Area Under The Receiver Operating Characteristic Research Articles

BiMPADR: A Deep Learning Framework for Predicting Adverse Drug Reactions in New Drugs

Detecting the unintended adverse reactions of drugs (ADRs) is a crucial concern in pharmacological research. The experimental validation of drug–ADR associations often entails expensive and time-consuming investigations. Thus, a computational model to predict ADRs from known associations is essential for enhanced efficiency and cost-effectiveness. Here, we propose BiMPADR, a novel model that integrates drug gene expression into adverse reaction features using a message passing neural network on a bipartite graph of drugs and adverse reactions, leveraging publicly available data. By combining the computed adverse reaction features with the structural fingerprints of drugs, we predict the association between drugs and adverse reactions. Our models obtained high AUC (area under the receiver operating characteristic curve) values ranging from 0.861 to 0.907 in an external drug validation dataset under differential experiment conditions. The case study on multiple BET inhibitors also demonstrated the high accuracy of our predictions, and our model’s exploration of potential adverse reactions for HWD-870 has contributed to its research and development for market approval. In summary, our method would provide a promising tool for ADR prediction and drug safety assessment in drug discovery and development.

Endocannabinoids and cortisol in plasma, aqueous and tear samples of primary angle closure glaucoma versus controls.

To compare the levels of endocannabinoids (EC) in plasma, aqueous humor and tears, cortisol in plasma and aqueous, in primary angle closure glaucoma (PACG) and controls, while comparing the quality of life in both groups. A total of 60 patients, ≥40years of age, with a diagnosis of PACG or cataract, 30 in each group were recruited. They were subjected to a detailed ophthalmic evaluation, a WHO Quality of Life Brief Version (WHOQOL-BREF) questionnaire answering and collection of tear, aqueous and blood samples. The levels of endocannabinoids-anandamide (AEA), 2-arachidonoylglycerol (2AG) in plasma, aqueous humor and tears; cortisol in plasma and aqueous humor; and WHO-QOL score in each group were noted. Plasma AEA (p = 0.01) and plasma 2-AG, (p = 0.002) levels were significantly higher in the control group as compared to the PACG group. WHO-QOL score was better in controls (p < 0.001). The EC were in undetectable levels in aqueous. Plasma and aqueous cortisol were significantly higher in PACG and both had the highest Area under the receiver operating characteristics (AUROC) curve value for differentiating PACG from controls. Tear 2AG and tear AEA had a moderately strong positive correlation with plasma 2-AG. Females had insignificantly higher levels of plasma and tear endocannabinoids. Tear endocannabinoids were determined for the first time in PACG and normal with no difference between the two groups. Plasma and aqueous cortisol levels are a differentiating factor between normal and glaucoma patients with plasma endocannabinoids being remarkably higher in normals. Quality of life in glaucoma patients with high cortisol levels is poorer.

Enhancing Selection of Alcohol Consumption Associated Genes by Random Forest.

Machine learning methods have been used in identifying omics markers for a variety of phenotypes. We aimed to examine whether a supervised machine learning algorithm can improve identification of alcohol-associated transcriptomic markers. In this study, we analyzed array-based, whole-blood derived expression data for 17,873 gene transcripts in 5,508 Framingham Heart Study participants. By using the Boruta algorithm, a supervised Random Forest (RF)-based feature selection method, we selected 25 alcohol-associated transcripts. In a testing set (30% of entire study participants), AUCs (area under the receiver operating characteristics curve) of these 25 transcripts were 0.73, 0.69, and 0.66 for nondrinkers vs. moderate drinkers, nondrinkers vs. heavy drinkers, and moderate drinkers vs. heavy drinkers, respectively. The AUCs of the selected transcripts by the Boruta method were comparable to those identified using conventional linear regression models, e.g., AUCs of 1,985 transcripts identified by conventional linear regression models (false discovery rate < 0.05) were 0.72, 0.68, and 0.68, respectively. With Bonferroni correction for the 25 Boruta method selected transcripts and three CVD risk factors (i.e., at P < 6.7e-4), we observed 13 transcripts were associated with obesity, 3 transcripts with type 2 diabetes, and 1 transcript with hypertension. For example, we observed that alcohol consumption was inversely associated with the expression of DOCK4, IL4R, and SORT1, and DOCK4 and SORT1 were positively associated with obesity and IL4R was inversely associated with hypertension. In conclusion, using a supervised machine learning method, the RF-based Boruta algorithm, we identified novel alcohol-associated gene transcripts.

An Interpretable Machine Learning Approach to Predict Survival Outcomes in Spinal and Sacropelvic Chordomas.

Retrospective, population-based cohort study. This study aimed to develop machine learning (ML) models to predict five-year and 10-year mortality in spinal and sacropelvic chordoma patients and integrate them into a web application for enhanced prognostication. Past research has uncovered factors influencing survival in spinal chordoma patients. While identifying individual predictors is important, personalized survival predictions are equally vital. Though prior efforts have resulted in nomograms aiming to serve this purpose, they cannot capture complex interactions within data and rely on statistical assumptions that may not fit real-world data. Adult spinal and sacropelvic chordoma patients were identified from the National Cancer Database. Sociodemographic, clinicopathologic, diagnostic, and treatment-related variables were utilized as predictive features. Five supervised ML algorithms (TabPFN, CatBoost, XGBoost, LightGBM, and Random Forest) were implemented to predict mortality at five and 10 years postdiagnosis. Model performance was primarily evaluated using the area under the receiver operating characteristic (AUROC). SHapley Additive exPlanations (SHAP) values and partial dependence plots provided feature importance and interpretability. The top models were integrated into a web application. From the NCDB, 1206 adult patients diagnosed with histologically confirmed spinal and sacropelvic chordomas were retrieved for the five-year mortality outcome [423 (35.1%) with five-year mortality] and 801 patients for the 10-year mortality outcome [588 (73.4%) with 10-year mortality]. Top-performing models for both of the outcomes were the models created with the CatBoost algorithm. The CatBoost model for five-year mortality predictions displayed a mean AUROC of 0.801, and the CatBoost model predicting 10-year mortality yielded a mean AUROC of 0.814. This study developed ML models that can accurately predict five-year to 10-year survival probabilities in spinal chordoma patients. Integrating these interpretable, personalized prognostic models into a web application provides quantitative survival estimates for a given patient. The local interpretability enables transparency into how predictions are influenced. Further external validation is warranted to support generalizability and clinical utility.

The Performance of a Survival Nomogram and Albumin–Bilirubin Grade as Prognostic Tools in Advanced Hepatocellular Carcinoma Treated with FOLFOX4

Background: The ability of the survival nomogram developed in the EACH study and albumin–bilirubin (ALBI) grade to predict the survival of advanced hepatocellular carcinoma (HCC) patients receiving oxaliplatin plus 5-fluorouracil/leucovorin (FOLFOX4) remains unvalidated. Here, we comprehensively evaluated these prognostic tools. Methods: The survival nomogram and ALBI grade of each patient were assessed, and the area under the receiver operating curve (AUROC) and Harrell’s C-index for the risk classification model were calculated. Results: Overall, 76 HCC patients who received FOLFOX4 between August 2017 and June 2023 were included. The survival nomogram classified patients into low-, intermediate-, and high-risk groups, with a median overall survival (OS) of 9.82, 10.64, and 3.70 months, respectively (p = 0.23). The AUROC was 0.621 and Harrell’s C-index was 0.589. However, the ALBI grade categorized all patients into grade 1, 2, and 3, with a median OS of 9.82, 6.83, and 1.58 months, respectively (p = 0.00024). The AUROC was 0.663 and Harrell’s C-index was 0.663. Conclusion: The ALBI grade can be a potential prognostic tool. However, the survival nomogram does not provide clear discrimination. Therefore, FOLFOX4 should be an option for patients with ALBI grade 1 who cannot receive immunotherapy or targeted therapy. Additional prospective studies with a larger cohort are warranted to validate the survival nomogram and ALBI grade as prognostic tools.

Validation of three health literacy screening questions compared with S-TOFHLA in a low-income diverse English- and Spanish-Speaking population.

Clinicians need a tool to gauge patients' ability to understand health conditions and treatment options. The Short-form Test of Functional Health Literacy in Adults (S-TOFHLA) is the gold standard for this, but its length is prohibitive for use in clinical settings. This study seeks to validate a novel three-item question set for predicting health literacy. This cross-sectional study utilized an in-person questionnaire alongside the S-TOFHLA. The sample included 2027 English- and Spanish-speaking adults (≥18years) recruited from primary care practices serving a low-income eastern Pennsylvania community. Most patients (57.7%) identified as Hispanic. Diagnostic accuracy of each question and aggregated scores were assessed against the validated survey by calculating the area under the receiver operating characteristic (AUROC) curve. Questions in the 'Problems Learning' and 'Help Reading' domains (AUROC 0.66 for each) performed better than the 'Confident Forms' question (AUROC 0.64). Summing all three scores resulted in an even higher AUROC curve (0.71). Cronbach's alpha of the combined items was 0.696. Study results suggest that any of the three questions are viable options for screening health literacy levels of diverse patients in primary care clinical settings. However, they perform better as a summed score than when used individually.

Deep learning to predict rapid progression of Alzheimer's disease from pooled clinical trials: A retrospective study.

The rate of progression of Alzheimer's disease (AD) differs dramatically between patients. Identifying the most is critical because when their numbers differ between treated and control groups, it distorts the outcome, making it impossible to tell whether the treatment was beneficial. Much recent effort, then, has gone into identifying RPs. We pooled de-identified placebo-arm data of three randomized controlled trials (RCTs), EXPEDITION, EXPEDITION 2, and EXPEDITION 3, provided by Eli Lilly and Company. After processing, the data included 1603 mild-to-moderate AD patients with 80 weeks of longitudinal observations on neurocognitive health, brain volumes, and amyloid-beta (Aβ) levels. RPs were defined by changes in four neurocognitive/functional health measures. We built deep learning models using recurrent neural networks with attention mechanisms to predict RPs by week 80 based on varying observation periods from baseline (e.g., 12, 28 weeks). Feature importance scores for RP prediction were computed and temporal feature trajectories were compared between RPs and non-RPs. Our evaluation and analysis focused on models trained with 28 weeks of observation. The models achieved robust internal validation area under the receiver operating characteristic (AUROCs) ranging from 0.80 (95% CI 0.79-0.82) to 0.82 (0.81-0.83), and the area under the precision-recall curve (AUPRCs) from 0.34 (0.32-0.36) to 0.46 (0.44-0.49). External validation AUROCs ranged from 0.75 (0.70-0.81) to 0.83 (0.82-0.84) and AUPRCs from 0.27 (0.25-0.29) to 0.45 (0.43-0.48). Aβ plasma levels, regional brain volumetry, and neurocognitive health emerged as important factors for the model prediction. In addition, the trajectories were stratified between predicted RPs and non-RPs based on factors such as ventricular volumes and neurocognitive domains. Our findings will greatly aid clinical trialists in designing tests for new medications, representing a key step toward identifying effective new AD therapies.

Crossectional Study on the Performance of Screening Questionnaires for Prediction of Moderate to Severe Obstructive Sleep Apnea in Women

Abstract Introduction The clinical manifestations of obstructive sleep apnea (OSA) are different between genders. Though there are several screening questionnaires for OSA, their performance in females is not fully understood, as women have been historically underrepresented in research studies. Objective To assess the performance of screening questionnaires and their capacity to identify a moderate to severe apnea-hypopnea index (AHI) in women. Materials and Methods The Epworth sleep scale (ESS), Berlin questionnaire, and STOP-BANG questionnaire (SBQ) were correlated with AHI. Also, the sensitivity (S), specificity (Sp), and area under the receiver operating characteristic (AUC-ROC) curve were calculated for each questionnaire and combinations thereof. Multiple regression models were used to identify ≥15 ev/h AHI. Results Our study included 5,344 patients: 1978 women (37.1%) aged 55.06 ± 14 years with body mass index (BMI): 32.6 ± 8.30 kg/m2, ESS: 7.69 ± 5.2 points, and high-risk Berlin score: 87.25%. An AHI ≥15 ev/h was found in 30.4% of women. In terms of the capacity to identify an ≥15 ev/h AHI in women, the AUC-ROC of ESS &gt;10 and high-risk Berlin was 0.53 and 0.58, respectively. Three components of SBQ in any combination showed: a S of 65.1% (95% CI: 61.2–68.9), a Sp: 61.5% (95% CI: 58.9–64.1), with the AUC-ROC: 0.67. Conclusions Questionnaires perform differently in women. Therefore, it is necessary to take a gender-specific approach. The SBQ showed a higher discriminative power and more specificity than the ESS and the Berlin questionnaire. The best performance was obtained with any combination of 3 SBQ components. Age, BMI, neck circumference, and hypertension were the strongest predictors.

Development and validation of prediction models for papillary thyroid cancer structural recurrence using machine learning approaches

BackgroundAlthough papillary thyroid cancer (PTC) patients are known to have an excellent prognosis, up to 30% of patients experience disease recurrence after initial treatment. Accurately predicting disease prognosis remains a challenge given that the predictive value of several predictors remains controversial. Thus, we investigated whether machine learning (ML) approaches based on comprehensive predictors can predict the risk of structural recurrence for PTC patients.MethodsA total of 2244 patients treated with thyroid surgery and radioiodine were included. Twenty-nine perioperative variables consisting of four dimensions (demographic characteristics and comorbidities, tumor-related variables, lymph node (LN)-related variables, and metabolic and inflammatory markers) were analyzed. We applied five ML algorithms—logistic regression (LR), support vector machine (SVM), extreme gradient boosting (XGBoost), random forest (RF), and neural network (NN)—to develop the models. The area under the receiver operating characteristic (AUC-ROC) curve, calibration curve, and variable importance were used to evaluate the models’ performance.ResultsDuring a median follow-up of 45.5 months, 179 patients (8.0%) experienced structural recurrence. The non-stimulated thyroglobulin, LN dissection, number of LNs dissected, lymph node metastasis ratio, N stage, comorbidity of hypertension, comorbidity of diabetes, body mass index, and low-density lipoprotein were used to develop the models. All models showed a greater AUC (AUC = 0.738 to 0.767) than did the ATA risk stratification (AUC = 0.620, DeLong test: P < 0.01). The SVM, XGBoost, and RF model showed greater sensitivity (0.568, 0.595, 0.676), specificity (0.903, 0.857, 0.784), accuracy (0.875, 0.835, 0.775), positive predictive value (PPV) (0.344, 0.272, 0.219), negative predictive value (NPV) (0.959, 0.959, 0.964), and F1 score (0.429, 0.373, 0.331) than did the ATA risk stratification (sensitivity = 0.432, specificity = 0.770, accuracy = 0.742, PPV = 0.144, NPV = 0.938, F1 score = 0.216). The RF model had generally consistent calibration compared with the other models. The Tg and the LNR were the top 2 important variables in all the models, the N stage was the top 5 important variables in all the models.ConclusionsThe RF model achieved the expected prediction performance with generally good discrimination, calibration and interpretability in this study. This study sheds light on the potential of ML approaches for improving the accuracy of risk stratification for PTC patients.Trial registrationRetrospectively registered at www.chictr.org.cn (trial registration number: ChiCTR2300075574, date of registration: 2023-09-08).

Neutrophil–Creatinine Index: A New Prognostic Factor for Severity of Acute Pancreatitis

Background and Objectives: Determining the severity of acute pancreatitis (AP) is the main goal in the early stage of AP. The aim of this study was to compare laboratory parameters and indices, including the neutrophil to lymphocyte ratio (NLR) and the neutrophil–creatinine index (NCI), at admission in order to predict the severity of AP. Materials and Methods: Data from 421 patients who were admitted with a diagnosis of AP were collected retrospectively. Disease severity was assessed using the Bedside Index of Severity in Acute Pancreatitis (BISAP) and the revised Atlanta classification (RAC). BISAP was graded as mild and severe, and RAC was graded as mild (MAP), moderately severe (MSAP), and severe (SAP). The laboratory parameters and indices, including the NLR and NCI, were compared. Results: Of the patients, 70 (16.6%) had severe AP according to BISAP; the AP subgroups according to the RAC were as follows: MAP (n = 213), MSAP (n = 158), and SAP (n = 50). The NCI had the highest area under the receiver operator characteristic (AUROC) curve value (0.862), demonstrating severe disease according to BISAP, with a sensitivity of 78.6% and a specificity of 79.8%. Age (OR:1.046), white blood cell count (WBC) (OR:1.141), hematocrit (OR:1.081), blood urea nitrogen (BUN) (OR:1.040), and NCI (OR:1.076) were independently associated with severe disease, according to the multivariate analysis results, and were determined as components of the newly developed nomogram. The AUROC of the nomogram (0.891) was superior to the AUROCs of all the components of the nomogram except the NCI. Moreover, the NCI was the only parameter to distinguish MSAP from MAP (OR:1.119, 95% CI: 1.015–1.235, p = 0.023) and SAP from MSAP (OR:1.095, 95% CI: 1.031–1.162, p = 0.003). Conclusions: The present study enabled the identification of the neutrophil–creatinine index as a new prognostic tool for the assessment of AP severity at hospital admission.

Abstract LB243: Deep learning-based molecular characterization of lung cancers from never smokers using hematoxylin and eosin-stained whole slide images

Abstract Purpose: This study employs cutting-edge deep learning techniques to comprehensively analyze hematoxylin and eosin-stained (H&amp;E) whole slide images (WSIs) to predict which tumors carry driver gene alterations, mutational signatures, and additional molecular features among lung cancers from never-smokers (LCINS)— the seventh leading cause of cancer death worldwide. Method: A total of 464 H&amp;E-stained WSIs of lung adenocarcinomas from the Sherlock-Lung study were utilized, with 325 WSIs for model training and 139 for testing. Alongside the WSIs, genomic data, including mutations (overall nonsynonymous mutations and driver mutations only), fusions, or copy numbers in driver genes (TP53, KRAS, EGFR, CDKN2A, MDM2, ALK, RBM10), mutational signatures (APOBEC), molecular features (Kataegies, Whole Genome Doubling (WGD) status, Tumor Mutational Burden (TMB)), and specific hotspot driver mutation in EGFR (p.L858R and p.E746_A750del) and KRAS (p.G12C, p.G12V, and p.G12D), were included for all tumors. Each WSI was assigned different labels based on the presence or absence of these genomic alterations or features. We used a customized multilabel binary deep learning model based on ResNet50, a residual convolutional neural network, for analyzing the data. The network was trained from scratch, with an initial epoch set to 100. Results: Our methodology demonstrated high predictive performance, measured by the area under the receiver operating characteristic (AUROC). In the presence or absence category, we achieved high AUROC for detecting any alterations in these driver genes: TP53 (0.81), KRAS (0.92), EGFR (0.93), CDKN2A deletion (0.88), MDM2 amplification (0.94), and ALK fusion (0.86). We also evaluated molecular features, obtaining AUROC scores of 0.79 for WGD status and 0.86 for Kataegies. Moderate AUROC scores were observed for tumors with RBM10 mutations (0.67), high TMB (0.64), and APOBEC signatures (0.56). When focusing on driver mutations only, we achieved high AUROC scores for EGFR (0.88), KRAS (0.84), TP53 (0.93), and RBM10 (0.78). In addition, our model successfully predicted tumors with specific mutations in EGFR (p.L858R = 0.79; p.E746_A750del = 0.77) and KRAS (p.G12C = 0.77). Performance was suboptimal for KRAS p.G12V (0.41) and KRAS p.G12D (0.47). Conclusions: Our deep learning network achieves high prediction scores in identifying tumors with critical driver gene alterations and actionable mutations, holding promise for potential clinical use. In the future, the model could be optimized as a screening assay to guide molecular testing and therapeutic management of patients with LCINS. Citation Format: Monjoy Saha, Tongwu Zhang, Praphulla Bhawsar, Wei Zhao, Jianxin Shi, Soo Ryum Yang, Jonas Almeida, Maria Teresa Landi. Deep learning-based molecular characterization of lung cancers from never smokers using hematoxylin and eosin-stained whole slide images [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(7_Suppl):Abstract nr LB243.

Relative band power in assessing temporary neurological dysfunction post- type A aortic dissection surgery: a prospective study

Temporary neurological dysfunction (TND), a common complication following surgical repair of Type A Aortic Dissection (TAAD), is closely associated with increased mortality and long-term cognitive impairment. Currently, effective treatment options for TND remain elusive. Therefore, we sought to investigate the potential of postoperative relative band power (RBP) in predicting the occurrence of postoperative TND, with the aim of identifying high-risk patients prior to the onset of TND. We conducted a prospective observational study between February and December 2022, involving 165 patients who underwent surgical repair for TAAD at our institution. Bedside Quantitative electroencephalography (QEEG) was utilized to monitor the post-operative brain electrical activity of each participant, recording changes in RBP (RBP Delta, RBP Theta, RBP Beta and RBP Alpha), and analyzing their correlation with TND. Univariate and multivariate analyses were employed to identify independent risk factors for TND. Subsequently, line graphs were generated to estimate the incidence of TND. The primary outcome of interest was the development of TND, while secondary outcomes included intensive care unit (ICU) admission and length of hospital stay. A total of 165 patients were included in the study, among whom 68 (41.2%) experienced TND. To further investigate the independent risk factors for postoperative TND, we conducted both univariate and multivariate logistic regression analyses on all variables. In the univariate regression analysis, we identified age (Odds Ratio [OR], 1.025; 95% CI, 1.002–1.049), age ≥ 60 years (OR, 2.588; 95% CI, 1.250–5.475), hemopericardium (OR, 2.767; 95% CI, 1.150–7.009), cardiopulmonary bypass (CPB) (OR, 1.007; 95% CI, 1.001–1.014), RBP Delta (OR, 1.047; 95% CI, 1.020–1.077), RBP Alpha (OR, 0.853; 95% CI, 0.794–0.907), and Beta (OR, 0.755; 95% CI, 0.649–0.855) as independent risk factors for postoperative TND. Further multivariate regression analyses, we discovered that CPB time ≥ 180 min (OR, 1.021; 95% CI, 1.011–1.032), RBP Delta (OR, 1.168; 95% CI, 1.105–1.245), and RBP Theta (OR, 1.227; 95% CI, 1.135–1.342) emerged as independent risk factors. TND patients had significantly longer ICU stays (p < 0.001), and hospital stays (p = 0.002). We obtained the simplest predictive model for TND, consisting of three variables (CPB time ≥ 180 min, RBP Delta, RBP Theta, upon which we constructed column charts. The areas under the receiver operating characteristic (AUROC) were 0.821 (0.755, 0.887). Our study demonstrates that postoperative RBP monitoring can detect changes in brain function in patients with TAAD during the perioperative period, providing clinicians with an effective predictive method that can help improve postoperative TND in TAAD patients. These findings have important implications for improving clinical care in this population.Trial registration ChiCTR2200055980. Registered 30th Jan. 2022. This trial was registered before the first participant was enrolled.

Development of automated neural network prediction for echocardiographic left ventricular ejection fraction

IntroductionThe echocardiographic measurement of left ventricular ejection fraction (LVEF) is fundamental to the diagnosis and classification of patients with heart failure (HF).MethodsThis paper aimed to quantify LVEF automatically and accurately with the proposed pipeline method based on deep neural networks and ensemble learning. Within the pipeline, an Atrous Convolutional Neural Network (ACNN) was first trained to segment the left ventricle (LV), before employing the area-length formulation based on the ellipsoid single-plane model to calculate LVEF values. This formulation required inputs of LV area, derived from segmentation using an improved Jeffrey’s method, as well as LV length, derived from a novel ensemble learning model. To further improve the pipeline’s accuracy, an automated peak detection algorithm was used to identify end-diastolic and end-systolic frames, avoiding issues with human error. Subsequently, single-beat LVEF values were averaged across all cardiac cycles to obtain the final LVEF.ResultsThis method was developed and internally validated in an open-source dataset containing 10,030 echocardiograms. The Pearson’s correlation coefficient was 0.83 for LVEF prediction compared to expert human analysis (p &amp;lt; 0.001), with a subsequent area under the receiver operator curve (AUROC) of 0.98 (95% confidence interval 0.97 to 0.99) for categorisation of HF with reduced ejection (HFrEF; LVEF&amp;lt;40%). In an external dataset with 200 echocardiograms, this method achieved an AUC of 0.90 (95% confidence interval 0.88 to 0.91) for HFrEF assessment.ConclusionThe automated neural network-based calculation of LVEF is comparable to expert clinicians performing time-consuming, frame-by-frame manual evaluations of cardiac systolic function.

Reducing Airway Occlusion Time Without Losing Accuracy to Predict Successful Mechanical Ventilator Liberation During the Measurement of the Timed Inspiratory Effort Index.

In 2013, a new predictor of successful mechanical ventilation liberation named timed inspiratory effort (TIE) index was devised with the normalization of the maximum inspiratory pressure (obtained within 60 s of unidirectional airway occlusion) with the time at which the value was reached. The aim of this study was to verify whether the presence of a sequence of a certain number of inspiratory effort values between 30-60 s > 1.0 cm H2O/s could predict weaning success in a performance comparable to the TIE index. This was a retrospective observational study using 4 databases of previous studies on the TIE index. All patients receiving mechanical ventilation for ≤ 24 h were eligible. Liberation from mechanical ventilation-extubation decisions was made based on performance with spontaneous breathing trials. P < .05 was considered significant. The performance of the TIE index was evaluated by calculating the area under the receiver operating characteristics (AUROC) curve. From 349 eligible patients, 165 subjects were selected for analysis. The AUROC for the TIE index in the studied sample was 0.92 (95% CI 0.87-0.97, P < .001). A sequence of ≤ 4 inspiratory efforts > 1.0 cm H2O/s was found in 51.5% of the subjects, with successful ventilatory liberation occurring in 95.3%. The highest specificity values belonged to the sequence of ≤ 4 and ≤ 5 inspiratory efforts > 1.0 cm H2O/s; the highest positive predictive value and positive likelihood ratio belonged to the sequence of ≤ 4 inspiratory efforts > 1.0 cm H2O/s. The mean time that could have been spared if the procedure were interrupted after the first sequence of 4 inspiratory efforts > 1.0 cm H2O/s was 23 ± 3 s. The presence of a sequence of ≤ 4 inspiratory efforts > 1.0 cm H2O/s during the TIE index measurement was a reliable predictor of weaning success, which could allow timely interruption of the procedure and entail a substantial reduction in airway occlusion time.

Early Vital Sign Thresholds Associated with 24-Hour Mortality among Trauma Patients: A Trauma Quality Improvement Program (TQIP) Study.

Identifying patients at imminent risk of death is critical in the management of trauma patients. This study measures the vital sign thresholds associated with death among trauma patients. This study included data from patients ≥15 years of age in the American College of Surgeons Trauma Quality Improvement Program (TQIP) database. Patients with vital signs of zero were excluded. Documented prehospital and emergency department (ED) vital signs included systolic pressure, heart rate, respiratory rate, and calculated shock index (SI). The area under the receiver operator curves (AUROC) was used to assess the accuracy of these variables for predicting 24-hour survival. Optimal thresholds to predict mortality were identified using Youden's Index, 90% specificity, and 90% sensitivity. Additional analyses examined patients 70+ years of age. There were 1,439,221 subjects in the 2019-2020 datasets that met inclusion for this analysis with <0.1% (10,270) who died within 24 hours. The optimal threshold for prehospital systolic pressure was 110, pulse rate was 110, SI was 0.9, and respiratory rate was 15. The optimal threshold for the ED systolic was 112, pulse rate was 107, SI was 0.9, and respiratory rate was 21. Among the elderly sub-analysis, the optimal threshold for prehospital systolic was 116, pulse rate was 100, SI was 0.8, and respiratory rate was 21. The optimal threshold for ED systolic was 121, pulse rate was 95, SI was 0.8, and respiratory rate was 0.8. Systolic blood pressure (SBP) and SI offered the best predictor of mortality among trauma patients. The SBP values predictive of mortality were significantly higher than the traditional 90mmHg threshold. This dataset highlights the need for better methods to guide resuscitation as initial vital signs have limited accuracy in predicting subsequent mortality.

Machine learning based predictive modeling of readmissions following extracorporeal membrane oxygenation hospitalizations

BackgroundDespite increasing utilization and survival benefit over the last decade, extracorporeal membrane oxygenation (ECMO) remains resource-intensive with significant complications and rehospitalization risk. We thus utilized machine learning (ML) to develop prediction models for 90-day nonelective readmission following ECMO. MethodsAll adult patients receiving ECMO who survived index hospitalization were tabulated from the 2016–2020 Nationwide Readmissions Database. Extreme Gradient Boosting (XGBoost) models were developed to identify features associated with readmission following ECMO. Area under the receiver operating characteristic (AUROC), mean Average Precision (mAP), and the Brier score were calculated to estimate model performance relative to logistic regression (LR). Shapley Additive Explanation summary (SHAP) plots evaluated the relative impact of each factor on the model. An additional sensitivity analysis solely included patient comorbidities and indication for ECMO as potential model covariates. ResultsOf ~22,947 patients, 4495 (19.6 %) were readmitted nonelectively within 90 days. The XGBoost model exhibited superior discrimination (AUROC 0.64 vs 0.49), classification accuracy (mAP 0.30 vs 0.20) and calibration (Brier score 0.154 vs 0.165, all P < 0.001) in predicting readmission compared to LR. SHAP plots identified duration of index hospitalization, undergoing heart/lung transplantation, and Medicare insurance to be associated with increased odds of readmission. Upon sub-analysis, XGBoost demonstrated superior disclination compared to LR (AUROC 0.61 vs 0.60, P < 0.05). Chronic liver disease and frailty were linked with increased odds of nonelective readmission. ConclusionsML outperformed LR in predicting readmission following ECMO. Future work is needed to identify other factors linked with readmission and further optimize post-ECMO care among this cohort.

Significance of Pretreatment Hemoglobin-Albumin-Lymphocyte-Platelet Index for the Prediction of Suboptimal Surgery in Epithelial Ovarian Cancer.

Epithelial ovarian cancer (EOC) is the leading cause of death in gynecological cancers in developed countries. In recent years, there has been a growing need for economical and accurate pretreatment laboratory investigations to assess the prognosis of patients with advanced EOC (AEOC). We aimed to investigate the role of the hemoglobin-albumin-lymphocyte-platelet (HALP) index in suboptimal cytoreduction and oncological outcomes. A prognostic prediction model for diagnosing suboptimal cytoreduction for patients with AEOC receiving neoadjuvant chemotherapy (NACT) was developed. Multivariate logistic regression analysis was performed to identify the independent predictors of suboptimal cytoreduction, with a P-value < 0.05, and then transformed into risk-scoring systems. Internal validation was performed using the bootstrapping procedure, and predictive cytoreduction (PSC) scores were compared using non-parametric receiver operating characteristic (ROC) regression. Survival analysis was performed using Kaplan-Meier estimation and Cox proportional regression. In total, 473 patients were analyzed, and the rate of suboptimal surgery was 43%. A scoring system in predicting suboptimal cytoreduction included age, cancer antigen (CA)-125 level before surgery, performance status, cycles of chemotherapy, peritoneal cancer index, and HALP index ≤ 22.6. The model had good discriminative ability (area under the ROC (AUROC), 0.80; 95% confidence interval (CI), 0.76 - 0.84), outperforming the PSC score (AUROC, 0.75; 95% CI, 0.71 - 0.80). The score was divided into the low-risk (positive predictive value (PPV), 22.4; 95% CI, 17.8 - 27.7), moderate-risk (PPV, 65.9; 95% CI, 56.9 - 74.0), and high-risk (PPV, 90.6; 95% CI, 79.3 - 96.9) groups. The HALP index score of ≤ 22.6 was independently associated with progression-free survival (hazard ratio (HR), 2.92; 95% CI, 1.58 - 5.40) and overall survival (HR, 2.66; 95% CI, 1.57 - 4.49). The HALP index is a newly predicted factor for suboptimal cytoreduction and oncological outcomes in patients with AEOC after NACT.

Machine learning-enabled maternal risk assessment for women with pre-eclampsia (the PIERS-ML model): a modelling study

Affecting 2-4% of pregnancies, pre-eclampsia is a leading cause of maternal death and morbidity worldwide. Using routinely available data, we aimed to develop and validate a novel machine learning-based and clinical setting-responsive time-of-disease model to rule out and rule in adverse maternal outcomes in women presenting with pre-eclampsia. We used health system, demographic, and clinical data from the day of first assessment with pre-eclampsia to predict a Delphi-derived composite outcome of maternal mortality or severe morbidity within 2 days. Machine learning methods, multiple imputation, and ten-fold cross-validation were used to fit models on a development dataset (75% of combined published data of 8843 patients from 11 low-income, middle-income, and high-income countries). Validation was undertaken on the unseen 25%, and an additional external validation was performed in 2901 inpatient women admitted with pre-eclampsia to two hospitals in south-east England. Predictive risk accuracy was determined by area-under-the-receiver-operator characteristic (AUROC), and risk categories were data-driven and defined by negative (-LR) and positive (+LR) likelihood ratios. Of 8843 participants, 590 (6·7%) developed the composite adverse maternal outcome within 2 days, 813 (9·2%) within 7 days, and 1083 (12·2%) at any time. An 18-variable random forest-based prediction model, PIERS-ML, was accurate (AUROC 0·80 [95% CI 0·76-0·84] vs the currently used logistic regression model, fullPIERS: AUROC 0·68 [0·63-0·74]) and categorised women into very low risk (-LR <0·1; eight [0·7%] of 1103 women), low risk (-LR 0·1 to 0·2; 321 [29·1%] women), moderate risk (-LR >0·2 and +LR <5·0; 676 [61·3%] women), high risk (+LR 5·0 to 10·0, 87 [7·9%] women), and very high risk (+LR >10·0; 11 [1·0%] women). Adverse maternal event rates were 0% for very low risk, 2% for low risk, 5% for moderate risk, 26% for high risk, and 91% for very high risk within 48 h. The 2901 women in the external validation dataset were accurately classified as being at very low risk (0% with outcomes), low risk (1%), moderate risk (4%), high risk (33%), or very high risk (67%). The PIERS-ML model improves identification of women with pre-eclampsia who are at lowest and greatest risk of severe adverse maternal outcomes within 2 days of assessment, and can support provision of accurate guidance to women, their families, and their maternity care providers. University of Strathclyde Diversity in Data Linkage Centre for Doctoral Training, the Fetal Medicine Foundation, The Canadian Institutes of Health Research, and the Bill & Melinda Gates Foundation.

Multimodal Machine Learning for Prediction of 30-day Readmission Risk in Elderly Population

Background: Readmission within 30 days is a prevalent issue among elderly patients, linked to unfavorable health outcomes. Our objective was to develop and validate multimodal machine learning models for predicting 30-day readmission risk in elderly patients discharged from internal medicine departments. Methods: This was a retrospective cohort study which included elderly patients aged 75 or older, who were hospitalized at the Hadassah Medical Center internal medicine departments between 2014 and 2020. Three machine learning algorithms were developed and employed to predict 30-day readmission risk. The primary measures were predictive model performance scores, specifically area under the receiver operator curve (AUROC), and average precision. Results: This study included 19,569 admissions. Of them, 3,258 (16.65%) resulted in 30-day readmission. Our three proposed models demonstrated high accuracy and precision on an unseen test set, with AUROC values of 0.87, 0.89, and 0.93, respectively, and average precision values of 0.76, 0.78, and 0.81. Feature importance analysis revealed that the number of admissions in the past year, history of 30-day readmission, Charlson score, and admission length were the most influential variables. Notably, the natural language processing score, representing the probability of readmission according to a textual-based model trained on social workers assessment letters during hospitalization, ranked among the top 10 contributing factors. Conclusions: Leveraging multimodal machine learning offers a promising strategy for identifying elderly patients who are at high risk for 30-day readmission. By identifying these patients, machine learning models may facilitate the effective execution of preventive actions to reduce avoidable readmission incidents.

Comparison of Lymphocyte/Monocyte ratio (LMR) with Child-Pugh Score and MELD Score in the Determination of Hepatic Dysfunction Severity and Outcome in Cirrhotic Patient

Background: Lymphocyte/monocyte ratio (LMR) is a simple and effective marker that has the potential to predict the severity of hepatic dysfunction in a cirrhotic patient, like the Model for End-Stage Liver Disease (MELD) score and the Child- Pugh (CP) score. However, the usefulness of this newer tool has not yet been tested. Objectives: Comparison of Lymphocyte/Monocyte ratio with Child-Pugh Score and MELD Score in the determination of hepatic dysfunction severity in patients with cirrhosis was the objective of the study. Materials and Methods: This cross-sectional analytical study was conducted on cases of cirrhotic patients in the Department of Gastroenterology, Bangabandhu Sheikh Mujib Medical University, dated from September 2018 to August 2019. A total of 40 compensated cases and 82 decompensated cases were included in the process of obtaining informed written consent. Information about clinical profile, laboratory parameters (complete blood count including ESR, serum total bilirubin, serum albumin, ALT, AST, serum creatinine, prothrombin time, INR, etc.) were collected. LMR, MELD score, and CP score were calculated both in compensated and decompensated cirrhotic patients to assess the relationship with the severity of hepatic dysfunction. The correlation between LMR and the CP/MELD score was established by the Pearson correlation test. The receiver operating characteristic (ROC) curve and cut-off values of LMR were obtained, and areas under the ROC (AUROC) curve were calculated to identify the best LMR and/or the MELD score or CPS for predicting hepatic decompensation. The Bland-Altman plot and the Heliey- MacNeil test were used for comparison of measurement techniques. The data were analyzed with the help of SPSS version 20. Observation and Results: The mean age of the decompensated group was higher than that of the compensated group (p&lt;0.05). The mean values for LMR and MELD and CP scores were 7.93±3.08, 5.25±0.43 &amp; 7.10 ±1.19, respectively. Average LMR was significantly higher in compensated cirrhosis patients while CP and MELD Score was lower in that group (p&lt;0.001). MELD and CP were positively correlated with each other (p&lt;0.001). LMR was negatively correlated to both MELD and CP scores (p&lt;0.001). Of all patients, only 4 died (3.3%) and rest 118 (96.7%) survived. The cut off value of LMR, CP and MELD were LMR ≤ 2.18 (sensitivity: 75%, specificity: 87.3%), CPS ≥13.5 (sensitivity: 75%, specificity: 98.3%) &amp; MELD ≥30.5 (sensitivity: 75%, specificity: 94.9%) respectively. Pairwise comparison showed that the difference between the AUCs of MELD and LMR was not statistically significant (0.953 vs 0.887; p=0.081). Patients in the low LMR group showed decreased survival than those in the high LMR group (p=0.000). The no survived group had lower LMR and higher MELD and CP scores than those of the survived group (p&lt;0.001). Conclusion: In comparison with Child-Pugh Score and MELD Score in the determination of hepatic dysfunction severity in cirrhotic patients, Lymphocyte/ Monocyte ratio is a useful tool.