions, with conferences, workshops, and courses, that clinicians have successfully started to lead AI efforts, even with software coding schools by clinicians 214 , 215 , 216 , 217 , 218 .

Neonatal critical conditions will be monitored by the human in the loop systems in the near future, and AI empowered risk classification systems may help clinicians prioritize critical care and allocate supplies precisely. Hence, AI could not replace neonatologists, but there would be a clinical decision support system in the critical and calls for prompt response environment of NICU.

Lack of imaging data and annotations and reproducibility problems

There is a rising interest in building deep learning approaches to predict neurological abnormalities using connectome data; however, their usage in preterm populations has been limited 81 , 88 , 89 , 90 , 91 . Similar to most DL applications, the training of such models often requires the use of big datasets 11 ; however, large neuroimaging datasets are either not accessible or difficult and expensive to acquire, especially in the pediatric world. Since the success of DL methods currently relies on well-labeled data and high-capacity models requiring several iterative updates across many labeled examples and obtaining millions of labeled examples, is an extreme challenge, there is not enough jump in the neonatal AI applications.

As a side note, accurate labeling always requires physician effort and time, which overcomplicates the current challenges. Unfortunately, there is no established collaboration between physicians and data scientists at a large scale that can ease some of the challenges (data gathering/sharing and labeling). Nonetheless, once these problems are addressed, DL can be used in prevention and diagnosis programs for optimal results, radically transforming clinical practice. In the following, we envision the potential of DL to transform other imaging modalities in the context of neonatology and child health.

The requirement for a massive volume of data is a significant barrier, as mentioned earlier. The quantity of data needed by an AI or ML system can grow in proportion to the sophistication of its underlying architecture; deep neural networks (DNN), for example, have particularly high volume of data needs. It’s not enough that the needed data just be sufficient; they also need to be of good quality in terms of data cleaning and data variability (both ANN and DNN tend to avoid overfitting data if the variability is high). It may be difficult to collect a substantial amount of clean, verified, and varied data for several uses in neonatology. For this reason, there is a data repository shared with neonatal researchers, including EHR 202 and clinical variables. Some approaches for addressing the lack of labeled, annotated, verified, and clean datasets include: (1) building and training a model with a very shallow network (only a few thousand parameters) and (2) data augmentation. Data augmentation techniques are not helpful in the medical imaging field or medical setting 219 .

In the field of neonatal imaging, high-quality labeling and medical imaging data are exceedingly uncommon. One of the other comparable available neonatal datasets the authors are aware of has just ten individuals 166 , 220 , 221 . This pattern holds even in more recent research, as detailed by the majority of studies involving little more than 20 individuals 167 . Regardless of sample size and technology, it is crucial to be able to generalize to new data in the field of image segmentation, especially considering the wide range of MRI contrasts and variations between scanners and sequences between institutions. Moreover, it is generally known that models based on DL have weak generalization skills on unseen data. This is especially crucial for the future translation of research into reality since (1) there is a shift between images obtained in various situations, and (2) the model must be retrained as these images become accessible. Adopting a strategy of continuous learning is the most practical way to handle this challenge. This method involves progressively retraining deep models while preventing any virtual memory loss on previously viewed data sets that may not be available during retraining. This field of endeavor will advance 169 .

Most of the studies did not release their algorithms as open source to the libraries. Even though algorithms are available, it should be known whether separate training and testing datasets exist. There is a strong expectation that studies should have clarified which validation method has been chosen. In terms of comparing algorithm success, reproducibility is a crucial point. Methodological bias is another issue with this system. Research is frequently based on databases and guidelines from other nations that may or may not have patient populations similar to ours 96 . A database that only contains data that is applicable to the specific problem that must be solved; however, obtaining the relevant information may be difficult due to the number of databases.

Lack of explanations and reasoning

The trustworthiness of algorithms is another obstacle 222 . The most widely used deep learning models use a black-box methodology, in which the model simply receives input and outputs a prediction without explaining its thought process. In high-stakes medical settings, this can be dangerous. Some models, on the other hand, incorporate human judgment (human-in-the-loop) or provide interpretability maps or explainability layers to illuminate the decision-making process. Especially in the field of neonatology, where AI is expected to have a significant impact, this trustworthiness is essential for its widespread adoption.

Lack of collaboration efforts (multi-institutions) and privacy concerns

New collaborations have been forged because of this information; early detection and treatment of diseases that affect children, who make up a large portion of the world’s population, will change treatment and follow-up status. Monitoring systems and knowing mortality and treatment activity with multi-site data will help. Considering the necessity for consent to the processing of personal health data by AI systems as an example of a subject related to the protection of privacy and security 96 . Efforts involving multiple institutions can facilitate training, but there are privacy concerns associated with the cross-site sharing of imaging data. Federated learning (FL) was introduced recently to address privacy concerns by facilitating distributed training without the transfer of imaging data 223 . Existing FL techniques utilize conditional reconstruction models to map from under sampled to fully-sampled acquisitions using explicit knowledge of the accelerated imaging operator 223 . Nevertheless, the data from various institutions is typically heterogeneous, which may diminish the efficacy of models trained using federated learning. SplitAVG is proposed as a novel heterogeneity-aware FL method to surmount the performance declines in federated learning caused by data heterogeneity 224 .

While AI has great promise for enhancing healthcare, it also presents significant ethical concerns. Ethical concerns in health AI include informed consent, bias, safety, transparency, patient privacy, and allocation, and their solutions are complicated to negotiate 225 . In neonatology, crucial decision-making is frequently accompanied by a complicated and challenging ethical component. Interdisciplinary approaches are required for progress 226 . The border of viability, life sustaining treatments 227 and the different regulations worldwide made AI utilization in neonatology more complicated. How an ethics framework is implemented in an AI in neonatology has not been reported yet, and there is a need for transparency for trustworthy AI.

The applications of AI in real-world contexts have the potential to result in a few potential benefits, including increased speed of execution; potential reduction in costs, both direct and indirect; improved diagnostic accuracy; increased healthcare delivery efficiency (“algorithms work without a break”); and the potential of supplying access to clinical information even to persons who would not normally be able to utilize healthcare due to geographic or economic constraints 4 .

To achieve an accurate diagnosis, it is planned to limit the number of extra invasive procedures. New DL technologies and easy-to-implement platforms will enable regular and complete follow-up of health data for patients unable to access their records owing to a physician shortage, hence reducing health costs.

The future of neonatal intensive care units and healthcare will likely be profoundly impacted by AI. This article’s objective is to provide neonatologists in the AI era with a reference guide to the information they might require. We defined AI, its levels, its techniques, and the distinctions between the approaches used in the medical field, and we examined the possible advantages, pitfalls, and challenges of AI. While also attempting to present a picture of its potential future implementation in standard neonatal practice. AI and pediatrics require clinicians’ support, and due to the fact that AI researchers with clinicians need to work together and cooperatively. As a result, AI in neonatal care is highly demanded, and there is a fundamental need for a human (pediatrician) to be involved in the AI-backed up applications, in contrast to systems that are more technically advanced and involve fewer healthcare professionals.

Literature review and search strategy

We used PubMed™, IEEEXplore™, Google Scholar™, and ScienceDirect™ to search for publications relating to AI, ML, and DL applications towards neonatology. We have done a varying combination of the keywords (i.e., one from technical keywords and one from clinical keywords) for the search. Clinical keywords were “infant,” “neonate,” “prematurity,” “preterm infant,” “hypoxic ischemic encephalopathy,” “neonatology,” “intraventricular hemorrhage,” “infant brain segmentation,” “NICU mortality,” “infant morbidity,” “ bronchopulmonary dysplasia,” “retinopathy of prematurity.” The inclusion criteria were (i) publication date between 1996–2022 and, (ii) being an artificial intelligence in neonatology study, (iii) written in English, (iv) published in a scholarly peer-reviewed journal, and (v) conducted an assessment of AI applications in neonatology objectively. Technical keywords were AI, DL, ML, and CNN. Review papers, commentaries, letters to the editor and papers with only technical improvement without any clinical background, animal studies, and papers that used statistical models like linear regression, studies written in any language other than English, dissertation thesis, posters, biomarker prediction studies, simulation-based studies, studies with infants are older than 28 days of life, perinatal death, and obstetric care studies were excluded. The preliminary investigation yielded a substantial collection of articles, amounting to approximately 9000 in total. Through a meticulous examination of the abstracts of the papers, a subset of 987 research was found (Fig. 4 ). Ultimately, 106 studies were selected for inclusion in our systematic review (Supplementary file). The evaluation encompassed diverse aspects, including sample size, methodology, data type, evaluation metrics, advantages, and limitations of the studies (Tables 2 – 7 ).

Data availability

Dr. E. Keles and Dr. U. Bagci have full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. All study materials are available from the corresponding author upon reasonable request.

Turing, A.M. & Haugeland, J. In The Turing Test: Verbal Behavior as the Hallmark of Intelligence , 29–56 (1950).

Padula, W. V. et al. Machine learning methods in health economics and outcomes research—the PALISADE checklist: a good practices report of an ISPOR task force. Value Health 25 , 1063–1080 (2022).

Article   PubMed   Google Scholar  

Bagci, U., Irmakci, I., Demir, U. & Keles, E. in AI in Clinical Medicine: A Practical Guide for Healthcare Professionals 56–65 (2023).

Burt, J. R. et al. Deep learning beyond cats and dogs: recent advances in diagnosing breast cancer with deep neural networks. Br. J. Radio. 91 , 20170545 (2018).

Article   Google Scholar  

Piccialli, F., Somma, V. D., Giampaolo, F., Cuomo, S. & Fortino, G. A survey on deep learning in medicine: Why, how and when? Inf. Fusion 66 , 111–137 (2021).

Rubinger, L., Gazendam, A., Ekhtiari, S. & Bhandari, M. Machine learning and artificial intelligence in research and healthcare. Injury 54 , S69–S73 (2023).

Sarker, I. H. Deep Learning: a comprehensive overview on techniques, taxonomy, applications and research directions. SN Comput. Sci. 2 , 420 (2021).

Article   PubMed   PubMed Central   Google Scholar  

Savadjiev, P. et al. Demystification of AI-driven medical image interpretation: past, present and future. Eur. Radio. 29 , 1616–1624 (2019).

Beam, A. L. & Kohane, I. S. Big data and machine learning in health care. JAMA 319 , 1317–1318 (2018).

Janiesch, C., Zschech, P. & Heinrich, K. Machine learning and deep learning. Electron. Mark. 31 , 685–695 (2021).

LeCun, Y., Bengio, Y. & Hinton, G. Deep learning. Nature 521 , 436–444 (2015).

Article   CAS   PubMed   Google Scholar  

Wiens, J. et al. Do no harm: a roadmap for responsible machine learning for health care. Nat. Med . 25 , 1337–1340 (2019).

Chen, P. C., Liu, Y. & Peng, L. How to develop machine learning models for healthcare. Nat. Mater. 18 , 410–414 (2019).

Futoma, J., Simons, M., Panch, T., Doshi-Velez, F. & Celi, L. A. The myth of generalisability in clinical research and machine learning in health care. Lancet Digit. Health 2 , e489–e492 (2020).

Nakaura, T., Higaki, T., Awai, K., Ikeda, O. & Yamashita, Y. A primer for understanding radiology articles about machine learning and deep learning. Diagn. Inter. Imaging 101 , 765–770 (2020).

Mortazi, A. & Bagci, U. Automatically designing CNN architectures for medical image segmentation. in Machine Learning in Medical Imaging: 9th International Workshop, MLMI 2018, Held in Conjunction with MICCAI 2018, Granada, Spain, September 16, 2018 , Proceedings 9 98–106 (Springer, 2018).

Perna, D. & Tagarelli, A. Deep auscultation: predicting respiratory anomalies and diseases via recurrent neural networks. in 2019 IEEE 32nd International Symposium on Computer-Based Medical Systems (CBMS) 50–55 (2019).

Murabito, F. et al. Deep recurrent-convolutional model for automated segmentation of craniomaxillofacial CT scans. in 2020 25th International Conference on Pattern Recognition (ICPR) 9062-9067 (IEEE, 2021).

Aytekin, I. et al. COVID-19 detection from respiratory sounds with hierarchical spectrogram transformers. arXiv https://arxiv.org/abs/2207.09529 (2022).

Ker, J., Wang, L., Rao, J. & Lim, T. Deep learning applications in medical image analysis. IEEE Access 6 , 9375–9389 (2018).

Demir, U. et al. Transformer Based Generative Adversarial Network for Liver Segmentation. in Image Analysis and Processing. ICIAP 2022 Workshops: ICIAP International Workshops, Lecce, Italy, May 23–27, 2022, Revised Selected Papers, Part II 340-347 (Springer, 2022).

Irmakci, I., Unel, Z. E., Ikizler-Cinbis, N. & Bagci, U. Multi-contrast MRI segmentation trained on synthetic images. in 2022 44th Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC) 5030–5034 (IEEE, 2022).

Kim, H. E. et al. Transfer learning for medical image classification: a literature review. BMC Med. Imaging 22 , 69 (2022).

Zhuang, F. et al. A comprehensive survey on transfer learning. Proc. IEEE 109 , 43–76 (2020).

Valverde, J. M. et al. Transfer learning in magnetic resonance brain imaging: a systematic review. J. Imaging 7 , 66 (2021).

Swati, Z. N. K. et al. Content-based brain tumor retrieval for MR images using transfer learning. IEEE Access 7 , 17809–17822 (2019).

LaLonde, R., Xu, Z., Irmakci, I., Jain, S. & Bagci, U. Capsules for biomedical image segmentation. Med. image Anal. 68 , 101889 (2021).

Zhang, X.-M., Liang, L., Liu, L. & Tang, M.-J. Graph neural networks and their current applications in bioinformatics. Front. Genet. 12 , 690049 (2021).

Article   CAS   PubMed   PubMed Central   Google Scholar  

Cheng, Z., Qu, A. & He, X. Contour-aware semantic segmentation network with spatial attention mechanism for medical image. Vis. Comput. 38 , 749–762 (2022).

Gonçalves, T., Rio-Torto, I., Teixeira, L. F. & Cardoso, J. S. A survey on attention mechanisms for medical applications: are we moving towards better algorithms? IEEE Access (2022).

Zhou, J. et al. Graph neural networks: a review of methods and applications. AI Open 1 , 57–81 (2020).

Fout, A., Byrd, J., Shariat, B. & Ben-Hur, A. Protein interface prediction using graph convolutional networks. in Advances in Neural Information Processing Systems 30 (2017).

Khalil, E., Dai, H., Zhang, Y., Dilkina, B. & Song, L. Learning combinatorial optimization algorithms over graphs. in Advances in Neural Information Processing Systems 30 (2017).

Gaggion, N., Mansilla, L., Mosquera, C., Milone, D. H. & Ferrante, E. Improving anatomical plausibility in medical image segmentation via hybrid graph neural networks: applications to chest X-ray analysis. IEEE Trans. Med. Imaging 42 , 546–556 (2023).

Liang, D., Cheng, J., Ke, Z. & Ying, L. Deep magnetic resonance image reconstruction: inverse problems meet neural networks. IEEE Signal Process Mag. 37 , 141–151 (2020).

Dar, S. U. H., Özbey, M., Çatlı, A. B. & Çukur, T. A transfer-learning approach for accelerated MRI using deep neural networks. Magn. Reson Med. 84 , 663–685 (2020).

Güngör, A. et al. Adaptive diffusion priors for accelerated MRI reconstruction. Med. Image Anal. 88 , 102872 (2023).

Monga, V., Li, Y. & Eldar, Y. C. Algorithm unrolling: Interpretable, efficient deep learning for signal and image processing. IEEE Signal Process. Mag. 38 , 18–44 (2021).

Yaman, B. et al. Self-supervised learning of physics-guided reconstruction neural networks without fully sampled reference data. Magn. Reson. Med. 84 , 3172–3191 (2020).

Akata, Z. et al. A research agenda for hybrid intelligence: augmenting human intellect with collaborative, adaptive, responsible, and explainable artificial intelligence. Computer 53 , 18–28 (2020).

RaviPrakash, H. & Anwar, S. M. In AI in Clinical Medicine: A Practical Guide for Healthcare Professionals 94–103 (2023).

Keles, E., Irmakci, I. & Bagci, U. Musculoskeletal MR image segmentation with artificial intelligence. Adv. Clin. Radiol. 4 , 179–188 (2022).

Hussein, S., Cao, K., Song, Q. & Bagci, U. Risk stratification of lung nodules using 3D CNN-based multi-task learning. in International Conference on Information Processing in Medical Imaging 249–260 (Springer, 2017).

Hussein, S., Kandel, P., Bolan, C. W., Wallace, M. B. & Bagci, U. Lung and pancreatic tumor characterization in the deep learning era: novel supervised and unsupervised learning approaches. IEEE Trans. Med. imaging 38 , 1777–1787 (2019).

Topol, E. J. High-performance medicine: the convergence of human and artificial intelligence. Nat. Med. 25 , 44–56 (2019).

Esteva, A. et al. A guide to deep learning in healthcare. Nat. Med. 25 , 24–29 (2019).

Sujith, A. V. L. N., Sajja, G. S., Mahalakshmi, V., Nuhmani, S. & Prasanalakshmi, B. Systematic review of smart health monitoring using deep learning and Artificial intelligence. Neuroscience Informatics 2 , 100028 (2022).

Stewart, J. E., Rybicki, F. J. & Dwivedi, G. Medical specialties involved in artificial intelligence research: is there a leader. Tasman Med. J. 2 , 20–27 (2020).

Google Scholar  

Mesko, B. & Gorog, M. A short guide for medical professionals in the era of artificial intelligence. NPJ Digit Med . 3 , 126 (2020).

Rajpurkar, P., Chen, E., Banerjee, O. & Topol, E. J. AI in health and medicine. Nat. Med. 28 , 31–38 (2022).

Hicks, S. A. et al. On evaluation metrics for medical applications of artificial intelligence. Sci. Rep. 12 , 5979 (2022).

Maier-Hein, L. et al. Metrics reloaded: Pitfalls and recommendations for image analysis validation. Preprint https://arxiv.org/abs/2206.01653 (2022).

McAdams, R. M. et al. Predicting clinical outcomes using artificial intelligence and machine learning in neonatal intensive care units: a systematic review. J. Perinatol. 42 , 1561–1575 (2022).

Kwok, T. N. C. et al. Application and potential of artificial intelligence in neonatal medicine. Semin. Fetal Neonatal Med. 27 , 101346 (2022).

Jeong, H. & Kamaleswaran, R. Pivotal challenges in artificial intelligence and machine learning applications for neonatal care. In Seminars in Fetal and Neonatal Medicine Vol. 27, 101393 (Elsevier, 2022)

Page, M. J. et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 372 , n71 (2021).

McGuinness, L. A. & Higgins, J. P. Risk-of-bias VISualization (robvis): an R package and Shiny web app for visualizing risk-of-bias assessments. Res. Synth. Methods 12 , 55–61 (2021).

Sounderajah, V. et al. A quality assessment tool for artificial intelligence-centered diagnostic test accuracy studies: QUADAS-AI. Nat. Med. 27 , 1663–1665 (2021).

Yang, B. et al. QUADAS-C: a tool for assessing risk of bias in comparative diagnostic accuracy studies. Ann. Intern Med. 174 , 1592–1599 (2021).

SDG Target 3.2: End Preventable Deaths of Newborns and Children under 5 Years of Age in 2021 ( https://www.who.int/data/gho/data/themes/theme-details/GHO/child-health ) (2022).

United Nations General Assembly. Resolution adopted by the General Assembly on 25 September 2015. 70/1. Transforming our world: the 2030 agenda for sustainable development New York, NY ( https://sdgs.un.org/goals ) (2015).

Townsend, D. & Frize, M. Complimentary artificial neural network approaches for prediction of events in the neonatal intensive care unit. In 2008 30th Annual International Conference of the IEEE Engineering in Medicine and Biology Society 4605-4608 (IEEE, 2008).

Ambalavanan, N. et al. Prediction of death for extremely low birth weight neonates. Pediatrics 116 , 1367–1373 (2005).

Nascimento, L. F. C. & Ortega, N. R. S. Fuzzy linguistic model for evaluating the risk of neonatal death. Rev. Saúde. Pública. 36 , 686–692 (2002).

Do, H. J., Moon, K. M. & Jin, H. S. Machine learning models for predicting mortality in 7472 very low birth weight infants using data from a nationwide neonatal network. Diagnostics 12 , 625 (2022).

Podda, M. et al. A machine learning approach to estimating preterm infants survival: development of the Preterm Infants Survival Assessment (PISA) predictor. Sci. Rep. 8 , 13743 (2018).

Mangold, C. et al. Machine learning models for predicting neonatal mortality: a systematic review. Neonatology 118 , 394–405 (2021).

Zernikow, B. et al. Artificial neural network for risk assessment in preterm neonates. Arch. Dis. Child.-Fetal Neonatal Ed. 79 , F129–F134 (1998).

Pearlman, S. A. Advancements in neonatology through quality improvement. J. Perinatol. 42 , 1277–1282 (2022).

Khurshid, F. et al. Comparison of multivariable logistic regression and machine learning models for predicting bronchopulmonary dysplasia or death in very preterm infants. Front Pediatr. 9 , 759776 (2021).

Hsu, J. F. et al. Machine learning algorithms to predict mortality of neonates on mechanical intubation for respiratory failure. Biomedicines 9 , 1377 (2021).

Moreira, A. et al. Development and validation of a mortality prediction model in extremely low gestational age neonates. Neonatology 119 , 418–427 (2022).

Shane, A. L., Sánchez, P. J. & Stoll, B. J. Neonatal sepsis. lancet 390 , 1770–1780 (2017).

Gomez, R., Garcia, N., Collantes, G., Ponce, F. & Redon, P. Development of a non-invasive procedure to early detect neonatal sepsis using HRV monitoring and machine learning algorithms. in 2019 IEEE 32nd International Symposium on Computer-Based Medical Systems (CBMS) 132–137 (2019).

Stocker, M. et al. Machine learning used to compare the diagnostic accuracy of risk factors, clinical signs and biomarkers and to develop a new prediction model for neonatal early-onset sepsis. Pediatr. Infect. Dis. J. 41 , 248–254 (2022).

Manuck, T. A. et al. Preterm neonatal morbidity and mortality by gestational age: a contemporary cohort. Am. J. Obstet. Gynecol. 215 , 103.e101–103.e114 (2016).

Volpe, J. J. Brain injury in premature infants: a complex amalgam of destructive and developmental disturbances. Lancet Neurol. 8 , 110–124 (2009).

Johnson, S. et al. Neurodevelopmental disability through 11 years of age in children born before 26 weeks of gestation. Pediatrics 124 , e249–e257 (2009).

Ment, L. R., Hirtz, D. & Hüppi, P. S. Imaging biomarkers of outcome in the developing preterm brain. Lancet Neurol. 8 , 1042–1055 (2009).

Ophelders, D. et al. Preterm brain injury, antenatal triggers, and therapeutics: timing is key. Cells 9 , 1871 (2020).

Rogers, C. E., Lean, R. E., Wheelock, M. D. & Smyser, C. D. Aberrant structural and functional connectivity and neurodevelopmental impairment in preterm children. J. Neurodev. Disord. 10 , 1–13 (2018).

Article   CAS   Google Scholar  

Smyser, C. D. et al. Resting-state network complexity and magnitude are reduced in prematurely born infants. Cereb. Cortex 26 , 322–333 (2016).

Vohr, B. R. Neurodevelopmental outcomes of premature infants with intraventricular hemorrhage across a lifespan. Semin. Perinatol. 46 , 151594 (2022).

Zernikow, B. et al. Artificial neural network for predicting intracranial haemorrhage in preterm neonates. Acta Paediatr. 87 , 969–975 (1998).

Turova, V. et al. Machine learning models for identifying preterm infants at risk of cerebral hemorrhage. PLoS ONE 15 , e0227419 (2020).

Keunen, K., Counsell, S. J. & Benders, M. J. The emergence of functional architecture during early brain development. Neuroimage 160 , 2–14 (2017).

Sripada, K. et al. Trajectories of brain development in school-age children born preterm with very low birth weight. Sci. Rep. 8 , 15553 (2018).

Smyser, C. D. et al. Prediction of brain maturity in infants using machine-learning algorithms. Neuroimage 136 , 1–9 (2016).

Gao, W., Lin, W., Grewen, K. & Gilmore, J. H. Functional connectivity of the infant human brain: plastic and modifiable. Neuroscientist 23 , 169–184 (2017).

Ball, G. et al. Machine-learning to characterise neonatal functional connectivity in the preterm brain. Neuroimage 124 , 267–275 (2016).

Chiarelli, A. M., Sestieri, C., Navarra, R., Wise, R. G. & Caulo, M. Distinct effects of prematurity on MRI metrics of brain functional connectivity, activity, and structure: Univariate and multivariate analyses. Hum. Brain Mapp. 42 , 3593–3607 (2021).

Shang, J. et al. A machine learning investigation of volumetric and functional MRI abnormalities in adults born preterm. Hum. Brain Mapp. 40 , 4239–4252 (2019).

Zimmer, V. A. et al. Learning and combining image neighborhoods using random forests for neonatal brain disease classification. Med. Image Anal. 42 , 189–199 (2017).

Song, Z., Awate, S. P., Licht, D. J. & Gee, J. C. Clinical neonatal brain MRI segmentation using adaptive nonparametric data models and intensity-based Markov priors. In International Conference on Medical Image Computing and Computer-assisted Intervention 883–890 (Springer, 2007).

Vassar, R. et al. Neonatal brain microstructure and machine-learning-based prediction of early language development in children born very preterm. Pediatr. Neurol. 108 , 86–92 (2020).

Schadl, K. et al. Prediction of cognitive and motor development in preterm children using exhaustive feature selection and cross-validation of near-term white matter microstructure. Neuroimage Clin. 17 , 667–679 (2018).

Wee, C. Y. et al. Neonatal neural networks predict children behavioral profiles later in life. Hum. Brain Mapp. 38 , 1362–1373 (2017).

Li, Y. et al. Brain connectivity based graph convolutional networks and its application to infant age prediction. IEEE Trans. Med Imaging 41 , 2764–2776 (2022).

Liu, M. et al. Deep learning of cortical surface features using graph-convolution predicts neonatal brain age and neurodevelopmental outcome. in 2020 IEEE 17th International Symposium on Biomedical Imaging (ISBI) 1335–1338 (IEEE, 2020).

Krishnan, M. L. et al. Machine learning shows association between genetic variability in PPARG and cerebral connectivity in preterm infants. Proc. Natl Acad. Sci. USA 114 , 13744–13749 (2017).

Karayiannis, N. B. et al. Automated detection of videotaped neonatal seizures of epileptic origin. Epilepsia 47 , 966–980 (2006).

Koolen, N. et al. Automated classification of neonatal sleep states using EEG. Clin. Neurophysiol. 128 , 1100–1108 (2017).

Mathieson, S. R. et al. Validation of an automated seizure detection algorithm for term neonates. Clin. Neurophysiol. 127 , 156–168 (2016).

Temko, A., Lightbody, G., Thomas, E. M., Boylan, G. B. & Marnane, W. Instantaneous measure of EEG channel importance for improved patient-adaptive neonatal seizure detection. IEEE Trans. Biomed. Eng. 59 , 717–727 (2012).

Temko, A., Thomas, E., Marnane, W., Lightbody, G. & Boylan, G. B. Performance assessment for EEG-based neonatal seizure detectors. Clin. Neurophysiol. 122 , 474–482 (2011).

Matic, V. et al. Improving reliability of monitoring background EEG dynamics in asphyxiated infants. IEEE Trans. Biomed. Eng. 63 , 973–983 (2016).

Ahmed, R., Temko, A., Marnane, W. P., Boylan, G. & Lightbody, G. Exploring temporal information in neonatal seizures using a dynamic time warping based SVM kernel. Comput Biol. Med. 82 , 100–110 (2017).

Thomas, E., Greene, B., Lightbody, G., Marnane, W. & Boylan, G. Seizure detection in neonates: improved classification through supervised adaptation. in 2008 30th Annual International Conference of the IEEE Engineering in Medicine and Biology Society 903-906 (IEEE, 2008).

Ansari, A. H. et al. Improvement of an automated neonatal seizure detector using a post-processing technique. In 2015 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) 5859–5862 (IEEE, 2015).

Schetinin, V. & Schult, J. The combined technique for detection of artifacts in clinical electroencephalograms of sleeping newborns. IEEE Trans. Inf. Technol. Biomed. 8 , 28–35 (2004).

Mohseni, H.R., Mirghasemi, H., Shamsollahi, M.B. & Zamani, M.R. Detection of rhythmic discharges in newborn EEG signals. in 2006 International Conference of the IEEE Engineering in Medicine and Biology Society 6577–6580 (IEEE, 2006).

Simayijiang, Z., Backman, S., Ulén, J., Wikström, S. & Åström, K. Exploratory study of EEG burst characteristics in preterm infants. in 2013 35th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) 4295–4298 (IEEE, 2013).

Navarro, X. et al. Multi-feature classifiers for burst detection in single EEG channels from preterm infants. J. Neural Eng. 14 , 046015 (2017).

Ahmed, R., Temko, A., Marnane, W., Lightbody, G. & Boylan, G. Grading hypoxic-ischemic encephalopathy severity in neonatal EEG using GMM supervectors and the support vector machine. Clin. Neurophysiol. 127 , 297–309 (2016).

Temko, A., Boylan, G., Marnane, W. & Lightbody, G. Robust neonatal EEG seizure detection through adaptive background modeling. Int. J. neural Syst. 23 , 1350018 (2013).

Stevenson, N. et al. An automated system for grading EEG abnormality in term neonates with hypoxic-ischaemic encephalopathy. Ann. Biomed. Eng. 41 , 775–785 (2013).

Clyman, R. I. Mechanisms regulating the ductus arteriosus. Biol. Neonate 89 , 330–335 (2006).

Sellmer, A. et al. Morbidity and mortality in preterm neonates with patent ductus arteriosus on day 3. Arch. Dis. Child Fetal Neonatal Ed. 98 , F505–510 (2013).

El-Khuffash, A., Rios, D. R. & McNamara, P. J. Toward a rational approach to patent ductus arteriosus trials: selecting the population of interest. J. Pediatr. 233 , 11–13 (2021).

de Waal, K., Phad, N., Stubbs, M., Chen, Y. & Kluckow, M. A randomized placebo-controlled pilot trial of early targeted nonsteroidal anti-inflammatory drugs in preterm infants with a patent ductus arteriosus. J. Pediatr. 228 , 82–86.e82 (2021).

El-Khuffash, A. et al. A pilot randomized controlled trial of early targeted patent ductus arteriosus treatment using a risk based severity score (The PDA RCT). J. Pediatr. 229 , 127–133 (2021).

Sung, S. I., Lee, M. H., Ahn, S. Y., Chang, Y. S. & Park, W. S. Effect of nonintervention vs oral ibuprofen in patent ductus arteriosus in preterm infants: a randomized clinical trial. JAMA Pediatr. 174 , 755–763 (2020).

Na, J. Y. et al. Artificial intelligence model comparison for risk factor analysis of patent ductus arteriosus in nationwide very low birth weight infants cohort. Sci. Rep. 11 , 22353 (2021).

Gomez-Quintana, S. et al. A framework for AI-assisted detection of patent ductus arteriosus from neonatal phonocardiogram. Healthcare 9 , 169 (2021).

Verder, H. et al. Bronchopulmonary dysplasia predicted at birth by artificial intelligence. Acta Paediatr. 110 , 503–509 (2021).

Ochab, M. & Wajs, W. Expert system supporting an early prediction of the bronchopulmonary dysplasia. Comput Biol. Med 69 , 236–244 (2016).

Dai, D. et al. Bronchopulmonary dysplasia predicted by developing a machine learning model of genetic and clinical information. Front Genet 12 , 689071 (2021).

Mikhno, A. & Ennett, C.bM. Prediction of extubation failure for neonates with respiratory distress syndrome using the MIMIC-II clinical database. in 2012 Annual international conference of the IEEE Engineering in Medicine and Biology Society 5094–5097 (IEEE, 2012).

Precup, D. et al. Prediction of extubation readiness in extreme preterm infants based on measures of cardiorespiratory variability. in 2012 Annual international conference of the IEEE Engineering in Medicine and Biology Society 5630–5633 (IEEE, 2012).

Mueller, M. et al. Predicting extubation outcome in preterm newborns: a comparison of neural networks with clinical expertise and statistical modeling. Pediatr. Res 56 , 11–18 (2004).

Hatzakis, G. E. & Davis, G. M. Fuzzy logic controller for weaning neonates from mechanical ventilation. in Proceedings of the AMIA Symposium 315 (American Medical Informatics Association, 2002).

Barrero-Castillero, A., Corwin, B. K., VanderVeen, D. K. & Wang, J. C. Workforce shortage for retinopathy of prematurity care and emerging role of telehealth and artificial intelligence. Pediatr. Clin. North Am. 67 , 725–733 (2020).

Rani, P. & Rajkumar, E. R. Classification of retinopathy of prematurity using back propagation neural network. Int. J. Biomed. Eng. Technol. 22 , 338–348 (2016).

Ataer-Cansizoglu, E. et al. Computer-based image analysis for plus disease diagnosis in retinopathy of prematurity: performance of the “i-ROP” system and image features associated with expert diagnosis. Transl. Vis. Sci. Technol. 4 , 5 (2015).

Reed, N. E., Gini, M., Johnson, P. E. & Moller, J. H. Diagnosing congenital heart defects using the Fallot computational model. Artif. Intell. Med. 10 , 25–40 (1997).

Li, L. et al. The use of fuzzy backpropagation neural networks for the early diagnosis of hypoxic ischemic encephalopathy in newborns. J. Biomed. Biotechnol. 2011 , 349490 (2011).

Taylor, J. A. et al. Use of a Smartphone App to Assess Neonatal Jaundice. Pediatrics 140 , e20170312 (2017).

Ferreira, D., Oliveira, A. & Freitas, A. Applying data mining techniques to improve diagnosis in neonatal jaundice. BMC Med. Inform. Decis. Mak. 12 , 1–6 (2012).

Ji, J. et al. A data-driven algorithm integrating clinical and laboratory features for the diagnosis and prognosis of necrotizing enterocolitis. PLoS ONE 9 , e89860 (2014).

Ambalavanan, N. et al. Prediction of neurologic morbidity in extremely low birth weight infants. J. Perinatol. 20 , 496–503 (2000).

Soleimani, F., Teymouri, R. & Biglarian, A. Predicting developmental disorder in infants using an artificial neural network. Acta Med. Iran. 51 , 347–352 (2013).

PubMed   Google Scholar  

Bartz-Kurycki, M. A. et al. Enhanced neonatal surgical site infection prediction model utilizing statistically and clinically significant variables in combination with a machine learning algorithm. Am. J. Surg. 216 , 764–777 (2018).

Reed, R. A. et al. Machine-learning vs. expert-opinion driven logistic regression modelling for predicting 30-day unplanned rehospitalisation in preterm babies: a prospective, population-based study (EPIPAGE 2). Front Pediatr. 8 , 585868 (2020).

Tsien, C. L., Kohane, I. S. & McIntosh, N. Multiple signal integration by decision tree induction to detect artifacts in the neonatal intensive care unit. Artif. Intell. Med. 19 , 189–202 (2000).

Cabrera-Quiros, L. et al. Prediction of late-onset sepsis in preterm infants using monitoring signals and machine learning. Crit. Care Explor. 3 , e0302 (2021).

Saria, S., Rajani, A. K., Gould, J., Koller, D. & Penn, A. A. Integration of early physiological responses predicts later illness severity in preterm infants. Sci. Transl. Med. 2 , 48ra65–48ra65 (2010).

Jalali, A., Simpao, A. F., Galvez, J. A., Licht, D. J. & Nataraj, C. Prediction of periventricular leukomalacia in neonates after cardiac surgery using machine learning algorithms. J. Med. Syst. 42 , 177 (2018).

Aucouturier, J. J., Nonaka, Y., Katahira, K. & Okanoya, K. Segmentation of expiratory and inspiratory sounds in baby cry audio recordings using hidden Markov models. J. Acoust. Soc. Am. 130 , 2969–2977 (2011).

Cano Ortiz, S. D., Escobedo Beceiro, D. I. & Ekkel, T. A radial basis function network oriented for infant cry classification. in Iberoamerican Congress on Pattern Recognition 374–380 (Springer, 2004).

Yassin, I. et al. Infant asphyxia detection using autoencoders trained on locally linear embedded-reduced Mel Frequency Cepstrum Coefficient (MFCC) features. J. Fundam. Appl. Sci. 9 , 716–729 (2017).

Hsu, K. P. et al. A newborn screening system based on service-oriented architecture embedded support vector machine. J. Med. Syst. 34 , 899–907 (2010).

Baumgartner, C. et al. Supervised machine learning techniques for the classification of metabolic disorders in newborns. Bioinformatics 20 , 2985–2996 (2004).

Chen, W. H. et al. Web-based newborn screening system for metabolic diseases: machine learning versus clinicians. J. Med. Internet Res . 15 , e98 (2013).

Zhang, W. et al. Deep convolutional neural networks for multi-modality isointense infant brain image segmentation. Neuroimage 108 , 214–224 (2015).

Hyun, D. & Brickson, L. Classification of Neonatal Brain Ultrasound Scans Using Deep Convolutional Neural Networks . (Stanford CS229, 2016).

Kelly, C. et al. Investigating brain structural maturation in children and adolescents born very preterm using the brain age framework. Neuroimage 247 , 118828 (2022).

Kim, K. Y., Nowrangi, R., McGehee, A., Joshi, N. & Acharya, P. T. Assessment of germinal matrix hemorrhage on head ultrasound with deep learning algorithms. Pediatr. Radio. 52 , 533–538 (2022).

Lei, H., Ashrafi, A., Chang, P., Chang, A. & Lai, W. Patent ductus arteriosus (PDA) detection in echocardiograms using deep learning. Intelligence-Based Med. 6 , 100054 (2022).

Li, H. et al. Automatic segmentation of diffuse white matter abnormality on T2-weighted brain MR images using deep learning in very preterm infants. Radio. Artif. Intell. 3 , e200166 (2021).

Ding, W., Abdel-Basset, M., Hawash, H. & Pedrycz, W. Multimodal infant brain segmentation by fuzzy-informed deep learning. IEEE Trans. Fuzzy Syst. 30 , 1088–1101 (2022).

Mostapha, M. & Styner, M. Role of deep learning in infant brain MRI analysis. Magn. Reson Imaging 64 , 171–189 (2019).

Makropoulos, A. et al. Automatic tissue and structural segmentation of neonatal brain MRI using expectation-maximization. MICCAI Gd. Chall. Neonatal Brain Segment. 2012 , 9–15 (2012).

Beare, R. J. et al. Neonatal brain tissue classification with morphological adaptation and unified segmentation. Front. Neuroinform. 10 , 12 (2016).

Liu, M. et al. Patch-based augmentation of Expectation–Maximization for brain MRI tissue segmentation at arbitrary age after premature birth. NeuroImage 127 , 387–408 (2016).

Moeskops, P. et al. Automatic segmentation of MR brain images of preterm infants using supervised classification. NeuroImage 118 , 628–641 (2015).

Weisenfeld, N. I. & Warfield, S. K. Automatic segmentation of newborn brain MRI. NeuroImage 47 , 564–572 (2009).

Kim, H., Lepage, C., Evans, A. C., Barkovich, A. J. & Xu, D. NEOCIVET: Extraction of cortical surface and analysis of neonatal gyrification using a modified CIVET pipeline. in Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015 (eds. Navab, N., Hornegger, J., Wells, W. M. & Frangi, A. F.) 571–579 (Springer International Publishing, 2015).

Wang, L. et al. 4D Multi-modality tissue segmentation of serial infant images. PLoS ONE 7 , e44596 (2012).

Ding, Y. et al. Using deep convolutional neural networks for neonatal brain image segmentation. Front Neurosci. 14 , 207 (2020).

Greenbury, S. F. et al. Identification of variation in nutritional practice in neonatal units in England and association with clinical outcomes using agnostic machine learning. Sci. Rep. 11 , 7178 (2021).

Brown, J. M. et al. Automated diagnosis of plus disease in retinopathy of prematurity using deep convolutional neural networks. JAMA Ophthalmol. 136 , 803–810 (2018).

Taylor, S. et al. Monitoring disease progression with a quantitative severity scale for retinopathy of prematurity using deep learning. JAMA Ophthalmol. 137 , 1022–1028 (2019).

Campbell, J. P. et al. Applications of artificial intelligence for retinopathy of prematurity screening. Pediatrics 147 , e2020016618 (2021).

Wang, B. et al. Application of a deep convolutional neural network in the diagnosis of neonatal ocular fundus hemorrhage. Biosci. Rep. 38 , BSR20180497 (2018).

Shabanian, M., Eckstein, E. C., Chen, H. & DeVincenzo, J. P. Classification of neurodevelopmental age in normal infants using 3D-CNN based on brain MRI. in 2019 IEEE International Conference on Bioinformatics and Biomedicine (BIBM) 2373–2378 (IEEE, 2019).

Saha, S. et al. Predicting motor outcome in preterm infants from very early brain diffusion MRI using a deep learning convolutional neural network (CNN) model. Neuroimage 215 , 116807 (2020).

He, L. et al. A multi-task, multi-stage deep transfer learning model for early prediction of neurodevelopment in very preterm infants. Sci. Rep. 10 , 15072 (2020).

Temple, M. W., Lehmann, C. U. & Fabbri, D. Predicting discharge dates from the NICU using progress note data. Pediatrics 136 , e395–405 (2015).

Wang, J. et al. Automated retinopathy of prematurity screening using deep neural networks. EBioMedicine 35 , 361–368 (2018).

Valikodath, N., Cole, E., Chiang, M. F., Campbell, J. P. & Chan, R. V. P. Imaging in retinopathy of prematurity. Asia Pac. J. Ophthalmol. 8 , 178–186 (2019).

CAS   Google Scholar  

Biten, H. et al. Diagnostic accuracy of ophthalmoscopy vs telemedicine in examinations for retinopathy of prematurity. JAMA Ophthalmol. 136 , 498–504 (2018).

Chiang, M. F. et al. Detection of clinically significant retinopathy of prematurity using wide-angle digital retinal photography: a report by the american academy of ophthalmology. Ophthalmology 119 , 1272–1280 (2012).

Ednick, M. et al. A review of the effects of sleep during the first year of life on cognitive, psychomotor, and temperament development. Sleep 32 , 1449–1458 (2009).

Ansari, A. H. et al. Quiet sleep detection in preterm infants using deep convolutional neural networks. J. Neural Eng. 15 , 066006 (2018).

Ansari, A. H. et al. A deep shared multi-scale inception network enables accurate neonatal quiet sleep detection with limited EEG. Channels IEEE J. Biomed. Health Inf. 26 , 1023–1033 (2022).

Werth, J., Radha, M., Andriessen, P., Aarts, R. M. & Long, X. Deep learning approach for ECG-based automatic sleep state classification in preterm infants. Biomed. Signal Process. Control 56 , 101663 (2020).

Hauptmann, A., Arridge, S., Lucka, F., Muthurangu, V. & Steeden, J. A. Real-time cardiovascular MR with spatio-temporal artifact suppression using deep learning-proof of concept in congenital heart disease. Magn. Reson Med. 81 , 1143–1156 (2019).

Ceschin, R. et al. A computational framework for the detection of subcortical brain dysmaturation in neonatal MRI using 3D Convolutional Neural Networks. Neuroimage 178 , 183–197 (2018).

Ornek, A. H. & Ceylan, M. Explainable artificial intelligence (XAI): classification of medical thermal images of neonates using class activation maps. Trait. Signal 38 , 1271–1279 (2021).

Ervural, S. & Ceylan, M. Classification of neonatal diseases with limited thermal Image data. Multimed. Tools Appl. 81 , 9247–9275 (2021).

Ervural, S. & Ceylan, M. Thermogram classification using deep siamese network for neonatal disease detection with limited data. Quant. InfraRed Thermogr. J. 19 , 312–330 (2022).

Ervural, S. & Ceylan, M. Convolutional neural networks-based approach to detect neonatal respiratory system anomalies with limited thermal image. Trait. Signal 38 , 437–442 (2021).

Xu, S. et al. Wireless skin sensors for physiological monitoring of infants in low-income and middle-income countries. Lancet Digit. Health 3 , e266–e273 (2021).

Hoshino, E. et al. An iPhone application using a novel stool color detection algorithm for biliary atresia screening. Pediatr. Surg. Int. 33 , 1115–1121 (2017).

Dong, Y. et al. Artificial intelligence algorithm-based computed tomography images in the evaluation of the curative effect of enteral nutrition after neonatal high intestinal obstruction operation. J. Health. Eng. 2021 , 7096286 (2021).

Mooney, C. et al. Predictive modelling of hypoxic ischaemic encephalopathy risk following perinatal asphyxia. Heliyon 7 , e07411 (2021).

Pavel, A. M. et al. A machine-learning algorithm for neonatal seizure recognition: a multicentre, randomised, controlled trial. Lancet Child Adolesc. Health 4 , 740–749 (2020).

Mathieson, S. et al. In-depth performance analysis of an EEG based neonatal seizure detection algorithm. Clin. Neurophysiol. 127 , 2246–2256 (2016).

Moeskops, P. et al. Prediction of cognitive and motor outcome of preterm infants based on automatic quantitative descriptors from neonatal MR brain images. Sci. Rep. 7 , 2163 (2017).

Bahado-Singh, R. O. et al. Precision cardiovascular medicine: artificial intelligence and epigenetics for the pathogenesis and prediction of coarctation in neonates. J. Matern Fetal Neonatal Med 35 , 457–464 (2022).

Sentner, T. et al. The Sleep Well Baby project: an automated real-time sleep-wake state prediction algorithm in preterm infants. Sleep 45 , zsac143 (2022).

Sirota, M. et al. Enabling precision medicine in neonatology, an integrated repository for preterm birth research. Sci. Data 5 , 180219 (2018).

Young, J., Macke, C. J. & Tsoukalas, L. H. Short-term acoustic forecasting via artificial neural networks for neonatal intensive care units. J. Acoust. Soc. Am. 132 , 3234–3239 (2012).

Reis, M., Ortega, N. & Silveira, P. S. P. Fuzzy expert system in the prediction of neonatal resuscitation. Braz. J. Med. Biol. Res. 37 , 755–764 (2004).

Saadah, L. M. et al. Palivizumab prophylaxis during nosocomial outbreaks of respiratory syncytial virus in a neonatal intensive care unit: predicting effectiveness with an artificial neural network model. Pharmacotherapy 34 , 251–259 (2014).

Kakarmath, S. et al. Best practices for authors of healthcare-related artificial intelligence manuscripts. NPJ Digit Med. 3 , 134 (2020).

Plana, D. et al. Randomized clinical trials of machine learning interventions in health care: a systematic review. JAMA Netw. Open 5 , e2233946–e2233946 (2022).

Caparros-Gonzalez, R. A., de la Torre-Luque, A., Diaz-Piedra, C., Vico, F. J. & Buela-Casal, G. Listening to relaxing music improves physiological responses in premature infants: a randomized controlled trial. Adv. Neonatal Care 18 , 58–69 (2018).

Pillai Riddell, R. & Fabrizi, L Rebooting Infant Pain Assessment: Using Machine Learning to Exponentially Improve Neonatal Intensive Care Unit Practice (BabyAI) ClinicalTrials.gov Identifier: NCT05579496. https://clinicaltrials.gov/study/NCT05579496?id=NCT05579496%20&rank=1#more-information , https://www.yorku.ca/lamarsh/rebooting-infant-pain-assessment-using-machine-learning-to-exponentially-improveneonatal-intensive-care-unit-practice (2022).

Roue, J. M., Morag, I., Haddad, W. M., Gholami, B. & Anand, K. J. S. Using sensor-fusion and machine-learning algorithms to assess acute pain in non-verbal infants: a study protocol. BMJ Open 11 , e039292 (2021).

Shalish, W. et al. Prediction of Extubation readiness in extremely preterm infants by the automated analysis of cardiorespiratory behavior: study protocol. BMC Pediatr. 17 , 167 (2017).

Janvier, A., et al. The ethics of family integrated care in the NICU: Improving care for families without causing harm. Seminars in Perinatology 46 , 151528 (2022).

Waddington, C., van Veenendaal, N. R., O’Brien, K. & Patel, N. Family integrated care: Supporting parents as primary caregivers in the neonatal intensive care unit. Pediatr. Investig. 5 , 148–154 (2021).

Morton, C. E., Smith, S. F., Lwin, T., George, M. & Williams, M. Computer programming: should medical students be learning it? JMIR Med. Educ. 5 , e11940 (2019).

Acosta, J. N., Falcone, G. J., Rajpurkar, P. & Topol, E. J. Multimodal biomedical AI. Nat. Med. 28 , 1773–1784 (2022).

Ahuja, A. S. The impact of artificial intelligence in medicine on the future role of the physician. PeerJ 7 , e7702 (2019).

Han, E.-R. et al. Medical education trends for future physicians in the era of advanced technology and artificial intelligence: an integrative review. BMC Med. Educ. 19 , 460 (2019).

Lozano, P. M. et al. Training the next generation of learning health system scientists. Learn. Health Syst. 6 , e10342 (2022).

Kawahara, J. et al. BrainNetCNN: convolutional neural networks for brain networks; towards predicting neurodevelopment. NeuroImage 146 , 1038–1049 (2017).

Alexander, B. et al. A new neonatal cortical and subcortical brain atlas: the Melbourne Children’s Regional Infant Brain (M-CRIB) atlas. NeuroImage 147 , 841–851 (2017).

Prastawa, M., Gilmore, J. H., Lin, W. & Gerig, G. Automatic segmentation of MR images of the developing newborn brain. Med. Image Anal. 9 , 457–466 (2005).

Cutillo, C. M. et al. Machine intelligence in healthcare—perspectives on trustworthiness, explainability, usability, and transparency. npj Digit. Med. 3 , 47 (2020).

Elmas, G. et al. Federated learning of generative image priors for MRI reconstruction. IEEE Trans. Med. Imaging 42 , 1996–2009 (2022).

Zhang, M., Qu, L., Singh, P., Kalpathy-Cramer, J. & Rubin, D. L. SplitAVG: a heterogeneity-aware federated deep learning method for medical imaging. IEEE J. Biomed. Health Inf. 26 , 4635–4644 (2022).

Katznelson, G. & Gerke, S. The need for health AI ethics in medical school education. Adv. Health Sci. Educ. 26 , 1447–1458 (2021).

Mercurio, M. R. & Cummings, C. L. Critical decision-making in neonatology and pediatrics: the I–P–O framework. J. Perinatol. 41 , 173–178 (2021).

Lin, M., Vitcov, G. G. & Cummings, C. L. Moral equivalence theory in neonatology. Semin. Perinatol. 46 , 151525 (2022).

Porcelli, P. J. & Rosenbloom, S. T. Comparison of new modeling methods for postnatal weight in ELBW infants using prenatal and postnatal data. J. Pediatr. Gastroenterol. Nutr. 59 , e2–8 (2014).

Temple, M. W., Lehmann, C. U. & Fabbri, D. Natural language processing for cohort discovery in a discharge prediction model for the neonatal ICU. Appl Clin. Inf. 7 , 101–115 (2016).

Moeskops, P. et al. Automatic segmentation of MR brain images with a convolutional neural network. IEEE Trans. Med. Imaging 35 , 1252–1261 (2016).

Download references

Acknowledgements

This work is partially supported by the NIH NCI funding: R01-CA246704 and R01-CA240639. Dr. E Keles is working as a senior clinical research associate in the Machine and Hybrid Intelligence Lab at the Northwestern University Feinberg School of Medicine, Department of Radiology. Dr. U. Bagci is director of the Machine and Hybrid Intelligence Lab and Associate Professor at the Department of Radiology, Northwestern University, Feinberg School of Medicine.

Author information

Authors and affiliations.

Northwestern University, Feinberg School of Medicine, Department of Radiology, Chicago, IL, USA

Elif Keles & Ulas Bagci

Northwestern University, Department of Biomedical Engineering, Chicago, IL, USA

Department of Electrical and Computer Engineering, Chicago, IL, USA

You can also search for this author in PubMed   Google Scholar

Contributions

Both authors contributed to the review design, data collection, interpretation of the data, analysis of data and drafting the report.

Corresponding author

Correspondence to Elif Keles .

Ethics declarations

Competing interests.

Dr. E. Keles has no COI. Dr. U. Bagci discloses Ther-AI LLC.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Keles, E., Bagci, U. The past, current, and future of neonatal intensive care units with artificial intelligence: a systematic review. npj Digit. Med. 6 , 220 (2023). https://doi.org/10.1038/s41746-023-00941-5

Download citation

Received : 29 January 2023

Accepted : 05 October 2023

Published : 27 November 2023

DOI : https://doi.org/10.1038/s41746-023-00941-5

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing: Translational Research newsletter — top stories in biotechnology, drug discovery and pharma.

•   OpenBU
• Theses & Dissertations
• Boston University Theses & Dissertations

Show simple item record

Reducing stress in the neonatal intensive care unit:an occupational therapy approach to preterm infant massage

This item appears in the following collection(s).

• Boston University Theses & Dissertations [9207]

Deposit Materials

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here .

Loading metrics

Open Access

Peer-reviewed

Research Article

Treatment outcome of neonatal sepsis and associated factors among neonates admitted to neonatal intensive care unit in public hospitals, Addis Ababa, Ethiopia, 2021. Multi-center cross-sectional study

Roles Data curation, Investigation, Writing – original draft

Affiliation Yekatit 12 Hospital, Addis Ababa, Ethiopia

Contributed equally to this work with: Tefera Mulugeta, Fikertemariam Abebe, Yitayal Guadie, Dires Birhanu, Esmelealem Mihretu

Roles Software, Supervision

Affiliation College of Health Science, Addis Ababa University, Addis Ababa, Ethiopia

Roles Methodology, Visualization, Writing – review & editing

Roles Data curation, Formal analysis, Methodology, Validation

Affiliation Colleges of Medicine and Health Science, Debre Markos University, Debre Markos, Ethiopia

Roles Methodology, Software, Validation

Roles Data curation, Formal analysis, Software, Writing – review & editing

Affiliation College of Health Science, Dila University, Dilla, Ethiopia

Roles Conceptualization, Methodology, Writing – review & editing

* E-mail: [email protected]

• Asalef Endazanaw, 
• Tefera Mulugeta, 
• Fikertemariam Abebe, 
• Yohannes Godie, 
• Yitayal Guadie, 
• Dires Birhanu, 
• Esmelealem Mihretu

• Published: May 30, 2023
• https://doi.org/10.1371/journal.pone.0284983
• Peer Review
• Reader Comments

Globally, neonatal sepsis is the leading cause of neonatal mortality and morbidity, particularly in developing countries. Despite studies that revealed the prevalence of neonatal sepsis in developing countries, the outcome of the diseases, barriers for poor outcomes were inconclusive. The aim of this study was to assess the treatment outcome of neonatal sepsis and its associated factors among neonates admitted to neonatal intensive care unit in public hospitals, Addis Ababa, Ethiopia, 2021.

A cross-sectional study was carried out from February 15 to May 10, 2021 on 308 neonates admitted to neonatal intensive care units of Addis Ababa city public hospitals. Hospitals and study participants were selected by lottery and systematic random sampling techniques, respectively. Data were collected through face-to-face interviews with a structured, pretested questionnaire and by reviewing both the maternal and newborn profile cards. Epi-data version 4.6 was used to enter the collected data, which was then exported to SPSS version 26 for analysis. The 95% CI odds ratio is used to determine the direction and strength of the association between the dependent and independent variables.

Among the total study 308 neonates, 75(24.4%) were died. Regarding the poor treatment outcome of neonatal sepsis, neonates whose mothers <37 weeks of gestational age (AOR = 4.87, 95% CI: 1.23–19.22), Grunting (AOR 6.94: 1.48–32.54), Meconium amniotic stained (AOR = 3.03, 95% CI: 1.02–9.01), Duration of rupture of membrane >18hours (AOR = 3.66, 95% CI: (1.20–11.15), Hypertensive PIH/ Eclampsia (AOR = 3.54, 95% CI: 1.24–10.09), Meropenum (AOR = 4.16, 95% CI: 1.22–14.21) and CRP positive result (AOR = 5.87, 95% CI: 1.53–22.56) were significantly associated with poor treatment outcome of neonatal sepsis.

Conclusion and recommendation

The treatment outcomes of neonates were 75.6% recovered and 24.4% died. In this setting, empirical treatment was the cornerstone for managing neonatal sepsis. Professionals who are working in labor and delivery ward screened for mothers preeclampsia and duration of rupture of membrane >18hrs /PROM/ treated with antihypertensive drug and antibiotics for the prevention of neonatal sepsis.

Citation: Endazanaw A, Mulugeta T, Abebe F, Godie Y, Guadie Y, Birhanu D, et al. (2023) Treatment outcome of neonatal sepsis and associated factors among neonates admitted to neonatal intensive care unit in public hospitals, Addis Ababa, Ethiopia, 2021. Multi-center cross-sectional study. PLoS ONE 18(5): e0284983. https://doi.org/10.1371/journal.pone.0284983

Editor: Sanjoy Kumer Dey, Bangabandhu Sheikh Mujib Medical University (BSMMU), BANGLADESH

Received: August 1, 2022; Accepted: April 12, 2023; Published: May 30, 2023

Copyright: © 2023 Endazanaw et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All relevant data are within the paper.

Funding: The author(s) received no specific funding for this work.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Neonatal sepsis is characterized as an infectious disease syndrome with a clinically suspected or culture-confirmed infection occurring within the first 28 days of life. It is a prevalent critical illness in the neonatal intensive care unit (NICU) [ 1 ]. One of the primary causes of morbidity and mortality in newborns is neonatal sepsis, a systemic infection that affects infants up to 28 days of age [ 2 ].

Around the world, newborn sepsis is thought to be the cause of 26% of fatalities in children under the age of five, with Sub-Saharan Africa (SSA) having the highest mortality rates. Neonatal mortality is unevenly distributed in Sub-Saharan Africa, where it is estimated that 49.6% of all under-five deaths occurred in 2013 [ 3 ]. Additionally, 30–50% of newborn mortality in underdeveloped nations and 15% of all neonatal deaths worldwide are attributable to neonatal sepsis. In Sub-Saharan Africa, neonatal sepsis is responsible for 17% of neonatal mortality. Neonatal sepsis, which causes over 37% of newborn mortality in Ethiopia and more than one-third of all neonatal deaths, is to blame [ 4 ].

One of the global initiatives of WHO is to reduce infant and under-five mortality in African countries to as low as 12/1000 and 25/1000, respectively, by the year 2030. The key to accomplishing this would be to improve the management and prevention of severe infections and preterm births [ 5 ].

In Ethiopia study shows that, small gestational age, being male, out born, not having breastfeed and lower Apgar score in the first 5 minute were identified as associated factors for the poor treatment outcome of preterm neonates admitted in NICU [ 6 ]. According to the 2019 Mini Ethiopian Demographic Health Survey (MEDHS) report, the neonatal mortality rate (NMR) is 30/1000 live births, a slight decrease from the 2011 EDHS report of 37/1000 live births. This high number of deaths is largely due to neonatal sepsis [ 5 , 7 ]. In a number of developing countries, identification of factors for neonatal sepsis and treatment of neonates with sepsis is not satisfactory. Moreover, reports from low income countries revealed inconsistencies in the prevalence, risk factors, and mortality from that of developed countries. Identification of risk factors and timely initiation of treatments can significantly decrease neonatal mortality and morbidity [ 8 ]. The outcomes of neonatal sepsis treatment vary between hospitals with different setups. Early detection and treatment are required to save the lives of our children and grandchildren. This may necessitate the use of expertise to identify common risk factors, antimicrobial use patterns, and clinical outcome treatment of neonatal sepsis. Therefore the purpose of this study was to assess treatment outcome and associated factors of neonatal sepsis among neonates admitted to neonatal intensive care unit in public hospitals, Addis Ababa, Ethiopia.

Methods and materials

Study design and setting.

This was a facility-based quantitative cross-sectional study carried out from February 15 to May 10, 2021 across Addis Ababa city public hospitals. Addis Ababa is the capital city of Ethiopia and seat of African Union and the United Nations World Economic Commission for Africa. It covers an area of 527 square kilometers and has 11 sub cities [ 9 ]. According to a population projection value for 2020 the city has an estimated population of 4.8 million.

The city has 12 public Hospitals among these, 11 hospitals having neonatal intensive care unit. Among these 6 were under Addis Ababa Health Bureau, 5 were under ministry of health and 1 was under Addis Ababa University (Tikur Anbessa Specialized Hospital) [ 10 ]. The study was conducted in four Addis Ababa public Hospitals (36%) selected by lottery method. These selected hospitals are Gandhi Memorial hospital (GMH), St peter Specialized hospital (SPSH), Tikur Anbessa special hospital (TASH) and Yekatit 12 hospital medical college (Y12HMC).

Sample size and sampling procedure

The sample size was determined using a single population–proportion formula by assuming a proportion (P) with culture-proven neonatal sepsis of 23.9%, the previous study in Ethiopia [ 11 ], a 5% margin of error (d), 10% nonresponse rate, n = ( Za /2) 2 p (1 − p )/( d ) 2 and the result was 308 neonates. Of the Eleven public hospitals in Addis Ababa city, four public hospitals (36%) were randomly selected for this study using a lottery method to.

Before the actual data collection started, the total number of neonates diagnosed as sepsis and admitted to NICU monthly was reviewed in each of the selected hospitals. Then the total sample size of the study was allocated proportionally to each selected hospital based on their previous number of neonates diagnosed as sepsis admitted to NICU. Neonates that fulfill eligibility criteria were recruited from respective study hospitals. Finally, 308 neonates diagnosed with sepsis were selected by using systematic random sampling technique from February 15, 2021 to May 10, 2021 with ‘k’ interval of 2. During the data collection, the first participant was selected by lottery method.

Study population.

During the study period, all neonate patients admitted to NICUs of Addis Ababa selected public hospitals were diagnosed with neonatal sepsis.

Inclusion criteria

The study included neonates with clinical diagnosis of sepsis based on the following two risk factors and/or clinical features of bacterial infections.

Risk factors include low birth weight (<2500 grams) or prematurity (<37 weeks of gestation age), febrile illness in the mother within 2 weeks prior to delivery, foul-smelling discharge and/or meconium stained amniotic liquid, prolonged rupture of membranes >18 hours, suspected chorioamnionitis, prolonged labor (> 24 hours), and perinatal asphyxia (Apgar score <4 at 1 minute).

Clinical features of sepsis include poor reflexes, lethargy, respiratory distress, bradycardia, apea, fever, convulsions, abdominal distension, and bleeding.

Data collection instrument and procedures

Data were gathered using a structured questionnaire and in face to face interviews. The data collection tools were adapted from various sources of information [ 3 , 10 – 13 ] and by reviewing both maternal and newborns profile cards. To ensure consistency, the questionnaire was created in English first, and then translated into Amharic and then back into English by language experts. The data was collected in day the day time shift from 8:00 a.m. to 5:30 p.m. after obtaining consent from each participant prior to data collection in the hospitals after finishing the services and returning home. Data was recruited by eight trained BSc nurses and they were supervised by four senior nurses having previous experience in data collection. Training was provided on data collection procedures; including how to conduct interviews, administer questionnaires, obtain consent, maintain confidentiality, and respect the rights of participants. Continuous follow up and supervision was made by principal investigator throughout the data collection period from February 15, 2021 to May 10, 202.

Data quality control

Supervisors and data collectors were trained on how and what information they should collect from the targeted data sources to ensure data quality. Expertise was given a tool to check the content’s validity and accuracy. It was pre-tested on 5% (n = 15) of similar mothers outside the study area at Zewditu Memorial Hospital to assess its completeness, clarity, length, skip patterns, and correctness of filled questioners. The questionnaire was modified based on the results of the pretest. Data was collected by trained health professionals from other units of the health facility.

Data analysis procedure

Data were entered into EPI Data version 4.6 and analyzed in SPSS Software version 26. Bi-variable analysis was used to examine the relationship between each independent variable and the outcome variable. To account for all potential confounders, all variables with p-values ≤ 0.25 were included in the multivariable model. The linear correlation among the independent variables was also examined using multi-co linearity. The degree of association between dependent and independent variables was determined using an odds ratio with a 95% confidence interval and a p-value ≤ 0.05.

Socio-demographic characteristics of respondents

All 308 required study participants were interviewed, with a 100% response rate. Among those who responded, 121 (39.3%) of the mothers were between the ages of 25 and 29. The mother’s average age was 29.37(±5.16) years. Almost all of the respondents, 296 (96.2%), were married. Around 267 (63.6%) of mothers had secondary or higher education. The average monthly household income of the study participants was 7501 Ethiopian Birr (136 (44.2%) of government employees and 84 (27.3%) of respondents. The mean and standard deviation of household income were 5775.8 (±3543.2) years.

Obstetric and neonatal health related factors of the study participants for their neonates

Almost all the respondents, 303 (98.4%), had an ANC follow-up. 221 (71.8%) of those who received ANC follow-up had four or more visits. The majority of mothers, 247 (80.2%), delivered their newborns in a hospital. Half of the 157 mothers with their newborns were 37 weeks gestational age ( Table 1 ).

• PPT PowerPoint slide
• PNG larger image
• TIFF original image

https://doi.org/10.1371/journal.pone.0284983.t001

In terms of birth weight, 163 (52.9%) of neonates weighed 2.5 kg. The majority of neonates, 262 (85.1%), were admitted when they were older than 30 minutes ( Table 2 ).

https://doi.org/10.1371/journal.pone.0284983.t002

Treatment and laboratory findings of neonatal sepsis

Regarding the treatment outcomes, the majority of neonates 233 (75.6%) recovered from their condition with improvement, and 75 (24.4%) died ( Table 3 ).

https://doi.org/10.1371/journal.pone.0284983.t003

Almost all neonates were administered the combination of ampicillin and gentamicin as a first line ( Fig 1 ).

Others **: fluconazole, esomeprazole, TTC, phototherapy.

https://doi.org/10.1371/journal.pone.0284983.g001

In this study, the possible causes of neonatal deaths are 35(46.67%), cardio respiratory arrest secondary to respiratory problems and 14(18.67%) sepsis ( Fig 2 ).

https://doi.org/10.1371/journal.pone.0284983.g002

Predictor variable of treatment outcome of neonatal sepsis

The association of the independent and dependent variable were first tested by using bi-variable analysis which (P≤0.25) were tested in the final multivariable analysis to see their significant association with their treatment outcome of neonatal sepsis. Accordingly, as shown in Table 4 below those bi-variable regression associated with the crude odds ratios (COR) treatment outcome of neonatal sepsis such as education status of the mothers, place of delivery, gestational age < 37weeks, birth weight of the neonate, age of the neonate admitted to NICU, sever chest in drawing, grunting, un able to feed, temperature, Chorioamnaties, meconium amniotic stained, PROM, Hypertensive PIH/ Eclampsia, meropenum, vancomycin, metronidazole, CBC and CRP result.

https://doi.org/10.1371/journal.pone.0284983.t004

In Multivariable analysis results showed that, there was statistically significance association found between) Poor treatment outcome of neonatal sepsis parameters which showed p-value of below 0.05 were Preterm babies admitted to NICU (gestational age), grunting, meconium amniotic stained, duration of rupture of membrane >18hours (PROM), Hypertensive PIH/ Eclampsia, meropenum and CRP result.

This study aimed to assess treatment outcome of neonatal sepsis and its associated factors among newborns delivered in public hospitals in Addis Ababa, Ethiopia. In this study, the prevalence of poor neonatal sepsis treatment outcome was 24.4% with a 95% CI among newborns admitted to NICU in public hospitals in Addis Ababa (19.5–29.2). This finding is higher compared to a study conducted in Bahir dar (16%) [ 12 ]. This deference could be attributed to the way neonatal sepsis has been used to confirmatory blood culture to assert neonatal sepsis to the study that is done in some facility. In addition to this, NICU services at Felegehowot referral hospital are organized with personnel and equipment based on newborn conditions(severity) and classified as level-1(basic), level-2 (specialty) and level-3 (subspecialty) and study design (retrospective follow up that all deaths might not be documented. Many of the complications of preterm birth that result in poor in-hospital outcomes are related to gestational age, with premature infants being more vulnerable. This finding is nearly identical to a study conducted in Debrezeyt, Ethiopia, where 26% [ 13 ]. This finding is lower compared to a study conducted in Nigeria were (34%) [ 14 ]. This could be related with use of advanced confirmatory blood culture to establish neonatal sepsis in the Nigeria, which is difficult to apply in this study area due to the lack of some facilities. Another possible explanation for this variation is the variation in health facility and sample size across studies.

Neonatal sepsis treatment outcomes were 5 times more likely to be bad in newborns under 37 weeks of gestational age compared to newborns who were 37 weeks or older. This is consistent with research from Gondar [ 15 ], Tikur Anbessa Specialized Hospital, Ethiopia [ 16 ] and Kenya [ 17 ], which found that preterm newborns had a higher mortality rate than term newborns. This result is comparable to that of an Australian study, which discovered that a gestational age of one week enhanced the neonatal survival rate by more than 5% [ 18 ].

The odds of neonates born from mothers with developing Hypertensive PIH/ Eclampsia were 4 times higher than those of neonates born from who did not develop Hypertensive PIH/ Eclampsia. This finding is consistent with previous research that found chronic hypertension to be a risk factor for neonatal sepsis in Ethiopia. This could be because maternal hypertensive problems have a direct impact on fetal wellbeing in the uterus, which contributes to neonatal sepsis at birth.

PROM was statistically significantly associated with a poor treatment outcome of sepsis. Mothers who gave birth to neonates with PROM were 4 times more likely to suffer from sepsis compared with those neonates born from women who had not developed PROM. This finding is comparable with studies conducted in Nepal [ 19 ], Mexico [ 20 ] and USA [ 21 ]. These could be caused by aerobic and anaerobic pathogens colonizing the birth canal, resulting in ascending amniotic fluid infection and neonate colonization at birth. Mother-to-fetus transmission of bacterial agents infecting the amniotic fluid and birth canal during labor and delivery may occur more frequently, resulting in neonatal sepsis (EONS) [ 22 ].

Meconium amniotic stained were 3 times developed poor treatment outcome when compared to those neonates without history of Meconium amniotic stained. Which is similar with a study in Bahir dar [ 12 ], Uganda [ 23 ], Ghana [ 3 ], India [ 24 ] and Nepal [ 19 ]. This is revealed that after meconium aspiration strict follow up is needed for neonates. This may be due to neonates delivered from women with meconium stained amniotic fluid are more liable to aspirate it and fill smaller air ways and alveoli in the lung. And it increases the multiplication of microbes that cause sepsis and predisposes to late onset neonatal sepsis (LONS) [ 25 ].

A neonate who has at risk of respiratory problem was significantly associated with poor outcome of neonatal sepsis. Those neonates developed with grunting were 7 times developed poor outcome compared to neonates without respiratory distress syndrome. This is similar with a study in Bahir dar [ 12 ]. This result comparable with studies done in Uganda [ 23 ] and Sudan [ 26 ]. This may be due to health workers’ ignorance of the syndrome’s poor early detection of signs, and another explanation may be due to mothers’ delay coming in coming to health facilities or institutions.

C-reactive protein levels were found to be significantly associated with a poor sepsis outcome. Positive CRP laboratory results of neonates were 6 times they developed poor treatment outcomes when compared to those of neonates’ negative CRP results. This finding is similar with studies done in Nepal [ 19 ]. This could be because CRP is the most sensitive and widely used test, but it is necessary to consider a sepsis panel of at least three tests, at least two of which must be positive for one to suspect septicemia with reasonable certainty.

In most developing countries, including Ethiopia, empirical treatment is the primary method of managing neonatal sepsis. When compared to neonates who did not receive Meropenum, those who received it had a fourfold worse treatment outcome. In this hospital, antimicrobial use is primarily empirical. This may promote the development of resistant bacteria, influencing future drug selection in the treatment of neonatal sepsis.

The study found both maternal and neonatal factors as possible independent risk factors to have a strong association with the risk of poor outcome of neonatal sepsis. Preterm babies admitted to NICU, grunting, meconium amniotic stained (MSAF), duration of rupture of membrane >18hours (PROM), Hypertensive PIH/ Eclampsia, meropenum and CRP result were significantly associated with poor treatment outcome of neonatal sepsis. Researchers who are interested in conducting research on neonatal sepsis should have to include neonates in the community, which may increase the external validity of the study. It is also better to do a meta-analysis since the previous findings about the factors causing neonatal sepsis were inconsistent.

Acknowledgments

We would like to Acknowledge Addis Ababa University, college of health sciences, school of nursing and Midwifery and department of nursing for giving me the chance. In addition, we would like to extend our thanks to the study participants, data collectors, supervisors for their contribution and commitment throughout the study period.

Ethics approval and consent to participate

Ethical clearance was obtained from the institutional review board of Addis Ababa University, College of Health Sciences, School of Nursing and Midwifery, Department of Nursing (Protocol number: aau/chs/chnsg/18/21). Formal letters were obtained from Addis Ababa Public Health Research and Emergency Management Core Process in order to get permission to carry out the study. After explaining the purpose and procedure of the study, each respondent (mothers/care givers) signed a written informed consent form. No name or other identifying information was included with the instrument. The eligible study participants were enrolled in the study only after they gave written informed consent and will not be forced to participate. All the information given by the respondents was used for research purposes only; Confidentiality and privacy were maintained by omitting the names of the respondents during the data collection procedure and after the data collection was completed.

• View Article
• Google Scholar
• 5. WHO, World health Statistics. Monitoring health for the SDGs, sustainable development goals. WHO Library Cataloguing-in-Publication Data 2016.
• 7. Rockville, Maryland, USA. Mini Demographic and Health Survey Ethiopian Public Health Institute. 2019.
• 8. UNICEF. Committing to Child Survival: A Promise Renewed World Health Organ Tech Rep Ser. 2014.
• 9. World Population projections. Ethiopia Demographics Profile. CIA World. 2019.
• 10. Addis Ababa Health Service Information. General Hospital. Addis Ababa health Bureau 2019/20 (Unpublished)
• PubMed/NCBI
• 26. Khair k. Prevalence and outcome in a tertiary neonatal unit in Sudan. 2014.

Open Access is an initiative that aims to make scientific research freely available to all. To date our community has made over 100 million downloads. It’s based on principles of collaboration, unobstructed discovery, and, most importantly, scientific progression. As PhD students, we found it difficult to access the research we needed, so we decided to create a new Open Access publisher that levels the playing field for scientists across the world. How? By making research easy to access, and puts the academic needs of the researchers before the business interests of publishers.

We are a community of more than 103,000 authors and editors from 3,291 institutions spanning 160 countries, including Nobel Prize winners and some of the world’s most-cited researchers. Publishing on IntechOpen allows authors to earn citations and find new collaborators, meaning more people see your work not only from your own field of study, but from other related fields too.

Brief introduction to this section that descibes Open Access especially from an IntechOpen perspective

Want to get in touch? Contact our London head office or media team here

Our team is growing all the time, so we’re always on the lookout for smart people who want to help us reshape the world of scientific publishing.

Home > Books > Neonatology

Selected Topics in Neonatal Care

Book metrics overview

19,816 Chapter Downloads

Impact of this book and its chapters

Total Chapter Downloads on intechopen.com

Total Chapter Views on intechopen.com

Overall attention for this book and its chapters

Book Citations

Total Chapter Citations

Academic Editor

Austral University of Chile , Chile

Published 27 June 2018

Doi 10.5772/65564

ISBN 978-1-78923-363-6

Print ISBN 978-1-78923-362-9

eBook (PDF) ISBN 978-1-83881-248-5

Copyright year 2018

Number of pages 248

Neonatology is one of the areas of greatest development and evolution within pediatrics. The technoscientific advances in this area have led to an increase in the survival of premature infants who sometimes require sophisticated care. However, there is essential care that must be included in all centers that care for high-risk babies. This book includes important topics related to neonatal care gr...

Neonatology is one of the areas of greatest development and evolution within pediatrics. The technoscientific advances in this area have led to an increase in the survival of premature infants who sometimes require sophisticated care. However, there is essential care that must be included in all centers that care for high-risk babies. This book includes important topics related to neonatal care grouped into four sections. In 14 chapters that address relevant issues about neonatal care, the book seeks to contribute to the clinical work of the health teams of neonatal units. Specialists in the field of neonatology from different countries have developed these chapters and through them they hope to share part of their experience.

By submitting the form you agree to IntechOpen using your personal information in order to fulfil your library recommendation. In line with our privacy policy we won’t share your details with any third parties and will discard any personal information provided immediately after the recommended institution details are received. For further information on how we protect and process your personal information, please refer to our privacy policy .

Cite this book

There are two ways to cite this book:

Edited Volume and chapters are indexed in

Table of contents.

By R. Mauricio Barría

By Teja Škodič Zakšek, Anita Jug Došler, Ana Polona Mivšek and Petra Petročnik

By Petja Fister and Štefan Grosek

By Štefan Grosek and Petja Fister

By Adauto Dutra Moraes Barbosa, Israel Figueiredo Júnior and Gláucia Macedo de Lima

By Selim Öncel

By Lidia Decembrino, Nunzia Decembrino and Mauro Stronati

By Rejane P. Neves, Ana Maria R. de Carvalho Parahym, Carolina M. da Silva, Danielle P.C. Macêdo, André F.G. Leal, Henrique J. Neves and Reginaldo G. Lima-Neto

By Esra Caliskan

By Carlos Baeza Herrera, Alin Villalobos Castillejos and Ricardo Cortés García

By Kriengkrai Srithanaviboonchai

By Koichi Miyake, Noriko Miyake and Takashi Shimada

By Maria Slomczynska, Malgorzata Grzesiak and Katarzyna Knapczyk- Stwora

By Hippolite O. Amadi and Mohammed B. Kawuwa

IMPACT OF THIS BOOK AND ITS CHAPTERS

19,816 Total Chapter Downloads

3,775 Total Chapter Views

10 Crossref Citations

4 Web of Science Citations

19 Dimensions Citations

3 Altmetric Score

Order a print copy of this book

Available on

Delivered by

£119 (ex. VAT)*

Hardcover | Printed Full Colour

FREE SHIPPING WORLDWIDE

* Residents of European Union countries need to add a Book Value-Added Tax Rate based on their country of residence. Institutions and companies, registered as VAT taxable entities in their own EU member state, will not pay VAT by providing IntechOpen with their VAT registration number. This is made possible by the EU reverse charge method.

As an IntechOpen contributor, you can buy this book for an Exclusive Author price with discounts from 30% to 50% on retail price.

Log in to your Author Panel to purchase a book at the discounted price.

For any assistance during ordering process, contact us at [email protected]

Related books

Update on critical issues on infant and neonatal care.

Edited by René Mauricio Barría

Topics on Critical Issues in Neonatal Care

Cohort studies in health sciences, selected topics in breastfeeding, neonatal bacterial infection.

Edited by Bernhard Resch

Neonatal Medicine

Edited by Antonina Chubarova

Pediatric and Neonatal Surgery

Edited by Joanne Baerg

Neonatal Care

Edited by Zoe Iliodromiti

Infant Feeding

Edited by Isam Jaber Al-Zwaini

Selected Topics on Infant Feeding

Call for authors, submit your work to intechopen.

• Open access
• Published: 08 April 2024

Parental satisfaction with neonatal intensive care unit services and associated factors in Ethiopia: systematic review and meta-analysis

• Eshetu Elfios 1 ,
• Nefsu Awoke 1 ,
• Temesgen Geta 1 ,
• Christian Kebede 1 &
• Abdulkerim Hassen 2  

BMC Nursing volume  23 , Article number:  234 ( 2024 ) Cite this article

139 Accesses

1 Altmetric

Metrics details

In the context of healthcare, satisfaction is the attainment of adequate or acceptable treatment and serves as both a benchmark for quality and the ultimate objective of providing care. In neonatal care facilities, parent satisfaction is a key measure of the quality of the services offered to the newborns and aids in improving healthcare delivery. This is the first systematic review aiming to address critical knowledge gaps regarding factors influencing parental satisfaction with neonatal intensive care unit services, and determine pooled prevalence in Ethiopia.

After comprehensive systematic search for full texts in the English language through an electronic web-based search strategy from databases of PubMed, CINAHL, Embase, African Journals Online, PsycINFO, and Google Scholar, included a total of 8 articles. Checklists from the Joanna Briggs Institute were used to assess the studies’ quality of methodology. STATA version 14 software for windows was used for all statistical analyses and meta-analysis was done using a random-effects method. Subgroup and sensitivity analyses were performed to clarify the source of heterogeneity.

Pooled national level of parents’ satisfaction with neonatal intensive unit service was 57.87% (95% CI (49.99, 65.75%)). Age of respondents and availability of chair were significantly associated with parents’ satisfaction with neonatal intensive unit service.

Conclusions and recommendation

In our review we found that nearly half of parents expressed dissatisfaction with neonatal intensive unit service, which is high. Age of respondents and availability of chair in neonatal intensive unit waiting area were significantly associated with neonatal intensive unit service. Efforts to enhance parental satisfaction with neonatal intensive unit services are crucial, given that nearly half of parents reported dissatisfaction. Necessary infrastructure should be fulfilled to increase parental satisfaction with neonatal intensive unit service.

(PROSPERO) International prospective register of systematic reviews:

CRD42023483474

Peer Review reports

Introduction

In neonatal care facilities, parental satisfaction is a key measure of the quality of the services offered to the newborns and aids in improving healthcare delivery [ 1 ]. Parental satisfaction is a way of responding to the expectation to meet the health needs of the people. Parental satisfaction is a belief and attitude of parents towards a specific service in an institution. It is an effective parameter in improving quality of care in neonatal intensive care unit (NICU) [ 2 , 3 ].

Parent satisfaction plays a pivotal role in enhancing the quality of neonatal care. The assessment of healthcare interventions is significantly influenced by parent satisfaction, offering vital insights for evaluating staff performance, enhancing health intervention systems, facilitating future planning, and devising effective strategies for neonatal care [ 4 ].

Patient satisfaction is related to medical services; however, it is not the only factor. In the particular case of neonatal patients who are hospitalized in the NICU and cannot express their own opinion with regard to satisfaction, information is collected indirectly, from their parents. The measurement of parental satisfaction is an important indicator of the quality of services offered by an NICU, as it contributes to the evaluation of the health care provided and its improvement, resulting in the maximization of parental satisfaction with the health care system [ 5 , 6 ].

Neonates are unable to communicate their health requirements, requests, experiences, opinions, or satisfaction, therefore parents play this role [ 7 , 8 ]. Length of the infant’s hospital stay, Age, education level, and income are the socioeconomic factors and the parents’ sex are the most significant factors in determining how satisfied parents are [ 6 ]. Researchers found that Parents who could not hold their baby had a lower level of satisfaction. Furthermore, the parents who felt unable to protect their baby from pain and painful procedures showed the lower the degree of satisfaction [ 9 ].

Complications and morbidity often stem from inadequate quality of care, notably driven by parental dissatisfaction with NICU services and a deficiency in professional care and treatment [ 10 ]. This parental dissatisfaction can lead to inefficiencies in infant care, stemming from a shortage of essential nursing, medical, and family-centered support [ 11 ].

Furthermore, the satisfaction of parents with healthcare is linked to enhancements in their child’s health and a reduction of symptoms, encompassing adherence to the therapeutic regimen and comprehension of medical information. Consequently, the extent of client satisfaction with healthcare can serve as a valuable proxy variable, representing a crucial aspect of care quality [ 12 ].

The determinants influencing parental satisfaction included having infants outside the infection isolation room, parents with infants that breast-fed [ 1 ], parents’ educational status, parents’ occupation, duration of hospital stay, adequacy of care, and adequacy of pain management were significant factors of parental satisfaction [ 13 ].

Pooled proportion is needed to determine which sociodemographic factors are connected to parents’ satisfaction with their newborn’s care in NCU. This study can advance our understanding of parents’ satisfaction and offer advice to healthcare facilities on how to improve parents’ satisfaction in this condition.

This is the first systematic review aiming to address critical knowledge gaps regarding factors influencing parental satisfaction with neonatal intensive care unit services, and determine pooled prevalence in Ethiopia. This study will help to be used as evidence to evaluate the goal of Sustainable development goals (SDGs) planned to reduce neonatal death by improving quality of NICU services. The findings of this review will inform policy and decision makers to monitor and improve the quality of NICU services in Ethiopia. Thus, this systematic review and meta-analysis was intended to answer the following question:

What is the level of parental satisfaction with NICU services in Ethiopia, and what are the associated factors influencing parental satisfaction with the NICU service?

Design and search strategy

The procedure for this systematic review and meta-analysis was developed in accordance with the Preferred Reporting Items for Systematic review and Meta-analysis Protocols (PRISMA-P) statement [ 14 ]. PRISMA- 2020 statement was used to report the findings [ 15 , 16 ]. This systematic review and meta-analysis was registered on PROSPERO with the registration number CRD42023483474.

We searched PubMed, CINAHL. Cochrane Library, Embase, Google Scholar, and PsycINFO database for studies reporting the level of parental satisfaction with NICU service from study conception to November 2023. We used EndNote (version X8) software to download, rearrange, review and cite the articles. Manual search was conducted for cross-references in order to find other related studies. A comprehensive search was conducted with the following search terms: “Parent satisfaction”, “satisfaction”, “determinants of parent satisfaction”, “Neonatal intensive care unit”, “NICU”, and “Ethiopia”. To combine search terms, we used Boolean operators like “AND” and “OR”.

Eligibility criteria

We included studies reporting the level of parental satisfaction with NICU service irrespective of the type of instrument used to estimate satisfaction, the level of satisfaction assessed, and scoring system used to generate the overall score of satisfaction, Studies from both published and gray literature reported in English language, findings from national research repository, a study that has been conducted in Ethiopia, and a study that reports associated factors of parental satisfaction with NICU service were included in this review.

Studies without full text and those which lack information on Parental satisfaction with NICU service and associated factors or studies for which unable to get the necessary detail information after contacting the authors were excluded. Three authors (E.E. N.A. CK) independently evaluated the eligibility of all retrieved studies, and other reviewer’s opinion (T.G) was requested to reach a general agreement with regard to potential in- or exclusion of studies.

Data extraction and quality assessment

Data was extracted on Microsoft Excel spread sheet. Three independent authors (EE, NA, AH) extracted the data independently. For each included article, the name of primary author, year of publication, the setting where the study was conducted or country, region, study design, study period, sample size, response rate, population, proportion of parental satisfaction and associated factors was recorded. During extraction, discrepancies between data extractors were discussed to reach agreement.

Two authors (EE, TG) independently conducted a critical appraisal of the included studies. Joanna Briggs Institute (JBI) checklists was used to assess the quality of the studies. The tool has nine parameters which have yes, no, unclear, and not applicable options (1). appropriate sampling frame (2), proper sampling technique (3), adequate sample size (4), study subject and setting description (5), sufficient data analysis (6), use of valid methods for the identified conditions (7), valid measurement for all participants (8), using appropriate statistical analysis and (9) adequate response rate [ 17 ]. The tools have yes, no, not applicable, and unknown options. One for yes responses and zero for unclear, not applicable, and no responses was scored. During the critical appraisal, whenever it was necessary other reviewers (NA, CK) were involved. Accordingly, studies that met the inclusion criteria were included and tabulated following a consensus and thorough discussion on data extraction and the completion of significant analyses by utilizing the JBI checklist. Prior to being chosen for a final review, articles undergo quality control checks. Research which scored quality index score of seven or above were categorized as low risk ( Table  1 ) .

Data analysis methods

The extracted data were imported to STATA 14 statistical software was used to carry out the pooled proportion of parental satisfaction with NICU service in Ethiopia. A meta-analysis of the level of parental satisfaction with NICU care was carried out using a random-effects method since it is the most common method in a meta-analysis to adjust for the observed variability [ 24 ].

To examine the possible risk of publication bias and small study effects, funnel plots and Egger’s test was used [ 25 , 26 ]. Cochrane Q-Static and I 2 were used to confirm heterogeneity between studies. Subgroup analyzes were performed to compare the pooled prevalence of parental satisfaction with NICU service and associated factors across regions. Pooled prevalence was presented in forest pilot format with 95% CI.

Search result and study characteristics

The electronic online search from database searching of PubMed, Google scholar, CINAHL, African Journals Online and manual search yielded 354 records, of which 46 duplicate records were identified and removed. Title and abstract screening resulted in the exclusion of 147 irrelevant articles. Then, remaining 162 articles underwent for full-text review. Among these, 154 articles were excluded based on the predetermined eligibility criteria. Finally, a total of 8 articles were included in the meta-analysis (Fig.  1 ).

PRISMA flow diagram of the selection process of studies on parental satisfaction with NICU service in Ethiopia, 2024

A total of 8 studies with 2255 participants were included in this meta-analysis. Of these four studies were conducted in Amhara region of Ethiopia [ 13 , 19 , 27 , 28 ], Two studies were conducted in Oromia region of Ethiopia [ 20 , 21 ], One study was conducted in Addis Ababa [ 22 ] and remaining one study was conducted in Southern Ethiopia [ 17 ]. All these included studies were cross-sectional in design and sample sizes ranged from 109 [ 21 ] to 401 [ 17 ] (Table  2 ).

Patient satisfaction with nursing care

The pooled effect size of parental satisfaction with NICU service using the fixed effect model showed significant heterogeneity across the studies. Therefore, we performed the analysis with a random effects model with 95% CI in order to adjust for the observed variability. Accordingly, the pooled national level of parents’ satisfaction with NICU service was 57.87% (95% CI (49.99, 65.75%)) with significant heterogeneity between studies (I 2  = 93.3, P  = 0.000) (Fig.  2 ).

Forest plot showing the pooled level of satisfied parents with NICU service

Based on the subgroup analysis by region, the highest level of parental satisfaction was observed in Oromia (65.43% (95% CI: 50.88, 79.98), I 2  = 82.1%) while, the lowest level of parental satisfaction was observed in Addis Ababa (41.80% (95% CI: 36.97, 46.63), I 2  = 0.0%) (Fig.  3 ).

Subgroup analysis by regions on the level of parental satisfaction with NICU service

Publication bias and heterogeneity

Presence of publication bias was examined using visual inspection of the funnel plot and Egger’s test. Visual inspection of the funnel plot suggested symmetrical distribution of included studies (Fig.  4 a). The result of Egger’s test was not statistically significant for the presence of publication bias ( P  = 0.174) (Fig.  4 b). Both tests confirm that there was no publication bias. The result of this meta-analysis revealed statistically significant heterogeneity among studies (I 2  = 93.3%), we performed a subgroup analysis by region to adjust and minimize heterogeneity in addition to exploring potential sources of heterogeneity and examining whether the effect size varies across different subgroups. (Fig. 3 ).

Additionally, in our efforts to pinpoint potential sources of heterogeneity, we conducted a meta-regression analysis, incorporating sample size and publication year as covariates. Nevertheless, our findings revealed that neither of these variables had a significant impact on the observed heterogeneity among studies (Table  3 ).

( a ) funnel plot and ( b ) Egger’s test of the study

Leave-out-one sensitivity analysis

The leave-out-one sensitivity analysis conducted t assess the effect of a single studies on the overall pooled proportion of parental satisfaction with NICU service. In this systematic review, each study was excluded from the analysis one at a time. The outcomes of this analysis indicated that the exclusion of any single study did not lead to a statistically significant alteration in the overall pooled associated factors with parental satisfaction with NICU service Ethiopia. The findings are visually represented in Fig.  5 , illustrating the stability of the overall pooled estimate even with the removal of specific studies from the analysis.

Sensitivity analysis of pooled prevalence for each study being removed at a time for systematic review and meta-analysis of turnover intention among nurses in Ethiopia

Associated factors with parental satisfaction with NICU service

In our review we found that two variables (age of respondents and availability of chair) were significantly associated with NICU service. Parents in the age group of 25–35 had 61% lower odds of satisfaction compared to the reference group of older age categories (OR = 0.39, p  = 0.009, I 2  = 85.2%) (Fig.  6 ). In our review, the availability of a chair for parents was significantly associated with 3.13 times increase in parental satisfaction (OR = 0.32, p  = 0.004, I 2  = 87.9%). (Fig.  7 ).

Pooled effect (OR) of the association between age and parental satisfaction on NICU service in Ethiopia, 2023

Pooled effect (OR) of the association between availability of chair and parental satisfaction on NICU service in Ethiopia, 2023

Using data from both published and unpublished studies, this meta-analysis was carried out to determine the degree of parents’ satisfaction with NICU service nationwide and to identify the factors associated to it. This meta-analysis showed that Ethiopian parents’ satisfaction with NICU service was 57.87% (95% CI (49.99, 65.75%)) with significant heterogeneity between studies (I 2  = 93.3, P  = 0.000). This result was consistent with earlier research carried out in London 56% [ 29 ]. However, our meta-analysis’s assessment of parents’ satisfaction with NICU service was lower than what other comparable studies reported USA(Massachusetts and California), 21 European union countries and Norway [ 8 , 30 , 31 ]. The difference in the hospital infrastructure and socioeconomic status could be the cause of this discrepancy. Developed nations offer higher quality healthcare than developing nations. Healthcare organizations in industrialized nations incorporate technologies into their operational structure and patient interaction strategy to enhance the overall parent satisfaction. The parental satisfaction in this review is higher than the study conducted in Greece 48.7% [ 1 ]. This difference might be due to variations in the measurement tools used.

According to our subgroup analysis by region, Oromia had the highest degree of parental satisfaction (65.43% (95% CI: 50.88, 79.98). This might be due to Oromia region has a greater nurse to neonate ratio than other parts of the nation. However, parents in Addis Ababa, the capital city, may be more knowledgeable and have higher expectations. If these expectations are not met, this could lead to a decrease in satisfaction [ 32 ].

In our review we found that age of parents and availability of chair were significantly associated with parental satisfaction with NICU service. Other socio-demographic as well as parent and newborn related factors were found non-significant in final meta-analysis.

On this systematic review and meta-analysis we found that parents in the age group of 25–35 had 61% lower odds of satisfaction compared to the reference group of older age categories This finding is in line with the study conducted in Greece [ 4 ]. Because they are less experienced, younger parents could have lower needs and expectations from the healthcare providers in NICU. While older parents having higher expectations from healthcare providers and when this expectation met, they satisfy with the care given to their newborns. In contrary to this finding, study conducted in Norway revealed that age of parents was not significantly associated with overall parental satisfaction [ 31 ].

Our review revealed a significant association between the availability of a chair for parents and parental satisfaction in neonatal care units. The odds ratio of 0.32 (95% CI: 0.15–0.68, p  = 0.004) indicates a substantial impact, suggesting that parents who had access to a chair were approximately 3.13 times more likely to report higher satisfaction levels compared to those without such amenities. This finding underscores the significance of environmental factors in healthcare settings, particularly those catering to neonatal care. Comfortable facilities, such as the availability of seating for parents, play a crucial role in shaping their overall satisfaction with the care provided. This aligns with previous studies emphasizing the importance of a supportive and accommodating environment in healthcare settings [ 33 ].

Conclusion and recommendation

Based on the outcomes of our systematic review and meta-analysis, which revealed a low proportion of parental satisfaction with NICU services, we recommend healthcare facilities to address a range of factors influencing parental satisfaction comprehensively. Beyond the provision of seating arrangements, it is crucial to explore thoroughly the impact of age and regional variations on parental satisfaction. Future research endeavors should focus on clarifying these complexities further, enabling the development of tailored interventions. By addressing existing knowledge gaps, we can enhance our understanding of the intricate dynamics influencing parental satisfaction in neonatal care units, ultimately contributing to more effective and targeted healthcare strategies.

The strength of this study lies in its extensive systematic review and meta-analysis, reaching across varied regions. It thoroughly explores the complex factors that impact parental satisfaction in neonatal care units, offering valuable insights for better interventions and policy enhancements. Our review had few limitations, even though it has given useful information and current evidence on the degree of parent satisfaction with NICU service in Ethiopia. Careful consideration must be given to the interpretation of the results as our aggregate estimations revealed considerable variation among the research.

Data availability

The datasets generated and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Abbreviations

African Journals Online

Cumulative Index to Nursing and Allied Health Literature

Joanna Briggs Institute

Medical Subject Headings

Neonatal Intensive Care Unit

Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols

Galanis P, Bilali A, Pouliou K, Matziou V. Factors associated with parents’ satisfaction with care provided in a neonatal intensive care unit in Greece. J Neonatal Nurs. 2016;22(4):177–84.

Article   Google Scholar  

Bjertnaes OA, Sjetne IS, Iversen HH. Overall patient satisfaction with hospitals: effects of patient-reported experiences and fulfilment of expectations. BMJ Qual Saf. 2012;21(1):39–46.

Article   PubMed   Google Scholar  

Ali MS, Mekonen EG, Workneh BS. Parental satisfaction and associated factors towards neonatal intensive care unit services at the University of Gondar Comprehensive Specialized Hospital, Northwest Ethiopia. Int J Afr Nurs Sci. 2021;15:100364.

Google Scholar  

Tsironi S, Bovaretos N, Tsoumakas K, Giannakopoulou M, Matziou V. Factors affecting parental satisfaction in the neonatal intensive care unit. J Neonatal Nurs. 2012;18(5):183–92.

Mitchell-DiCenso A, Guyatt G, Paes B, Blatz S, Kirpalani H, Fryers M, et al. A new measure of parent satisfaction with medical care provided in the neonatal intensive care unit. J Clin Epidemiol. 1996;49(3):313–8.

Article   CAS   PubMed   Google Scholar  

Demographic NR. Health Survey 2014–15-Final Report. Rwanda: National Institute of Statistics of Rwanda (NISR). 2015:640.

Auslander GK, Netzer D, Arad I. Parents’ satisfaction with care in the neonatal intensive care unit: the role of sociocultural factors. Children’s Health Care. 2003;32(1):17–36.

McCormick MC, Escobar GJ, Zheng Z, Richardson DK. Factors influencing parental satisfaction with neonatal intensive care among the families of moderately premature infants. Pediatrics. 2008;121(6):1111–8.

Voulgaridou A, Paliouras D, Deftereos S, Skarentzos K, Tsergoula E, Miltsakaki I, et al. Hospitalization in neonatal intensive care unit: parental anxiety and satisfaction. Pan Afr Med J. 2023;44:55.

Article   PubMed   PubMed Central   Google Scholar  

Hug L, Alexander M, You D, Alkema L. National, regional, and global levels and trends in neonatal mortality between 1990 and 2017, with scenario-based projections to 2030: a systematic analysis. Lancet Global Health. 2019;7(6):e710–20.

Fidanci BE, Arslan F, Fidanci K. Parents’ expectations and satisfaction on pediatrics clinic. Translational Med. 2014;4(137):2161.

Ammentorp J, Mainz J, Sabroe S. Determinants of priorities and satisfaction in pediatric care. Pediatr Nurs. 2006;32(4).

Alle YF, Akenaw B, Seid S, Bayable SD. Parental satisfaction and its associated factors towards neonatal intensive care unit service: a cross-sectional study. BMC Health Serv Res. 2022;22(1):1266.

Moher D, Shamseer L, Clarke M, Ghersi D, Liberati A, Petticrew M, et al. Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. Syst Reviews. 2015;4(1):1–9.

Moher D, Liberati A, Tetzlaff J, Altman DG, Group* P. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Med. 2009;151(4):264–9.

Moher D, Shamseer L, Clarke M, Ghersi D, Liberati A, Petticrew M et al. Group, P.-P.(2015). Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement.

Sileshi E, Mohammed B, Eshetu D, Dure A, Bante A, Mersha A et al. Parental Satisfaction towards Care Given at Neonatal Intensive Care Unit and Associated Factors in Comprehensive and Referral Hospitals of Southern Ethiopia. Journal of Pregnancy. 2023;2023.

Alemu A, Zeleke B, Gessesse FG, Meseret F, Wondimeneh F, Desalew A. Parental satisfaction and its associated factors with neonatal intensive care unit services at public hospitals in Bahir Dar, Northwest Ethiopia. Int J Afr Nurs Sci. 2022;17:100435.

Mekonnen WN, Desalegn AM. Quality of Service and Associated Factors in neonatal intensive care unit at Debre Berhan Referral Hospital, Debre Berhan Ethiopia: A Cross Sectional Study. Med J Dr DY Patil Univ. 2018;11(5):412–9.

Adal Z, Atomsa G, Tulu G. Parental Satisfaction with Neonatal Intensive Care Unit Services and Associated Factors in Jimma University Medical Center, Ethiopia. Research and Reports in Neonatology. 2022:1–10.

Fikadu L, Adamu K, Haile Y. Evaluation of Quality of Neonatal Intensive Care Unit Service in Primary Hospitals of Jimma Zone, Southwest Ethiopia. 2022.

Endale H. Parental Satisfaction Concerning their Neonatal Care and Associated Factors in Selected Governmental Hospital, Addis Ababa, Ethiopia, 2017: Addis Ababa University; 2017.

Sileshi E, Mohammed B, Eshetu D, Dure A, Bante A. Parental satisfaction towards Care given at Neonatal Intensive Care Unit and Associated factors in Comprehensive and Referral hospitals of Southern Ethiopia. 2023;2023:3338929.

McFarland LV. Meta-analysis of probiotics for the prevention of antibiotic associated diarrhea and the treatment of Clostridium difficile disease. Am J Gastroenterol. 2006;101(4):812–22.

Egger M, Smith GD. Meta-analysis: potentials and promise. BMJ. 1997;315(7119):1371–4.

Article   CAS   PubMed   PubMed Central   Google Scholar  

Tura G, Fantahun M, Worku A. The effect of health facility delivery on neonatal mortality: systematic review and meta-analysis. BMC Pregnancy Childbirth. 2013;13:18.

Seid Ali M, Getaneh Mekonen E, Shetie Workneh B. Parental satisfaction and associated factors towards neonatal intensive care unit services at the University of Gondar Comprehensive Specialized Hospital, Northwest Ethiopia. Int J Afr Nurs Sci. 2021;15:100364.

Alemu A, Zeleke B, Girma Gessesse F, Meseret F, Wondimeneh F, Desalew A. Parental satisfaction and its associated factors with neonatal intensive care unit services at public hospitals in Bahir Dar, Northwest Ethiopia. Int J Afr Nurs Sci. 2022;17:100435.

Sakonidou S, Andrzejewska I, Webbe J. Interventions to improve quantitative measures of parent satisfaction in neonatal care: a systematic review. 2020;4(1):e000613.

Bleich SN, Ozaltin E, Murray CK. How does satisfaction with the health-care system relate to patient experience? Bull World Health Organ. 2009;87(4):271–8.

Hagen IH, Iversen VC, Nesset E, Orner R, Svindseth MF. Parental satisfaction with neonatal intensive care units: a quantitative cross-sectional study. BMC Health Serv Res. 2019;19(1):37.

Berhane A, Enquselassie F. Patient expectations and their satisfaction in the context of public hospitals. Patient Prefer Adherence. 2016;10:1919–28.

Salehi Z, Nouri JM, Khademolhoseini M, Ebadi A. Survey of parents satisfaction of infants admitted in the NICU. Iran J Crit Care Nurs. 2015;7(4):245–52.

Download references

Acknowledgements

Not appilicable.

No funding was received.

Author information

Authors and affiliations.

Department of Nursing, School of Nursing, College of Medicine and Health Sciences, Wolaita Sodo University, Wolaita Sodo, Ethiopia

Eshetu Elfios, Nefsu Awoke, Temesgen Geta & Christian Kebede

Department of Nursing, School of Nursing, College of Medicine and Health Sciences, Samara University, Samara, Ethiopia

Abdulkerim Hassen

You can also search for this author in PubMed   Google Scholar

Contributions

EE: Designed the study and conceived the data, extracted data, performed the analysis, interpretation of data, drafted the manuscript and approved it. TG, NA, CK, and AH assisted in designing the study and data, interpretation, prepared figures, searched and screened the studies, and critically reviewed the manuscript. All authors read and approved final manuscript.

Corresponding author

Correspondence to Eshetu Elfios .

Ethics declarations

Ethics approval.

Ethical approval and informed consent are not required, as this study is a systematic review and meta-analysis that only involved the use of previously published data.

Consent for publication

Not applicable.

Competing of interest

The authors declare that they have no competing interests.

Additional information

Publisher’s note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ . The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Cite this article.

Elfios, E., Awoke, N., Geta, T. et al. Parental satisfaction with neonatal intensive care unit services and associated factors in Ethiopia: systematic review and meta-analysis. BMC Nurs 23 , 234 (2024). https://doi.org/10.1186/s12912-024-01902-3

Download citation

Received : 29 November 2023

Accepted : 29 March 2024

Published : 08 April 2024

DOI : https://doi.org/10.1186/s12912-024-01902-3

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Parental satisfaction
• Neonatal intensive care unit
• Systematic review

BMC Nursing

ISSN: 1472-6955

• General enquiries: [email protected]

Message placeholder

• S .K. Folomkina, Teaching reading in a foreign language in a non-linguistic institution (Moscow: Vysshaya Shkola Publ., 2005) [Google Scholar]
• Z.K. Klychnikova, Psychology of reading (Moscow: Prosveshchenie Publ., 1975) [Google Scholar]
• Maley, A. Duff The Inward Ear: Poetry in the Language Classroom (Cambridge: CUP, 1993) [Google Scholar]
• S.F. Shatilov, Actual methodological problems of teaching Russian to foreign students (Leningrad: LSU Publ., 1985) [Google Scholar]
• N.N. Gavrilenko, Linguistic and methodical basics of training translators from foreign languages into Russian in the field of science and technology (by the example of translation from French into Russian): Doc. Sci. (Philology) thesis. (Moscow, 2006). [Google Scholar]
• T.G. Agapitova, Teaching foreign-language informative reading in the conditions of forming the professional culture of future foreign-language teachers : Cand. Sci. (Pedagogics) thesis, (Perm, 2002) [Google Scholar]
• N.A. Ippolitova, Reading in the teacher’s professional activity , (Pedagogical rhetoric: Workbook. Moscow, 2001) [Google Scholar]
• S.V. Ivanov, Methods of teaching discourse reading in a foreign language to students of sociopolitical specialties : Cand. Sci. (Pedagogics) thesis, (St. Petersburg, 2011) [Google Scholar]
• I.K. Gapochka, The goals and content of abstractive reading in the Russian (foreign) language, Teaching Russian to foreign graduate students . (Moscow, 1981) [Google Scholar]
• W. Grabe, F. Stoller Teaching and Researching Reading (Longman: Pearson Education, 2004) [Google Scholar]
• N.F. Koryakovtseva, Text as a material for teaching philological reading techniques. Text in the learning process: Col. Papers 282 (Moscow: MSLU Publ., 1987) [Google Scholar]
• Ye. L. Maryanovskaya, The interconnected formation of communicative and educational competencies in teaching home reading to students in a language university : Cand. Sci. (Pedagogics) thesis, (Ryazan, 2007) [Google Scholar]
• O.V. Perlova, Methodology of teaching philological reading as a component of professional training (the English language, language university) : Cand. Sci. (Pedagogics) thesis, (Moscow, 2009) [Google Scholar]
• A.V. Rubtsova, Proc. of the Inter. Sci. Conf. (St. Petersburg (St. Petersburg: Remone Publ., 2012) [Google Scholar]
• L.A. Konstantinova, Stylistics and literary editing . (Tula, 2007) [Google Scholar]
• E. Malushko, O. Maletina, V. Tsybaneva, ASSEHR 97 , 175–180 (2017) [Google Scholar]
• E. Yu. Malushko, International Multidisciplinary Scientific Conferences on Social Sciences and Arts , 1155–1160 (2015) [Google Scholar]
• E. Malushko, O. Maletina, V. Lizunkov, V. Tsybaneva, International Multidisciplinary Scientific Conferences on Social Sciences and Arts , 481–487 (2016) [Google Scholar]

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.

Decomposition Rates and Community Structure of Arthropods in the Litter of Invasive Solidago gigantea Do Not Support the Home-Field Advantage Hypothesis

• Published: 19 July 2022
• Volume 53 , pages 328–334, ( 2022 )

Cite this article

• E. N. Ustinova 1 ,
• M. N. Maslov 1 ,
• S. N. Lysenkov 1 &
• A. V. Tiunov 2  

166 Accesses

Explore all metrics

Decomposition rates of an invasive plant litter in native-species communities can be slower, since decomposers are not adapted to the litter of the invasive species. We have compared rates of plant decomposition and the structure of arthropod communities during the incubation of the litter of the invasive giant goldenrod Solidago gigantea (Asteraceae) and three native species ( Urtica dioica , Cirsium arvense , and Chamaenerion angustifolium ) in the biotopes with dominance of local and invasive plant species. Our results suggest that the arthropod community involved in decomposition of S. gigantea and other species is not species specific and does not provide a higher or lower rate of decomposition of the invasive species. Neither the rate of litter decomposition, nor the structure and diversity of arthropod communities support the home-field advantage hypothesis.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price includes VAT (Russian Federation)

Instant access to the full article PDF.

Rent this article via DeepDyve

Institutional subscriptions

Similar content being viewed by others

Exotic tree and shrub invasions alter leaf-litter microflora and arthropod communities

Gerald R. Woodworth, Jennifer N. Ward & David E. Carr

Increased litter decomposition rates of exotic invasive species Hieracium pilosella (Asteraceae) in Southern Patagonia, Argentina

Karen Braun, M. B. Collantes, … J. A. Anchorena

Do fire and seasonality affect the establishment and colonisation of litter arthropods?

Diego Anjos, Estevao Alves-Silva & Sérvio Pontes Ribeiro

Gholz, H.L., Wedin, D.A., Smitherman, S.M., et al., Long-term dynamics of pine and hardwood litter in contrasting environments: toward a global model of decomposition, GCB , 2000, vol. 6, no. 7, pp. 751–765.

Google Scholar  

Ayres, E., Steltzer, H., Simmons, B.L., et al., Home-field advantage accelerates leaf litter decomposition in forests, Soil Biol. Biochem ., 2009, vol. 41, no. 3, pp. 606–610.

Article   CAS   Google Scholar  

Veen, G.F., Freschet, G.T., Ordonez, A., and Wardle, D.A., Litter quality and environmental controls of home-field advantage effects on litter decomposition, Oikos , 2015, vol. 124, no. 2, pp. 187–195.

Article   Google Scholar  

Ehrenfeld, J.G., Effects of exotic plant invasions on soil nutrient cycling processes, Ecosystems , 2003, vol. 6, no. 6, pp. 503–523.

Allison, S.D. and Vitousek, P.M., Rapid nutrient cycling in leaf litter from invasive plants in Hawai’i, Oecologia , 2004, vol. 141, no. 4, pp. 612–619.

Ashton, I.W., Hyatt, L.A., Howe, K.M., et al., Invasive species accelerate decomposition and litter nitrogen loss in a mixed deciduous forest, Ecol. Appl ., 2005, vol. 15, no. 4, pp. 1263–1272.

Samye opasnye invazionnye vidy Rossii (TOP-100) (The most Dangerous Invasive Species in Russia (TOP-100)), Dgebuadze, Yu.Yu., Petrosyan, V.G., and Khlyap, L.A., Eds., Moscow: KMK, 2018.

Berg, B. and McClaugherty, C., Plant Litter: Decomposition, Humus Formation, Carbon Sequestration , Berlin: Springer-Verlag, 2008.

Book   Google Scholar  

R Core Team. R: A language and environment for statistical computing, R Foundation for Statistical Computing , Vienna, 2020.

Perez, G., Aubert, M., Decaëns, T., et al., Home-Field Advantage: A matter of interaction between litter biochemistry and decomposer biota, Soil Biol. Biochem ., 2013, vol. 67, pp. 245–254.

Ayres, E., Steltzer, H., Berg, S., and Wall, D.H., Soil biota accelerate decomposition in high-elevation forests by specializing in the breakdown of litter produced by the plant species above them, J. Ecol ., 2009, vol. 97, no. 5, pp. 901–912.

Canty, A. and Ripley, B., Boot: Bootstrap R (S-Plus) functions (R package version 1.3-9), R Foundation for Statistical Computing , Vienna, 2013.

Oksanen, J., Kindt, R., Legendre, P., et al., The vegan package, Community Ecology Package , 2007, vol. 10, pp. 631–637.

McTee, M.R., Lekberg, Y., Mummey, D., et al., Do invasive plants structure microbial communities to accelerate decomposition in intermountain grasslands?, Ecol. Evol ., 2017, vol. 7, no. 24, pp. 11227–11235.

Gießelmann, U.C., Martins, K.G., Brändle, M., et al., Lack of home-field advantage in the decomposition of leaf litter in the Atlantic Rainforest of Brazil, Appl. Soil Ecol ., 2011, vol. 49, pp. 5–10.

John, M.G.S., Orwin, K.H., and Dickie, I.A., No ‘home’ versus ‘away’ effects of decomposition found in a grassland–forest reciprocal litter transplant study, Soil Biol. Biochem ., 2011, vol. 43, no. 7, pp. 1482–1489.

Gartner, T.B. and Cardon, Z.G., Decomposition dynamics in mixed-species leaf litter, Oikos , 2004, vol. 104, no. 2, pp. 230–246.

Hättenschwiler, S., Tiunov, A.V., and Scheu, S., Biodiversity and litter decomposition in terrestrial ecosystems, Annu. Rev. Ecol. Evol. Syst ., 2005, vol. 36, pp. 191–218.

Shilenkova, O.L. and Tiunov, A.V., Soil-litter nitrogen transfer and changes in δ13C and δ15N values in decomposing leaf litter during laboratory incubation, Pedobiologia , 2013, vol. 56, no. 3, pp. 147–152.

Gergócs, V., Rétháti, G., and Hufnagel, L., Litter quality indirectly influences community composition, reproductive mode and trophic structure of oribatid mite communities: a microcosm experiment, Exp. Appl. Acarol ., 2015, vol. 67, no. 3, pp. 335–356.

Stefanowicz, A.M., Stanek, M., Nobis, M., and Zubek, S., Species-specific effects of plant invasions on activity, biomass, and composition of soil microbial communities, Biol. Fertil. Soils , 2016, vol. 52, no. 6, pp. 841–852.

Wang, C., Jiang, K., Zhou, J., and Wu, B., Solidago canadensis invasion affects soil N-fixing bacterial communities in heterogeneous landscapes in urban ecosystems in East China, Sci. Total Environ ., 2018, vol. 631, pp. 702–713.

Zhang, C.B., Wang, J., Qian, B.Y., and Li, W.H., Effects of the invader Solidago canadensis on soil properties, Appl. Soil Ecol ., 2009, vol. 43, nos. 2–3, pp. 163–169.

Scharfy, D., Güsewell, S., Gessner, M.O., and Venterink, H.O., Invasion of Solidago gigantea in contrasting experimental plant communities: effects on soil microbes, nutrients and plant-soil feedbacks, J. Ecol ., 2010, vol. 98, no. 6, pp. 1379–1388.

Sterzyńska, M., Shrubovych, J., and Nicia, P., Impact of plant invasion ( Solidago gigantea L.) on soil mesofauna in a riparian wet meadows, Pedobiologia , 2017, vol. 64, pp. 1–7.

Ustinova, E.N., Schepetov, D.M., Lysenkov, S.N., and Tiunov, A.V., Soil arthropod communities are not affected by invasive Solidago gigantea Aiton (Asteraceae), based on morphology and metabarcoding analyses, Soil Biol. Biochem ., 2021, vol. 159, art. ID 108288.

Download references

ACKNOWLEDGMENTS

We thank the experts who carried out taxonomic identification of soil arthropods: Collembola—A.Yu. Korotkevich (Moscow State Pedagogical University, Zoology and Ecology Department); Oribatida—V.D. Leonov (Institute of Ecology and Evolution, Russian Academy of Sciences); Mesostigmata—M.S. Bizin (Institute of Ecology and Evolution, Russian Academy of Sciences).

Author information

Authors and affiliations.

Moscow State University, 119991, Moscow, Russia

E. N. Ustinova, M. N. Maslov & S. N. Lysenkov

Institute for Ecology and Evolution, Russian Academy of Sciences, 119071, Moscow, Russia

A. V. Tiunov

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to E. N. Ustinova .

Additional information

Translated by T. Kuznetsova

Rights and permissions

Reprints and permissions

About this article

Ustinova, E.N., Maslov, M.N., Lysenkov, S.N. et al. Decomposition Rates and Community Structure of Arthropods in the Litter of Invasive Solidago gigantea Do Not Support the Home-Field Advantage Hypothesis. Russ J Ecol 53 , 328–334 (2022). https://doi.org/10.1134/S1067413622040063

Download citation

Received : 28 December 2021

Revised : 03 February 2022

Accepted : 07 February 2022

Published : 19 July 2022

Issue Date : August 2022

DOI : https://doi.org/10.1134/S1067413622040063

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• decomposition of plant residues
• home-field advantage
• invasive plants
• field experiment
• Find a journal
• Publish with us
• Track your research

• Share full article

Advertisement

Supported by

Guest Essay

Why Is Biden Struggling? Because America Is Broken.

By Damon Linker

Mr. Linker, a former columnist at The Week, writes the newsletter Notes From the Middleground.

Seven months away from a rematch election pitting President Biden against former President Donald Trump, the incumbent is struggling. Mr. Biden suffers from persistently low approval ratings, he barely manages to tie Mr. Trump in national head-to-head polls and he lags behind the former president in most of the swing states where the election will be decided (despite some recent modestly encouraging movement in his direction).

The question is why.

When Mr. Biden’s defenders seek to answer the question, most of them tick off declining rates of inflation, historically low unemployment, strong economic growth, a list of legislative accomplishments and other evidence of a successful presidency. This suggests the problem is primarily a failure of communication — the thing flailing administrations always blame first, since it implies the path to improvement requires little more than doing a better job of “getting the message out” about how great the president is doing.

It’s usually wiser to listen to what voters are saying — beyond the obvious concerns about the president’s age.

Recently, Gallup released the latest edition of its longstanding survey measuring “satisfaction with the way things are going in the U.S.” Three out of four Americans (75 percent) claimed to be dissatisfied. The long-term trend tells a clear story: From the mid-1990s to late 2004, the level of satisfaction bounced around between 39 percent and 71 percent. But in the aftermath of the George W. Bush administration’s failure to find weapons of mass destruction in Iraq and during a yearslong violent insurgency challenging American military occupation of the country, numbers began to slide. They would reach a low of 9 percent satisfaction in October 2008, in the midst of the worst financial crisis since the Great Depression.

What followed was a very slow 12-year recovery of satisfaction across almost the entirety of the Obama and Trump administrations, with a post-2004 high of 45 percent reached in February 2020, on the eve of the outbreak of the Covid-19 pandemic. By January 2021, the level of satisfaction was back down to 11 percent, just two points off its historical low. Under Joe Biden, Americans briefly became somewhat more upbeat — but figures have sunk again from the mid-30s to the high teens and low 20s in recent months.

These findings mirror what other pollsters have found when they asked respondents about whether they think the country is on the right or wrong track, and about their trust in government and confidence in American institutions . The latter number has been slowly falling since the 1960s, but it, too, really began to collapse in 2004, eventually reaching the low 30s by 2007. In 2023, just 26 percent of Americans expressed confidence in our institutions.

In January 2021, Alana Newhouse published an essay in Tablet, “Everything Is Broken,” that gave voice to this incredibly widespread (but underreported) sentiment. Why did so many people in the United States believe that, as Ms. Newhouse put it in a follow-up essay , “whole parts of American society were breaking down before our eyes”?

The examples are almost too numerous to list: a disastrous war in Iraq; a ruinous financial crisis followed by a decade of anemic growth when most of the new wealth went to those who were already well off; a shambolic response to the deadliest pandemic in a century; a humiliating withdrawal from Afghanistan; rising prices and interest rates; skyrocketing levels of public and private debt; surging rates of homelessness and the spread of tent encampments in American cities; undocumented migrants streaming over the southern border; spiking rates of gun violence, mental illness, depression, addiction, suicide, chronic illness and obesity, coupled with a decline in life expectancy.

That’s an awful lot of failure over the past 20-odd years. Yet for the most part, the people who run our institutions have done very little to acknowledge or take responsibility for any of it, let alone undertake reforms that aim to fix what’s broken. That’s no doubt why angry anti-establishment populism has become so prominent in our politics over the past decade — with Mr. Trump, a political outsider, taking over the Republican Party in 2016 by running against the elites of both parties, and Senator Bernie Sanders giving the establishment favorite Hillary Clinton a run for her money that same year by taking on the banking and finance sectors of the economy, along with their Democratic and Republican enablers.

Mr. Biden has never been that kind of politician. Most of the time he speaks and acts as if he thinks American institutions are doing perfectly fine — at least so long as Mr. Trump doesn’t get his hands on them. Part of that is undoubtedly because Mr. Biden is an incumbent, and incumbents always find themselves having to defend what they’ve done in office, which isn’t compatible with acting like an insurgent going to war against the system.

Then there’s the fact that Mr. Biden has worked within our elected institutions since the Nixon administration, making him deeply invested in them (and implicated in their failures). Finally, as a Democrat who came of age during the heyday of mid-20th-century liberalism, Mr. Biden is wedded to the idea of using a functional, competent and capable federal government to improve people’s lives — whether or not more recent history validates that faith.

This places him badly out of step with the national mood, speaking a language very far removed from the talk of a broken country that suffuses Mr. Trump’s meandering and often unhinged remarks on the subject. The more earnest statements of the third-party candidates Robert F. Kennedy Jr. , Cornel West and Jill Stein also speak to aspects of our brokenness, taking ample and often nostalgic note of what’s gone wrong and promising bold, if vague, action to begin an effort of repair.

That leaves Mr. Biden as the lone institutionalist defender of the status quo surrounded by a small army of brokenists looking for support from an electorate primed to respond to their more downcast message.

There may be limits to what Mr. Biden can do to respond. For one thing, his 81-year-old frailty can’t help appearing to mirror the fragile state of our public institutions. For another, in an era of political bad feeling, when presidential approval ratings sink quickly and never recover, incumbents from both parties may no longer enjoy the kind of advantage in seeking re-election that they once did, at least at the national level.

Still, there are things the Biden campaign could do to help the president better connect with voters.

First, he should stop being so upbeat — about the economy in particular — and making the election entirely about the singular awfulness of his opponent. While the latter sounds evasive, the former makes the president seem hopelessly out of touch and risks antagonizing people who aren’t in the mood for a chipper message.

Mr. Biden should instead try to meet Americans where they are. He should admit Washington has gotten a lot of things wrong over the past two decades and sound unhappy about and humbled by it. He could make the argument that all governments make mistakes because they are run by fallible human beings — but also point out that elected representatives in a democracy should be upfront about error and resolve to learn from mistakes so that they avoid them in the future. Just acknowledging how much in America is broken could generate a lot of good will from otherwise skeptical and dismissive voters.

Even better would be an effort to develop a reform agenda: Mr. Biden could declare it’s long past time for America to put its house in order, to begin cleaning up the messes of the past two decades, to face our problems and return to our own best national self. He might even think of adapting and repurposing for the center-left a few lines from Ronald Reagan’s first Inaugural Address : “It’s not my intention to do away with government. It is rather to make it work — work with us, not over us; to stand by our side, not ride on our back. Government can and must provide opportunity, not smother it; foster productivity, not stifle it.”

In concrete terms, this means pledging to reform existing institutions and programs, not promising to build new ones on top of the ambitious legislation and substantial spending Congress passed during Mr. Biden’s first two years in office. It means, instead, a commitment to pause and begin assessing what government has been doing at all levels, under both Republican and Democratic leadership, over the past two decades.

It means, more specifically, a resolution to continue and expand existing reviews into what worked and what didn’t during the pandemic — in red states and blue states, in cities, suburbs and small towns — in order to prepare for a better response the next time we confront a public-health emergency. It means talking honestly about the surging and unsustainable national debt and what it will take to begin reining it in. It means trying to help government function better, including a concerted effort to increase state capacity , eliminate regulations that constrain the nation’s housing supply and build on the administration’s attempts at permitting reform to streamline or remove regulations that slow down and increase the cost of private as well as public development.

These projects will far outlast a second Biden term. But the president can promise to get them started, with the remaining work to be completed by presidents and generations to come.

Taking this approach may help to neutralize the populist advantages Mr. Trump enjoys (at least when he isn’t running as an incumbent). However much voters appreciate his denunciations of a corrupt and rigged system, as well as his management of the economy over the first three years of his presidency, they have no love for the G.O.P.’s obsession with pairing cuts to entitlement programs and upper-income tax rates with draconian restrictions on abortion — not to mention Mr. Trump’s focus on personal grievances and legal recklessness. That leaves plenty of room for Mr. Biden to make a case for himself as the guy who can enact the sweeping reforms American needs, and without all the unnecessary and dangerous drama a second Trump administration would surely bring.

Everything is broken — or so it feels to many of our fellow citizens. Denying this reality only empowers populist candidates whose message acquires its potency by pointing to an entrenched political establishment unwilling or unable to learn from (or even admit) its myriad mistakes. That shirking needs to stop. And it should do so with Joe Biden.

Damon Linker, who writes the newsletter “ Notes From the Middleground ,” is a senior lecturer in the department of political science at the University of Pennsylvania and a senior fellow at the Open Society Project at the Niskanen Center.

The Times is committed to publishing a diversity of letters to the editor. We’d like to hear what you think about this or any of our articles. Here are some tips . And here’s our email: [email protected] .

Follow the New York Times Opinion section on Facebook , Instagram , TikTok , WhatsApp , X and Threads .

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Int J Environ Res Public Health

Nursing Interventions That Promote Sleep in Preterm Newborns in the Neonatal Intensive Care Units: An Integrative Review

Catarina firmino.

1 Escola Superior de Saúde Egas Moniz, Caparica, 2829-511 Almada, Portugal

Marlene Rodrigues

Sofia franco, judicília ferreira, ana rita simões, cidália castro.

2 Centro de Investigação Interdisciplinar Egas Moniz (CiiEM), 2829-511 Almada, Portugal

Júlio Belo Fernandes

3 Grupo de Patologia Médica, Nutrição e Exercício Clínico (PaMNEC), 2829-511 Almada, Portugal

Associated Data

The data presented in this study are available upon request from J.B.F.

Sleep is a crucial factor for the psychological and physiological well-being of any human being. In Neonatal Intensive Care Units, preterm newborns’ sleep may be at risk due to medical and nursing care, environmental stimuli and manipulation. This review aims to identify the nurses’ interventions that promote sleep in preterm newborns in the Neonatal Intensive Care Units. An integrative review was conducted following Whittemore and Knafl’s methodology and the 2020 Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) statement. The research was carried out on the electronic databases PubMed, Cochrane Central Register of Controlled Trials, Cochrane Database of Systematic Reviews, and ScienceDirect, with a timeframe from 2010 to 2021. A total of 359 articles were initially identified. After selection and analysis, five studies were included in the sample. Interventions by nursing staff that promote sleep in preterm newborns in the Neonatal Intensive Care Units fall within three categories: environmental management, relaxation techniques and therapeutic positioning. Nurses play a vital role in implementing interventions that promote preterm newborns’ sleep. They can positively affect preterm newborns’ sleep by controlling environmental stimuli and applying relaxation techniques and therapeutic positioning to their care practices.

1. Introduction

Worldwide, it is estimated that 5 million children are born preterm each year, and one million of these newborns will not survive the first year of life due to complications associated with prematurity [ 1 ].

In addition, complications associated with preterm delivery are the leading cause of death in children under five years of age [ 2 ]. It is estimated that three-quarters of these deaths can be prevented using the available interventions. These interventions should focus on using incubators, nasogastric tubes, and ventilators, in addition to implementing communication strategies aiming to reach at-risk populations and educating healthcare professionals and others in healthcare delivery systems to play key roles in reducing the risk of preterm birth. Furthermore, worldwide, the incidence of preterm births is rising with the result that prematurity is considered a major public health problem [ 3 ].

Preterm is defined as babies born alive before 37 weeks of pregnancy are completed [ 1 ]. Prematurity is the primary reason for hospitalization in neonatal intensive care units [ 4 , 5 , 6 ].

In addition to a lower birth weight, gestational age, and susceptibility to maturation and prematurity complications, preterm newborns in neonatal intensive care units are exposed to several stressors, including separation from the mother, frequent nursing or medical interventions, and excessive sound or light intensity [ 7 , 8 , 9 ].

In neonatal intensive care units, preterm newborns can be exposed to stressful events, such as painful clinical procedures, as many as 5 to 15 painful clinical procedures daily [ 10 ]. This stress may overwhelm the newborns and result in autonomic instability leading to adverse cardiac and respiratory changes and reduced levels of oxygen saturation. For example, pain can lead to significant changes in cardiac dynamics and loss of complexity of heart rate fluctuations. These changes include the altered fractal organization of heart rate variability that resemble some life-threatening conditions [ 11 , 12 , 13 ]. Stress has been associated with potentially long-lasting effects on newborns’ brain organization and neuroendocrine stress responses. Furthermore, current studies reported epigenetic changes in preterm newborns exposed to high-stress levels during the neonatal period [ 14 , 15 ]. Stress may also predispose preterm newborns to sleep disturbances caused by frequent interruptions or noise [ 13 ].

Sleep is the recurring physiologic state characterized by altered consciousness and is of great importance for optimal development [ 16 ].

Sleep is a critical issue in current neonatal intensive care unit practice for preterm newborns [ 17 ]. Fetuses and newborns spend most of their time sleeping. In the fetus, sleep and sleep cycles develop between 26 and 28 weeks of gestational age, and rapid eye movement sleep develops at around 28 to 30 weeks [ 18 ]. Even though quiet sleep and active sleep are present in newborns, for reasons not known, they do not establish a circadian rhythm for sleep. As a result, newborns are predisposed to have inefficient and easily interrupted sleep cycles [ 19 ].

In contrast, adults’ sleep occurs in two qualitatively different phases that display cyclicity: the active phase known as rapid eye movement sleep and the non-active phase called non-rapid eye movement sleep [ 16 ].

Newborns have different sleep patterns than infants and adults. The literature describes that, in newborns, sleep occurs in distinct phases, namely active sleep, quiet sleep, and indeterminate sleep [ 20 ]. In the active sleep phase, there is rapid eye movement as in older children and adults. This is the main sleep phase. In addition to rapid eye movement, it is also characterized by irregular breathing, sporadic motor movements, and a continuous electroencephalography pattern. In the quiet sleep phase, there is non-rapid eye movement, regular breathing patterns, an absence of motor movements, and a discontinuous electroencephalography pattern. Finally, in the indeterminate sleep phase, sleep characteristics cannot be clearly classified as active or quiet sleep [ 20 , 21 ].

The newborn sleep pattern organized into distinct phases allows normal neurodevelopmental outcomes. In addition, evidence shows that sleep cyclicity is key to the maintenance of brain plasticity, specifically the ability of the brain to reorganize its neural pathways in the face of environmental stimuli [ 22 ].

In preterm newborns, the lack of rest or sleep is a significant stressor and may harm the infant’s overall development [ 13 ]. Studies revealed that a longer sleep duration, regularity, and quality of sleep are associated with improved attention span, behavior, cognitive functioning, emotional regulation, and physical health in children [ 23 ]. Although the effects of sleep disturbance in preterm newborns are not known with certainty, sleep deprivation is associated with physiologic instability and less than optimal developmental outcomes. Current studies support that sleep plays a vital role in preterm newborns’ overall growth and development [ 17 , 24 , 25 , 26 ]. Consequently, it is considered a crucial action, similar to breathing and nutrition [ 25 ].

Preterm newborns receive multisensory stimulation in neonatal intensive care units, which can lead to difficulty establishing circadian rhythm [ 27 ]. In addition, newborns are exposed to continuous stimuli, leading to sleep-wake transition disruptions. A past study showed that newborns’ sleep is interrupted about 234 times in 24 h [ 28 ]. The main neonatal intensive care units’ environmental factors that disturb newborns’ sleep are sound and light intensity. When preterm newborns are admitted to the neonatal intensive care units, their optimal development may be at risk due to environmental stimuli, and medical and nursing care [ 25 ]. Therefore, special attention should be paid to preterm newborns’ sleep in the neonatal intensive care units.

When caring for preterm newborns, nurses should be aware of sleep stages, the benefits of sleep, and its disturbing and inducing factors in planning and performing practices that cause minimal distress to preterm newborns. Consequently, developing nursing interventions that promote sleep for preterm newborns in neonatal intensive care units is an important area of research. This review aims to identify the nurses’ interventions that promote sleep in preterm newborns in neonatal intensive care units.

2.1. Design

The present integrative review was drawn based on the 2020 Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) statement [ 29 ]. We used the methodological approach proposed by Whittemore and Knafl [ 30 ], involving five stages: (1) problem identification, (2) literature search, (3) data evaluation, (4) data analysis, and (5) presentation.

The question that guided this integrative review was defined in accordance with population, concept, and context (PCC) questions. What are nursing interventions (C) that promote sleep in the preterm newborn (P) hospitalized in neonatal intensive care units (C)?

2.2. Search Methods

The literature search was conducted using PubMed, Cochrane Central Register of Controlled Trials, Cochrane Database of Systematic Reviews, and ScienceDirect databases. The final search was performed on 7 October 2021.

Both DeCS and MeSH health sciences descriptors were combined with Boolean operators using the following search string:

(“Newborn” OR “Preterm”) AND (“Sleep”) AND (“Neonatal Intensive Care Unit”).

The selection criteria were: documents written in Portuguese and English, published between 2010 and 2021, which addressed or referred to the nursing interventions that promote sleep for preterm newborns (born at less than 37 weeks of gestation) receiving care in the NICU. All documents that did not meet the selection criteria were excluded from the review.

2.3. Study Selection

To increase consistency, the search, selection, and extraction of data were carried out independently by two researchers. After duplicate elimination, researchers proceeded with a selection process that enrolled three phases. In the first phase, researchers screened the titles, followed by abstract analysis. Finally, researchers obtained the full text of relevant documents and read them thoroughly. This process allowed verifying the relevance and appropriateness of the selected documents according to the inclusion and exclusion criteria and the research question. If there was disagreement, a third reviewer made the final decision.

2.4. Search Outcomes

The initial database search identified 359 articles. After duplicate articles were removed, 352 titles and abstracts were reviewed, of which 9 were considered suitable for a full-text review. At the end of the screening process, five studies met the eligibility criteria and were included in this review. The flow chart describing the screening process is presented in Figure 1 .

PRISMA flow diagram for study selection.

2.5. Quality Appraisal

Researchers defined the quality of the selected documents using the Joanna Briggs Institute levels of evidence and grading, ranging from 1c to 4b.

The Joanna Briggs Institute Critical Appraisal Checklist was applied to each study.

The checklist aims to assess the study’s research design and the validity of its results using a sequence of appraisal questions with four possible answers (“yes”, “no”, “unclear”, or “not applicable”).

For scores below 49%, the study should be considered as having a high risk of bias, between 50% and 69% a moderate risk of bias, and more than 70% a low risk of bias.

The methodological rigor of the five studies selected by researchers ranged from 82% to 100%, which was considered a low risk of bias.

2.6. Data Extraction and Synthesis

A data extraction form was developed to guide the data extraction.

This instrument allowed the extraction of the following data: authors, publication year, study title, study design, aim, and findings. All data items extracted were cross-checked.

This integrative literature review allowed the identification of five articles that focus on sleep promotion in preterm newborns in the Neonatal Intensive Care Units. Out of the five studies, there were two studies conducted in the United States [ 13 , 31 ], one in Greece [ 32 ], one in Turkey [ 33 ], and another in Brazil [ 34 ]. Research studies were primarily involved randomized controlled trials [ 32 , 33 ], one randomized crossover pilot study [ 13 ], one systematic review of reviews [ 34 ], and one cross-sectional study [ 31 ].

A summary of the included articles with an overview of their key characteristics is provided in Table 1 .

Data extraction and synthesis.

Data analysis revealed several nursing interventions that promote sleep in preterm newborns. Using an inductive analysis process, we grouped the different interventions into three categories based on the differences and similarities found between them. Each category is detailed below.

3.1. Category 1: Environmental Management

A study developed by Boutopoulou, Effrossine, Despoina, Konstatntinos, and Matziou [ 32 ] aimed to measure neonatal non-rapid eye movement sleep duration and how it was affected by the implementation of improved nursing conditions verified that, by reducing sound or light intensity, the duration of non-rapid eye movement sleep increased significantly. In this study, researchers applied specific earplugs (Minimuffs Neonatal Noise attenuators, Natus) to reduce the sound intensity and minimize ambient auditory stimuli reaching neonates’ ears. This intervention allowed to reduce the perceived volume by about 30 dB. For the reduction in light intensity, researchers used incubator covers.

Researchers concluded that providing low noise and light levels within neonatal intensive care units improves the structural organization of sleep with more prolonged non-rapid eye movement periods.

3.2. Category 2: Relaxation Techniques

In the category relaxation techniques, pediatric massage emerged as an intervention that promotes sleep in newborns. Two articles address this intervention. First, the study conducted by Yates, Mitchell, Booth, Williams, Lowe, and Hall [ 13 ] aimed to determine whether massage therapy can be used as an adjunct intervention to induce sleep in infants born preterm. Researchers adapted the massage protocol originally published by Field et al. [ 35 ], which resulted in an overall massage time of approximately 10 min.

Baby lotion (Johnson & Johnson) was used to assist with ease of skin-to-skin contact during moderate pressure massage. The infant was undressed to the diaper and covered with a blanket during massage to maintain warmth. Each infant received two repetitions of a series of defined strokes to five body areas. Massage occurred over 1 min intervals with the application of 12 strokes lasting approximately five seconds each for each of the body areas receiving massage. The following sequence was performed: (1) in the prone position, the infant was stroked from the top of the head to the neck and back to the top of the head and back to the neck; (2) from the neck across the shoulders; (3) from the upper back to the waist and back to the upper back; (4) from the thigh to the foot and back to the thigh on both legs; and (5) from the shoulder to the hand and back to the shoulder on both arms.

The review performed by de Britto Pereira, Mendes Abdala, Portella, Ghelman, and Schveitzer [ 34 ] analyzed 38 reviews that evaluated pediatrics’ massage as an intervention in several health outcomes. The outcomes were divided into four major groups: physical and metabolic effects; vitality, well-being, and quality of life; mental health; and management. This review showed the positive effects of massage in promoting quality of sleep.

Another intervention that promotes sleep in newborns that encompasses the category relaxation techniques is tub bathing.

In the study developed by Taşdemir and Efe [ 33 ], tub bathing emerged as an intervention with positive results in promoting sleep in the preterm newborn. Bathing was performed anywhere from 6 to 48 h post-birth, between 8 a.m. and 11 a.m., with minimal nutritional and other interventions provided to the infant during this period.

Bathing was performed at room temperature measured at 25–26 °C and 40% humidity. The water level in the bath was set at approximately 9–12 cm or deep enough to cover the baby’s shoulders. The bathwater temperature was controlled using a water thermometer and set at 37–38 °C. A folded cloth towel was placed into the tub before bathing. During bathing, infants were spoken to softly, and their bodies were cleaned in a slow, rhythmical motion. The infant’s face was washed and dried before immersion. First, the lower part of the body was immersed in a tub before immersion up to the neck. The infant was held securely; the head and neck were supported on the researcher’s forearm, and the shoulder was grasped using the researcher’s thumb and finger. Cleaning was performed using a soft cloth and an infant skin cleaner. The front and back areas were cleaned without turning the infant. Bathing took approximately 3.64 ± 0.77 min. Then, the baby was safely removed from the water and wrapped in a clean, soft towel. The body was quickly dried with gentle movements, baby oil applied, and the nappy put on.

3.3. Category 3: Therapeutic Positioning

A study developed by Zarem, Crapnell, Tiltges, Madlinger, Reynolds, Lukas, and Pineda [ 31 ], aiming to determine perceptions about positioning for preterm infants in the neonatal intensive care units, identified that, in comparison to other positioning methods (i.e., Sleep Sack, Snuggle Up, nesting, boundaries, and swaddling), the Dandle Roo is the easiest to use and most beneficial. The Dandle Roo is a device made of stretchable cotton that provides containment, allowing the infant to move the extremities into extension, followed by recoil to flexion and midline orientation [ 31 ]. Although the participants recognized the benefits of this positioning method, they acknowledged that good positioning might be achieved in various ways. For example, the use of nesting and blankets can still facilitate positive results for the preterm infant.

4. Discussion

The current review provides a comprehensive understanding of nursing interventions that promote sleep in preterm newborns hospitalized in the neonatal intensive care units. A total of three categories of interventions ( environmental management, relaxation techniques, and therapeutic positioning) were identified from five studies. The results show that nurses can implement different strategies to promote sleep in preterm newborns hospitalized in neonatal intensive care units.

Preterm newborns are exposed to various stimuli in neonatal intensive care units, causing frequent sleep–wake transition disruptions and leading to sleep disorganization. In addition, neonatal intensive care units are excessively sensory environments with high sound and light intensity levels. These two factors are the most common environmental factors that disturb neonatal sleep [ 36 , 37 , 38 , 39 ].

In most neonatal intensive care units, the light comes from artificial sources, such as examination lights, phototherapy lamps, and ambient space light, varying in intensity according to the unit’s needs during the 24 h day. The sound is mainly produced by healthcare workers and medical devices, such as monitors, respiratory equipment, and double-walled incubators.

Several studies associate the light and sound intensity with autonomic nervous system disorders, increased heart rate and vasoconstriction, delay in obtaining complete enteral nutrition, and uncontrolled circadian cycle [ 40 ]. There is also evidence that newborns’ non-rapid eye movement sleep increases when sound or light intensity is reduced [ 32 ]. Therefore, improving these parameters of nursing practices may facilitate the newborns’ sleep duration and result in better neurodevelopmental outcomes [ 41 ].

Despite the positive conclusions drawn by Boutopoulou, Effrossine, Despoina, Konstatntinos, and Matziou [ 32 ] regarding the use of earplugs to promote newborns’ sleep, the authors considered that this method cannot be regarded as a practical clinical approach for noise reduction. Instead, the authors suggest alternative nursing practices such as lowering the tone of conversations and reducing the intensity of noises produced by alarms of monitoring devices.

Concerning incubator covers, this is already a method instituted by many neonatal intensive care units to minimize the intensity of light affecting newborns [ 42 ].

Among the relaxation techniques to help to promote sleep, pediatric massage is an inexpensive intervention that should be incorporated into nursing practices.

For long, the introduction of massage therapy into nursing practices has been delayed due to fear of overstimulating the infant. However, evidence supports its safety and shows that the significant benefits outweigh the minimal risks [ 34 ]. In the study conducted by Yates, Mitchell, Booth, Williams, Lowe, and Hall [ 13 ], although newborns did not demonstrate induction of sleep immediately after the massage, their response of increased wakefulness may be enlightened by evidence that massage therapy enhances the electrical activity and brain maturation in preterm newborns [ 43 , 44 ]. The review performed by de Britto Pereira, Mendes Abdala, Portella, Ghelman, and Schveitzer [ 34 ] revealed the benefits of implementing massage as a sleep-promoting strategy in preterm newborns. In addition, other benefits of massage therapy include the stabilization of the autonomic nervous system, promotion of growth and development, and shorter hospital stays [ 45 , 46 , 47 ].

Current evidence shows that massage therapy is a low-cost, safe practice associated with multiple benefits for preterm newborns and, therefore, should be incorporated into nurses’ practice.

Another relaxation technique identified in this review was tub bathing. The study developed by Taşdemir and Efe [ 33 ] suggests that tub bathing can be effective in reducing infant crying and helping them to sleep. In addition, when compared with sponge bathing, tub bathing had a more significant stress reduction effect in preterm newborns.

Bathing can be an extremely stressful agent for preterm newborns, which may lead to some behaviors, such as agitation, crying, and hiccoughing [ 48 ]. However, in Taşdemir and Efe’s [ 33 ] study, tub bathing positively affected the newborns’ comfort and helped them to maintain their regular heart rate. Consequently, tub bathing is a stress reduction intervention that promotes sleep in preterm newborns in neonatal intensive care units. The evidence provided by this study supports the need to emphasize the nurses’ role in the care provided to preterm newborns and the integration of tub bathing in nurses’ practice. Furthermore, the results of this study indicate that tub bathing helps to prevent thermodynamic instability, stress, and impairments in physiological parameters. For this reason, it is an important method to be promoted to obtain better health outcomes, specifically, sleep promotion.

Finally, the use of positioning aids was identified as a nurse intervention that promotes sleep in preterm newborns. The last pregnancy trimester encourages the development of physiologic flexion and midline orientation [ 49 , 50 ], preparing the fetus for later function, supporting neurodevelopment, and promoting self-soothing [ 51 ]. By being born preterm, newborns are deprived of this critical experience, frequently resulting in low muscle tone and strength that cause them to maintain their bodies in extended positions [ 52 , 53 ]. This position may affect the newborns’ development, inhibit self-regulation [ 54 ], and interfere with their ability to interact and attach to their caregivers [ 31 ].

Neonatal intensive care unit nurses are responsible for multiple aspects of care. Therapeutic positioning is one of these aspects of care that can have critical developmental effects on preterm newborns. To minimize the sequelae of prematurity, nurses try to encourage adequate positioning using various methods. The study developed by Zarem, Crapnell, Tiltges, Madlinger, Reynolds, Lukas, and Pineda [ 31 ] identified that the Dandle Roo was considered the most accessible positioning aid and the most beneficial one. However, other aids, such as the sleep sack, snuggle up, nesting, boundaries, and swaddling, are also effective in promoting the newborn’s adequate positioning and sleep.

Nurses must be aware of the importance of sleep for preterm newborns as a necessary neurodevelopmental process and also be acquainted with interventions that promote sleep in this population. The findings from this review allowed us to understand that nurse practice has a predominant effect on preterm newborn development. Several interventions were identified to promote sleep in preterm newborns in the neonatal intensive care units. In addition, it is also essential to educate and include parents to promote appropriate care to meet the newborns after discharge.

This review identifies several studies that report sufficient detail of its interventions permitting replication. This state of play allows nurses to implement those interventions in their daily practice. Nonetheless, this research has several limitations. First, limiting the search to four databases and imposing time limits may have excluded some relevant studies. Second, we have to consider the exclusion of the literature written in languages other than English and Portuguese. Third, the low number of studies included in the review. While there has been an increasing body of evidence focusing on preterm newborns, this review shows a gap regarding the nursing interventions for promoting sleep in preterm newborns as only five studies were identified. Based on the possible multiple benefits of sleep promotion care practices for preterm newborns and the scarcity of studies focusing on this subject, further research is needed.

5. Conclusions

Sleep is generally considered vital for brain development and growth during the neonatal period. In addition, identifying and implementing appropriate nursing care practices that ensure physiologic stability, growth and development, rest, sleep, and the ability to cope with procedures and other numerous stressors are believed to positively affect the newborns’ sleep organization. Therefore, sleep promotion is essential for preterm newborns’ optimal development.

The findings of this review will contribute to the advancement in nursing care practice addressing a critical issue of improving care of preterm newborns in the NICU impacting their early development. By managing the environment by reducing sound and light intensity, nurses ensure that the ambient stimuli that reach newborns are minimized and, therefore, promote their sleep duration and result in better neurodevelopmental outcomes. Better sleep outcomes can also be achieved by introducing relaxation techniques, such as pediatric massage or tub bathing in nursing practice. In addition, therapeutic positioning, with or without positioning aids, can also positively affect preterm newborns’ sleep.

Acknowledgments

This publication was financed by national funds through the FCT—Foundation for Science and Technology, I.P., under the project UIDB/04585/2020. The researchers would like to thank the Centro de Investigação Interdisciplinar Egas Moniz (CiiEM) for the support provided for the publication of this article.

Funding Statement

This research received no external funding.

Author Contributions

Conceptualization: A.R.S., C.F., J.F., M.R., S.F., C.C. and J.B.F.; Formal analysis: A.R.S., C.F., J.F., M.R., S.F., C.C. and J.B.F.; Investigation: A.R.S., C.F., J.F., M.R., S.F., C.C. and J.B.F.; Methodology: A.R.S., C.F., J.F., M.R., S.F., C.C. and J.B.F.; Writing—original draft preparation: A.R.S., C.F., J.F., M.R., S.F., C.C. and J.B.F.; Project administration: C.C. and J.B.F.; Data curation: C.C. and J.B.F. All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Data availability statement, conflicts of interest.

The authors declare no conflict of interest.

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.