ICU follow-up services and their impact on post-intensive care syndrome: a scoping review protocol

INTRODUCTION

Post-intensive care syndrome (PICS) seriously affects the quality of life of intensive care unit (ICU) survivors, their ability to return to work and society and the quality of life of their families, increasing overall care costs and healthcare expenditures. ICU follow-up services have important potential to improve PICS. However, the best clinical practice model of ICU follow-up service has not been fully defined and its benefits for ICU survivors are not clear. This review will synthesise and map the current types of follow-up services for ICU survivors and summarise the impact of follow-up services on PICS.

METHODS AND ANALYSIS

This scoping review will be conducted by applying the five-stage protocol proposed by Arksey and O'Malley in an updated version of the Joanna Briggs Institute. Eight academic databases including the Cochrane Library, MEDLINE, Web of Science, Embase, EBSCO Academic, CINAHL, PsycInfo and SinoMed (China Biology Medicine) will be systematically searched from inception to the present. Peer-reviewed literature and grey literature will be included. Qualitative, quantitative and mixed methods studies will be included. Studies published in English or Chinese will be included. There will be no time restriction. Two reviewers will screen and select the articles independently and if there is any disagreement, the two reviewers will discuss or invite a third reviewer to make decisions together. Descriptive analysis will be used to conduct an overview of the literature. The results will be presented in a descriptive format in response to the review questions accompanied by the necessary tables or charts.

ETHICS AND DISSEMINATION

Ethical approval is not required for this scoping review because data could be obtained by reviewing published primary study results and do not involve human participants. Findings should be disseminated at scientific meetings and published in peer-reviewed journals.

B7-1 and PlGF-1 are two possible new biomarkers to identify fracture-associated trauma patients at higher risk of developing complications: a cohort study

BACKGROUND

Around 10% of fractures lead to complications. With increasing fracture incidences in recent years, this poses a serious burden on the healthcare system, with increasing costs for treatment. In the present study, we aimed to identify potential 'new' blood markers to predict the development of post-surgical complications in trauma patients following a fracture.

METHODS

A total of 292 trauma patients with a complete three-month follow-up were included in this cohort study. Blood samples were obtained from 244 of these patients. Two complication groups were distinguished based on the Clavien-Dindo (CD) classification: CD grade I and CD grade III groups were compared to the controls (CD 0). The Mann-Whitney U test was used to compare the complication groups to the control group.

RESULTS

Analysis of the patients' data revealed that risk factors are dependent on sex. Both, males and females who developed a CD III complication showed elevated blood levels of B7-1 (p = 0.015 and p = 0.018, respectively) and PlGF-1 (p = 0.009 and p = 0.031, respectively), with B7-1 demonstrating greater sensitivity (B7-1: 0.706 (male) and 0.692 (female), PlGF-1: 0.647 (male) and 0.615 (female)). Further analysis of the questionnaires and medical data revealed the importance of additional risk factors. For males (CD 0: 133; CD I: 12; CD III: 18 patients) alcohol consumption was significantly increased for CD I and CD III compared to control with p = 0.009 and p = 0.007, respectively. For females (CD 0: 107; CD I: 10; CD III: 12 patients) a significantly increased average BMI [kg/m2] from 25.5 to 29.7 with CD III was observed, as well as an elevation from one to three comorbidities (p = 0.003).

CONCLUSIONS

These two potential new blood markers hold promise for predicting complication development in trauma patients. Nevertheless, further studies are necessary to evaluate the diagnostic utility of B7-1 and PlGF-1 in predicting complications in trauma patients and consider sex differences before their possible use as routine clinical screening tools.

Predicting hospital length of stay using machine learning on a large open health dataset

BACKGROUND

Governments worldwide are facing growing pressure to increase transparency, as citizens demand greater insight into decision-making processes and public spending. An example is the release of open healthcare data to researchers, as healthcare is one of the top economic sectors. Significant information systems development and computational experimentation are required to extract meaning and value from these datasets. We use a large open health dataset provided by the New York State Statewide Planning and Research Cooperative System (SPARCS) containing 2.3 million de-identified patient records. One of the fields in these records is a patient's length of stay (LoS) in a hospital, which is crucial in estimating healthcare costs and planning hospital capacity for future needs. Hence it would be very beneficial for hospitals to be able to predict the LoS early. The area of machine learning offers a potential solution, which is the focus of the current paper.

METHODS

We investigated multiple machine learning techniques including feature engineering, regression, and classification trees to predict the length of stay (LoS) of all the hospital procedures currently available in the dataset. Whereas many researchers focus on LoS prediction for a specific disease, a unique feature of our model is its ability to simultaneously handle 285 diagnosis codes from the Clinical Classification System (CCS). We focused on the interpretability and explainability of input features and the resulting models. We developed separate models for newborns and non-newborns.

RESULTS

The study yields promising results, demonstrating the effectiveness of machine learning in predicting LoS. The best R2 scores achieved are noteworthy: 0.82 for newborns using linear regression and 0.43 for non-newborns using catboost regression. Focusing on cardiovascular disease refines the predictive capability, achieving an improved R2 score of 0.62. The models not only demonstrate high performance but also provide understandable insights. For instance, birth-weight is employed for predicting LoS in newborns, while diagnostic-related group classification proves valuable for non-newborns.

CONCLUSION

Our study showcases the practical utility of machine learning models in predicting LoS during patient admittance. The emphasis on interpretability ensures that the models can be easily comprehended and replicated by other researchers. Healthcare stakeholders, including providers, administrators, and patients, stand to benefit significantly. The findings offer valuable insights for cost estimation and capacity planning, contributing to the overall enhancement of healthcare management and delivery.

al-BERT: a semi-supervised denoising technique for disease prediction

BACKGROUND

Medical records are a valuable source for understanding patient health conditions. Doctors often use these records to assess health without solely depending on time-consuming and complex examinations. However, these records may not always be directly relevant to a patient's current health issue. For instance, information about common colds may not be relevant to a more specific health condition. While experienced doctors can effectively navigate through unnecessary details in medical records, this excess information presents a challenge for machine learning models in predicting diseases electronically. To address this, we have developed 'al-BERT', a new disease prediction model that leverages the BERT framework. This model is designed to identify crucial information from medical records and use it to predict diseases. 'al-BERT' operates on the principle that the structure of sentences in diagnostic records is similar to regular linguistic patterns. However, just as stuttering in speech can introduce 'noise' or irrelevant information, similar issues can arise in written records, complicating model training. To overcome this, 'al-BERT' incorporates a semi-supervised layer that filters out irrelevant data from patient visitation records. This process aims to refine the data, resulting in more reliable indicators for disease correlations and enhancing the model's predictive accuracy and utility in medical diagnostics.

METHOD

To discern noise diseases within patient records, especially those resembling influenza-like illnesses, our approach employs a customized semi-supervised learning algorithm equipped with a focused attention mechanism. This mechanism is specifically calibrated to enhance the model's sensitivity to chronic conditions while concurrently distilling salient features from patient records, thereby augmenting the predictive accuracy and utility of the model in clinical settings. We evaluate the performance of al-BERT using real-world health insurance data provided by Taiwan's National Health Insurance.

RESULT

In our study, we evaluated our model against two others: one based on BERT that uses complete disease records, and another variant that includes extra filtering techniques. Our findings show that models incorporating filtering mechanisms typically perform better than those using the entire, unfiltered dataset. Our approach resulted in improved outcomes across several key measures: AUC-ROC (an indicator of a model's ability to distinguish between classes), precision (the accuracy of positive predictions), recall (the model's ability to find all relevant cases), and overall accuracy. Most notably, our model showed a 15% improvement in recall compared to the current best-performing method in the field of disease prediction.

CONCLUSION

The conducted ablation study affirms the advantages of our attention mechanism and underscores the crucial role of the selection module within al-BERT.

PSO-XnB: a proposed model for predicting hospital stay of CAD patients

Coronary artery disease poses a significant challenge in decision-making when predicting the length of stay for a hospitalized patient. This study presents a predictive model-a Particle Swarm Optimized-Enhanced NeuroBoost-that combines the deep autoencoder with an eXtreme gradient boosting model optimized using particle swarm optimization. The model uses a fuzzy set of rules to categorize the length of stay into four distinct classes, followed by data preparation and preprocessing. In this study, the dimensionality of the data is reduced using deep neural autoencoders. The reconstructed data obtained from autoencoders is given as input to an eXtreme gradient boosting model. Finally, the model is tuned with particle swarm optimization to obtain optimal hyperparameters. With the proposed technique, the model achieved superior performance with an overall accuracy of 98.8% compared to traditional ensemble models and past research works. The model also scored highest in other metrics such as precision, recall, and particularly F1 scores for all categories of hospital stay. These scores validate the suitability of our proposed model in medical healthcare applications.

Adverse Events of Surgical Drain Placement: An Analysis of the NSQIP Database

BACKGROUND

Surgical site drainage is important to prevent hematoma, seroma, and abscess formation. However, the placement of drain placement also predispose patients to several postoperative complications. The aim of this study is to clarify the risk-benefit profile of surgical drain placement.

METHODS

The American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) Procedure-Targeted Databases were used to identify patients who underwent hepatectomy, pancreatectomy, nephrectomy, cystectomy, and prostatectomy. Patients who underwent each procedure were divided into 2 groups based on intraoperative drain placement. Propensity score-matched cohorts of these 2 groups were compared in terms of postoperative adverse events, readmission, reoperation, and length of stay.

RESULTS

Hepatectomy patients with drains experienced organ space infections (P < .001), sepsis (P < .001), and readmission (P = .021) more often than patients without drains. Patients who underwent pancreatectomy and had drains placed experienced wound dehiscence less frequently than those without drains (P = .04). For hepatectomy, pancreatectomy, nephrectomy, and prostatectomy populations, patients with drains had longer lengths of stay (P < .05). Matched populations across all procedures did not differ in terms of reoperation rate.

DISCUSSION

Prophylactic surgical drain placement may be associated with increased infectious complications and prolonged length of stay. Further studies are needed to elucidate the complete adverse event profile of surgical drains. Nonetheless, outcomes may be improved with better patient selection or advancements in drain technology.

Multi-modal learning for inpatient length of stay prediction

Predicting inpatient length of stay (LoS) is important for hospitals aiming to improve service efficiency and enhance management capabilities. Patient medical records are strongly associated with LoS. However, due to diverse modalities, heterogeneity, and complexity of data, it becomes challenging to effectively leverage these heterogeneous data to put forth a predictive model that can accurately predict LoS. To address the challenge, this study aims to establish a novel data-fusion model, termed as DF-Mdl, to integrate heterogeneous clinical data for predicting the LoS of inpatients between hospital discharge and admission. Multi-modal data such as demographic data, clinical notes, laboratory test results, and medical images are utilized in our proposed methodology with individual "basic" sub-models separately applied to each different data modality. Specifically, a convolutional neural network (CNN) model, which we termed CRXMDL, is designed for chest X-ray (CXR) image data, two long short-term memory networks are used to extract features from long text data, and a novel attention-embedded 1D convolutional neural network is developed to extract useful information from numerical data. Finally, these basic models are integrated to form a new data-fusion model (DF-Mdl) for inpatient LoS prediction. The proposed method attains the best R2 and EVAR values of 0.6039 and 0.6042 among competitors for the LoS prediction on the Medical Information Mart for Intensive Care (MIMIC)-IV test dataset. Empirical evidence suggests better performance compared with other state-of-the-art (SOTA) methods, which demonstrates the effectiveness and feasibility of the proposed approach.

Treatable mortality and health care related factors across European countries

Introduction

Despite the improvements in European health systems, a large number of premature deaths are attributable to treatable mortality. Men make up the majority of these deaths, with a significant gap existing between women and men's treatable mortality rate in the EU.

Aim

This study aims to identify the healthcare-related factors, including health expenditures, human and physical resources, and hospital services use associated with treatable mortality in women and men across European countries during the period 2011-2019.

Methods

We use Eurostat data for 28 EU countries in the period 2011-2019. We estimate a panel data linear regression with country fixed effects and quantile linear regression for men and women.

Results

The results found (i) differences in drivers for male and female treatable mortality, but common drivers hold the same direction for both sexes; (ii) favorable drivers are GDP per capita, health expenditures, number of physicians per capita, and (only for men) the average length of a hospital stay, (iii) unfavorable drivers are nurses and beds per capita, although nurses are not significant for explaining female mortality.

Conclusion

Policy recommendations may arise that involve an improvement in hospital bed management and the design of more specific policies aimed at healthcare professionals.

Cross-sectional evaluation of pharmaceutical care competences in nurse education: how well do curricula prepare students of different educational levels?

BACKGROUND

Nurses play an important role in interprofessional pharmaceutical care. Curricula related to pharmaceutical care, however, vary a lot. Mapping the presence of pharmaceutical care related domains and competences in nurse educational programs can lead to a better understanding of the extent to which curricula fit expectations of the labour market. The aim of this study was to describe 1) the presence of pharmaceutical care oriented content in nursing curricula at different educational levels and 2) nursing students' perceived readiness to provide nurse pharmaceutical care in practice.

METHODS

A quantitative cross-sectional survey design was used. Nursing schools in 14 European countries offering educational programs for levels 4-7 students were approached between January and April 2021. Through an online survey final year students had to indicate to what extent pharmaceutical care topics were present in their curriculum.

RESULTS

A total of 1807 students participated, of whom 8% had level 4-5, 80% level 6, 12% level 7. Up to 84% of the students indicated that pharmaceutical care content was insufficiently addressed in their curriculum. On average 14% [range 0-30] felt sufficiently prepared to achieve the required pharmaceutical care competences in practice. In level 5 curricula more pharmaceutical care domains were absent compared with other levels.

CONCLUSIONS

Although several pharmaceutical care related courses are present in current curricula of level 4-7 nurses, its embedding should be extended. Too many students perceive an insufficient preparation to achieve pharmaceutical care competences required in practice. Existing gaps in pharmaceutical care should be addressed to offer more thoroughly prepared nurses to the labour market.

The Effect of Listening to Holy Quran Recital on Pain and Length of Stay Post-CABG: A Randomized Control Trial

Background

Nearly, 75% of patients post-CABG complain of moderate to severe pain during their hospital stay. Nonpharmacological interventions have been investigated; however, the effect of Holy Quran recital post-CABG is still not well studied, especially in developing Islamic countries.

Objective

To investigate the effect of listening to the Holy Quran recital on pain and length of stay post-CABG.

Methods

This was a randomized control trial on 132 patients recruited from four hospitals in Amman, Jordan. The intervention group listened to the Holy Quran recited for 10 minutes twice daily while the control group received the usual care. Data were analyzed using paired and independent samples t-tests.

Results

Paired t-test testing showed that there was a significant reduction in the pain level, (M [SD], 6.82 [2.27] vs. 4.65 [2.18], t = 23.65, p < 0.001) for the intervention group. In addition, the intervention group had shorter LoS in the ICU and in the hospital compared to the control group, (M [SD], 5.0 [4.02] vs. 6.58 [4.18], t = -2.1, p < 0.05), (M [SD], 10.15 [9.21] vs. 15.01 [13.14], t = -2.6, p < 0.05), respectively.

Conclusions

Listening to the Quran was significantly effective in improving pain intensity among post-CABG patients and shortening their hospital/ICU stay. This trial is registered with NCT05419554.

Improving Operating Room Efficiency Through Reducing First Start Delays in an Academic Center

BACKGROUND

Delays in operating room (OR) first-case start times can cause additional costs for hospitals, healthcare team frustration and delay in patient care. Here, a novel process improvement strategy to improving first-case start times is presented.

METHODS

First case in room start times were recorded for ORs at an academic medical center. Three interventions-automatic preoperative orders, dot phrases to permit re-creation of unavailable consent forms, and improved H&P linking to the surgical encounter-were implemented to target documentation-related delays. Monthly percentages of first-case on-time starts (FCOTS) and time saved were compared with the "preintervention" time period, and total cost savings were estimated.

RESULTS

During the first 3-months after implementation of the interventions, the percentage of FCOTS improved from an average of 36.7%-52.7%. Total time savings across all ORs over the same time period was found to be 55.63 hours, which is estimated to have saved a total of $121,834.52 over the 3-month interventional period.

CONCLUSIONS

By implementing multiple quality improvement interventions, delays to first start in room OR cases can be meaningfully reduced. Quality improvement protocols targeted toward root causes of OR delays can be a significant driver to reduce healthcare costs.

Characteristics, clinical course and outcome of ventilated patients at a non-surgical intensive care unit in Germany: a single-centre, retrospective observational cohort analysis

OBJECTIVES

The objective of this study was to evaluate epidemiological characteristics, clinical course and outcome of mechanically ventilated non-surgical intensive care unit (ICU) patients, with the aim of improving the strategic planning of ICU capacities.

DESIGN

We conducted a retrospective observational cohort analysis. Data from mechanically ventilated intensive care patients were obtained by investigating electronic health records. The association between clinical parameters and ordinal scale data of clinical course was evaluated using Spearman correlation and Mann-Whitney U test. Relations between clinical parameters and in-hospital mortality rates were examined using binary logistic regression analysis.

SETTING

A single-centre study at the non-surgical ICU of the University Hospital of Frankfurt, Germany (tertiary care-level centre).

PARTICIPANTS

All cases of critically ill adult patients in need of mechanical ventilation during the years 2013-2015 were included. In total, 932 cases were analysed.

RESULTS

From a total of 932 cases, 260 patients (27.9%) were transferred from peripheral ward, 224 patients (24.1%) were hospitalised via emergency rescue services, 211 patients (22.7%) were admitted via emergency room and 236 patients (25.3%) via various transfers. In 266 cases (28.5%), respiratory failure was the reason for ICU admission. The length of stay was higher in non-geriatric patients, patients with immunosuppression and haemato-oncological disease or those in need of renal replacement therapy. 431 patients died, which corresponds to an all-cause in-hospital mortality rate of 46.2%. 92 of 172 patients with presence of immunosuppression (53.5%), 111 of 186 patients (59.7%) with pre-existing haemato-oncological disease, 27 of 36 patients (75.0%) under extracorporeal membrane oxygenation (ECMO) therapy, and 182 of 246 patients (74.0%) undergoing renal replacement therapy died. In logistic regression analysis, these subgroups and older age were significantly associated with higher mortality rates.

CONCLUSIONS

Respiratory failure was the main reason for ventilatory support at this non-surgical ICU. Immunosuppression, haemato-oncological diseases, the need for ECMO or renal replacement therapy and older age were associated with higher mortality.

Prediction of Diagnosis-Related Groups for Appendectomy Patients Using C4.5 and Neural Network

Due to the increasing cost of health insurance, for decades, many countries have endeavored to constrain the cost of insurance by utilizing a DRG payment system. In most cases, under the DRG payment system, hospitals cannot exactly know which DRG code inpatients are until they are discharged. This paper focuses on the prediction of what DRG code appendectomy patients will be classified with when they are admitted to hospital. We utilize two models (or classifiers) constructed using the C4.5 algorithm and back-propagation neural network (BPN). We conducted experiments with the data collected from two hospitals. The results show that the accuracies of these two classification models can be up to 97.84% and 98.70%, respectively. According to the predicted DRG code, hospitals can effectively arrange medical resources with certainty, then, in turn, improve the quality of the medical care patients receive.

Development, validation, and feature extraction of a deep learning model predicting in-hospital mortality using Japan's largest national ICU database: a validation framework for transparent clinical Artificial Intelligence (cAI) development

OBJECTIVE

While clinical Artificial Intelligence (cAI) mortality prediction models and relevant studies have increased, limitations including the lack of external validation studies and inadequate model calibration leading to decreased overall accuracy have been observed. To combat this, we developed and evaluated a novel deep neural network (DNN) and a validation framework to promote transparent cAI development.

METHODS

Data from Japan's largest ICU database was used to develop the DNN model, predicting in-hospital mortality including ICU and post-ICU mortality by days since ICU discharge. The most important variables to the model were extracted with SHapley Additive exPlanations (SHAP) to examine the DNN's efficacy as well as develop models that were also externally validated.

MAIN RESULTS

The area under the receiver operating characteristic curve (AUC) for predicting ICU mortality was 0.94 [0.93-0.95], and 0.91 [0.90-0.92] for in-hospital mortality, ranging between 0.91-0.95 throughout one year since ICU discharge. An external validation using only the top 20 variables resulted with higher AUCs than traditional severity scores.

CONCLUSIONS

Our DNN model consistently generated AUCs between 0.91-0.95 regardless of days since ICU discharge. The 20 most important variables to our DNN, also generated higher AUCs than traditional severity scores regardless of days since ICU discharge. To our knowledge, this is the first study that predicts ICU and in-hospital mortality using cAI by post-ICU discharge days up to over a year. This finding could contribute to increased transparency on cAI applications.

Two-Step Approach for Occupancy Estimation in Intensive Care Units Based on Bayesian Optimization Techniques

Due to the high occupational pressure suffered by intensive care units (ICUs), a correct estimation of the patients' length of stay (LoS) in the ICU is of great interest to predict possible situations of collapse, to help healthcare personnel to select appropriate treatment options and to predict patients' conditions. There has been a high amount of data collected by biomedical sensors during the continuous monitoring process of patients in the ICU, so the use of artificial intelligence techniques in automatic LoS estimation would improve patients' care and facilitate the work of healthcare personnel. In this work, a novel methodology to estimate the LoS using data of the first 24 h in the ICU is presented. To achieve this, XGBoost, one of the most popular and efficient state-of-the-art algorithms, is used as an estimator model, and its performance is optimized both from computational and precision viewpoints using Bayesian techniques. For this optimization, a novel two-step approach is presented. The methodology was carefully designed to execute codes on a high-performance computing system based on graphics processing units, which considerably reduces the execution time. The algorithm scalability is analyzed. With the proposed methodology, the best set of XGBoost hyperparameters are identified, estimating LoS with a MAE of 2.529 days, improving the results reported in the current state of the art and probing the validity and utility of the proposed approach.

Rate and risk factors of in-hospital and early post-discharge mortality in patients admitted to an internal medicine ward

BACKGROUND

We sought to quantify in-hospital and early post-discharge mortality rates in hospitalised patients.

METHODS

Consecutive adult patients admitted to an internal medicine ward were prospectively enrolled. The rates of in-hospital and 4-month post-discharge mortality and their possible associated sociodemographic and clinical factors (eg Cumulative Illness Rating Scale [CIRS], body mass index [BMI], polypharmacy, Barthel Index) were assessed.

RESULTS

1,451 patients (median age 80 years, IQR 69-86; 53% female) were included. Of these, 93 (6.4%) died in hospital, while 4-month post-discharge mortality was 15.9% (191/1,200). Age and high dependency were associated (p<0.01) with a higher risk of in-hospital (OR 1.04 and 2.15) and 4-month (HR 1.04 and 1.65) mortality, while malnutrition and length of stay were associated (p<0.01) with a higher risk of 4-month mortality (HR 2.13 and 1.59).

CONCLUSIONS

Several negative prognostic factors for early mortality were found. Interventions addressing dependency and malnutrition could potentially decrease early post-discharge mortality.

The magnitude of mortality and its determinants in Ethiopian adult intensive care units: A systematic review and meta-analysis

Introduction

Despite mortality in intensive care units being a global burden, it is higher in low-resource countries, including Ethiopia. A sufficient number of evidence is not yet established regarding mortality in the intensive care unit and its determinants. This study intended to determine the prevalence of ICU mortality and its determinants in Ethiopia.

Methods

PubMed, Google Scholar, The Cochrane Library, HINARI, and African Journals Online (AJOL) databases were systematically explored for potentially eligible studies on mortality prevalence and determinants reported by studies done in Ethiopia. Using a Microsoft Excel spreadsheet, two reviewers independently screen, select, review, and extract data for further analysis using STATA/MP version 17. A meta-analysis using a random-effects model was performed to calculate the pooled prevalence and odds ratio with a 95% confidence interval. In addition, using study region and sample size, subgroup analysis was also performed.

Results

9799 potential articles were found after removing duplicates and screening for eligibility, 14 were reviewed. Ethiopia's pooled national prevalence of adult intensive care unit mortality was 39.70% (95% CI: 33.66, 45.74). Mechanical ventilation, length of staying more than two weeks, GCS below 9, and acute respiratory distress syndrome were major predictors of mortality in intensive care units of Ethiopia.

Conclusion

Mortality in adult ICU is high in Ethiopia. We strongly recommend that all health care professionals and other stakeholders should act to decrease the high mortality among critically ill patients in Ethiopia.

A Machine Learning Model to Predict Length of Stay and Mortality among Diabetes and Hypertension Inpatients

Background and Objectives: Taiwan is among the nations with the highest rates of Type 2 Diabetes Mellitus (T2DM) and Hypertension (HTN). As more cases are reported each year, there is a rise in hospital admissions for people seeking medical attention. This creates a burden on hospitals and affects the overall management and administration of the hospitals. Hence, this study aimed to develop a machine learning (ML) model to predict the Length of Stay (LoS) and mortality among T2DM and HTN inpatients. Materials and Methods: Using Taiwan's National Health Insurance Research Database (NHIRD), this cohort study consisted of 58,618 patients, where 25,868 had T2DM, 32,750 had HTN, and 6419 had both T2DM and HTN. We analyzed the data with different machine learning models for the prediction of LoS and mortality. The evaluation was done by plotting descriptive statistical graphs, feature importance, precision-recall curve, accuracy plots, and AUC. The training and testing data were set at a ratio of 8:2 before applying ML algorithms. Results: XGBoost showed the best performance in predicting LoS (R2 0.633; RMSE 0.386; MAE 0.123), and RF resulted in a slightly lower performance (R2 0.591; RMSE 0.401; MAE 0.027). Logistic Regression (LoR) performed the best in predicting mortality (CV Score 0.9779; Test Score 0.9728; Precision 0.9432; Recall 0.9786; AUC 0.97 and AUPR 0.93), closely followed by Ridge Classifier (CV Score 0.9736; Test Score 0.9692; Precision 0.9312; Recall 0.9463; AUC 0.94 and AUPR 0.89). Conclusions: We developed a robust prediction model for LoS and mortality of T2DM and HTN inpatients. Linear Regression showed the best performance for LoS, and Logistic Regression performed the best in predicting mortality. The results showed that ML algorithms can not only help healthcare professionals in data-driven decision-making but can also facilitate early intervention and resource planning.

Evaluation of factors that influenced the length of hospital stay using data mining techniques

Background

length of stay (LOS) is the time between hospital admission and discharge. LOS has an impact on hospital management and hospital care functions.

Methods

A descriptive, retrospective study was designed on about 27,500 inpatients between March 2019 and 2020. Required data were collected from six wards (CCU, ICU, NICU, General, Maternity, and Women) in a teaching hospital. Clinical data such as demographic characteristics (age, sex), type of ward, and duration of hospital stay were analyzed by the R-studio program. Violin plots, bar charts, mosaic plots, and tree-based models were used to demonstrate the results.

Results

The mean age of the population was 40.8 ± 19.2 years. The LOS of the study population was 2.43 ± 4.13 days. About 60% of patients were discharged after staying one day in the hospital. After staying one day in the hospital, 67% of women were discharged. However, 23% of men were discharged within this time frame. The majority of LOS in the CCU, ICU, and NICU ranged from 5 to 9 days.; In contrast, LOS was one day in General, Maternity, and Woman wards. Due to the tree plot, there was a different LOS pattern between Maternity-Women and the CCU-General-ICU-NICU wards group.

Conclusion

We observed that patients with more severe diseases hospitalized in critical care wards had a longer LOS than those not admitted to critical care wards. The older patient had longer hospital LOS than the younger. By excluding Maternity and Woman wards, LOS in the hospital was comparable between males and females and demonstrated a similar pattern.

Integrated multimodal artificial intelligence framework for healthcare applications

Artificial intelligence (AI) systems hold great promise to improve healthcare over the next decades. Specifically, AI systems leveraging multiple data sources and input modalities are poised to become a viable method to deliver more accurate results and deployable pipelines across a wide range of applications. In this work, we propose and evaluate a unified Holistic AI in Medicine (HAIM) framework to facilitate the generation and testing of AI systems that leverage multimodal inputs. Our approach uses generalizable data pre-processing and machine learning modeling stages that can be readily adapted for research and deployment in healthcare environments. We evaluate our HAIM framework by training and characterizing 14,324 independent models based on HAIM-MIMIC-MM, a multimodal clinical database (N = 34,537 samples) containing 7279 unique hospitalizations and 6485 patients, spanning all possible input combinations of 4 data modalities (i.e., tabular, time-series, text, and images), 11 unique data sources and 12 predictive tasks. We show that this framework can consistently and robustly produce models that outperform similar single-source approaches across various healthcare demonstrations (by 6–33%), including 10 distinct chest pathology diagnoses, along with length-of-stay and 48 h mortality predictions. We also quantify the contribution of each modality and data source using Shapley values, which demonstrates the heterogeneity in data modality importance and the necessity of multimodal inputs across different healthcare-relevant tasks. The generalizable properties and flexibility of our Holistic AI in Medicine (HAIM) framework could offer a promising pathway for future multimodal predictive systems in clinical and operational healthcare settings.

Time-to-event modeling for hospital length of stay prediction for COVID-19 patients

Providing timely patient care while maintaining optimal resource utilization is one of the central operational challenges hospitals have been facing throughout the pandemic. Hospital length of stay (LOS) is an important indicator of hospital efficiency, quality of patient care, and operational resilience. Numerous researchers have developed regression or classification models to predict LOS. However, conventional models suffer from the lack of capability to make use of typically censored clinical data. We propose to use time-to-event modeling techniques, also known as survival analysis, to predict the LOS for patients based on individualized information collected from multiple sources. The performance of six proposed survival models is evaluated and compared based on clinical data from COVID-19 patients.

A Review of Artificial Intelligence and Robotics in Transformed Health Ecosystems

Health care is shifting toward become proactive according to the concept of P5 medicine-a predictive, personalized, preventive, participatory and precision discipline. This patient-centered care heavily leverages the latest technologies of artificial intelligence (AI) and robotics that support diagnosis, decision making and treatment. In this paper, we present the role of AI and robotic systems in this evolution, including example use cases. We categorize systems along multiple dimensions such as the type of system, the degree of autonomy, the care setting where the systems are applied, and the application area. These technologies have already achieved notable results in the prediction of sepsis or cardiovascular risk, the monitoring of vital parameters in intensive care units, or in the form of home care robots. Still, while much research is conducted around AI and robotics in health care, adoption in real world care settings is still limited. To remove adoption barriers, we need to address issues such as safety, security, privacy and ethical principles; detect and eliminate bias that could result in harmful or unfair clinical decisions; and build trust in and societal acceptance of AI.

Artificial intelligence and Big Data in neurology

Recent advances in technology have allowed us access to a multitude of datasets pertaining to various dimensions in neurology. Together with the enormous opportunities, we also face challenges related to data quality, ethics and intrinsic difficulties related to the application of data science in healthcare. In this article we will describe the main advances in the field of artificial intelligence and Big Data applied to neurology with a focus on neurosciences based on medical images. Real-World Data (RWD) and analytics related to large volumes of information will be described as well as some of the most relevant scientific initiatives at the time of this writing.

Patterns of presentation, prevalence and associated factors of mortality in ICU among adult patients during the pandemic of COVID 19: A retrospective cross-sectional study

Background

There is concern that patients admitted to the intensive care unit (ICU) with Corona Virus Disease 2019 (COVID-19) have variable prevalence reports of mortality. The survival rates are also inconsistently reported due to varying follow-up periods. Even if data on outcomes and baseline characteristics of ICU patients with COVID-19 is essential for action planning to manage complications, it is still left undisclosed in our study setting.

Materials and method

This cross-sectional study was conducted on 402 samples using a retrospective chart review of patient's data who were admitted in the past 2 years of the adult ICUs. All the data were entered and analyzed with SPSS version 21. A multivariable Logistic regression analysis was used to identify the association between outcome variables with independent factors and a p-value of less than 0.05 was taken as statistically significant with a 95% confidence interval. We used text, tables, and figures for the result.

Result

The overall prevalence of mortality among adult patients admitted to ICU during COVID-19 pandemics was 67.4%. From the multivariable logistic regression analysis, factors that were shown to have an association with an increase in ICU patient mortality were; lack of Vasopressor support, patients who had confirmed COVID 19 infection, core body temperature at admission greater than 37.5 °c, SPO2 at admission less than 90%, patients who had diagnosed ischemic heart disease (IHD), patients with acute respiratory distress syndrome (ARDS), patients who were intubated and mechanically ventilated (MV), and patient's ICU length of stay longer than two weeks.

Conclusion

The prevalence of ICU mortality in adult patients was higher in Debre Tabor Comprehensive specialized hospital. Therefore, clinicians need to minimize factors that maximize patient mortalities like ARDS, hyperthermia, Desaturation, Covid infection, IHD, intubation and MV, lack of Vasopressor use, and prolonged ICU stay.

Burden of Mental Illness among Primary HIV Discharges: A Retrospective Analysis of Inpatient Data

Background: Empirical evidence demonstrates the substantial burden of mental illness among people living with HIV and AIDS (PLWHA). Current literature also notes the co-morbidity of these two illnesses and its impact on quality of life and mortality. However, little evidence exists on patient outcomes, such as hospital length of stay or post-discharge status. Methods: A retrospective analysis of National Inpatient Sample data was conducted. The study population was defined as discharges having a primary diagnosis of HIV based on International Classification of Disease, 10th Revision, Clinical Modification (ICD-10-CM) codes in primary diagnosis field. Clinical Classification Software (CCS) codes are used to identify comorbid mental illness. Length of stay was defined as number of days between hospital admission and discharge. Disposition (or post-discharge status) was defined as routine versus not routine. Patient and hospital characteristics were used as control variables. All regression analyses were survey-weighted and adjusted for control variables. Results: The weighted population size (N) for this study was 26,055 (n = 5211). Among primary HIV discharges, presence of any mental illness as a secondary discharge was associated with 12% higher LOS, when compared to a lack of such comorbidity (incidence rate ratio [IRR] = 1.12, 95% confidence interval [CI] = 1.05, 1.22, p < 0.01). Likewise, among primary HIV discharges, those with mental illness had a 21% lower routine disposition, when compared to those without any mental illness (OR = 0.79, 95% CI = 0.68, 0.91, p < 0.001). Conclusion: Our results highlight the need for improved mental health screening and coordinated care to reduce the burden of mental illness among HIV discharges.

A systematic review of the prediction of hospital length of stay: Towards a unified framework

Hospital length of stay of patients is a crucial factor for the effective planning and management of hospital resources. There is considerable interest in predicting the LoS of patients in order to improve patient care, control hospital costs and increase service efficiency. This paper presents an extensive review of the literature, examining the approaches employed for the prediction of LoS in terms of their merits and shortcomings. In order to address some of these problems, a unified framework is proposed to better generalise the approaches that are being used to predict length of stay. This includes the investigation of the types of routinely collected data used in the problem as well as recommendations to ensure robust and meaningful knowledge modelling. This unified common framework enables the direct comparison of results between length of stay prediction approaches and will ensure that such approaches can be used across several hospital environments. A literature search was conducted in PubMed, Google Scholar and Web of Science from 1970 until 2019 to identify LoS surveys which review the literature. 32 Surveys were identified, from these 32 surveys, 220 papers were manually identified to be relevant to LoS prediction. After removing duplicates, and exploring the reference list of studies included for review, 93 studies remained. Despite the continuing efforts to predict and reduce the LoS of patients, current research in this domain remains ad-hoc; as such, the model tuning and data preprocessing steps are too specific and result in a large proportion of the current prediction mechanisms being restricted to the hospital that they were employed in. Adopting a unified framework for the prediction of LoS could yield a more reliable estimate of the LoS as a unified framework enables the direct comparison of length of stay methods. Additional research is also required to explore novel methods such as fuzzy systems which could build upon the success of current models as well as further exploration of black-box approaches and model interpretability.

Machine-learning based prediction of prognostic risk factors in patients with invasive candidiasis infection and bacterial bloodstream infection: a singled centered retrospective study

Background

Invasive candidal infection combined with bacterial bloodstream infection is one of the common nosocomial infections that is also the main cause of morbidity and mortality. The incidence of invasive Candidal infection with bacterial bloodstream infection is increasing year by year worldwide, but data on China is still limited.

Methods

We included 246 hospitalised patients who had invasive candidal infection combined with a bacterial bloodstream infection from January 2013 to January 2018; we collected and analysed the relevant epidemiological information and used machine learning methods to find prognostic factors related to death (training set and test set were randomly allocated at a ratio of 7:3).

Results

Of the 246 patients with invasive candidal infection complicated with a bacterial bloodstream infection, the median age was 63 years (53.25–74), of which 159 (64.6%) were male, 109 (44.3%) were elderly patients (> 65 years), 238 (96.7%) were hospitalised for more than 10 days, 168 (68.3%) were admitted to ICU during hospitalisation, and most patients had records of multiple admissions within 2 years (167/246, 67.9%). The most common blood index was hypoproteinemia (169/246, 68.7%), and the most common inducement was urinary catheter use (210/246, 85.4%). Moreover, the most frequently infected fungi and bacteria were Candida parapsilosis and Acinetobacter baumannii, respectively. The main predictors of death prognosis by machine learning method are serum creatinine level, age, length of stay, stay in ICU during hospitalisation, serum albumin level, C-Reactive protein (CRP), leukocyte count, neutrophil count, Procalcitonin (PCT), and total bilirubin level.

Conclusion

Our results showed that the most common candida and bacteria infections were caused by Candida parapsilosis and Acinetobacter baumannii, respectively. The main predictors of death prognosis are serum creatinine level, age, length of stay, stay in ICU during hospitalisation, serum albumin level, CRP, leukocyte count, neutrophil count, PCT and total bilirubin level.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12879-022-07125-8.

Towards accurate prediction of patient length of stay at emergency department: a GAN-driven deep learning framework

Recently, the hospital systems face a high influx of patients generated by several events, such as seasonal flows or health crises related to epidemics (e.g., COVID'19). Despite the extent of the care demands, hospital establishments, particularly emergency departments (EDs), must admit patients for medical treatments. However, the high patient influx often increases patients' length of stay (LOS) and leads to overcrowding problems within the EDs. To mitigate this issue, hospital managers need to predict the patient's LOS, which is an essential indicator for assessing ED overcrowding and the use of the medical resources (allocation, planning, utilization rates). Thus, accurately predicting LOS is necessary to improve ED management. This paper proposes a deep learning-driven approach for predicting the patient LOS in ED using a generative adversarial network (GAN) model. The GAN-driven approach flexibly learns relevant information from linear and nonlinear processes without prior assumptions on data distribution and significantly enhances the prediction accuracy. Furthermore, we classified the predicted patients' LOS according to time spent at the pediatric emergency department (PED) to further help decision-making and prevent overcrowding. The experiments were conducted on actual data obtained from the PED in Lille regional hospital center, France. The GAN model results were compared with other deep learning models, including deep belief networks, convolutional neural network, stacked auto-encoder, and four machine learning models, namely support vector regression, random forests, adaboost, and decision tree. Results testify that deep learning models are suitable for predicting patient LOS and highlight GAN's superior performance than the other models.

Predicting next-day discharge via electronic health record access logs

OBJECTIVE

Hospital capacity management depends on accurate real-time estimates of hospital-wide discharges. Estimation by a clinician requires an excessively large amount of effort and, even when attempted, accuracy in forecasting next-day patient-level discharge is poor. This study aims to support next-day discharge predictions with machine learning by incorporating electronic health record (EHR) audit log data, a resource that captures EHR users' granular interactions with patients' records by communicating various semantics and has been neglected in outcome predictions.

MATERIALS AND METHODS

This study focused on the EHR data for all adults admitted to Vanderbilt University Medical Center in 2019. We learned multiple advanced models to assess the value that EHR audit log data adds to the daily prediction of discharge likelihood within 24 h and to compare different representation strategies. We applied Shapley additive explanations to identify the most influential types of user-EHR interactions for discharge prediction.

RESULTS

The data include 26 283 inpatient stays, 133 398 patient-day observations, and 819 types of user-EHR interactions. The model using the count of each type of interaction in the recent 24 h and other commonly used features, including demographics and admission diagnoses, achieved the highest area under the receiver operating characteristics (AUROC) curve of 0.921 (95% CI: 0.919-0.923). By contrast, the model lacking user-EHR interactions achieved a worse AUROC of 0.862 (0.860-0.865). In addition, 10 of the 20 (50%) most influential factors were user-EHR interaction features.

CONCLUSION

EHR audit log data contain rich information such that it can improve hospital-wide discharge predictions.

Mortality in children aged <5 years with severe acute respiratory illness in a high HIV-prevalence urban and rural areas of South Africa, 2009-2013

BACKGROUND

Severe acute respiratory illness (SARI) is an important cause of mortality in young children, especially in children living with HIV infection. Disparities in SARI death in children aged <5 years exist in urban and rural areas.

OBJECTIVE

To compare the factors associated with in-hospital death among children aged <5 years hospitalized with SARI in an urban vs. a rural setting in South Africa from 2009-2013.

METHODS

Data were collected from hospitalized children with SARI in one urban and two rural sentinel surveillance hospitals. Nasopharyngeal aspirates were tested for ten respiratory viruses and blood for pneumococcal DNA using polymerase chain reaction. We used multivariable logistic regression to identify patient and clinical characteristics associated with in-hospital death.

RESULTS

From 2009 through 2013, 5,297 children aged <5 years with SARI-associated hospital admission were enrolled; 3,811 (72%) in the urban and 1,486 (28%) in the rural hospitals. In-hospital case-fatality proportion (CFP) was higher in the rural hospitals (6.9%) than the urban hospital (1.3%, p<0.001), and among HIV-infected than the HIV-uninfected children (9.6% vs. 1.6%, p<0.001). In the urban hospital, HIV infection (odds ratio (OR):11.4, 95% confidence interval (CI):5.4-24.1) and presence of any other underlying illness (OR: 3.0, 95% CI: 1.0-9.2) were the only factors independently associated with death. In the rural hospitals, HIV infection (OR: 4.1, 95% CI: 2.3-7.1) and age <1 year (OR: 3.7, 95% CI: 1.9-7.2) were independently associated with death, whereas duration of hospitalization ≥5 days (OR: 0.5, 95% CI: 0.3-0.8) and any respiratory virus detection (OR: 0.4, 95% CI: 0.3-0.8) were negatively associated with death.

CONCLUSION

We found that the case-fatality proportion was substantially higher among children admitted to rural hospitals and HIV infected children with SARI in South Africa. While efforts to prevent and treat HIV infections in children may reduce SARI deaths, further efforts to address health care inequality in rural populations are needed.

Improving clinical outcome predictions using convolution over medical entities with multimodal learning

Early prediction of mortality and length of stay (LOS) of a patient is vital for saving a patient's life and management of hospital resources. Availability of Electronic Health Records (EHR) makes a huge impact on the healthcare domain and there are several works on predicting clinical problems. However, many studies did not benefit from the clinical notes because of the sparse, and high dimensional nature. In this work, we extract medical entities from clinical notes and use them as additional features besides time-series features to improve proposed model predictions. The proposed convolution based multimodal architecture, which not only learns effectively combining medical entities and time-series Intensive Care Unit (ICU) signals of patients but also allows to compare the effect of different embedding techniques such as Word2vec and FastText on medical entities. Results show that the proposed deep multimodal method outperforms all other baseline models including multimodal architectures and improves the mortality prediction performance for Area Under the Receiver Operating Characteristics (AUROC) and Area Under Precision-Recall Curve (AUPRC) by around 3%. For LOS predictions, there is an improvement of around 2.5% over the time-series baseline. The code for the proposed method is available at https://github.com/tanlab/ConvolutionMedicalNer.

Methods and measures to quantify ICU patient heterogeneity

Patients in intensive care units are heterogeneous and the daily prediction of their days to discharge (DTD) a complex task that practitioners and computers are not always able to solve satisfactorily. In order to make more precise DTD predictors, it is necessary to have tools for the analysis of the heterogeneity of the patients. Unfortunately, the number of publications in this field is almost non-existent. In order to alleviate this lack of tools, we propose four methods and their corresponding measures to quantify the heterogeneity of intensive patients in the process of determining the DTD. These new methods and measures have been tested with patients admitted over four years to a tertiary hospital in Spain. The results deepen the understanding of the intensive patient and can serve as a basis for the construction of better DTD predictors.

Noninvasive Real-Time Mortality Prediction in Intensive Care Units Based on Gradient Boosting Method: Model Development and Validation Study

BACKGROUND

Monitoring critically ill patients in intensive care units (ICUs) in real time is vitally important. Although scoring systems are most often used in risk prediction of mortality, they are usually not highly precise, and the clinical data are often simply weighted. This method is inefficient and time-consuming in the clinical setting.

OBJECTIVE

The objective of this study was to integrate all medical data and noninvasively predict the real-time mortality of ICU patients using a gradient boosting method. Specifically, our goal was to predict mortality using a noninvasive method to minimize the discomfort to patients.

METHODS

In this study, we established five models to predict mortality in real time based on different features. According to the monitoring, laboratory, and scoring data, we constructed the feature engineering. The five real-time mortality prediction models were RMM (based on monitoring features), RMA (based on monitoring features and the Acute Physiology and Chronic Health Evaluation [APACHE]), RMS (based on monitoring features and Sequential Organ Failure Assessment [SOFA]), RMML (based on monitoring and laboratory features), and RM (based on all monitoring, laboratory, and scoring features). All models were built using LightGBM and tested with XGBoost. We then compared the performance of all models, with particular focus on the noninvasive method, the RMM model.

RESULTS

After extensive experiments, the area under the curve of the RMM model was 0.8264, which was superior to that of the RMA and RMS models. Therefore, predicting mortality using the noninvasive method was both efficient and practical, as it eliminated the need for extra physical interventions on patients, such as the drawing of blood. In addition, we explored the top nine features relevant to real-time mortality prediction: invasive mean blood pressure, heart rate, invasive systolic blood pressure, oxygen concentration, oxygen saturation, balance of input and output, total input, invasive diastolic blood pressure, and noninvasive mean blood pressure. These nine features should be given more focus in routine clinical practice.

CONCLUSIONS

The results of this study may be helpful in real-time mortality prediction in patients in the ICU, especially the noninvasive method. It is efficient and favorable to patients, which offers a strong practical significance.

The relationship of hospital stay and readmission with employment status

The Inpatient Clinico-Occupational Survey collected data from 3.76 million patients, showing that the average length of stay declined by 16.1 d in FY2008 and by 14.1 d in FY2015. In this study, we assessed the length of hospital stay and readmission, stratified by ICD-10 and employment status. A cross-sectional study was conducted on data from FY2008, including those from 65,806 first hospitalizations and 16,653 readmissions in FY2008, where 62,260 first admissions and 29,242 readmissions in FY 2015. The length of hospital stay was longest in those admitted due to external influences (24.8 d), followed by musculoskeletal disorders (22.5 d). This remained unchanged in FY2015, however, lengths of stay of those were reduced by 20.1 and 20.0 d, respectively. The length of hospital stay for most diseases was longer upon readmission than on first admission, and longer for those who were unemployed. It is necessary to give attention to patients who need to be discharged early due to work, or plan for frequent hospitalization in order to reduce the length of each hospital stay because of the expected increase in the number of elderly workers brought on by a declining birth rate and an aging population.

Factors associated with length of stay and death in tube-fed patients: A cross-sectional multicentre study

OBJECTIVES

To analyse the factors associated with length of stay (LOS) and death in nasogastric/nasoenteric tube (NG/NET)-fed patients.

DESIGN

A cross-sectional multicentre study.

METHOD

Data collection took place from October 2017-April 2019, and the sample consisted of 365 participants from seven Brazilian hospitals. Demographic, clinical and therapeutic data were collected from the patients' medical records. Data analysis was performed using bivariate and multivariate tests, considering a significance level of p<.05.

RESULTS

Most patients were male, older adults, with high risk of death and highly dependent on nursing care. The LOS was associated with age, patient care complexity and length of NG/NET use. Death was associated with patient age. In the multivariate analysis, patients highly dependent on nursing care, and intensive and semi-intensive care had a greater chance of dying when compared with patients receiving minimal care. Screening for factors affecting LOS and death is important to plan effective nursing care.

Factors associated with mechanical device-related complications in tube fed patients: A multicenter prospective cohort study

AIMS

To identify the types of nasogastric/nasoenteric tube (NGT/NET)-related adverse events and to analyze the degree of harm and the factors associated with mechanical device-related complications.

MATERIALS AND METHODS

A prospective cohort study was conducted from October 2017 to April 2019 in seven Brazilian hospitals. Data from 447 adult patients with NGT/NET were collected through electronic forms. Three methods were used to assess the NGT/NET-related adverse events: (1) encouraging spontaneous reports; (2) regular visits to the wards; and (3) review of medical records. The events were classified as mechanical device-related complications and other events. The degree of harm was classified according to the World Health Organization's International Classification for Patient Safety. Data were analyzed using the R program, version 3.5.3. The following tests were applied to identify associations between the explanatory and response variables: Cochran-Armitage Chi-Square test, Fisher's exact test, and Linear-by-linear Chi-Square test. Logistic regression analysis was performed to verify the predictors of mechanical device-related complications. All analyses were performed considering a 5% significance level.

RESULTS

191 NGT/NET-related adverse events were identified in 116 patients; the majority were mechanical device-related complications and resulted in mild harm to the patient. At the moment of the event, patients had a mean of 3.27 comorbidities, were highly dependent on nursing care, with high risk of death and altered level of consciousness. There was no association between the degree of harm and the care complexity, disease severity or level of consciousness. Intensive care was the strongest predictor for mechanical device-related complications and critical patients had a four times greater likelihood of presenting this type of event when compared to patients receiving minimal care.

CONCLUSION

Intensive care patients should receive special attention as the complexity of care is an important predictor for mechanical device-related complications in tube fed patients.

Malpractice and socioeconomic aspects in neurosurgery: a developing-country reality

OBJECTIVE

Neurosurgery occupies a prominent place in medical malpractice, but cases are still underreported in Brazil. This study describes the socioeconomic issues of medical malpractice in neurosurgery procedures and how they culminate in unfavorable outcomes in a developing country.

METHODS

The authors analyzed 112 neurosurgical procedures listed in the Brazilian Hospital Information System (Sistema de Informações Hospitalares do Sistema Único de Saúde [SIHSUS]) records in the DATASUS (Departamento de Informática do SUS) database between January 2008 and February 2020. Malpractice data were collected using the JusBrasil platform, with the authors searching the name of each of the 112 neurosurgical procedures plus “medical malpractice” among the jurisprudence records for January 2008 to February 2020. A simple linear regression analysis was performed using appropriate software. Analyses were considered statistically significant at p < 0.05.

RESULTS

According to DATASUS, 842,041 neurosurgical procedures were performed by the Brazilian Unified Health System between January 2008 and February 2020. The mean hospitalization cost for neurosurgical procedures was $714.06, and the average amount paid to professionals per procedure was $145.28 with variations according to the type of practice (public or private) in which they were performed, the complexity of the procedure, and the Brazilian region. The mortality rate and mean length of stay for neurosurgical procedures were 11.37% and 10.15 days, respectively. There were 79 medical malpractice lawsuits in the studied period. In these lawsuits, 26.58% of the court decisions were unfavorable to the neurosurgeons, with a mean compensation per procedure 15 times higher than the median value paid for all professionals in a neurosurgical procedure. The spine subspecialty had more lawsuits, and the brain tumor subspecialty had the most expensive compensation.

A lack of resources in public healthcare negatively impacts inpatient care. The mortality rate was 1.5 times higher in public practice than in private practice and was inversely proportional to the MTCs paid for the neurosurgical procedure. Patients with the lower educational levels associated with limited access to good medical care could reflect the lower plaintiff motivation in regions with a low gross domestic product and Human Development Index. In most cases, there is no understanding from either the patient or his family about the health-disease process, nor that there was medical malpractice committed by the physician to be sued.

CONCLUSIONS

The socioeconomic inequalities and the population’s low awareness of their rights could explain the few malpractice cases reported in Brazil. The authors recommend better decisions regarding the investments to be made in neurosurgical procedures to reduce malpractice lawsuits.

Machine Learning in the Prediction of Medical Inpatient Length of Stay

BACKGROUND

Length of stay (LOS) estimates are important for patients, doctors and hospital administrators. However, making accurate estimates of LOS can be difficult for medical patients.

AIMS

This review was conducted with the aim of identifying and assessing previous studies on the application of machine learning to the prediction of total hospital inpatient LOS for medical patients.

METHODS

A review of machine learning in the prediction of total hospital LOS for medical inpatients was conducted using the databases PubMed, EMBASE and Web of Science.

RESULTS

Of the 673 publications returned by the initial search, 21 articles met inclusion criteria. Of these articles the most commonly represented medical specialty was cardiology. Studies were also identified that had specifically evaluated machine learning LOS prediction in patients with diabetes and tuberculosis. The performance of the machine learning models in the identified studies varied significantly depending on factors including differing input datasets and different LOS thresholds and outcome metrics. Common methodological shortcomings included a lack of reporting of patient demographics and lack of reporting of clinical details of included patients.

CONCLUSIONS

The variable performance reported by the studies identified in this review supports the need for further research of the utility of machine learning in the prediction of total inpatient LOS in medical patients. Future studies should follow and report a more standardised methodology to better assess performance and to allow replication and validation. In particular, prospective validation studies and studies assessing the clinical impact of such machine learning models would be beneficial. This article is protected by copyright. All rights reserved.

Improving preoperative risk-of-death prediction in surgery congenital heart defects using artificial intelligence model: A pilot study

BACKGROUND

Congenital heart disease accounts for almost a third of all major congenital anomalies. Congenital heart defects have a significant impact on morbidity, mortality and health costs for children and adults. Research regarding the risk of pre-surgical mortality is scarce.

OBJECTIVES

Our goal is to generate a predictive model calculator adapted to the regional reality focused on individual mortality prediction among patients with congenital heart disease undergoing cardiac surgery.

METHODS

Two thousand two hundred forty CHD consecutive patients' data from InCor's heart surgery program was used to develop and validate the preoperative risk-of-death prediction model of congenital patients undergoing heart surgery. There were six artificial intelligence models most cited in medical references used in this study: Multilayer Perceptron (MLP), Random Forest (RF), Extra Trees (ET), Stochastic Gradient Boosting (SGB), Ada Boost Classification (ABC) and Bag Decision Trees (BDT).

RESULTS

The top performing areas under the curve were achieved using Random Forest (0.902). Most influential predictors included previous admission to ICU, diagnostic group, patient's height, hypoplastic left heart syndrome, body mass, arterial oxygen saturation, and pulmonary atresia. These combined predictor variables represent 67.8% of importance for the risk of mortality in the Random Forest algorithm.

CONCLUSIONS

The representativeness of "hospital death" is greater in patients up to 66 cm in height and body mass index below 13.0 for InCor's patients. The proportion of "hospital death" declines with the increased arterial oxygen saturation index. Patients with prior hospitalization before surgery had higher "hospital death" rates than who did not required such intervention. The diagnoses groups having the higher fatal outcomes probability are aligned with the international literature. A web application is presented where researchers and providers can calculate predicted mortality based on the CgntSCORE on any web browser or smartphone.

What factors predict length of stay in the intensive care unit? Systematic review and meta-analysis

PURPOSE

Studies have shown that a small percentage of ICU patients have prolonged length of stay (LoS) and account for a large proportion of resource use. Therefore, the identification of prolonged stay patients can improve unit efficiency. In this study, we performed a systematic review and meta-analysis to understand the risk factors of ICU LoS.

MATERIALS AND METHODS

We searched MEDLINE, Embase and Scopus databases from inception to November 2018. The searching process focused on papers presenting risk factors of ICU LoS. A meta-analysis was performed for studies reporting appropriate statistics.

RESULTS

From 6906 citations, 113 met the eligibility criteria and were reviewed. A meta-analysis was performed for six factors from 28 papers and concluded that patients with mechanical ventilation, hypomagnesemia, delirium, and malnutrition tend to have longer stay, and that age and gender were not significant factors.

CONCLUSIONS

This work suggested a list of risk factors that should be considered in prediction models for ICU LoS, as follows: severity scores, mechanical ventilation, hypomagnesemia, delirium, malnutrition, infection, trauma, red blood cells, and PaO2:FiO2. Our findings can be used by prediction models to improve their predictive capacity of prolonged stay patients, assisting in resource allocation, quality improvement actions, and benchmarking analysis.

Adverse Events Related to Partial Splenic Embolization for the Treatment of Hypersplenism: A Systematic Review

Partial splenic embolization is a common procedure that reduces thrombocytopenia in patients with hypersplenism. The present review evaluated the adverse event profile of partial splenic embolization detailed in 30 articles. Although the technical success rate of the procedure in these papers is high, many patients experienced postprocedural complications. Minor complications such as postembolization syndrome occurred frequently. Major complications were less frequent but sometimes resulted in mortality. Underlying liver dysfunction and high infarction rates may be risk factors leading to major complications. Interventional radiologists should be aware of the complication profile of this procedure and further advance research in techniques dealing with hypersplenism.

Cerebral Salt Wasting in Traumatic Brain Injury Is Associated with Increased Morbidity and Mortality

INTRODUCTION

In the setting of cerebral injury, cerebral salt wasting (CSW) is a potential cause of hyponatremia, which contributes to adverse effects and mortality.

OBJECTIVE

The primary objective of this study was to evaluate the clinical outcomes of severe traumatic brain injury (TBI) patients complicated by CSW.

METHODS

A retrospective data analysis was performed on data collected from patients with TBI with an abbreviated injury scale (AIS) greater than 3. Data was divided into 2 groups of patients with CSW and those without. The primary endpoint was incidence of adverse effects of CSW in regard to injury severity score (ISS), hospital length of stay (HLOS), ventilator days, ICU length of stay (ICU LOS) and survival to discharge. Data was analyzed using a one-way analysis of variance (ANOVA).

RESULTS

A total of 310 consecutive patients with severe head injury (anatomic injury score 3 or greater) were evaluated over a 3-year period. A total of 125 of the 310 patients (40%) were diagnosed with cerebral salt wasting as defined by hyponatremia with appropriate urinary output and salt replacement. Patients with CSW had poorer outcomes in regard to ISS (21.8 vs 14.2, p<0.0001), HLOS (14.1 vs 3.5, p<0.0001), ventilator days (5.0 vs 0.45, p<0.0001), ICU LOS (8.5 vs 1.6, p<0.0001), and survival to discharge (88% vs 99%, p<0.0001).

DISCUSSION

Common adverse effects of CSW were noted in this study. Patients with TBI have a predilection towards development of CSW and consequently have poorer outcomes including increased morbidity and mortality. Data is sparse on the duration of CSW and degree of hyponatremia over time. Larger, comparative studies need to be performed to investigate the hyponatremic patient population and the clinical outcomes of those who present with CSW.

Semi-quantitative visual assessment of chest radiography is associated with clinical outcomes in critically ill patients

Background

Respiratory pathology is a major driver of mortality in the intensive care unit (ICU), even in the absence of a primary respiratory diagnosis. Prior work has demonstrated that a visual scoring system applied to chest radiographs (CXR) is associated with adverse outcomes in ICU patients with Acute Respiratory Distress Syndrome (ARDS). We hypothesized that a simple, semi-quantitative CXR score would be associated with clinical outcomes for the general ICU population, regardless of underlying diagnosis.

Methods

All individuals enrolled in the Registry of Critical Illness at Brigham and Women’s Hospital between June 2008 and August 2018 who had a CXR within 24 h of admission were included. Each patient’s CXR was assigned an opacification score of 0–4 in each of four quadrants with the total score being the sum of all four quadrants. Multivariable negative binomial, logistic, and Cox regression, adjusted for age, sex, race, immunosuppression, a history of chronic obstructive pulmonary disease, a history of congestive heart failure, and APACHE II scores, were used to assess the total score’s association with ICU length of stay (LOS), duration of mechanical ventilation, in-hospital mortality, 60-day mortality, and overall mortality, respectively.

Results

A total of 560 patients were included. Higher CXR scores were associated with increased mortality; for every one-point increase in score, in-hospital mortality increased 10% (OR 1.10, CI 1.05–1.16, p < 0.001) and 60-day mortality increased by 12% (OR 1.12, CI 1.07–1.17, p < 0.001). CXR scores were also independently associated with both ICU length of stay (rate ratio 1.06, CI 1.04–1.07, p < 0.001) and duration of mechanical ventilation (rate ratio 1.05, CI 1.02–1.07, p < 0.001).

Conclusions

Higher values on a simple visual score of a patient’s CXR on admission to the medical ICU are associated with increased in-hospital mortality, 60-day mortality, overall mortality, length of ICU stay, and duration of mechanical ventilation.