Risk factors for length of NICU stay of newborns: A systematic review

Frontiers in Pediatrics, Journal Year: 2023, Volume and Issue: 11

Published: March 13, 2023

Background The improvement in survival of preterm infants is accompanied by an increase neonatal intensive care unit (NICU) admissions. Prolonged length stay the NICU (LOS-NICU) increases incidence complications and even mortality places a significant economic burden on families strain healthcare systems. This review aims to identify risk factors influencing LOS-NICU newborns provide basis for interventions shorten avoid prolonged LOS-NICU. Methods A systematic literature search was conducted PubMed, Web Science, Embase, Cochrane library studies that were published English from January 1994 October 2022. PRISMA guidelines followed all phases this review. Quality Prognostic Studies (QUIPS) tool used assess methodological quality. Results Twenty-three included, 5 which high quality 18 moderate quality, with no low-quality literature. reported 58 possible six broad categories (inherent factors; antenatal treatment maternal diseases adverse conditions newborn; clinical scores laboratory indicators; organizational factors). Conclusions We identified several most critical affecting LOS-NICU, including birth weight, gestational age, sepsis, necrotizing enterocolitis, bronchopulmonary dysplasia, retinopathy prematurity. As only few high-quality are available at present, well-designed more extensive prospective investigating still needed future.

Language: Английский

---

Prognosis of COVID-19 severity using DERGA, a novel machine learning algorithm

European Journal of Internal Medicine, Journal Year: 2024, Volume and Issue: 125, P. 67 - 73

Published: March 8, 2024

Language: Английский

---

Early prediction of COVID-19 outcome using artificial intelligence techniques and only five laboratory indices

Clinical Immunology, Journal Year: 2022, Volume and Issue: 246, P. 109218 - 109218

Published: Dec. 29, 2022

Language: Английский

---

A forecasting approach for hospital bed capacity planning using machine learning and deep learning with application to public hospitals

Healthcare Analytics, Journal Year: 2023, Volume and Issue: 4, P. 100245 - 100245

Published: Aug. 15, 2023

Hospital Bed Capacity (HBC) planning affects economic and social sustainability in healthcare through bed capacity efficiency medical treatment accessibility. Conventionally, this problem is solved using programming or simulation models with assumptions limits. Forecasting the HBC time series data on occupancy has been considered but not factors such as Number of Hospitalized Patients (NHP) patient's length stay (LOS). This study proposes a data-driven methodology to forecast Machine Learning (ML) Deep (DL). The LOS classification performed several ML techniques, including Bayesian network, K-nearest neighbor, support vector machine, decision tree, Linear regression. Also, seasonal autoregressive integrated moving average, linear regression Long short-term memory neural network are applied for NHP forecasting. forecasting descriptive analysis outputs based classes directly simple mathematical model predict required capacity. case heart ward at public hospital. set includes 51231 records, DL algorithms developed Python. Results show that ward's must be raised from 45 137 by 2026. In addition, managerial recommendations formulated.

Language: Английский

---

WT-LSTM for Intensive Care Unit Length-of-Stay Prediction with Real-Time Signal (Preprint)

Published: Jan. 13, 2025

Efficient allocation of healthcare resources is essential for sustainable hospital operation. Effective intensive care unit (ICU) management alleviating the financial strain on systems. Accurate prediction length-of-stay in ICUs vital optimizing capacity planning and resource allocation, with challenge achieving early, real-time predictions. This study aims to develop a predictive model, namely WT-LSTM, ICU using only sign data. The model designed urgent settings where demographic historical patient data or lab results may be unavailable; leverages inputs deliver early accurate proposed integrates discrete wavelet transformation Long Short-Term Memory (LSTM) neural networks filter noise from patients’ series improve accuracy. Model performance was evaluated eICU database, focusing ten common admission diagnoses database. demonstrate that WT-LSTM consistently outperforms baseline models, including linear regression, LSTM, BiLSTM, predicting data, significant improvements Mean Squared Error (MSE). Specifically, component enhances overall WT-LSTM. Removing this an average decrease 3.3% MSE; such phenomenon particularly pronounced specific cohorts. model's adaptability highlighted through predictions 3-hour, 6-hour, 12-hour, 24-hour input Using three hours delivers competitive across most diagnoses, often outperforming APACHE IV, leading outcome system currently implemented clinical practice. effectively captures patterns signs recorded during initial patient’s stay, making it promising tool optimization ICU. Our based offers solution prediction. Its high accuracy capabilities hold potential enhancing practice, supporting critical administrative decisions management.

Language: Английский

---

Machine Learning First Response to COVID-19: A Systematic Literature Review of Clinical Decision Assistance Approaches during Pandemic Years from 2020 to 2022

Electronics, Journal Year: 2024, Volume and Issue: 13(6), P. 1005 - 1005

Published: March 7, 2024

Background: The declaration of the COVID-19 pandemic triggered global efforts to control and manage virus impact. Scientists researchers have been strongly involved in developing effective strategies that can help policy makers healthcare systems both monitor spread mitigate impact pandemic. Machine Learning (ML) Artificial Intelligence (AI) applied several fronts fight. Foremost is diagnostic assistance, encompassing patient triage, prediction ICU admission mortality, identification mortality risk factors, discovering treatment drugs vaccines. Objective: This systematic review aims identify original research studies involving actual data construct ML- AI-based models for clinical decision support early response during years. Methods: Following PRISMA methodology, two large academic publication indexing databases were searched investigate use ML-based technologies their applications combat Results: literature search returned more than 1000 papers; 220 selected according specific criteria. illustrate usefulness ML with respect supporting professionals (1) triage patients depending on disease severity, (2) predicting hospital or Intensive Care Units (ICUs), (3) new repurposed treatments (4) factors. Conclusion: ML/AI community was able propose develop a wide variety solutions hospitalizations recommendations diagnostic, opening door further integration practices fighting this forecoming pandemics. However, translation practice impeded by heterogeneity datasets methodological computational approaches. lacks robust model validations desired translation.

Language: Английский

---

Development and Validation of a Machine Learning Model That Predicts Short Inpatient Stays Among Urgent Admissions

Emergency Care and Medicine, Journal Year: 2025, Volume and Issue: 2(1), P. 11 - 11

Published: Feb. 25, 2025

Background/Objectives: This study aimed to explore the feasibility of predicting short stays among urgent admissions an acute hospital in Singapore. With increase average length stay (LOS) hospitals recent years, accurately could enable better manage inpatient demand and reduce emergency department (ED) overcrowding. Methods: was a retrospective Changi General Hospital, Singapore, from 1 January 2016 30 June 2022. To identify potential stayers, total 25 features comprising demographic characteristics, admission clinical healthcare utilization history were analyzed for each admitted patient at point when ED physician decided admit patient. The dataset further split into development external validation based on year admission. A CatBoost classifier trained using 75% dataset. Apart reporting model’s prediction accuracy, we conducted various analyses simulations effects crucial output. Results: accuracy model evaluated both test (25%) On former, area under receiver operating characteristic (AUROC) precision-recall curve (AUPRC) 0.803 (95% CI: 0.799, 0.808) 0.755 0.749, 0.762), respectively, with precision = 0.700 0.694, 0.707) recall 0.692 0.685, 0.699). dataset, performance similar. diagnosis whether required surgical procedure most important making prediction. Conclusions: LOS help providers stayers early course their journeys so they make interventions overall beds.

Language: Английский

---

A Predictive Model for PICC-related Thrombosis in Sepsis Patients Using XGBoost Algorithm

Research Square (Research Square), Journal Year: 2025, Volume and Issue: unknown

Published: May 9, 2025

Background Percutaneous insertion of central venous catheters (PICC) is critical for the management sepsis patients requiring prolonged intravenous therapy; however, it poses significant complications, including thrombosis. Identifying risk factors PICC-related thrombosis can enhance clinical and patient outcomes. This study aimed to develop a predictive model in using XGBoost algorithm. Methods We analyzed data from 8,128 ICU diagnosed with PICC Medical Information Mart Intensive Care IV version 3.1 (MIMIC-IV 3.1) database. Patients were divided into training set (70%, n = 5,690) validation (30%, 2,438). Variables included demographic, laboratory, potentially associated An was developed validated, performance assessed area under receiver operating characteristic curve (AUC) SHAP analysis interpretability. Decision confirmed utility model. Results The achieved an AUC 0.761 (95% CI: 0.734–0.787) 0.766 0.731–0.801) set. calibration demonstrated good model, indicating that predicted probabilities closely aligned observed outcome. utility, yielding net benefit 0.31 at 20% threshold, outperforming treat-all/none strategies. Key predictors, white blood cell count, hemoglobin levels, age, creatinine platelet identified thrombosis, top ten predictors significantly contributing model's performance. Conclusions effective predictor among patients, its potential role guiding decision-making high-risk patients.

Language: Английский

---

Prediction of mortality risk and duration of hospitalization of COVID-19 patients with chronic comorbidities based on machine learning algorithms

Digital Health, Journal Year: 2023, Volume and Issue: 9

Published: Jan. 1, 2023

Background The severity of coronavirus (COVID-19) in patients with chronic comorbidities is much higher than other patients, which can lead to their death. Machine learning (ML) algorithms as a potential solution for rapid and early clinical evaluation the disease help allocating prioritizing resources reduce mortality. Objective objective this study was predict mortality risk length stay (LoS) COVID-19 history using ML algorithms. Methods This retrospective conducted by reviewing medical records from March 2020 January 2021 Afzalipour Hospital Kerman, Iran. outcome hospitalization recorded discharge or filtering technique used score features well-known were applied LoS patients. Ensemble Learning methods also used. To evaluate performance models, different measures including F1, precision, recall, accuracy calculated. TRIPOD guideline assessed transparent reporting. Results performed on 1291 900 alive 391 dead Shortness breath (53.6%), fever (30.1%), cough (25.3%) three most common symptoms Diabetes mellitus(DM) (31.3%), hypertension (HTN) (27.3%), ischemic heart (IHD) (14.2%) Twenty-six important factors extracted each patient's record. Gradient boosting model 84.15% best predicting multilayer perceptron (MLP) rectified linear unit function (MSE = 38.96) LoS. among these DM HTN IHD (14.2%). hyperlipidemia, diabetes, asthma, cancer, shortness breath. Conclusion results showed that use be good tool based physiological conditions, symptoms, demographic information MLP quickly identify at death long-term notify physicians do appropriate interventions.

Language: Английский

---

Prognostic Model of ICU Admission Risk in Patients with COVID-19 Infection Using Machine Learning

Diagnostics, Journal Year: 2022, Volume and Issue: 12(9), P. 2144 - 2144

Published: Sept. 3, 2022

With the onset of COVID-19 pandemic, number critically sick patients in intensive care units (ICUs) has increased worldwide, putting a burden on ICUs. Early prediction ICU requirement is crucial for efficient resource management and distribution. Early-prediction scoring systems ill using mathematical models are available, but not generalized Non-COVID patients. This study aims to develop reliable prognostic model admission both non-COVID-19 best feature combination from patient data at admission. A retrospective cohort was conducted dataset collected pulmonology department Moscow City State Hospital between 20 April 2020 5 June 2020. The contains ten clinical features 231 patients, whom 100 were transferred 131 stable (non-ICU) There 156 COVID positive 75 non-COVID Different selection techniques investigated, stacking machine learning proposed compared with eight different classification algorithms detect risk need combined alone. C-reactive protein (CRP), chest computed tomography (CT), lung tissue affected (%), age, hospital, fibrinogen parameters hospital found be important ICU-requirement prediction. performance produced by approach, weighted precision, sensitivity, F1-score, specificity, overall accuracy 84.45%, 84.48%, 83.64%, 84.47%, respectively, types 85.34%, 85.35%, 85.11%, only. work can help doctors improve through early during as used

Language: Английский

---