ABSTRACT
BACKGROUND: Severe heart failure (HF) has a higher mortality during vulnerable period while targeted predictive tools, especially based on drug exposures, to accurately assess its prognoses remain largely unexplored. Therefore, this study aimed to utilize drug information as the main predictor to develop and validate survival models for severe HF patients during this period. METHODS: We extracted severe HF patients from the MIMIC-IV database (as training and internal validation cohorts) as well as from the MIMIC-III database and local hospital (as external validation cohorts). Three algorithms, including Cox proportional hazards model (CoxPH), random survival forest (RSF), and deep learning survival prediction (DeepSurv), were applied to incorporate the parameters (partial hospitalization information and exposure durations of drugs) for constructing survival prediction models. The model performance was assessed mainly using area under the receiver operator characteristic curve (AUC), brier score (BS), and decision curve analysis (DCA). The model interpretability was determined by the permutation importance and Shapley additive explanations values. RESULTS: A total of 11,590 patients were included in this study. Among the 3 models, the CoxPH model ultimately included 10 variables, while RSF and DeepSurv models incorporated 24 variables, respectively. All of the 3 models achieved respectable performance metrics while the DeepSurv model exhibited the highest AUC values and relatively lower BS among these models. The DCA also verified that the DeepSurv model had the best clinical practicality. CONCLUSIONS: The survival prediction tools established in this study can be applied to severe HF patients during vulnerable period by mainly inputting drug treatment duration, thus contributing to optimal clinical decisions prospectively.
Subject(s)
Heart Failure , Proportional Hazards Models , Humans , Heart Failure/mortality , Heart Failure/drug therapy , Female , Male , Aged , Reproducibility of Results , Prognosis , Survival Analysis , Middle Aged , ROC Curve , Algorithms , Area Under Curve , Databases, Factual , Deep Learning , Severity of Illness IndexABSTRACT
BACKGROUND: Nosocomial infections (NIs) frequently occur and adversely impact prognosis for hospitalized patients with cirrhosis. This study aims to develop and validate two machine learning models for NIs and in-hospital mortality risk prediction. METHODS: The Prediction of Nosocomial Infection and Prognosis in Cirrhotic patients (PIPC) study included hospitalized patients with cirrhosis at the Qingchun Campus of the First Affiliated Hospital of Zhejiang University. We then assessed several machine learning algorithms to construct predictive models for NIs and prognosis. We validated the best-performing models with bootstrapping techniques and an external validation dataset. The accuracy of the predictions was evaluated through sensitivity, specificity, predictive values, and likelihood ratios, while predictive robustness was examined through subgroup analyses and comparisons between models. RESULTS: We enrolled 1,297 patients into derivation cohort and 496 patients into external validation cohort. Among the six algorithms assessed, the Random Forest algorithm performed best. For NIs, the PIPC-NI model achieved an area under the curve (AUC) of 0.784 (95% confidence interval [CI] 0.741-0.826), a sensitivity of 0.712, and a specificity of 0.702. For in-hospital mortality, the PIPC- mortality model achieved an AUC of 0.793 (95% CI 0.749-0.836), a sensitivity of 0.769, and a specificity of 0.701. Moreover, our PIPC models demonstrated superior predictive performance compared to the existing MELD, MELD-Na, and Child-Pugh scores. CONCLUSIONS: The PIPC models showed good predictive power and may facilitate healthcare providers in easily assessing the risk of NIs and prognosis among hospitalized patients with cirrhosis.
Subject(s)
Cross Infection , Hospital Mortality , Liver Cirrhosis , Machine Learning , Humans , Cross Infection/mortality , Liver Cirrhosis/complications , Liver Cirrhosis/mortality , Male , Female , Middle Aged , Prognosis , Aged , Hospitalization , Algorithms , Risk Assessment/methods , Risk Factors , Area Under CurveABSTRACT
OBJECTIVES: Diagnostic prediction models exist to assess the probability of bacterial meningitis (BM) in paediatric patients with suspected meningitis. To evaluate the diagnostic accuracy of these models in a broad population of children suspected of a central nervous system (CNS) infection, we performed external validation. METHODS: We performed a systematic literature review in Medline to identify articles on the development, refinement or validation of a prediction model for BM, and validated these models in a prospective cohort of children aged 0-18 years old suspected of a CNS infection. PRIMARY AND SECONDARY OUTCOME MEASURES: We calculated sensitivity, specificity, predictive values, the area under the receiver operating characteristic curve (AUC) and evaluated calibration of the models for diagnosis of BM. RESULTS: In total, 23 prediction models were validated in a cohort of 450 patients suspected of a CNS infection included between 2012 and 2015. In 75 patients (17%), the final diagnosis was a CNS infection including 30 with BM (7%). AUCs ranged from 0.69 to 0.94 (median 0.83, interquartile range [IQR] 0.79-0.87) overall, from 0.74 to 0.96 (median 0.89, IQR 0.82-0.92) in children aged ≥28 days and from 0.58 to 0.91 (median 0.79, IQR 0.75-0.82) in neonates. CONCLUSIONS: Prediction models show good to excellent test characteristics for excluding BM in children and can be of help in the diagnostic workup of paediatric patients with a suspected CNS infection, but cannot replace a thorough history, physical examination and ancillary testing.
Subject(s)
Central Nervous System Infections , Meningitis, Bacterial , Humans , Meningitis, Bacterial/diagnosis , Child , Prospective Studies , Central Nervous System Infections/diagnosis , Child, Preschool , Infant , Adolescent , Infant, Newborn , Area Under Curve , ROC Curve , Predictive Value of Tests , Sensitivity and SpecificityABSTRACT
BACKGROUND/AIMS: Hepatocellular carcinoma (HCC) is one of the cancers with the highest incidence and mortality rates. This study aims to explore the diagnostic and prognostic utility of methyltransferase like 13 (METTL13) in patients with HCC via bioinformatics analysis. MATERIALS AND METHODS: We obtained mRNA data of HCC from the database of the Cancer Genome Atlas (TCGA), drawing survival curve by R 4.2.1 software. Cox regression analysis was conducted based on tumor stage and METTL13 expression. The GSE114564 dataset was chosen from the Gene Expression Omnibus. The differences in serum METTL13 levels between the groups of early HCC (eHCC) and non-cancer controls were evaluated. Using a receiver operating characteristic curve, we calculated the area under the curve (AUC) of serum METTL13 for diagnosing eHCC. RESULTS: A total of 225 cases with HCC were screened from TCGA, and 29 cases were normal controls. The results showed that the METTL13 expression in the HCC group was higher than that in the normal controls (P <.001). The univariate [hazard ratio (HR) = 1.895, P = .006] and multivariate Cox regression (HR = 1.702, P = .037) analyses showed that high METTL13 expression reduced overall survival in HCC. Serum METTL13 levels were higher in the eHCC group than in the non-cancer controls (P = .008). The optimum AUC for predicting eHCC by serum METTL13 was 0.7091. CONCLUSION: Serum METTL13 has a moderate diagnostic value for eHCC. High METTL13 expression is correlated with a worse prognosis in patients with HCC. Methyltransferase like 13 possesses the potential to be a novel biomarker for HCC.
Subject(s)
Biomarkers, Tumor , Carcinoma, Hepatocellular , Computational Biology , Liver Neoplasms , Methyltransferases , Humans , Carcinoma, Hepatocellular/genetics , Carcinoma, Hepatocellular/blood , Carcinoma, Hepatocellular/mortality , Carcinoma, Hepatocellular/diagnosis , Liver Neoplasms/genetics , Liver Neoplasms/blood , Liver Neoplasms/mortality , Liver Neoplasms/diagnosis , Female , Prognosis , Male , Biomarkers, Tumor/blood , Biomarkers, Tumor/genetics , Middle Aged , Methyltransferases/genetics , Methyltransferases/blood , ROC Curve , Proportional Hazards Models , RNA, Messenger/blood , Area Under Curve , Case-Control Studies , AgedABSTRACT
Purpose: To develop and validate machine learning (ML) models for predicting cycloplegic refractive error and myopia status using noncycloplegic refractive error and biometric data. Methods: Cross-sectional study of children aged five to 18 years who underwent biometry and autorefraction before and after cycloplegia. Myopia was defined as cycloplegic spherical equivalent refraction (SER) ≤-0.5 Diopter (D). Models were evaluated for predicting SER using R2 and mean absolute error (MAE) and myopia status using area under the receiver operating characteristic (ROC) curve (AUC). Best-performing models were further evaluated using sensitivity/specificity and comparison of observed versus predicted myopia prevalence rate overall and in each age group. Independent data sets were used for training (n = 1938) and validation (n = 1476). Results: In the validation dataset, ML models predicted cycloplegic SER with high R2 (0.913-0.935) and low MAE (0.393-0.480 D). The AUC for predicting myopia was high (0.984-0.987). The best-performing model for SER (XGBoost) had high sensitivity and specificity (91.1% and 97.2%). Random forest (RF), the best-performing model for myopia, had high sensitivity and specificity (92.2% and 96.9%). Within each age group, difference between predicted and actual myopia prevalence was within 4%. Conclusions: Using noncycloplegic refractive error and ocular biometric data, ML models performed well for predicting cycloplegic SER and myopia status. When measuring cycloplegic SER is not feasible, ML may provide a useful tool for estimating cycloplegic SER and myopia prevalence rate in epidemiological studies. Translational Relevance: Using ML to predict cycloplegic refraction based on noncycloplegic data is a powerful tool for large, population-based studies of refractive error.
Subject(s)
Machine Learning , Mydriatics , Myopia , Refraction, Ocular , Humans , Child , Cross-Sectional Studies , Male , Female , Myopia/epidemiology , Myopia/diagnosis , Adolescent , Child, Preschool , Mydriatics/administration & dosage , Refraction, Ocular/physiology , China/epidemiology , Biometry/methods , Refractive Errors/epidemiology , Refractive Errors/diagnosis , ROC Curve , Prevalence , Area Under Curve , Students , East Asian PeopleABSTRACT
BACKGROUND AND OBJECTIVES: Esmethadone (dextromethadone; d-methadone; S-methadone (+)-methadone; REL-1017) is a low potency N-methyl-D-aspartate (NMDA) receptor channel blocker that showed a rapid and sustained adjunctive antidepressant effects in patients with major depressive disorder with inadequate response to ongoing serotonergic antidepressant treatment. Previous studies indicated that esmethadone is partially excreted by the kidney (53.9% of the dose) and by the liver (39.1% of the dose). METHODS: Here we studied the pharmacokinetics and safety of esmethadone after a single oral dose of 25 mg in subjects with different stages of kidney and liver impairment. RESULTS: In subjects with a mild and moderate decrease in glomerular fraction rate (GFR), esmethadone Cmax and AUC0-inf values did not differ compared with healthy subjects. In patients with severe renal impairment, the ratios of Cmax and AUC0-inf values compared with healthy subjects were above 100% (138.22-176.85%) and, while modest, these increases reached statistical significance. In subjects with end stage renal disease (ESRD) undergoing intermittent hemodialysis (IHD), Cmax and AUC0-inf values were not statistically different compared with healthy subjects. IHD did not modified plasma total esmethadone concentrations in blood exiting versus entering the dialyzer. Dose adjustment is not warranted in subjects with mild-to-moderate impaired renal function. Dose reduction may be considered for select patients with severe renal disfunction. In subjects with mild-or-moderate hepatic impairment, Cmax and AUC0-inf were approximately 20-30% lower compared with healthy controls. The drug free fraction increased with the severity of hepatic impairment, from 5.4% in healthy controls to 8.3% in subjects with moderate hepatic impairment. CONCLUSION: Mild and moderate hepatic impairment has a minimal to modest impact on exposure to total or unbound esmethadone and dose adjustments are not warranted in subjects with mild and moderate hepatic impairment. Administration of esmethadone was well tolerated in healthy adult subjects, in subjects with mild or moderate hepatic impairment, and in subjects with mild moderate or severe renal impairment, including patients with ESRF undergoing dialysis.
Subject(s)
Methadone , Renal Insufficiency, Chronic , Humans , Male , Female , Middle Aged , Adult , Methadone/pharmacokinetics , Methadone/administration & dosage , Methadone/adverse effects , Renal Insufficiency, Chronic/therapy , Liver Diseases , Aged , Area Under Curve , Young Adult , Administration, Oral , Glomerular Filtration Rate/drug effectsABSTRACT
Purpose: Compare the use of optic disc and macular optical coherence tomography measurements to predict glaucomatous visual field (VF) worsening. Methods: Machine learning and statistical models were trained on 924 eyes (924 patients) with circumpapillary retinal nerve fiber layer (cp-RNFL) or ganglion cell inner plexiform layer (GC-IPL) thickness measurements. The probability of 24-2 VF worsening was predicted using both trend-based and event-based progression definitions of VF worsening. Additionally, the cp-RNFL and GC-IPL predictions were combined to produce a combined prediction. A held-out test set of 617 eyes was used to calculate the area under the curve (AUC) to compare cp-RNFL, GC-IPL, and combined predictions. Results: The AUCs for cp-RNFL, GC-IPL, and combined predictions with the statistical and machine learning models were 0.72, 0.69, 0.73, and 0.78, 0.75, 0.81, respectively, when using trend-based analysis as ground truth. The differences in performance between the cp-RNFL, GC-IPL, and combined predictions were not statistically significant. AUCs were highest in glaucoma suspects using cp-RNFL predictions and highest in moderate/advanced glaucoma using GC-IPL predictions. The AUCs for the statistical and machine learning models were 0.63, 0.68, 0.69, and 0.72, 0.69, 0.73, respectively, when using event-based analysis. AUCs decreased with increasing disease severity for all predictions. Conclusions: cp-RNFL and GC-IPL similarly predicted VF worsening overall, but cp-RNFL performed best in early glaucoma stages and GC-IPL in later stages. Combining both did not enhance detection significantly. Translational Relevance: cp-RNFL best predicted trend-based 24-2 VF progression in early-stage disease, while GC-IPL best predicted progression in late-stage disease. Combining both features led to minimal improvement in predicting progression.
Subject(s)
Disease Progression , Glaucoma , Optic Disk , Retinal Ganglion Cells , Tomography, Optical Coherence , Visual Fields , Humans , Tomography, Optical Coherence/methods , Female , Optic Disk/diagnostic imaging , Optic Disk/pathology , Male , Visual Fields/physiology , Middle Aged , Glaucoma/diagnostic imaging , Glaucoma/physiopathology , Retinal Ganglion Cells/pathology , Machine Learning , Aged , Nerve Fibers/pathology , Area Under Curve , Macula Lutea/diagnostic imaging , Macula Lutea/pathology , Vision Disorders/physiopathology , Vision Disorders/diagnostic imaging , Vision Disorders/diagnosisABSTRACT
BACKGROUND: Sjögren's Syndrome (SS) is a rare chronic autoimmune disorder primarily affecting adult females, characterized by chronic inflammation and salivary and lacrimal gland dysfunction. It is often associated with systemic lupus erythematosus, rheumatoid arthritis and kidney disease, which can lead to increased mortality. Early diagnosis is critical, but traditional methods for diagnosing SS, mainly through histopathological evaluation of salivary gland tissue, have limitations. METHODS: The study used 100 labial gland biopsy, creating whole-slide images (WSIs) for analysis. The proposed model, named Cell-tissue-graph-based pathological image analysis model (CTG-PAM) and based on graph theory, characterizes single-cell feature, cell-cell feature, and cell-tissue feature. Building upon these features, CTG-PAM achieves cellular-level classification, enabling lymphocyte recognition. Furthermore, it leverages connected component analysis techniques in the cell graph structure to perform SS diagnosis based on lymphocyte counts. FINDINGS: CTG-PAM outperforms traditional deep learning methods in diagnosing SS. Its area under the receiver operating characteristic curve (AUC) is 1.0 for the internal validation dataset and 0.8035 for the external test dataset. This indicates high accuracy. The sensitivity of CTG-PAM for the external dataset is 98.21%, while the accuracy is 93.75%. In comparison, the sensitivity and accuracy for traditional deep learning methods (ResNet-50) are lower. The study also shows that CTG-PAM's diagnostic accuracy is closer to skilled pathologists compared to beginners. INTERPRETATION: Our findings indicate that CTG-PAM is a reliable method for diagnosing SS. Additionally, CTG-PAM shows promise in enhancing the prognosis of SS patients and holds significant potential for the differential diagnosis of both non-neoplastic and neoplastic diseases. The AI model potentially extends its application to diagnosing immune cells in tumor microenvironments.
Subject(s)
Sjogren's Syndrome , Sjogren's Syndrome/diagnosis , Sjogren's Syndrome/pathology , Humans , Female , Cohort Studies , ROC Curve , Image Processing, Computer-Assisted/methods , Middle Aged , Deep Learning , Area Under Curve , Adult , AutomationABSTRACT
Diagnosing patients in the medical emergency department is complex and this is expected to increase in many countries due to an ageing population. In this study we investigate the feasibility of training machine learning algorithms to assist physicians handling the complex situation in the medical emergency departments. This is expected to reduce diagnostic errors and improve patient logistics and outcome. We included a total of 9,190 consecutive patient admissions diagnosed and treated in two hospitals in this cohort study. Patients had a biochemical workup including blood and urine analyses on clinical decision totaling 260 analyses. After adding nurse-registered data we trained 19 machine learning algorithms on a random 80% sample of the patients and validated the results on the remaining 20%. We trained algorithms for 19 different patient outcomes including the main outcomes death in 7 (Area under the Curve (AUC) 91.4%) and 30 days (AUC 91.3%) and safe-discharge(AUC 87.3%). The various algorithms obtained areas under the Receiver Operating Characteristics -curves in the range of 71.8-96.3% in the holdout cohort (68.3-98.2% in the training cohort). Performing this list of biochemical analyses at admission also reduced the number of subsequent venipunctures within 24 h from patient admittance by 22%. We have shown that it is possible to develop a list of machine-learning algorithms with high AUC for use in medical emergency departments. Moreover, the study showed that it is possible to reduce the number of venipunctures in this cohort.
Subject(s)
Emergency Service, Hospital , Machine Learning , Humans , Female , Male , Aged , Middle Aged , Algorithms , ROC Curve , Cohort Studies , Aged, 80 and over , Adult , Area Under CurveABSTRACT
Introduction: Identifying heart failure (HF) in acute exacerbation of chronic obstructive pulmonary disease (AECOPD) can be challenging. Lung ultrasound sonography (LUS) B-lines quantification has recently gained a large place in the diagnosis of HF, but its diagnostic performance in AECOPD remains poorly studied. Purpose: This study aimed to assess the contribution of LUS B-lines score (LUS score) in the diagnosis of HF in AECOPD patients. Patients and methods: This is a prospective cross-sectional multicenter cohort study including patients admitted to the emergency department for AECOPD. All included patients underwent LUS. A lung ultrasound score (LUS score) based on B-lines calculation was assessed. A cardiac origin of dyspnea was retained for a LUS score greater than 15. HF diagnosis was based on clinical examination, pro-brain natriuretic peptide levels, and echocardiographic findings. The LUS score diagnostic performance was assessed by receiver operating characteristic (ROC) curve, sensitivity, specificity, and likelihood ratio at the best cutoffs. Results: We included 380 patients, mean age was 68±11.6 years, sex ratio (M/F) 1.96. Patients were divided into two groups: the HF group [n=157 (41.4%)] and the non-HF group [n=223 (58.6%)]. Mean LUS score was higher in the HF group (26.8±8.4 vs 15.3±7.1; p<0.001). The mean LUS score in the HF patients with reduced LVEF was 29.2±8.7, and was 24.5±7.6 in the HF patients with preserved LVEF. LUS score area under ROC curve for the diagnosis of HF was 0.71 [0.65-0.76]. The best sensitivity (89% [85.9-92,1]) was observed at the threshold of 5; the best specificity (85% [81.4-88.6]) was observed at the threshold of 30. Correlation between LUS score and E/E' ratio was good (R=0.46, p=0.0001). Conclusion: Our results suggest that LUS score could be helpful and should be considered in the diagnostic approach of HF in AECOPD patients, at least as a ruling in test.
Subject(s)
Disease Progression , Heart Failure , Lung , Predictive Value of Tests , Pulmonary Disease, Chronic Obstructive , Ultrasonography , Humans , Pulmonary Disease, Chronic Obstructive/diagnostic imaging , Pulmonary Disease, Chronic Obstructive/physiopathology , Pulmonary Disease, Chronic Obstructive/complications , Male , Female , Heart Failure/diagnostic imaging , Heart Failure/physiopathology , Aged , Prospective Studies , Cross-Sectional Studies , Lung/diagnostic imaging , Lung/physiopathology , Middle Aged , Area Under Curve , ROC Curve , Aged, 80 and over , Natriuretic Peptide, Brain/blood , Reproducibility of Results , Prognosis , Peptide FragmentsABSTRACT
Purpose: We aimed to preliminarily compare the glaucoma detection accuracy of a head-mounted binocular visual perimeter "imo" screening program (ISP) with that of frequency doubling technology (FDT). Methods: This multicenter, diagnostic accuracy study based on prospectively collected data included 76 non-glaucoma (including pre-perimetric glaucoma) eyes and 92 glaucomatous eyes from patients visiting two hospitals. Patients underwent ISP and FDT (C-20-1 screening program) on the same day. Diagnostic efficacy was evaluated using receiver operating characteristic curves and areas under the curve (AUCs). In addition, we compared the ISP and FDT testing times. Results: AUC values for ISP versus FDT were as follows: (1) mild-stage glaucoma (mean deviation [MD] > -6 dB), 0.82 (95% confidence interval [CI], 0.75-0.88) versus 0.76 (95% CI, 0.68-0.83); moderate-stage glaucoma (-6 dB ≥ MD ≥ -12 dB), 0.98 (95% CI, 0.95-1.00) versus 0.96 (95% CI, 0.93-1.00); and advanced-stage glaucoma (-12 dB > MD), 1.00 (95% CI, 1.00-1.00) versus 0.99 (95% CI, 0.98-1.00). In addition, mild-stage glaucoma was classified into two stages (MD > -3 D) and (-3 D ≥ MD > -6 D). AUC values were 0.81 (95% CI, 0.73-0.88) versus 0.76 (95% CI, 0.68-0.84) for MD > -3 D and 0.86 (95% CI, 0.77-0.94) versus 0.73 (95% CI, 0.61-0.86) for -3 D ≥ MD > -6 D. The testing time for the ISP was significantly shorter than that of FDT for all glaucoma stages (P < 0.001). Conclusions: The ISP demonstrates non-inferiority in detecting glaucoma and has a shorter testing time compared with FDT. These findings provide evidence for applied further studies on large-scale population-based glaucoma screening. Translational Relevance: Our study provides a non-inferior and quicker glaucoma screening than existing tools.
Subject(s)
Glaucoma , Visual Field Tests , Humans , Female , Male , Visual Field Tests/methods , Visual Field Tests/instrumentation , Middle Aged , Glaucoma/diagnosis , Aged , Prospective Studies , ROC Curve , Visual Fields/physiology , Area Under Curve , Vision, Binocular/physiology , Adult , Intraocular Pressure/physiologyABSTRACT
Diagnosing liver lesions is crucial for treatment choices and patient outcomes. This study develops an automatic diagnosis system for liver lesions using multiphase enhanced computed tomography (CT). A total of 4039 patients from six data centers are enrolled to develop Liver Lesion Network (LiLNet). LiLNet identifies focal liver lesions, including hepatocellular carcinoma (HCC), intrahepatic cholangiocarcinoma (ICC), metastatic tumors (MET), focal nodular hyperplasia (FNH), hemangioma (HEM), and cysts (CYST). Validated in four external centers and clinically verified in two hospitals, LiLNet achieves an accuracy (ACC) of 94.7% and an area under the curve (AUC) of 97.2% for benign and malignant tumors. For HCC, ICC, and MET, the ACC is 88.7% with an AUC of 95.6%. For FNH, HEM, and CYST, the ACC is 88.6% with an AUC of 95.9%. LiLNet can aid in clinical diagnosis, especially in regions with a shortage of radiologists.
Subject(s)
Carcinoma, Hepatocellular , Cholangiocarcinoma , Deep Learning , Hemangioma , Liver Neoplasms , Tomography, X-Ray Computed , Humans , Liver Neoplasms/diagnostic imaging , Carcinoma, Hepatocellular/diagnostic imaging , Tomography, X-Ray Computed/methods , Male , Hemangioma/diagnostic imaging , Cholangiocarcinoma/diagnostic imaging , Cholangiocarcinoma/pathology , Female , Liver/diagnostic imaging , Liver/pathology , Middle Aged , Focal Nodular Hyperplasia/diagnostic imaging , Adult , Aged , Area Under Curve , Cysts/diagnostic imagingABSTRACT
BACKGROUND: Acute respiratory distress syndrome (ARDS) after cardiac surgery is a severe respiratory complication with high mortality and morbidity. Traditional clinical approaches may lead to under recognition of this heterogeneous syndrome, potentially resulting in diagnosis delay. This study aims to develop and external validate seven machine learning (ML) models, trained on electronic health records data, for predicting ARDS after cardiac surgery. METHODS: This multicenter, observational cohort study included patients who underwent cardiac surgery in the training and testing cohorts (data from Nanjing First Hospital), as well as those patients who had cardiac surgery in a validation cohort (data from Shanghai General Hospital). The number of important features was determined using the sliding windows sequential forward feature selection method (SWSFS). We developed a set of tree-based ML models, including Decision Tree, GBDT, AdaBoost, XGBoost, LightGBM, Random Forest, and Deep Forest. Model performance was evaluated using the area under the receiver operating characteristic curve (AUC) and Brier score. The SHapley Additive exPlanation (SHAP) techinque was employed to interpret the ML model. Furthermore, a comparison was made between the ML models and traditional scoring systems. ARDS is defined according to the Berlin definition. RESULTS: A total of 1996 patients who had cardiac surgery were included in the study. The top five important features identified by the SWSFS were chronic obstructive pulmonary disease, preoperative albumin, central venous pressure_T4, cardiopulmonary bypass time, and left ventricular ejection fraction. Among the seven ML models, Deep Forest demonstrated the best performance, with an AUC of 0.882 and a Brier score of 0.809 in the validation cohort. Notably, the SHAP values effectively illustrated the contribution of the 13 features attributed to the model output and the individual feature's effect on model prediction. In addition, the ensemble ML models demonstrated better performance than the other six traditional scoring systems. CONCLUSIONS: Our study identified 13 important features and provided multiple ML models to enhance the risk stratification for ARDS after cardiac surgery. Using these predictors and ML models might provide a basis for early diagnostic and preventive strategies in the perioperative management of ARDS patients.
Subject(s)
Cardiac Surgical Procedures , Machine Learning , Respiratory Distress Syndrome , Humans , Respiratory Distress Syndrome/etiology , Male , Female , Middle Aged , Cohort Studies , Cardiac Surgical Procedures/adverse effects , Aged , ROC Curve , Area Under CurveABSTRACT
RATIONALE: Area under expiratory flow-volume curve (AEX) has been shown to be a valuable functional measurement in respiratory physiology. Area under inspiratory flow-volume loop (AIN) also shows promise in characterizing upper and/or lower airflow obstruction. OBJECTIVES: we aimed here to develop normative reference values for AIN, able to ascertain deviations from normal. METHODS: We analyzed AIN in 4,980 spirometry tests recorded in non-smoking, healthy individuals in the Pulmonary Function Testing Laboratory. RESULTS: The mean (95% confidence interval, CI), standard deviation and median (25th-75th interquartile range) AIN were 16.05 (15.79-16.31), 9.08 and 14.72 (9.12-21.42) L2·sec-1, respectively. The mean (95% CI) and standard deviation of the best-trial measurements for square root of AIN (Sqrt AIN) were 3.84 (3.81-3.87) and 1.14; 4.15 (4.12-4.18) and 1.03 in men, and 2.68 (2.63-2.72) and 0.72 L·sec-1/2 in women. The mean (standard deviation) of pre- and post-bronchodilator Sqrt AIN were 3.71 (1.17) and 3.81 (1.19) L·sec-1/2, respectively. The mean (95% CI), standard deviation and lowest 5th percentile (lower limit of normal, LLN) of Sqrt AIN/Sqrt AEX (%) were 101.3 (100.82-101.88), 18.7, and 71.8%; stratified by gender, it was 102.2 (101.6-102.8), 18.6, and 72.8% in men, and 98 (96.9-99.2), 18.8, and 68.6% in women, respectively. CONCLUSIONS: The availability of area under the inspiratory flow-volume curve (AIN) and the derived indices offers a promising opportunity to assess upper airway disease (e.g., involvement of larynx, trachea or major bronchi), especially because some of these measurements appear to be independent of age, race, height, and weight.
Subject(s)
Spirometry , Humans , Male , Female , Adult , Middle Aged , Spirometry/methods , Spirometry/standards , Reference Values , Aged , Young Adult , Respiratory Function Tests/methods , Respiratory Function Tests/standards , Inhalation/physiology , Adolescent , Area Under CurveABSTRACT
BACKGROUND: Daptomycin is a high-use intravenous antimicrobial agent affording the convenience of once-daily dosing. Prior studies suggest an opportunity to use a more operationally convenient fixed rather than weight-based dosing but this approach has not been studied prospectively. METHODS: This study quantified the probability of toxicity and efficacy end points by prospectively testing a fixed dose regimen of daptomycin (750 mg) in obese and non-obese adults. At least, three daptomycin concentrations were measured at steady-state for each patient. A population pharmacokinetic model was constructed to evaluate concentration-time profiles and investigate covariates of daptomycin clearance. Simulations were performed to evaluate the probability of achieving efficacy (24-h area under the curve (AUC0-24) ≥ 666 mgâh/L) and toxicity (minimum concentration (C min) ≥24.3 mg/L) targets for fixed (500-1000 mg) and weight-based (6-12 mg/kg) daptomycin doses. RESULTS: Thirty-one patients (16 females, 15 males) with median (interquartile range (IQR)) age of 50 (30, 62) years and weight of 74 (54, 156) kg were included in the final analysis. Fixed dose daptomycin (750 mg) resulted in similar exposure across weights with a median (IQR) AUC0-24 of 819 (499, 1501) mgâh/L and 749 (606, 1265) mgâh/L in patients weighing ≤74 kg and >74 kg, respectively. Overall, male sex and increased kidney function necessitate higher fixed and weight-based doses to achieve efficacy. Creatine phosphokinase elevation was observed in two patients (6.5%) and predicted to be lower with fixed versus weight-based regimens. CONCLUSIONS: Fixed daptomycin dosing adjusted for sex and kidney function is expected to improve the efficacy-to-toxicity ratio, transitions of care, and costs compared to weight-based doses. However, no empiric dosing approach is predicted to achieve ≥90% efficacy while minimizing the risk of toxicity, so therapeutic drug monitoring should be considered on a patient-specific basis.
Subject(s)
Anti-Bacterial Agents , Daptomycin , Staphylococcal Infections , Daptomycin/pharmacokinetics , Daptomycin/administration & dosage , Daptomycin/pharmacology , Humans , Male , Female , Anti-Bacterial Agents/pharmacokinetics , Anti-Bacterial Agents/administration & dosage , Anti-Bacterial Agents/pharmacology , Middle Aged , Adult , Staphylococcal Infections/drug therapy , Prospective Studies , Area Under Curve , Dose-Response Relationship, Drug , Models, Biological , Obesity/drug therapy , Staphylococcus aureus/drug effects , Body Weight , Microbial Sensitivity TestsABSTRACT
BACKGROUND AND OBJECTIVE: Venlafaxine hydrochloride extended-release (ER) capsules are commonly used to treat depression and anxiety disorders. Evaluation of the bioequivalence of generic formulations with reference products is essential to ensure therapeutic equivalence. The objective of this study was to evaluate the bioequivalence, safety, and tolerability of Chinese-manufactured venlafaxine hydrochloride extended-release capsules compared with USA-manufactured EFFEXOR® XR in healthy Chinese volunteers under fed conditions. METHODS: A randomized, open-label, single-dose, crossover study was conducted. Subjects were randomly assigned to receive the test formulation (one 150-mg ER capsule manufactured in China) or the reference formulation (one 150-mg ER capsule manufactured in the USA). The bioequivalence of the two drugs was assessed using the area under the plasma concentration-time curve from time zero to the last sampling time (AUC0-t) and the maximum observed concentration (Cmax). RESULTS: A total of 28 subjects were enrolled and randomly assigned to receive a single dose of either the test or reference capsule. All the subjects completed the study and were included in the pharmacokinetic (PK) and safety analyses. The mean AUC0-t and Cmax of venlafaxine and its active metabolite O-desmethylvenlafaxine were comparable between the test and reference products with both parameters close to 100% and the corresponding 90% confidence intervals within the specified 80-125% bioequivalence boundary. Safety was also assessed between the two products and all adverse events (AEs) in this study were mild in severity. CONCLUSIONS: Both the test and reference venlafaxine hydrochloride ER capsules were bioequivalent and showed a similar safety and tolerability profile in the population studied. CLINICAL TRIALS REGISTRATION: This study was registered at the Drug Clinical Trial Registration and Information Publicity Platform ( http://www.chinadrugtrials.org.cn/index.html ) with registration number CTR20211243, date: June 1, 2021.
Subject(s)
Capsules , Cross-Over Studies , Delayed-Action Preparations , Healthy Volunteers , Therapeutic Equivalency , Venlafaxine Hydrochloride , Humans , Venlafaxine Hydrochloride/pharmacokinetics , Venlafaxine Hydrochloride/administration & dosage , Venlafaxine Hydrochloride/adverse effects , Male , Adult , Delayed-Action Preparations/pharmacokinetics , Female , Young Adult , Area Under Curve , Drugs, Generic/pharmacokinetics , Drugs, Generic/administration & dosage , Drugs, Generic/adverse effects , Asian People , China , Middle Aged , East Asian PeopleABSTRACT
BACKGROUND: Considering the rise of new SARS-CoV-2 variants that have reduced the efficacy of COVID-19 vaccines, the development of new antiviral medications for the disease has become increasingly necessary. In this study, ASC10, a novel antiviral prodrug, was studied in a phase 1 trial in healthy Chinese participants. RESEARCH DESIGN AND METHODS: Part 1 involved 60 participants, receiving 50-800 mg ASC10 or placebo twice daily for 5.5 days. Part 2, with 12 participants, explored ASC10 dosing in the fed/fasting states. RESULTS: ASC10-A, the main pharmacologically active metabolite, rapidly appeared in plasma (Tmax: 1.00-2.00 h) and decreased (t1/2: 1.10-3.04 h) without accumulation. The Cmax and area under the plasma concentration - time curve (AUC) of ASC10-A increased dose-dependently (50-800 mg BID) over 5.5 days, with no accumulation. The Tmax was slightly delayed in the fed state; however, the Cmax and AUC were similar between the fed and fasting states. Adverse events (AEs) were comparable (ASC10/placebo, 66.7%) and mostly mild (95%). CONCLUSION: ASC10 was demonstrated to be safe and well tolerated and exhibited dose-proportional exposure and minimal food effects. CLINICAL TRIAL REGISTRATION: www.clinicaltrials.gov identifier is NCT05523141.
Subject(s)
Antiviral Agents , Prodrugs , Adult , Female , Humans , Male , Middle Aged , Young Adult , Administration, Oral , Antiviral Agents/administration & dosage , Antiviral Agents/adverse effects , Antiviral Agents/pharmacokinetics , Area Under Curve , Asian People , China , COVID-19 , COVID-19 Drug Treatment , Cytidine/administration & dosage , Cytidine/adverse effects , Cytidine/analogs & derivatives , Cytidine/pharmacokinetics , Dose-Response Relationship, Drug , Double-Blind Method , Healthy Volunteers , Prodrugs/adverse effects , Prodrugs/administration & dosageABSTRACT
GST-HG171 is a potent, broad-spectrum, orally bioavailable small-molecule 3C-like (3CL) protease inhibitor that was recently approved for treating mild to moderate coronavirus disease 2019 patients in China. Since cytochrome P450 (CYP) enzymes, primarily CYP3A, are the main metabolic enzymes of GST-HG171, hepatic impairment may affect its pharmacokinetic (PK) profile. Aiming to guide clinical dosing for patients with hepatic impairment, this study, using a non-randomized, open-label, single-dose design, assessed the impact of hepatic impairment on the PK, safety, and tolerability of GST-HG171. Patients with mild and moderate hepatic impairment along with healthy subjects were enrolled (n = 8 each), receiving a single oral dose of 150 mg GST-HG171, with concurrent administration of 100 mg ritonavir to sustain CYP3A inhibition before and after GST-HG171 administration (-12, 0, 12, and 24 hours). Compared to subjects with normal hepatic function, the geometric least-squares mean ratios (90% confidence intervals) for GST-HG171's maximum plasma concentration (Cmax), area under the concentration-time curve up to the last quantifiable time (AUC0-t), and area under the plasma concentration-time curve from time 0 extrapolated to infinity (AUC0-∞) in subjects with mild hepatic impairment were 1.14 (0.99, 1.31), 1.07 (0.88, 1.30), and 1.07 (0.88, 1.29), respectively. For moderate hepatic impairment, the ratios were 0.87 (0.70, 1.07), 0.82 (0.61, 1.10), and 0.82 (0.61, 1.10), respectively. Hepatic impairment did not significantly alter GST-HG171's peak time (Tmax) and elimination half-life (T1/2). GST-HG171 exhibited good safety and tolerability in the study. Taken together, mild to moderate hepatic impairment minimally impacted GST-HG171 exposure, suggesting no need to adjust GST-HG171 dosage for patients with mild to moderate hepatic impairment in the clinic.Clinical TrialsRegistered at ClinicalTrials.gov (NCT06106113).
Subject(s)
Cytochrome P-450 CYP3A Inhibitors , Liver , Protease Inhibitors , Adult , Aged , Female , Humans , Male , Middle Aged , Area Under Curve , China , COVID-19 Drug Treatment , Cytochrome P-450 CYP3A/metabolism , Cytochrome P-450 CYP3A Inhibitors/adverse effects , Cytochrome P-450 CYP3A Inhibitors/pharmacokinetics , East Asian People , Liver/drug effects , Liver Diseases , Protease Inhibitors/adverse effects , Protease Inhibitors/pharmacokinetics , Ritonavir/adverse effects , Ritonavir/pharmacokineticsABSTRACT
Ventricular septal rupture (VSR) is a mechanical complication of acute myocardial infarction (AMI), and its mortality has not decreased significantly in recent decades. However, no clinical model has been developed to predict short-term mortality in patients with post-infarction VSR (PIVSR). This study aimed to develop a nomogram to predict the 30-day mortality by using the clinical characteristics of hospitalized patients with PIVSR. The least absolute shrinkage and selection operator (LASSO) and multivariate logistic regression analysis was used to construct a nomogram by R. The model was evaluated by the area under the curve (AUC), calibration curve and decision curve analysis (DCA). The bootstrap method was used to validate the model internally. As a result, a nomogram was constructed by using six variables, including CRRT, mechanical ventilation, PPCI, WBC, PASP and methods of treatment. The AUC of the prediction model was 0.96 (0.93, 0.98). The prediction model was well calibrated. The DCA showed that if the threshold probability was between 15% and 95%, the nomogram model would provide a net benefit. The well-constructed and evaluated nomogram can be beneficial to clinicians to predict the risk of death within 30 days in patients with PIVSR.
Subject(s)
Myocardial Infarction , Nomograms , Ventricular Septal Rupture , Humans , Ventricular Septal Rupture/etiology , Ventricular Septal Rupture/mortality , Male , Female , Myocardial Infarction/mortality , Myocardial Infarction/complications , Middle Aged , Aged , Area Under Curve , Prognosis , Risk FactorsABSTRACT
Ramipril is an angiotensin-converting enzyme inhibitor used for hypertension and heart failure management. To date, scarce literature is available on pharmacogenetic associations affecting ramipril. The goal of this study was to investigate the effect of 120 genetic variants in 34 pharmacogenes (i.e., genes encoding for enzymes like CYPs or UGTs and transporters like ABC or SLC) on ramipril pharmacokinetic variability and adverse drug reaction (ADR) incidence. Twenty-nine healthy volunteers who had participated in a single-dose bioequivalence clinical trial of two formulations of ramipril were recruited. A univariate and multivariate analysis searching for associations between genetic variants and ramipril pharmacokinetics was performed. SLCO1B1 and ABCG2 genotype-informed phenotypes strongly predicted ramipril exposure. Volunteers with the SLCO1B1 decreased function (DF) phenotype presented around 1.7-fold higher dose/weight-corrected area under the curve (AUC/DW) than volunteers with the normal function (NF) phenotype (univariate p-value [puv] < 0.001, multivariate p-value [pmv] < 0.001, ß = 0.533, R2 = 0.648). Similarly, volunteers with ABCG2 DF + poor function (PF) phenotypes presented around 1.6-fold higher AUC/DW than those with the NF phenotype (puv = 0.011, pmv < 0.001, ß = 0.259, R2 = 0.648). Our results suggest that SLCO1B1 and ABCG2 are important transporters to ramipril pharmacokinetics, and their genetic variation strongly alters its pharmacokinetics. Further studies are required to confirm these associations and their clinical relevance.