Artificial Intelligence to Improve Health Outcomes in the NICU and PICU: A Systematic Review

2021 ◽  
Author(s):  
Claudette O. Adegboro ◽  
Avishek Choudhury ◽  
Onur Asan ◽  
Michelle M. Kelly
Keyword(s):  
Artificial Intelligence ◽  
Health Outcome ◽  
Intensive Care ◽  
Health Outcomes ◽  
Data Extraction ◽  
Quality Care ◽  
Experimental Studies ◽  
Receiver Operating Curve ◽  
Critically Ill Children ◽  
Statistical Measures

CONTEXT: Artificial intelligence (AI) technologies are increasingly used in pediatrics and have the potential to help inpatient physicians provide high-quality care for critically ill children. OBJECTIVE: We aimed to describe the use of AI to improve any health outcome(s) in neonatal and pediatric intensive care. DATA SOURCE: PubMed, IEEE Xplore, Cochrane, and Web of Science databases. STUDY SELECTION: We used peer-reviewed studies published between June 1, 2010, and May 31, 2020, in which researchers described (1) AI, (2) pediatrics, and (3) intensive care. Studies were included if researchers assessed AI use to improve at least 1 health outcome (eg, mortality). DATA EXTRACTION: Data extraction was conducted independently by 2 researchers. Articles were categorized by direct or indirect impact of AI, defined by the European Institute of Innovation and Technology Health joint report. RESULTS: Of the 287 publications screened, 32 met inclusion criteria. Approximately 22% ( n = 7) of studies revealed a direct impact and improvement in health outcomes after AI implementation. Majority were in prototype testing, and few were deployed into an ICU setting. Among the remaining 78% ( n = 25) AI models outperformed standard clinical modalities and may have indirectly influenced patient outcomes. Quantitative assessment of health outcomes using statistical measures, such as area under the receiver operating curve (56%; n = 18) and specificity (38%; n = 12), revealed marked heterogeneity in metrics and standardization. CONCLUSIONS: Few studies have revealed that AI has directly improved health outcomes for pediatric critical care patients. Further prospective, experimental studies are needed to assess AI’s impact by using established implementation frameworks, standardized metrics, and validated outcome measures.

scholarly journals High Red Cell Distribution Width as a Biomarker of Mortality in Critically Ill Paediatric Patients

2020 ◽  
Vol 40 (3) ◽  
pp. 210-216
Author(s):  
Eva Gauchan ◽  
Sahisnuta Basnet
Keyword(s):  
Intensive Care ◽  
Critically Ill ◽  
Prognostic Indicator ◽  
Red Cell Distribution Width ◽  
Receiver Operating Curve ◽  
Critically Ill Children ◽  
Red Cell ◽  
Cell Distribution ◽  
Distribution Width ◽  
Complicated Course

Introduction: Red cell distribution width (RDW) is a frequently overlooked parameter in routine haematological reports. It is a simple and inexpensive test which has been found by many adult studies to be a prognostic indicator of mortality in intensive care units. The objective of this study was to see if high RDW could be used as a marker to predict mortality in critically ill children. Methods: This was a prospective observational study conducted in the paediatric intensive care unit (PICU) of a tertiary hospital of Western Nepal over a period of one year. Study subjects were selected by purposive sampling method. RDW at admission and relative change in RDW (ΔRDW) was compared to see if they had any role in predicting mortality in this group of children. Receiver operating curve analysis was plotted to find an optimal cut-off point to define high and low RDW and various outcome parameters were analysed. Results: Out of 131 children, there were 12 (9.1%) mortalities. Admission RDW was higher in the death group as compared to the survivor group (17 vs 14.6; p = 0.012). Similar finding was seen with ΔRDW (0.45 vs 0.00; p = 0.006). ΔRDW above the cut-off value of 0.15 was found to be associated with a generally more complicated course during hospitalisation as well as had more risk of mortality. Conclusions: Both RDW and ΔRDW above the cut-off value were found to be associated with mortality. In addition, high ΔRDW was also found to predict a more complicated course during hospitalisation.

Effects of weight-neutral approaches compared with traditional weight-loss approaches on behavioral, physical, and psychological health outcomes: a systematic review and meta-analysis

2019 ◽  
Vol 78 (1) ◽  
pp. 39-55
Author(s):  
Jaslyn A Dugmore ◽  
Copeland G Winten ◽  
Hannah E Niven ◽  
Judy Bauer
Keyword(s):  
Weight Loss ◽  
Health Outcomes ◽  
Psychological Health ◽  
Data Extraction ◽  
Meta Analysis ◽  
Behavioral Outcomes ◽  
Experimental Studies ◽  
Final Analysis ◽  
Evaluation Tool ◽  
Level Of Evidence

Abstract Context Weight-neutral approaches for health are emerging therapeutic alternatives to traditional weight-loss approaches. The existing literature base comparing these approaches has not yet been systematically evaluated by a meta-analysis. Objective This review aims to determine if weight-neutral approaches are valid alternatives to weight-loss approaches for improving physical, psychological, and behavioral health outcomes. Data Sources Embase, Scopus, PsycINFO, PubMed, CINAHL, and the University of Queensland Library databases were searched. Study Selection Peer-reviewed, experimental, or quasi-experimental studies that included weight-neutral and weight-loss arms and reported physical, psychological, or behavioral outcomes were eligible. A total of 525 studies were identified through initial database searches, with 10 included in the final analysis after exclusion criteria were applied. Data Extraction Screening and eligibility assessment of studies followed the PRISMA protocol. The following outcomes were extracted: weight, body mass index, lipid and glucose variables, blood pressure, eating behavior, self-esteem, depression, quality of life, physical activity, and diet quality. Data Analysis Studies were graded per the National Health and Medical Research Council (NHMRC) level-of-evidence tool and the Academy of Nutrition and Dietetics quality-evaluation tool. Effect sizes were examined as a meta-analysis of standardized and mean differences using a random-effects inverse-variance model with 95%CIs. Practice recommendations for each outcome were graded per NHMRC body-of-evidence guidelines. Conclusions Weight-neutral approaches resulted in greater improvement in bulimia (P = 0.02), but no significant differences were observed for any other outcome. Weight-neutral approaches may be as effective as weight-loss methods for improving physical, psychological, and behavioral outcomes. Limitations include inconsistent definitions of both approaches and variable time frames of follow-up.

Presepsin - New Marker of Sepsis Romanian Neonatal Intensive Care Unit Experience

2019 ◽  
Vol 70 (8) ◽  
pp. 3008-3013
Author(s):  
Silvia Maria Stoicescu ◽  
Ramona Mohora ◽  
Monica Luminos ◽  
Madalina Maria Merisescu ◽  
Gheorghita Jugulete ◽  
...  
Keyword(s):  
Intensive Care Unit ◽  
Intensive Care ◽  
Neonatal Intensive Care Unit ◽  
Neonatal Sepsis ◽  
Predictive Value ◽  
Neonatal Intensive Care ◽  
Complete Blood Count ◽  
Clinical Manifestations ◽  
Receiver Operating Curve ◽  
Predictive Values

Difficulties in establishing the onset of neonatal sepsis has directed the medical research in recent years to the possibility of identifying early biological markers of diagnosis. Overdiagnosing neonatal sepsis leads to a higher rate and duration in the usage of antibiotics in the Neonatal Intensive Care Unit (NICU), which in term leads to a rise in bacterial resistance, antibiotherapy complications, duration of hospitalization and costs.Concomitant analysis of CRP (C Reactive Protein), procalcitonin, complete blood count, presepsin in newborn babies with suspicion of early or late neonatal sepsis. Presepsin sensibility and specificity in diagnosing neonatal sepsis. The study group consists of newborns admitted to Polizu Neonatology Clinic between 15th February- 15th July 2017, with suspected neonatal sepsis. We analyzed: clinical manifestations and biochemical markers values used for diagnosis of sepsis, namely the value of CRP, presepsin and procalcitonin on the onset day of the disease and later, according to evolution. CRP values may be influenced by clinical pathology. Procalcitonin values were mainly influenced by the presence of jaundice. Presepsin is the biochemical marker with the fastest predictive values of positive infection. Presepsin can be a useful tool for early diagnosis of neonatal sepsis and can guide the antibiotic treatment. Presepsin value is significantly higher in neonatal sepsis compared to healthy newborns (939 vs 368 ng/mL, p [ 0.0001); area under receiver operating curve (AUC) for presepsine was 0.931 (95% confidence interval 0.86-1.0). PSP has a greater sensibility and specificity compared to classical sepsis markers, CRP and PCT respectively (AUC 0.931 vs 0.857 vs 0.819, p [ 0.001). The cut off value for presepsin was established at 538 ng/mLwith a sensibility of 79.5% and a specificity of 87.2 %. The positive predictive value (PPV) is 83.8 % and negative predictive value (NPV) is 83.3%.

Artificial Intelligence for Skin Cancer Detection: A Scoping Review (Preprint)

2020 ◽  
Author(s):  
Abdulrahman Takiddin ◽  
Jens Schneider ◽  
Yin Yang ◽  
Alaa Abd-Alrazaq ◽  
Mowafa Househ
Keyword(s):  
Artificial Intelligence ◽  
Skin Cancer ◽  
Scoping Review ◽  
Performance Metrics ◽  
Data Extraction ◽  
Characteristic Curve ◽  
Evaluation Metrics ◽  
Cancer Type ◽  
Computer Algorithms ◽  
Scoping Reviews

BACKGROUND Skin cancer is the most common cancer type affecting humans. Traditional skin cancer diagnosis methods are costly, require a professional physician, and take time. Hence, to aid in diagnosing skin cancer, Artificial Intelligence (AI) tools are being used, including shallow and deep machine learning-based techniques that are trained to detect and classify skin cancer using computer algorithms and deep neural networks. OBJECTIVE The aim of this study is to identify and group the different types of AI-based technologies used to detect and classify skin cancer. The study also examines the reliability of the selected papers by studying the correlation between the dataset size and number of diagnostic classes with the performance metrics used to evaluate the models. METHODS We conducted a systematic search for articles using IEEE Xplore, ACM DL, and Ovid MEDLINE databases following the PRISMA Extension for Scoping Reviews (PRISMA-ScR) guidelines. The study included in this scoping review had to fulfill several selection criteria; to be specifically about skin cancer, detecting or classifying skin cancer, and using AI technologies. Study selection and data extraction were conducted by two reviewers independently. Extracted data were synthesized narratively, where studies were grouped based on the diagnostic AI techniques and their evaluation metrics. RESULTS We retrieved 906 papers from the 3 databases, but 53 studies were eligible for this review. While shallow techniques were used in 14 studies, deep techniques were utilized in 39 studies. The studies used accuracy (n=43/53), the area under receiver operating characteristic curve (n=5/53), sensitivity (n=3/53), and F1-score (n=2/53) to assess the proposed models. Studies that use smaller datasets and fewer diagnostic classes tend to have higher reported accuracy scores. CONCLUSIONS The adaptation of AI in the medical field facilitates the diagnosis process of skin cancer. However, the reliability of most AI tools is questionable since small datasets or low numbers of diagnostic classes are used. In addition, a direct comparison between methods is hindered by a varied use of different evaluation metrics and image types.

scholarly journals Paediatric admissions to a district general hospital adult intensive care unit over a six year period: A single centre retrospective case series, with review of practitioner confidence and competence in paediatric critical care

2021 ◽  
pp. 175114372110121
Author(s):  
Stephen A Spencer ◽  
Joanna S Gumley ◽  
Marcin Pachucki
Keyword(s):  
Critical Care ◽  
Intensive Care ◽  
Intensive Care Units ◽  
Case Series ◽  
Critically Ill Children ◽  
Single Centre ◽  
Retrospective Case ◽  
Retrospective Case Series ◽  
Paediatric Critical Care ◽  
Doctors And Nurses

Background Critically ill children presenting to district general hospitals (DGH) are admitted to adult intensive care units (AICUs) for stabilisation prior to transfer to paediatric intensive care units (PICUs). Current training in PICU for adult intensive care physicians is only three months. This single centre retrospective case series examines the case mix of children presenting to a DGH AICU and a multidisciplinary survey assesses confidence and previous experience, highlighting continued training needs for DGH AICU staff. Methods all paediatric admissions to AICU and paediatric retrievals were reviewed over a 6-year period (2014-2019). Cases were identified from the Electronic Patient Record (EPR) and from data provided by the regional paediatric retrieval service. A questionnaire survey was sent to AICU doctors and nurses to assess confidence and competence in paediatric critical care. Results Between 2014-2019, 284 children were managed by AICU. In total 35% of cases were <1 y, 48% of cases were <2 y and 64% of cases were <5 y, and 166/284 (58%) children were retrieved. Retrieval reduced with increasing age (OR 0.49 [0.40-0.60], p < 0.0001). The survey had an 82% response rate, and highlighted that only 13% of AICU nurses and 50% of doctors had received prior PICU training. Conclusion At least one critically unwell child presents to the AICU each week. Assessment, stabilisation and management of critically unwell children are vital skills for DGH AICU staff, but confidence and competence are lacking. Formalised strategies are required to develop and maintain paediatric competencies for AICU doctors and nurses.

scholarly journals Pediatric Multi-Organ Dysfunction Syndrome: Analysis by an Untargeted “Shotgun” Lipidomic Approach Reveals Low-Abundance Plasma Phospholipids and Dynamic Recovery over 8-Day Period, a Single-Center Observational Study

Nutrients ◽  
2021 ◽  
Vol 13 (3) ◽  
pp. 774
Author(s):  
Mara L. Leimanis-Laurens ◽  
Karen Ferguson ◽  
Emily Wolfrum ◽  
Brian Boville ◽  
Dominic Sanfilippo ◽  
...  
Keyword(s):  
Mass Spectrometry ◽  
Intensive Care ◽  
Critically Ill ◽  
Organ Dysfunction ◽  
Lipid Profiles ◽  
Critically Ill Children ◽  
P Value ◽  
Time Points ◽  
Logistic Organ Dysfunction ◽  
Multi Organ Dysfunction Syndrome

Lipids are molecules involved in metabolism and inflammation. This study investigates the plasma lipidome for markers of severity and nutritional status in critically ill children. Children with multi-organ dysfunction syndrome (MODS) (n = 24) are analyzed at three time-points and cross-referenced to sedation controls (n = 4) for a total of N = 28. Eight of the patients with MODS, needed veno-arterial extracorporeal membrane oxygenation (VA ECMO) support to survive. Blood plasma lipid profiles are quantified by nano-electrospray (nESI), direct infusion high resolution/accurate mass spectrometry (MS), and tandem mass spectrometry (MS/MS), and compared to nutritional profiles and pediatric logistic organ dysfunction (PELOD) scores. Our results show that PELOD scores were not significantly different between MODS and ECMO cases across time-points (p = 0.66). Lipid profiling provides stratification between sedation controls and all MODS patients for total lysophosphatidylserine (lysoPS) (p-value = 0.004), total phosphatidylserine (PS) (p-value = 0.015), and total ether-linked phosphatidylethanolamine (ether-PE) (p-value = 0.03) after adjusting for sex and age. Nutrition intake over time did not correlate with changes in lipid profiles, as measured by caloric and protein intake. Lipid measurement in the intensive care environment shows dynamic changes over an 8-day pediatric intensive care unit (PICU) course, suggesting novel metabolic indicators for defining critically ill children.

scholarly journals Is there a sex difference in mortality rates for deaths occurring in paediatric intensive care units? Systematic literature review protocol

BMJ Open ◽  
2021 ◽  
Vol 11 (2) ◽  
pp. e046794
Author(s):  
Ofran Almossawi ◽  
Amanda Friend ◽  
Luigi Palla ◽  
Richard Feltbower ◽  
Bianca De Stavola
Keyword(s):  
Intensive Care ◽  
Literature Review ◽  
Intensive Care Units ◽  
Systematic Literature Review ◽  
Data Extraction ◽  
Linkage Study ◽  
Paediatric Intensive Care ◽  
Admission Rates ◽  
Ethical Approval ◽  
Paediatric Intensive Care Units

IntroductionIn the general population, female children have been reported to have a survival advantage. For children admitted to paediatric intensive care units (PICUs), mortality has been reported to be lower in males despite the higher admission rates for males into intensive care. This apparent sex reversal in PICU mortality is not well studied. To address this, we propose to conduct a systematic literature review to summarise the available evidence. Our review will study the reported differences in mortality between males and females aged 0–17, who died in a PICU, to examine if there is a difference between the two sexes in PICU mortality, and if so, to describe the magnitude and direction of this difference.Methods and analysisStudies that directly or indirectly addressed the association between sex and mortality in children admitted to intensive care will be eligible for inclusion. Studies that directly address the association will be eligible for data extraction. The search strings were based on terms related to the population (children in intensive care), the exposure (sex) and the outcome (mortality). We used the databases MEDLINE (1946–2020), Embase (1980–2020) and Web of Science (1985–2020) as these cover relevant clinical publications. We will assess the reliability of included studies using the risk of bias in observational studies of exposures tool. We will consider a pooled effect if we have at least three studies with similar periods of follow up and adjustment variables.Ethics and disseminationEthical approval is not required for this review as it will synthesise data from existing studies. This manuscript is a part of a larger data linkage study, for which Ethical approval was granted. Dissemination will be via peer-reviewed journals and via public and patient groups.PROSPERO registration numberCRD42020203009.

scholarly journals Validation of the pediatric refractory septic shock definition: post hoc analysis of a controlled trial

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Luc Morin ◽  
Karthik Narayanan Ramaswamy ◽  
Muralidharan Jayashree ◽  
Arun Bansal ◽  
Karthi Nallasamy ◽  
...  
Keyword(s):  
Septic Shock ◽  
Intensive Care ◽  
Neonatal Intensive Care ◽  
Controlled Trial ◽  
Receiver Operating Curve ◽  
Double Blind ◽  
Post Hoc Analysis ◽  
Prognostic Accuracy ◽  
Refractory Septic Shock ◽  
Post Hoc

Abstract Background The European Society of Pediatric and Neonatal Intensive Care (ESPNIC) developed and validated a definition of pediatric refractory septic shock (RSS), based on two septic shock scores (SSS). Both bedside SSS (bSSS) and computed SSS (cSSS) were found to be strongly associated with mortality. We aimed at assessing the accuracy of the RSS definition on a prospective cohort from India. Methods Post hoc analysis of a cohort issued from a double-blind randomized trial that compared first-line vasoactive drugs in children with septic shock. Sequential bSSS and cSSS from 60 children (single-center study, 53% mortality) were analyzed. The prognostic value of the ESPNIC RSS definition was tested for 28-day all-cause mortality. Results In this septic shock cohort, RSS was diagnosed in 35 patients (58.3%) during the first 24 h. Death occurred in 30 RSS patients (85.7% mortality) and in 2 non-RSS patients (8% mortality), OR = 60.9 [95% CI: 10.5–676.2], p < 0.001 with a median delay from sepsis onset of 3 days [1.0–6.7]. Among patients diagnosed with RSS, the mortality was not significantly different according to vasopressors randomization. Diagnosis of RSS with bSSS and cSSS had a high discrimination for death with an area under the receiver operating curve of 0.916 [95% CI: 0.843–0.990] and 0.925 [95% CI: 0.845–1.000], respectively. High prognostic accuracy of the bSSS was found in the first hours following intensive care admission. The best interval of prognostication occurs after the 12th hour following treatment initiation (AUC 0.973 [95% CI: 0.925–1.000]). Conclusions The ESPNIC refractory septic shock definition accurately identifies, within the first 6 h of septic shock management, children with lethal outcome.

Artificial Intelligence and Ethics in Dentistry: A Scoping Review

2021 ◽  
pp. 002203452110138
Author(s):  
C.M. Mörch ◽  
S. Atsu ◽  
W. Cai ◽  
X. Li ◽  
S.A. Madathil ◽  
...  
Keyword(s):  
Artificial Intelligence ◽  
Scoping Review ◽  
Ethical Issues ◽  
Data Extraction ◽  
Cochrane Library ◽  
Data Sets ◽  
Ethical Challenges ◽  
Internal Validation ◽  
Neural Learning ◽  
Learning Techniques

Dentistry increasingly integrates artificial intelligence (AI) to help improve the current state of clinical dental practice. However, this revolutionary technological field raises various complex ethical challenges. The objective of this systematic scoping review is to document the current uses of AI in dentistry and the ethical concerns or challenges they imply. Three health care databases (MEDLINE [PubMed], SciVerse Scopus, and Cochrane Library) and 2 computer science databases (ArXiv, IEEE Xplore) were searched. After identifying 1,553 records, the documents were filtered, and a full-text screening was performed. In total, 178 studies were retained and analyzed by 8 researchers specialized in dentistry, AI, and ethics. The team used Covidence for data extraction and Dedoose for the identification of ethics-related information. PRISMA guidelines were followed. Among the included studies, 130 (73.0%) studies were published after 2016, and 93 (52.2%) were published in journals specialized in computer sciences. The technologies used were neural learning techniques for 75 (42.1%), traditional learning techniques for 76 (42.7%), or a combination of several technologies for 20 (11.2%). Overall, 7 countries contributed to 109 (61.2%) studies. A total of 53 different applications of AI in dentistry were identified, involving most dental specialties. The use of initial data sets for internal validation was reported in 152 (85.4%) studies. Forty-five ethical issues (related to the use AI in dentistry) were reported in 22 (12.4%) studies around 6 principles: prudence (10 times), equity (8), privacy (8), responsibility (6), democratic participation (4), and solidarity (4). The ratio of studies mentioning AI-related ethical issues has remained similar in the past years, showing that there is no increasing interest in the field of dentistry on this topic. This study confirms the growing presence of AI in dentistry and highlights a current lack of information on the ethical challenges surrounding its use. In addition, the scarcity of studies sharing their code could prevent future replications. The authors formulate recommendations to contribute to a more responsible use of AI technologies in dentistry.

Māori health outcomes in an intensive care unit in Aotearoa New Zealand

2021 ◽  
pp. 0310057X2198971
Author(s):  
M Atif Mohd Slim ◽  
Hamish M Lala ◽  
Nicholas Barnes ◽  
Robert A Martynoga
Keyword(s):  
Intensive Care Unit ◽  
Intensive Care ◽  
New Zealand ◽  
Renal Replacement Therapy ◽  
Confidence Interval ◽  
Health Outcomes ◽  
Odds Ratio ◽  
Adjusted Odds Ratio ◽  
One Year ◽  
Maori Health

Māori are the indigenous people of New Zealand, and suffer disparate health outcomes compared to non-Māori. Waikato District Health Board provides level III intensive care unit services to New Zealand’s Midland region. In 2016, our institution formalised a corporate strategy to eliminate health inequities for Māori. Our study aimed to describe Māori health outcomes in our intensive care unit and identify inequities. We performed a retrospective audit of prospectively entered data in the Australian and New Zealand Intensive Care Society database for all general intensive care unit admissions over 15 years of age to Waikato Hospital from 2014 to 2018 ( n = 3009). Primary outcomes were in–intensive care unit and in-hospital mortality. The secondary outcome was one-year mortality. In our study, Māori were over-represented relative to the general population. Compared to non-Māori, Māori patients were younger (51 versus 61 years, P < 0.001), and were more likely to reside outside of the Waikato region (37.2% versus 28.0%, P < 0.001) and in areas of higher deprivation ( P < 0.001). Māori had higher admission rates for trauma and sepsis ( P < 0.001 overall) and required more renal replacement therapy ( P < 0.001). There was no difference in crude and adjusted mortality in–intensive care unit (16.8% versus 16.5%, P = 0.853; adjusted odds ratio 0.98 (95% confidence interval 0.68 to 1.40)) or in-hospital (23.7% versus 25.7%, P = 0.269; adjusted odds ratio 0.84 (95% confidence interval 0.60 to 1.18)). One-year mortality was similar (26.1% versus 27.1%, P=0.6823). Our study found significant ethnic inequity in the intensive care unit for Māori, who require more renal replacement therapy and are over-represented in admissions, especially for trauma and sepsis. These findings suggest upstream factors increasing Māori risk for critical illness. There was no difference in mortality outcomes.

Sign in / Sign up

Export Citation Format

Share Document