Search

Objectives To evaluate the ability of an artificial intelligence (AI) model to predict the risk of cancer in patients referred from primary care based on routine blood tests. Results obtained with the AI model are compared to results based on logistic regression (LR). Methods An analytical profile consisting of 25 predefined routine laboratory blood tests was introduced to general practitioners (GPs) to be used for patients with non-specific symptoms, as an additional tool to identify individuals at increased risk of cancer. Consecutive analytical profiles ordered by GPs from November 29th 2011 until March 1st 2020 were included. AI and LR analysis were performed on data from 6,592 analytical profiles for their ability to detect cancer. Cohort I for model development included 5,224 analytical profiles ordered by GP’s from November 29th 2011 until the December 31st 2018, while 1,368 analytical profiles included from January 1st 2019 until March 1st 2020 constituted the “out of time” validation test Cohort II. The main outcome measure was a cancer diagnosis within 90 days. Results The AI model based on routine laboratory blood tests can provide an easy-to use risk score to predict cancer within 90 days. Results obtained with the AI model were comparable to results from the LR model. In the internal validation Cohort IB, the AI model provided slightly better results than the LR analysis both in terms of the area under the receiver operating characteristics curve (AUC) and PPV, sensitivity/specificity while in the “out of time” validation test Cohort II, the obtained results were comparable. Conclusions The AI risk score may be a valuable tool in the clinical decision-making. The score should be further validated to determine its applicability in other populations.

Objective To evaluate heterogeneity of treatment effects (HTE) in a trial of exercise-based interventions for knee osteoarthritis (OA). Methods Participants (n=350) were randomized to standard physical therapy (PT; n=140), Internet-Based Exercise Training (IBET; n=142), or wait list control (WL; n=68). We applied QUalitative INteraction Trees (QUINT), a sequential partitioning method, and Generalized Unbiased Interaction Detection and Estimation (GUIDE), a regression tree approach, to identify subgroups with greater improvements in Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) score over 4-months. Predictors included 24 demographic, clinical and psychosocial characteristics. We conducted internal validation to estimate optimism (bias) in the range of mean outcome differences among arms. Results Both QUINT and GUIDE indicated that for participants with lower body mass index (BMI), IBET was better than PT (improvements of WOMAC ranged from 6.3 to 9.1 points lower) and for those with higher BMI and longer duration of knee OA, PT was better than IBET (WOMAC improvement was 6.3 points). In GUIDE analyses comparing PT or IBET to WL, participants not employed had improvements in WOMAC ranging from 1.8 to 6.8 points lower with PT or IBT vs. WL. From internal validation, there were large corrections to the mean outcome differences among arms; however, after correction some differences remained in the clinically meaningful range. Conclusion Results suggest there may be subgroups who experience greater improvement in symptoms from PT or IBET, and this could guide referrals and future trials. However, uncertainty persists for specific treatment effect size estimates and how they apply beyond this study sample.

Purpose To develop and evaluate the performance of a 3-dimensional (3D) deep-learning–based automated digital gonioscopy system (DGS) in detecting 2 major characteristics in eyes with suspected primary angle-closure glaucoma (PACG): (1) narrow iridocorneal angles (static gonioscopy, Task I) and (2) peripheral anterior synechiae (PAS) (dynamic gonioscopy, Task II) on OCT scans. Design International, cross-sectional, multicenter study. Participants A total of 1.112 million images of 8694 volume scans (2294 patients) from 3 centers were included in this study (Task I, training/internal validation/external testing: 4515, 1101, and 2222 volume scans, respectively; Task II, training/internal validation/external testing: 378, 376, and 102 volume scans, respectively). Methods For Task I, a narrow angle was defined as an eye in which the posterior pigmented trabecular meshwork was not visible in more than 180° without indentation in the primary position captured in the dark room from the scans. For Task II, PAS was defined as the adhesion of the iris to the trabecular meshwork. The diagnostic performance of the 3D DGS was evaluated in both tasks with gonioscopic records as reference. Main Outcome Measures The area under the curve (AUC), sensitivity, and specificity of the 3D DGS were calculated. Results In Task I, 29.4% of patients had a narrow angle. The AUC, sensitivity, and specificity of 3D DGS on the external testing datasets were 0.943 (0.933–0.953), 0.867 (0.838–0.895), and 0.878 (0.859–0.896), respectively. For Task II, 13.8% of patients had PAS. The AUC, sensitivity, and specificity of 3D DGS were 0.902 (0.818–0.985), 0.900 (0.714–1.000), and 0.890 (0.841–0.938), respectively, on the external testing set at quadrant level following normal clinical practice; and 0.885 (0.836–0.933), 0.912 (0.816–1.000), and 0.700 (0.660–0.741), respectively, on the external testing set at clock-hour level. Conclusions The 3D DGS is effective in detecting eyes with suspected PACG. It has the potential to be used widely in the primary eye care community for screening of subjects at high risk of developing PACG.

Objective To develop an artificial intelligence (AI)–based tool to detect cardiac amyloidosis (CA) from a standard 12-lead electrocardiogram (ECG). Methods We collected 12-lead ECG data from 2541 patients with light chain or transthyretin CA seen at Mayo Clinic between 2000 and 2019. Cases were nearest neighbor matched for age and sex, with 2454 controls. A subset of 2997 (60%) cases and controls were used to train a deep neural network to predict the presence of CA with an internal validation set (n=999; 20%) and a randomly selected holdout testing set (n=999; 20%). We performed experiments using single-lead and 6-lead ECG subsets. Results The area under the receiver operating characteristic curve (AUC) was 0.91 (CI, 0.90 to 0.93), with a positive predictive value for detecting either type of CA of 0.86. By use of a cutoff probability of 0.485 determined by the Youden index, 426 (84%) of the holdout patients with CA were detected by the model. Of the patients with CA and prediagnosis electrocardiographic studies, the AI model successfully predicted the presence of CA more than 6 months before the clinical diagnosis in 59%. The best single-lead model was V5 with an AUC of 0.86 and a precision of 0.78, with other single leads performing similarly. The 6-lead (bipolar leads) model had an AUC of 0.90 and a precision of 0.85. Conclusion An AI-driven ECG model effectively detects CA and may promote early diagnosis of this life-threatening disease.

BACKGROUND Pseudomyxoma peritonei (PMP) is a rare disease, with the rate of overall survival (OS) influenced by many factors. The present study aimed to define independent predictors and establish a nomogram for individual risk prediction in PMP patients. METHODS One hundred forty-seven PMP patients were consecutively included between June 1, 2013, and November 22, 2019. The log-rank test was used to compare the OS rate between groups; subsequently, variables with p

The completeness of cytoreduction is one of the most important prognostic factors for patients with pseudomyxoma peritonei (PMP). To date, no nomograms have been established to predict incomplete cytoreduction (IC) for patients with PMP. The current study therefore proposed a nomogram to predict individual IC risk for PMP patients. Between 1 June 2013, and 22 November 2019, 144 consecutive PMP patients who underwent cytoreductive surgery (CRS) combined with hyperthermic intraperitoneal chemotherapy (HIPEC) for the first time in our center were included in a retrospective study. Possible predictors of cytoreducibility were analyzed using logistic regression modeling to predict IC for PMP patients. A nomogram was developed based on the multivariate analysis and further investigated for internal validation. After CRS, the 144 participants were divided into complete CRS (CCRS) (n = 46) and IC (n = 98) subgroups. Four independent predictors (sex, disease duration, anemia, and carbohydrate antigen 19-9 (CA 199)) were included in the prediction model. Then, a nomogram predicting IC was established based on the aforementioned variables, which demonstrated good predictive accuracy (C-index, 0.837; 95 % confidence interval [CI], 0.764–0.894). The predicted probability was close to the actual observed outcome according to the calibration plot. The current work led to the development of a nomogram capable of predicting IC for PMP patients who demonstrated good performance. Risk stratification by the established nomogram had ability to optimize individual IC prediction and help physicians to establish meticulous preoperative plans.

Background Cystoscopy plays an important role in bladder cancer (BCa) diagnosis and treatment, but its sensitivity needs improvement. Artificial intelligence has shown promise in endoscopy, but few cystoscopic applications have been reported. We report a Cystoscopy Artificial Intelligence Diagnostic System (CAIDS) for BCa diagnosis. Methods In total, 69 204 images from 10 729 consecutive patients from 6 hospitals were collected and divided into training, internal validation, and external validation sets. The CAIDS was built using a pyramid scene parsing network and transfer learning. A subset (n = 260) of the validation sets was used for a performance comparison between the CAIDS and urologists for complex lesion detection. The diagnostic accuracy, sensitivity, specificity, and positive and negative predictive values and 95% confidence intervals (CIs) were calculated using the Clopper-Pearson method. Results The diagnostic accuracies of the CAIDS were 0.977 (95% CI = 0.974 to 0.979) in the internal validation set and 0.990 (95% CI = 0.979 to 0.996), 0.982 (95% CI = 0.974 to 0.988), 0.978 (95% CI = 0.959 to 0.989), and 0.991 (95% CI = 0.987 to 0.994) in different external validation sets. In the CAIDS vs urologists’ comparisons, the CAIDS showed high accuracy and sensitivity (accuracy = 0.939, 95% CI = 0.902 to 0.964; sensitivity = 0.954, 95% CI = 0.902 to 0.983) with a short latency of 12 seconds, much more accurate and quicker than the expert urologists. Conclusions The CAIDS achieved accurate BCa detection with a short latency. The CAIDS may provide many clinical benefits, from increasing the diagnostic accuracy for BCa, even for commonly misdiagnosed cases such as flat cancerous tissue (carcinoma in situ), to reducing the operation time for cystoscopy.

Xiaoqiang Liu,1 Shaoya He,2 Xi Yao,1 Tianyang Hu3 1Department of Orthopedic Surgery, Anyue County People’s Hospital, Sichuan, People’s Republic of China; 2Department of Gastroenterology, Anyue County People’s Hospital, Sichuan, People’s Republic of China; 3Department of Cardiology, The Second Affiliated Hospital of Chongqing Medical University, Chongqing, People’s Republic of ChinaCorrespondence: Tianyang HuDepartment of Cardiology, The Second Affiliated Hospital of Chongqing Medical University, 74 Linjiang Road, Yuzhong District, Chongqing, 400010, People’s Republic of ChinaTel +8623-63693222Email hutianyang@stu.cqmu.edu.cnBackground: The aim of the current study was to construct prognostic nomograms for individual risk prediction in elderly patients with osteosarcoma.Methods: Data for 816 elderly patients (≥ 40 years old) with osteosarcoma between 2004 and 2016 from the Surveillance, Epidemiology, and End Results (SEER) database were randomly assigned to training (N=573) and internal validation (N=243) sets. The essential clinical predictors were identified based on least absolute shrinkage and selection operator (Lasso) Cox regression. Nomograms were constructed to predict the 1-, 3-, and 5-year cancer-specific survival (CSS) and overall survival (OS).Results: Our LASSO regression analyses of the training set yielded five clinicopathological features (age, chemotherapy, surgery, AJCC stage, and summary stage) in the training cohort for the prognosis of elderly patients with osteosarcoma, while grade was only associated with OS and M stage was only associated with CSS. Construction of nomograms based on these predictors was performed to evaluate the prognosis of elderly patients with osteosarcoma. The C-index, calibration and decision curve analysis also showed the satisfactory performance of these nomograms for prognosis prediction.Conclusion: The constructed nomograms are helpful tools for exactly predicting the prognosis of elderly patients with osteosarcoma, which could enable patients to be more accurately managed in clinical practice.Keywords: osteosarcoma, elderly, SEER, nomograms, prognosis, risk management

Diabetes mellitus (DM) is considered as ”the mother of all diseases" because it causes many complications. Knowing and measuring medication adherence may have a greater effect on DM patients. Several studies in Indonesia used a questionnaire scale to measure adherence, however they do not validate the study population, so it could still be found anomalous correlation analysis between adherence and clinical data even though it measured in the same country and scale. This study measure the adherence level of type 2 diabetes patients, evaluates the validity of the medication adherence scale, and analyze the correlation with the clinical outcome of type 2 diabetes patients in four health centers in Banyumas district. The study uses a cross-sectional design in Prolanis type 2 DM patients of January -April 2020. The adherence is measured by MARS-10, backward-forward translation method followed by content and internal validation. Clinical outcome is evaluated based on fasting blood glucose measurement. The results of the MARS-10 Gregory index analysis showed content validity in the high category (IG ≥ 0.8). The content validity showed the results of 9 questions with the value of r count> r table (n = 30, r table = 0.361). Reliability analysis showed Cronbach's Alpha 0.747> 0.6. The measurement showed 80.3% was adherent patients and 19.3% was non-adherent patients. Correlation analysis showed that there was no significant relationship (p> 0.05) between patient adherence and clinical outcome. Those results showed that type 2 diabetes mellitus patients in 4 health centers were categorized as adherent but not correlated with the clinical outcome. This was enable due to the clinical outcome was simultaneously influenced by several factors: general factors, individual factors, and unpredictable factors.

Item does not contain fulltext OBJECTIVE: The aim of this study was to develop and validate a clinical prediction model to predict overall survival in patients with nonmetastatic, resected gallbladder cancer (GBC). BACKGROUND: Although several tools are available, no optimal method has been identified to assess survival in patients with resected GBC. METHODS: Data from a Dutch, nation-wide cohort of patients with resected GBC was used to develop a prediction model for overall survival. The model was internally validated and a cohort of Australian GBC patients who underwent resection was used for external validation. The performance of the American Joint Committee on Cancer (AJCC) staging system and the present model were compared. RESULTS: In total, 446 patients were included; 380 patients in the development cohort and 66 patients in the validation cohort. In the development cohort median survival was 22 months (median follow-up 75 months). Age, T/N classification, resection margin, differentiation grade, and vascular invasion were independent predictors of survival. The externally validated C-index was 0.75 (95%CI: 0.69-0.80), implying good discriminatory capacity. The discriminative ability of the present model after internal validation was superior to the ability of the AJCC staging system (Harrell C-index 0.71, [95%CI: 0.69-0.72) vs. 0.59 (95% CI: 0.57-0.60)]. CONCLUSION: The proposed model for the prediction of overall survival in patients with resected GBC demonstrates good discriminatory capacity, reasonable calibration and outperforms the authoritative AJCC staging system. This model can be a useful tool for physicians and patients to obtain information about survival after resection and is available from https:// gallbladderresearch.shinyapps.io/Predict_GBC_survival/.

Lichang Chen,1,&ast; Lingling Wu,1,&ast; Dongzhu Lu,1 Mei Zi,2 Huapeng Yu1 1Department of Respiratory Medicine, Zhujiang Hospital, Southern Medical University, Guangzhou, People’s Republic of China; 2Department of Respiratory and Critical Care Medicine, The Third People’s Hospital of Shenzhen, Shenzhen, People’s Republic of China&ast;These authors contributed equally to this workCorrespondence: Huapeng YuDepartment of Respiratory Medicine, Zhujiang Hospital, Southern Medical University, Guangzhou, 510220, People’s Republic of ChinaTel +86 020 61643175Fax +86 020 84306143Email huapengyu@aliyun.comPurpose: To evaluate the contribution of fractional exhaled nitric oxide (FeNO) and impulse oscillometry (IOS) and spirometric parameters in predicting bronchial hyperresponsiveness (BHR) in adults with chronic cough.Patients and Methods: In total, 112 patients with chronic cough were enrolled in this prospective diagnostic study. Receiver operating characteristic (ROC) curves were generated to assess the diagnostic efficiency and optimal cut-off values of FeNO and IOS and spirometric parameters in predicting BHR. Optimal combinations of FeNO and IOS and spirometric parameters for BHR prediction were investigated using univariate and multivariate logistic regression models. Bootstrapping was employed for internal validation. Model discrimination and calibration were assessed using indices and calibration plots.Results: Rhinitis and values of FeNO, IOS parameters (resonant frequency (Fres), reactance at 5 Hz (X5), and integrated area of low-frequency X (AX)) and spirometric parameters (FEV1, PEF, MEF75, MEF50, MEF25, MMEF) were significantly different between patients with BHR and those without BHR (P < 0.05). After adjusting for rhinitis, logistic analyses showed that FeNO combined with Fres, FeNO combined with MMEF, or the combination of FeNO, Fres and MMEF had high predictive value in diagnosing BHR; the areas under the ROC curves (AUCs) of the corresponding three models were 0.914, 0.919 and 0.927, respectively. In addition, the three models displayed good discrimination, with high C-index values and good calibration.Conclusion: FeNO combined with Fres or MMEF or a combination of these three parameters may be conveniently used as indicators in BHR prediction.Keywords: maximum mid-expiratory flow, resonant frequency, cough variant asthma, ROC curve, diagnosis

PURPOSE The potential to compute volume metrics of emphysema from planar scout images was investigated in this study. The successful implementation of this concept will have a wide impact in different fields, and specifically, maximize the diagnostic potential of the planar medical images. METHODS We investigated our premise using a well-characterized chronic obstructive pulmonary disease (COPD) cohort. In this cohort, planar scout images from computed tomography (CT) scans were used to compute lung volume and percentage of emphysema. Lung volume and percentage of emphysema were quantified on the volumetric CT images and used as the "ground truth" for developing the models to compute the variables from the corresponding scout images. We trained two classical convolutional neural networks (CNNs), including VGG19 and InceptionV3, to compute lung volume and the percentage of emphysema from the scout images. The scout images (n = 1,446) were split into three subgroups: (1) training (n = 1,235), (2) internal validation (n = 99), and (3) independent test (n = 112) at the subject level in a ratio of 8:1:1. The mean absolute difference (MAD) and R-square (R2) were the performance metrics to evaluate the prediction performance of the developed models. RESULTS The lung volumes and percentages of emphysema computed from a single planar scout image were significantly linear correlated with the measures quantified using volumetric CT images (VGG19: R2 = 0.934 for lung volume and R2 = 0.751 for emphysema percentage, and InceptionV3: R2 = 0.977 for lung volume and R2 = 0.775 for emphysema percentage). The mean absolute differences (MADs) for lung volume and percentage of emphysema were 0.302 ± 0.247L and 2.89 ± 2.58%, respectively, for VGG19, and 0.366 ± 0.287L and 3.19 ± 2.14, respectively, for InceptionV3. CONCLUSIONS Our promising results demonstrated the feasibility of inferring volume metrics from planar images using CNNs.

BACKGROUND: Comorbidities influence the outcomes of injured patients, yet a lack of consensus exists regarding how to quantify that association. This study details the development and internal validation of a trauma comorbidity index (TCI) designed for use with trauma registry data and compares its performance to other existing measures to estimate the association between comorbidities and mortality. METHODS: Indiana state trauma registry data (2013–2015) were used to compare the TCI with the Charlson and Elixhauser comorbidity indices, a count of comorbidities, and comorbidities as separate variables. The TCI approach utilized a randomly selected training cohort and was internally validated in a distinct testing cohort. The C-statistic of the adjusted models was tested using each comorbidity measure in the testing cohort to assess model discrimination. C-statistics were compared using a Wald test, and stratified analyses were performed based on predicted risk of mortality. Multiple imputation was used to address missing data. RESULTS: The study included 84,903 patients (50% each in training and testing cohorts). The Indiana TCI model demonstrated no significant difference between testing and training cohorts (p = 0.33). It produced a C-statistic of 0.924 in the testing cohort, which was significantly greater than that of models using the other indices (p < 0.05). The C-statistics of models using the Indiana TCI and the inclusion of comorbidities as separate variables—the method used by the American College of Surgeons Trauma Quality Improvement Program—were comparable (p = 0.11) but use of the TCI approach reduced the number of comorbidity-related variables in the mortality model from 19 to one. CONCLUSIONS: When examining trauma mortality, the TCI approach using Indiana state trauma registry data demonstrated superior model discrimination and/or parsimony compared to other measures of comorbidities.

Objectives To evaluate the performance of a novel convolutional neural network (CNN) for the classification of typical perifissural nodules (PFN). Methods Chest CT data from two centers in the UK and The Netherlands (1668 unique nodules, 1260 individuals) were collected. Pulmonary nodules were classified into subtypes, including “typical PFNs” on-site, and were reviewed by a central clinician. The dataset was divided into a training/cross-validation set of 1557 nodules (1103 individuals) and a test set of 196 nodules (158 individuals). For the test set, three radiologically trained readers classified the nodules into three nodule categories: typical PFN, atypical PFN, and non-PFN. The consensus of the three readers was used as reference to evaluate the performance of the PFN-CNN. Typical PFNs were considered as positive results, and atypical PFNs and non-PFNs were grouped as negative results. PFN-CNN performance was evaluated using the ROC curve, confusion matrix, and Cohen’s kappa. Results Internal validation yielded a mean AUC of 91.9% (95% CI 90.6–92.9) with 78.7% sensitivity and 90.4% specificity. For the test set, the reader consensus rated 45/196 (23%) of nodules as typical PFN. The classifier-reader agreement (k = 0.62–0.75) was similar to the inter-reader agreement (k = 0.64–0.79). Area under the ROC curve was 95.8% (95% CI 93.3–98.4), with a sensitivity of 95.6% (95% CI 84.9–99.5), and specificity of 88.1% (95% CI 81.8–92.8). Conclusion The PFN-CNN showed excellent performance in classifying typical PFNs. Its agreement with radiologically trained readers is within the range of inter-reader agreement. Thus, the CNN-based system has potential in clinical and screening settings to rule out perifissural nodules and increase reader efficiency. Key Points • Agreement between the PFN-CNN and radiologically trained readers is within the range of inter-reader agreement. • The CNN model for the classification of typical PFNs achieved an AUC of 95.8% (95% CI 93.3–98.4) with 95.6% (95% CI 84.9–99.5) sensitivity and 88.1% (95% CI 81.8–92.8) specificity compared to the consensus of three readers.

An image-based deep learning model was developed to detect different morphological patterns of diabetic macular edema based on optical coherence tomography images with high accuracy and transparency. This model could help ophthalmologists to make personalized therapeutic strategies for patients with diabetic macular edema., Purpose: To develop a deep learning (DL) model to detect morphologic patterns of diabetic macular edema (DME) based on optical coherence tomography (OCT) images. Methods: In the training set, 12,365 OCT images were extracted from a public data set and an ophthalmic center. A total of 656 OCT images were extracted from another ophthalmic center for external validation. The presence or absence of three OCT patterns of DME, including diffused retinal thickening, cystoid macular edema, and serous retinal detachment, was labeled with 1 or 0, respectively. A DL model was trained to detect three OCT patterns of DME. The occlusion test was applied for the visualization of the DL model. Results: Applying 5-fold cross-validation method in internal validation, the area under the receiver operating characteristic curve for the detection of three OCT patterns (i.e., diffused retinal thickening, cystoid macular edema, and serous retinal detachment) was 0.971, 0.974, and 0.994, respectively, with an accuracy of 93.0%, 95.1%, and 98.8%, respectively, a sensitivity of 93.5%, 94.5%, and 96.7%, respectively, and a specificity of 92.3%, 95.6%, and 99.3%, respectively. In external validation, the area under the receiver operating characteristic curve was 0.970, 0.997, and 0.997, respectively, with an accuracy of 90.2%, 95.4%, and 95.9%, respectively, a sensitivity of 80.1%, 93.4%, and 94.9%, respectively, and a specificity of 97.6%, 97.2%, and 96.5%, respectively. The occlusion test showed that the DL model could successfully identify the pathologic regions most critical for detection. Conclusion: Our DL model demonstrated high accuracy and transparency in the detection of OCT patterns of DME. These results emphasized the potential of artificial intelligence in assisting clinical decision-making processes in patients with DME.

Background Nomograms applying the 8th edition of the TNM staging system aimed at predicting overall (OS), disease-specific (DSS), locoregional recurrence-free (LRRFS) and distant recurrence-free survivals (DRFS) for oral tongue squamous cell carcinoma (OTSCC) are still lacking. Methods A training cohort of 438 patients with OTSCC was retrospectively enrolled from a single institution. An external validation set of 287 patients was retrieved from two independent institutions. Results Internal validation of the multivariable models for OS, DSS, DRFS and LRRFS showed a good calibration and discrimination results with optimism-corrected c-indices of 0.74, 0.75, 0.77 and 0.70, respectively. The external validation confirmed the good performance of OS, DSS and DRFS models (c-index 0.73 and 0.77, and 0.73, respectively) and a fair performance of the LRRFS model (c-index 0.58). Conclusions The nomograms herein presented can be implemented as useful tools for prediction of OS, DSS, DRFS and LRRFS in OTSCC.

Background The Chronic Kidney Disease in Children Study (CKiD) equation for children and the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation for adults are recommended serum creatinine (SCr)-based calculations for estimating glomerular filtration rate (GFR). However, these equations, as well as their combination, have limitations, notably the problem of implausible changes in GFR during the transition from adolescence to adulthood and overestimation of GFR in young adults. The full age spectrum (FAS) equation addresses these issues but overestimates GFR when SCr levels are low. Objective To develop and validate a modified FAS SCr-based equation combining design features of the FAS and CKD-EPI equations. Design Cross-sectional analysis with separate pooled data sets for development and validation. Setting Research and clinical studies (n = 13) with measured GFR available. Patients 11 251 participants in 7 studies (development and internal validation data sets) and 8378 participants in 6 studies (external validation data set). Measurements Clearance of an exogenous marker (reference method), SCr level, age, sex, and height were used to develop a new equation to estimate GFR. Results The new European Kidney Function Consortium (EKFC) equation is a FAS equation with low bias (-1.2 mL/min/1.73 m2 [95% CI, -2.7 to 0.0 mL/min/1.73 m2] in children and -0.9 mL/min/1.73 m2 [CI, -1.2 to -0.5 mL/min/1.73 m2] in adults) across the FAS (2 to 90 years) and SCr range (40 to 490 µmol/L [0.45 to 5.54 mg/dL]) and with fewer estimation errors exceeding 30% (6.5% [CI, 3.8% to 9.1%] in children and 3.1% [CI, 2.5% to 3.6%] in adults) compared with the CKiD and CKD-EPI equations. Limitation No Black patients were included. Conclusion The new EKFC equation shows improved accuracy and precision compared with commonly used equations for estimating GFR from SCr levels. Primary funding source Swedish Research Council (Vetenskapsradet).

The notion of personalized medicine demands proper prognostic biomarkers to guide the optimal therapy for an invasive breast cancer patient. However, various risk prediction models based on conventional clinicopathological factors and emergent molecular assays have been frequently limited by either a low strength of prognosis or restricted applicability to specific types of patients. Therefore, there is a critical need to develop a strong and general prognosticator. Methods: We observed five large-scale tumor-associated collagen signatures (TACS4-8) obtained by multiphoton microscopy at the invasion front of the breast primary tumor, which contrasted with the three tumor-associated collagen signatures (TACS1-3) discovered by Keely and coworkers at a smaller scale. Highly concordant TACS1-8 classifications were obtained by three independent observers. Using the ridge regression analysis, we obtained a TACS-score for each patient based on the combined TACS1-8 and established a risk prediction model based on the TACS-score. In a blind fashion, consistent retrospective prognosis was obtained from 995 breast cancer patients in both a training cohort (n= 431) and an internal validation cohort (n = 300) collected from one clinical center, and in an external validation cohort (n = 264) collected from a different clinical center. Results: TACS1-8 model alone competed favorably with all reported models in predicting disease-free survival (AUC: 0.838, [0.800-0.872]; 0.827, [0.779-0.868]; 0.807, [0.754-0.853] in the three cohorts) and stratifying low- and high-risk patients (HR 7.032, [4.869-10.158]; 6.846, [4.370-10.726], 4.423, [2.917-6.708]). The combination of these factors with the TACS-score into a nomogram model further improved the prognosis (AUC: 0.865, [0.829-0.896]; 0.861, [0.816-0.898]; 0.854, [0.805-0.894]; HR 7.882, [5.487-11.323]; 9.176, [5.683-14.816], and 5.548, [3.705-8.307]). The nomogram identified 72 of 357 (~20%) patients with unsuccessful 5-year disease-free survival that might have been undertreated postoperatively. Conclusions: The risk prediction model based on TACS1-8 considerably outperforms the contextual clinical model and may thus convince pathologists to pursue a TACS-based breast cancer prognosis. Our methodology identifies a significant portion of patients susceptible to undertreatment (high-risk patients), in contrast to the multigene assays that often strive to mitigate overtreatment. The compatibility of our methodology with standard histology using traditional (non-tissue-microarray) formalin-fixed paraffin-embedded (FFPE) tissue sections could simplify subsequent clinical translation.

Background: Our research group has previously reported a noninvasive model that estimates phosphate removal within a 4-h hemodialysis (HD) treatment. The aim of this study was to modify the original model and validate the accuracy of the new model of phosphate removal for HD and hemodiafiltration (HDF) treatment. Methods: A total of 109 HD patients from 3 HD centers were enrolled. The actual phosphate removal amount was calculated using the area under the dialysate phosphate concentration time curve. Model modification was executed using second-order multivariable polynomial regression analysis to obtain a new parameter for dialyzer phosphate clearance. Bias, precision, and accuracy were measured in the internal and external validation to determine the performance of the modified model. Results: Mean age of the enrolled patients was 63 ± 12 years, and 67 (61.5%) were male. Phosphate removal was 19.06 ± 8.12 mmol and 17.38 ± 6.75 mmol in 4-h HD and HDF treatments, respectively, with no significant difference. The modified phosphate removal model was expressed as Tpo4 = 80.3 × C45 − 0.024 × age + 0.07 × weight + β × clearance − 8.14 (β = 6.231 × 10−3 × clearance − 1.886 × 10−5 × clearance2 – 0.467), where C45 was the phosphate concentration in the spent dialysate measured at the 45th minute of HD and clearance was the phosphate clearance of the dialyzer. Internal validation indicated that the new model was superior to the original model with a significantly smaller bias and higher accuracy. External validation showed that R2, bias, and accuracy were not significantly different than those of internal validation. Conclusions: A new model was generated to quantify phosphate removal by 4-h HD and HDF with a dialyzer surface area of 1.3–1.8 m2. This modified model would contribute to the evaluation of phosphate balance and individualized therapy of hyperphosphatemia.

ContextThe diagnostic work-up of primary aldosteronism (PA) includes screening and confirmation steps. Case confirmation is time-consuming, expensive, and there is no consensus on tests and thresholds to be used. Diagnostic algorithms to avoid confirmatory testing may be useful for the management of patients with PA.ObjectiveDevelopment and validation of diagnostic models to confirm or exclude PA diagnosis in patients with a positive screening test.Design, Patients, and SettingWe evaluated 1024 patients who underwent confirmatory testing for PA. The diagnostic models were developed in a training cohort (n = 522), and then tested on an internal validation cohort (n = 174) and on an independent external prospective cohort (n = 328).Main Outcome MeasureDifferent diagnostic models and a 16-point score were developed by machine learning and regression analysis to discriminate patients with a confirmed diagnosis of PA.ResultsMale sex, antihypertensive medication, plasma renin activity, aldosterone, potassium levels, and the presence of organ damage were associated with a confirmed diagnosis of PA. Machine learning-based models displayed an accuracy of 72.9%–83.9%. The Primary Aldosteronism Confirmatory Testing (PACT) score correctly classified 84.1% at training and 83.9% or 81.1% at internal and external validation, respectively. A flow chart employing the PACT score to select patients for confirmatory testing correctly managed all patients and resulted in a 22.8% reduction in the number of confirmatory tests.ConclusionsThe integration of diagnostic modeling algorithms in clinical practice may improve the management of patients with PA by circumventing unnecessary confirmatory testing.

OBJECTIVES To demonstrate the efficacy and safety of mini-percutaneous nephrolithotomy, micro-percutaneous nephrolithotomy, and flexible ureteroscopy for pediatric upper urinary tract calculi and to develop nomograms predicting surgical outcomes. METHODS A prospectively managed database containing children who were diagnosed with upper urinary tract calculi and treated with mini-percutaneous nephrolithotomy, micro-percutaneous nephrolithotomy, and flexible ureteroscopy between June 2014 and April 2019 was analysed. Patient demographics, intraoperative data, stone characteristics, stone-free rate, and complication rate were analysed and compared. Nomograms predicting the postoperative stone-free rate and complication rate were established based on predictors, and internal validation was performed. Calibration curves and decision curves were generated to assess the predictive efficacy and clinical benefit. RESULTS Forty-three children underwent mini-percutaneous nephrolithotomy on 56 sides in 47 operations, 30 children underwent micro-percutaneous nephrolithotomy on 30 sides in 30 operations, and 275 children underwent flexible ureteroscopy on 320 sides in 288 operations. The stone-free rates were 88.5% (282/320) for flexible ureteroscopy, 89.3% (50/56) for mini-percutaneous nephrolithotomy, and 90.0% (27/30) for micro-percutaneous nephrolithotomy (P = 0.94). And the complication rates were 19.8% (57/288), 36.2% (17/47), and 33.3% (10/30), respectively (P = 0.02). Nomograms based on stone characteristics, operation duration, and the physical condition of the child were shown to have good discrimination and calibration. The area under the curve of the models was 81% for stone-free rate and 73% for complication rate. The calibration curves showed that the nomogram might underestimate the probability of stone-free rate when the threshold was below 82% and might overestimate the risk of complication rate when the threshold was over 25%. The decision curves demonstrated that the Capital Medical University nomograms improved clinical risk prediction against threshold probabilities of stone-free rate ≤20% and complication rate ≤10%. CONCLUSIONS Both the percutaneous nephrolithotomy and flexible ureteroscopy procedures could have acceptable stone-free rates when treating pediatric stones. The Capital Medical University nomograms performed well in helping to predict stone-free and complication rates.

OBJECTIVE To develop and validate a prediction model for in-hospital mortality in patients with hip fracture 85 years of age or older undergoing surgery. DESIGN A multicenter prospective cohort study. SETTING Six Dutch trauma centers, level 2 and 3. PARTICIPANTS Patients with hip fracture 85 years of age or older undergoing surgery. INTERVENTION Hip fracture surgery. MAIN OUTCOME MEASUREMENTS In-hospital mortality. RESULTS The development cohort consisted of 1014 patients. In-hospital mortality was 4%. Age, male sex, American Society of Anesthesiologists classification, and hemoglobin levels at presentation were independent predictors of in-hospital mortality. The bootstrap adjusted performance showed good discrimination with a c-statistic of 0.77. CONCLUSION Age, male sex, higher American Society of Anesthesiologists classification, and lower hemoglobin levels at presentation are robust independent predictors of in-hospital mortality in patients with geriatric hip fracture and were incorporated in a simple prediction model with good accuracy and no lack of fit. LEVEL OF EVIDENCE Prognostic Level II. See Instructions for Authors for a complete description of levels of evidence.

Objective To develop and internally validate a multivariable predictive model for days with new-onset migraine headaches based on patient self-prediction and exposure to common trigger factors. Background Accurate real-time forecasting of one's daily risk of migraine attack could help episodic migraine patients to target preventive medications for susceptible time periods and help decrease the burden of disease. Little is known about the predictive utility of common migraine trigger factors. Methods We recruited adults with episodic migraine through online forums to participate in a 90-day prospective daily-diary cohort study conducted through a custom research application for iPhone. Every evening, participants answered questions about migraine occurrence and potential predictors including stress, sleep, caffeine and alcohol consumption, menstruation, and self-prediction. We developed and estimated multivariable multilevel logistic regression models for the risk of a new-onset migraine day vs a healthy day and internally validated the models using repeated cross-validation. Results We had 178 participants complete the study and qualify for the primary analysis which included 1870 migraine events. We found that a decrease in caffeine consumption, higher self-predicted probability of headache, a higher level of stress, and times within 2 days of the onset of menstruation were positively associated with next-day migraine risk. The multivariable model predicted migraine risk only slightly better than chance (within-person C-statistic: 0.56, 95% CI: 0.54, 0.58). Conclusions In this study, episodic migraine attacks were not predictable based on self-prediction or on self-reported exposure to common trigger factors. Improvements in accuracy and breadth of data collection are needed to build clinically useful migraine prediction models.

Digital dermatitis (DD) is linked to severe lameness, infertility, and decreased milk production in cattle. Early detection of DD provides an improved prognosis for treatment and recovery; however, this is extremely challenging on commercial dairy farms. Computer vision (COMV) models can help facilitate early DD detection on commercial dairy farms. The aim of this study was to develop and implement a novel COMV tool to identify DD lesions on a commercial dairy farm. Using a database of more than 3,500 DD lesion images, a model was trained using the YOLOv2 architecture to detect the M-stages of DD. The YOLOv2 COMV model detected DD with an accuracy of 71%, and the agreement was quantified as "moderate" by Cohen's kappa when compared with a human evaluator for the internal validation. In the external validation, the YOLOv2 COMV model detected DD with an accuracy of 88% and agreement was quantified as "fair" by Cohen's kappa. Implementation of COMV tools for DD detection provides an opportunity to identify cows for DD treatment, which has the potential to lower DD prevalence and improve animal welfare on commercial dairy farms.

To develop and internally validate nomogram predicting postoperative blood loss risk among pediatric patients with pulmonary atresia (PA) undergoing cardiopulmonary bypass (CPB). All patients aged from 6 months to 6 years with PA who underwent surgery at Fuwai Hospital from June 2015 to December 2019 were selected. And the prediction nomogram model was developed in the training set based on the selected patients. The demographic characteristics and laboratory data from each enrolled patient were gathered. Postoperative blood loss was defined as a blood loss exceeding 20.0 ml/kg within the first 24 postoperative hours. The least absolute shrinkage and selection operator (LASSO) method was used to optimize feature selection for multivariate logistic regression analysis that was applied to build a nomogram composed of all the features selected in the LASSO algorithm. The concordance index (C-index), calibration plot, and decision curve analysis (DCA) were used to evaluate the discrimination, calibration, and clinical net benefit of the nomogarm, respectively. Finally, internal validation was performed using the bootstrap technique. Of the 66 pediatric patients in the training set, 21 (31.82%) and 45 (68.18%) patients were assigned into bleeding group and non-bleeding group, respectively. The first postoperative 24-h blood loss in the bleeding group was significantly higher than that in the non-bleeding group during ICU stay (P = 0.000). Multivariate logistic regression analysis showed that, the immediate postoperative prothrombin time (odds ratio = 1.419, 95% confidence interval: 1.094-1.841, P = 0.008), the immediate postoperative platelet count (odds ratio = 0.985, 95% confidence interval: 0.973-0.997, P = 0.015) and the immediate postoperative red blood cell (RBC) count (odds ratio = 0.335, 95% confidence interval: 0.166-0.667, P = 0.002) were independent predictors of postoperative blood loss risk. The model presented favorable calibration and good discrimination with satisfactory calibration curve and a C-index of 0.858 (95% confidence interval: 0.758-0.958). High C-index value of 0.837 was achieved in the internal validation. The DCA revealed that the nomogram was great clinical effect when intervention was decided among nearly the entire range of threshold probabilities. We developed and internally validated an accurate nomogram to assist in the clinical decision-making concerning the presence of postoperative blood loss in pediatric patients with PA undergoing CPB. However, the nomogram should be endorsed by external validation before it can be recommended for routine practice.

Imaging technology and machine learning algorithms for disease classification set the stage for high-throughput phenotyping and promising new avenues for genome-wide association studies (GWAS). Despite emerging algorithms, there has been no successful application in GWAS so far. We establish machine learning-based phenotyping in genetic association analysis as misclassification problem. To evaluate chances and challenges, we performed a GWAS based on automatically classified age-related macular degeneration (AMD) in UK Biobank (images from 135,500 eyes; 68,400 persons). We quantified misclassification of automatically derived AMD in internal validation data (4,001 eyes; 2,013 persons) and developed a maximum likelihood approach (MLA) to account for it when estimating genetic association. We demonstrate that our MLA guards against bias and artifacts in simulation studies. By combining a GWAS on automatically derived AMD and our MLA in UK Biobank data, we were able to dissect true association (ARMS2/HTRA1, CFH) from artifacts (near HERC2) and identified eye color as associated with the misclassification. On this example, we provide a proof-of-concept that a GWAS using machine learning-derived disease classification yields relevant results and that misclassification needs to be considered in analysis. These findings generalize to other phenotypes and emphasize the utility of genetic data for understanding misclassification structure of machine learning algorithms.

AimsEGFR tyrosine kinase inhibitors (TKIs) are first-line molecularly targeted therapies in patients with advanced non-small cell lung cancer (NSCLC) who carry sensitising EGFR mutations, due to its superior survival outcomes compared with conventional chemotherapy regimens. In this study, we sought to identify clinical, immune and biochemical variables with prognostic significance in this patient subgroup and incorporate them into a nomogram-based risk score.MethodsA total of 199 patients with EGFR mutation-positive, advanced NSCLC (defined as stage IV at initial diagnosis or incurable disease recurrence) treated with first-line EGFR TKI therapy were retrospectively profiled. Univariable and multivariable survival analyses were conducted, with variables from the multivariable model with the highest Harrell’s Concordance (C) Index selected for inclusion in the subsequent survival nomogram. Internal validation and internal calibration of our prognostic nomogram were also performed.ResultsSerum lactate dehydrogenase (LDH) and lung/pleural metastasis were independent predictors of unfavourable overall survival in all three multivariable models. A survival nomogram was generated based on the multivariable model with the highest Harrell’s C Index, incorporating the following 11 variables: white cell count, haemoglobin, LDH, neutrophil/lymphocyte ratio, ethnicity (Chinese vs non-Chinese), Karnofsky-Performance Status (score of ‘90–100’ or ‘70–80’ vs ‘0–60’), Charlson Comorbidity Index (≥3, or 2, or 1 vs 0), neurological symptoms, brain, lung/pleural and adrenal metastases.ConclusionWe identified serum LDH as an independent predictor of unfavourable clinical outcomes in patients with advanced, EGFR mutation-positive NSCLC. We further developed a robust nomogram-based risk score that incorporates clinical, biochemical and immune variables that can provide more targeted prognostication and management in this patient subgroup.

We sought to validate a risk model to predict venous thromboembolism (VTE) in pediatric trauma through an analysis of a contemporary cohort in the National Trauma Data Bank (NTDB).Prospective internal validation was performed in 10 randomly stratified samples of children (age 0-17 years) from the NTDB 2013-2016. Model discrimination was determined by calculation of the c-statistic (AUC), and calibration was evaluated through analysis of observed to expected (O:E) ratio. Recalibration was performed with application of a mixed-effects logistic regression. Model parameters were reestimated based on recalibration.Retrospective review identified 481,485 pediatric trauma patients with 729 (0.2%) episodes of VTE. Discriminatory ability of the model in all random cohorts was significant with AUC 0.93 (p 0.001). Inadequate calibration was noted in 4 of 10 cohorts and the entire dataset (p 0.001) with an O:E ratio of 1.79. Model recalibration resulted in similar discrimination (AUC = 0.95) with improved calibration (O:E ratio = 1.33, p 0.0001).Pediatric trauma prediction models can provide useful data for VTE risk stratification in injured children, but these models must be validated and calibrated prior to use. Recalibration of the model in question resulted in improved accuracy in a contemporary NTDB dataset. These data provide an appropriately calibrated and validated model for clinical use.II - Prospective internal validation of a multivariable prediction model.

Background Timely prediction of AKI in children can allow for targeted interventions, but the wealth of data in the electronic health record poses unique modeling challenges. Methods We retrospectively reviewed the electronic medical records of all children younger than 18 years old who had at least two creatinine values measured during a hospital admission from January 2014 through January 2018. We divided the study population into derivation, and internal and external validation cohorts, and used five feature selection techniques to select 10 of 720 potentially predictive variables from the electronic health records. Model performance was assessed by the area under the receiver operating characteristic curve in the validation cohorts. The primary outcome was development of AKI (per the Kidney Disease Improving Global Outcomes creatinine definition) within a moving 48-hour window. Secondary outcomes included severe AKI (stage 2 or 3), inpatient mortality, and length of stay. Results Among 8473 encounters studied, AKI occurred in 516 (10.2%), 207 (9%), and 27 (2.5%) encounters in the derivation, and internal and external validation cohorts, respectively. The highest-performing model used a machine learning-based genetic algorithm, with an overall receiver operating characteristic curve in the internal validation cohort of 0.76 [95% confidence interval (CI), 0.72 to 0.79] for AKI, 0.79 (95% CI, 0.74 to 0.83) for severe AKI, and 0.81 (95% CI, 0.77 to 0.86) for neonatal AKI. To translate this prediction model into a clinical risk-stratification tool, we identified high- and low-risk threshold points. Conclusions Using various machine learning algorithms, we identified and validated a time-updated prediction model of ten readily available electronic health record variables to accurately predict imminent AKI in hospitalized children.

Purpose Venoarterial extracorporeal membrane oxygenation (VA-ECMO) is an increasingly used treatment option for patients in need of mechanical cardiopulmonary support, while available outcome data is limited. The aim of this study was to identify predictors for 30-day in-hospital mortality. Material and methods We analyzed baseline characteristics and outcomes of 8351 VA-ECMO procedures performed in Germany from 2007 to 2015. Using a multivariable model, we identified the ten most important variables to allow for prediction of 30-day in-hospital mortality. Based on these variables, we created a mortality prediction score (ECMO-ACCEPTS score) to enhance decision making in patients considered for or treated with VA-ECMO support. Results Of 8351 patients (71.7% male) 3567 had prior CPR. Mean age was 62 years in the present cohort. The overall 30-day in-hospital mortality was 61%. The ECMO-ACCEPTS score, derived among randomly selected 4175 patients, included ten independent predictors for in-hospital mortality. Internal validation in the remaining 4176 patients confirmed strong differentiation between low and high risk of 30-day in-hospital mortality (r = 0.97 for correlation of predicted with observed mortality, p Conclusions The ECMO-ACCEPTS score might help clinicians to improve risk prediction among VA-ECMO patients for refractory cardiogenic shock.

OBJECTIVES As evidence has proven that sublobar resection is oncologically contraindicated by tumour spread through air spaces (STAS), its preoperative recognition is vital in customizing surgical strategies. We aimed to assess the value of radiomics in predicting STAS in stage I lung adenocarcinoma. METHODS We retrospectively reviewed the patients with stage I lung adenocarcinoma, who accepted curative resection in our institution between January 2011 and December 2013. Using ‘PyRadiomics’ package, 88 radiomics features were extracted from computed tomography (CT) images and a prediction model was consequently constructed using Naïve Bayes machine-learning approach. The accuracy of the model was assessed through receiver operating curve analysis, and the performance of the model was validated both internally and externally. RESULTS A total of 233 patients were included as the training cohort with 69 (29.6%) patients being STAS (+). Patients with STAS had worse recurrence-free survival and overall survival (P CONCLUSIONS CT-based radiomics is valuable in preoperatively predicting STAS in stage I lung adenocarcinoma, which may aid surgeons in determining the optimal surgical approach.

Objective To develop a practicable nomogram aimed at predicting the risk of severe exacerbations in COPD patients at three and five years. Methods COPD patients with prospective follow-up data were extracted from Subpopulations and Intermediate Outcome Measures in COPD Study (SPIROMICS) obtained from National Heart, Lung and Blood Institute (NHLBI) Biologic Specimen and Data Repository Information Coordinating Center. We comprehensively considered the demographic characteristics, clinical data and inflammation marker of disease severity. Cox proportional hazard regression was performed to identify the best combination of predictors on the basis of the smallest Akaike Information Criterion. A nomogram was developed and evaluated on discrimination, calibration, and clinical efficacy by the concordance index (C-index), calibration plot and decision curve analysis, respectively. Internal validation of the nomogram was assessed by the calibration plot with 1000 bootstrapped resamples. Results Among 1711 COPD patients, 523 (30.6%) suffered from at least one severe exacerbation during follow-up. After stepwise regression analysis, six variables were determined including BMI, severe exacerbations in the prior year, comorbidity index, post-bronchodilator FEV1% predicted, and white blood cells. Nomogram to estimate patients' likelihood of severe exacerbations at three and five years was established. The C-index of the nomogram was 0.74 (95%CI: 0.71-0.76), outperforming ADO, BODE and DOSE risk score. Besides, the calibration plot of three and five years showed great agreement between nomogram predicted possibility and actual risk. Decision curve analysis indicated that implementation of the nomogram in clinical practice would be beneficial and better than aforementioned risk scores. Conclusion Our new nomogram was a useful tool to assess the probability of severe exacerbations at three and five years for COPD patients and could facilitate clinicians in stratifying patients and providing optimal therapies.

Purpose: Deep learning is a newer and advanced subfield in artificial intelligence (AI). The aim of our study is to validate a machine-based algorithm developed based on deep convolutional neural networks as a tool for screening to detect referable diabetic retinopathy (DR). Methods: An AI algorithm to detect DR was validated at our hospital using an internal dataset consisting of 1,533 macula-centered fundus images collected retrospectively and an external validation set using Methods to Evaluate Segmentation and Indexing Techniques in the field of Retinal Ophthalmology (MESSIDOR) dataset. Images were graded by two retina specialists as any DR, prompt referral (moderate nonproliferative diabetic retinopathy (NPDR) or above or presence of macular edema) and sight-threatening DR/STDR (severe NPDR or above) and compared with AI results. Sensitivity, specificity, and area under curve (AUC) for both internal and external validation sets for any DR detection, prompt referral, and STDR were calculated. Interobserver agreement using kappa value was calculated for both the sets and two out of three agreements for DR grading was considered as ground truth to compare with AI results. Results: In the internal validation set, the overall sensitivity and specificity was 99.7% and 98.5% for Any DR detection and 98.9% and 94.84%for Prompt referral respectively. The AUC was 0.991 and 0.969 for any DR detection and prompt referral respectively. The agreement between two observers was 99.5% and 99.2% for any DR detection and prompt referral with a kappa value of 0.94 and 0.96, respectively. In the external validation set (MESSIDOR 1), the overall sensitivity and specificity was 90.4% and 91.0% for any DR detection and 94.7% and 97.4% for prompt referral, respectively. The AUC was. 907 and. 960 for any DR detection and prompt referral, respectively. The agreement between two observers was 98.5% and 97.8% for any DR detection and prompt referral with a kappa value of 0.971 and 0.980, respectively. Conclusion: With increasing diabetic population and growing demand supply gap in trained resources, AI is the future for early identification of DR and reducing blindness. This can revolutionize telescreening in ophthalmology, especially where people do not have access to specialized health care.

International audience; Background: It has previously been demonstrated that a fraction of patients with hepatocellular carcinoma (HCC) > 10 cm can benefit from liver resection. However, there is still a lack of effective decision-making tools to inform intervention in these patients.Methods: We analysed a comprehensive set of clinical data from 234 patients who underwent liver resection for HCC >10 cm at the National Cancer Institute of Peru between 1990 and 2015, monitored their survival, and constructed a nomogram to predict the surgical outcome based on preoperative variables.Results: We identified cirrhosis, multifocality, macroscopic vascular invasion, and spontaneous tumour rupture as independent predictors of survival and integrated them into a nomogram model. The nomogram's ability to forecast survival at 1, 3, and 5 years was subsequently confirmed with high concordance using an internal validation. Through applying this nomogram, we stratified three groups of patients with different survival probabilities.Conclusion: We constructed a preoperative nomogram to predict long-term survival in patients with HCC >10 cm. This nomogram is useful in determining whether a patient with large HCC might truly benefit from liver resection, which is paramount in low- and middle-income countries where HCC is often diagnosed at advanced stages.

IntroductionAtrial fibrillation (AF) is a major cardiovascular health problem: it is common, chronic and incurs substantial healthcare expenditure because of stroke. Oral anticoagulation reduces the risk of thromboembolic stroke in those at higher risk; but for a number of patients, stroke is the first manifestation of undetected AF. There is a rationale for the early diagnosis of AF, before the first complication occurs, but population-based screening is not recommended. Previous prediction models have been limited by their data sources and methodologies. An accurate model that uses existing routinely collected data is needed to inform clinicians of patient-level risk of AF, inform national screening policy and highlight predictors that may be amenable to primary prevention.Methods and analysisWe will investigate the application of a range of deep learning techniques, including an adapted convolutional neural network, recurrent neural network and Transformer, on routinely collected primary care data to create a personalised model predicting the risk of new-onset AF over a range of time periods. The Clinical Practice Research Datalink (CPRD)-GOLD dataset will be used for derivation, and the CPRD-AURUM dataset will be used for external geographical validation. Both comprise a sizeable representative population and are linked at patient-level to secondary care databases. The performance of the deep learning models will be compared against classic machine learning and traditional statistical predictive modelling methods. We will only use risk factors accessible in primary care and endow the model with the ability to update risk prediction as it is presented with new data, to make the model more useful in clinical practice.Ethics and disseminationPermissions for CPRD-GOLD and CPRD-AURUM datasets were obtained from CPRD (ref no: 19_076). The CPRD ethical approval committee approved the study. The results will be submitted as a research paper for publication to a peer-reviewed journal and presented at peer-reviewed conferences.Trial registration detailsA systematic review to incorporate within the overall project was registered on PROSPERO (registration number CRD42021245093). The study was registered on ClinicalTrials.gov (NCT04657900).

Background. In a retrospective cohort study, we previously distinguished the factors associated with coronavirus 2019 (COVID-19) or dengue from those associated with other febrile illnesses (OFIs). In this study, we developed a scoring system to discriminate both infectious diseases. Methods. Predictors of both infections were sought using multinomial logistic regression models (OFIs as controls) in all subjects suspected of COVID-19 who attended the SARS-CoV-2 testing center of Saint-Pierre teaching hospital, Reunion Island, between March 23 and May 10, 2020. Two COVIDENGUE scores were developed and internally validated by bootstrapping for predicting each infection after weighting the odd ratios according to a predefined rule. The discriminative ability of each score was assessed using the area under the receiver operating characteristic curve (AUC). Their calibration was assessed using goodness-of-fit statistics. Results. Over 49 days, 80 COVID-19, 60 non-severe dengue and 872 OFI cases were diagnosed. The translation of the best fit model yielded two COVIDENGUE scores composed of 11 criteria: contact with a COVID-19 positive case (+3 points for COVID-19; 0 point for dengue), return from travel abroad within 15 days (+3/-1), previous individual episode of dengue (+1/+3), active smoking (-3/0), body ache (0/+5), cough (0/-2), upper respiratory tract infection symptoms (-1/-1), anosmia (+7/-1), headache (0/+5), retro-orbital pain (-1/+5), and delayed presentation (>3 days) to hospital (+1/0). The AUC was of 0.79 (95%CI 0.76-0.82) for COVID-19 score and of 0.88 (95%CI 0.85-0.90) for dengue score. Calibration was satisfactory for COVID-19 score and excellent for dengue score. For predicting COVID-19, sensitivity was of 97% at the 0-point cut-off and specificity approximated 99% at the 10-point cut-off. For predicting dengue, sensitivity approximated 97% at the 3-point cut-off and specificity 98% at the 11-point cut-off. Conclusions. In conclusion, the COVIDENGUE scores proved discriminant to differentiate COVID-19 and dengue from other febrile illnesses in the context of SARS-CoV-2 testing center during a co-epidemic. Further studies are needed to validate or refine these scores in other settings.

The aim of this pilot study was to develop logistic regression (LR) and support vector machine (SVM) models that differentiate low from high risk for prolonged hospital length of stay (LOS) in a South African cohort of 383 colorectal cancer patients who underwent surgical resection with curative intent. Additionally, the impact of 10-fold cross-validation (CV), Monte Carlo CV, and bootstrap internal validation methods on the performance of the two models was evaluated. The median LOS was 9 days, and prolonged LOS was defined as greater than 9 days post-operation. Preoperative factors associated with prolonged LOS were a prior history of hypertension and an Eastern Cooperative Oncology Group score between 2 and 4. Postoperative factors related to prolonged LOS were the need for a stoma as part of the surgical procedure and the development of post-surgical complications. The risk of prolonged LOS was higher in male patients and in any patient with lower preoperative hemoglobin. The highest area under the receiving operating characteristics (AU-ROC) was achieved using LR of 0.823 (CI = 0.798–0.849) and SVM of 0.821 (CI = 0.776–0.825), with each model using the Monte Carlo CV method for internal validation. However, bootstrapping resulted in models with slightly lower variability. We found no significant difference between the models across the three internal validation methods. The LR and SVM algorithms used in this study required incorporating important features for optimal hospital LOS predictions. The factors identified in this study, especially postoperative complications, can be employed as a simple and quick test clinicians may flag a patient at risk of prolonged LOS.

BackgroundThe diagnosis of adult GH deficiency (GHD) relies on a reduced GH response to provocative tests. Their diagnostic accuracy, however, is not perfect, and a reliable estimation of pre-test GHD probability could be helpful for a better interpretation of their results.MethodsEighty patients showing concordant GH response to two provocative tests, i.e. the insulin tolerance test and the GHRH + arginine test, were enrolled. Data on IGF-I values and on the presence/absence of other pituitary deficits were collected and integrated for the estimation of GHD probability prior to stimulation tests.ResultsAn independent statistically significant association with the diagnosis of GHD was found both for IGF-I SDS (OR 0.34, 95%-CI 0.18-0.65, p=0.001) and for the presence of other pituitary deficits (OR 6.55, 95%-CI 2.06-20.83, p=0.001). A low ( +0.91 in the presence of other pituitary deficits or IGF-I SDS > -0.52 in the absence of other pituitary deficits. A high (>75%) pre-test GHD probability could be predicted when IGF-I SDS < -0.82 in the presence of other pituitary deficits or IGF-I SDS < -2.26 in the absence of other pituitary deficits.ConclusionThis is the first study that proposes a quantitative estimation of GHD probability prior to stimulation tests. Our risk class stratification represents a simple tool that could be adopted for a Bayesian interpretation of stimulation test results, selecting patients who may benefit from a second stimulation test and possibly reducing the risk of wrong GHD diagnosis.

Purpose To systematically review the literature evaluating clinical utility of imaging metrics derived from baseline fluorine-18 fluorodeoxyglucose positron emission tomography/computed tomography (PET/CT) for prediction of progression-free (PFS) and overall survival (OS) in patients with classical Hodgkin lymphoma (HL) and diffuse large B cell lymphoma (DLBCL). Methods A search of MEDLINE/PubMed, Web of Science, Cochrane, Scopus and clinicaltrials.gov databases was undertaken for articles evaluating PET/CT imaging metrics as outcome predictors in HL and DLBCL. PRISMA guidelines were followed. Risk of bias was assessed using the Quality in Prognosis Studies (QUIPS) tool. Results Forty-one articles were included (31 DLBCL, 10 HL). Significant predictive ability was reported in 5/20 DLBCL studies assessing SUVmax (PFS: HR 0.13–7.35, OS: HR 0.83–11.23), 17/19 assessing metabolic tumour volume (MTV) (PFS: HR 2.09–11.20, OS: HR 2.40–10.32) and 10/13 assessing total lesion glycolysis (TLG) (PFS: HR 1.078–11.21, OS: HR 2.40–4.82). Significant predictive ability was reported in 1/4 HL studies assessing SUVmax (HR not reported), 6/8 assessing MTV (PFS: HR 1.2–10.71, OS: HR 1.00–13.20) and 2/3 assessing TLG (HR not reported). There are 7/41 studies assessing the use of radiomics (4 DLBCL, 2 HL); 5/41 studies had internal validation and 2/41 included external validation. All studies had overall moderate or high risk of bias. Conclusion Most studies are retrospective, underpowered, heterogenous in their methodology and lack external validation of described models. Further work in protocol harmonisation, automated segmentation techniques and optimum performance cut-off is required to develop robust methodologies amenable for clinical utility.

Simple Summary Exosomal miRNAs are associated with colorectal cancer liver metastasis (CRLM)-related biological behavior and prognosis. However, an exosomal miRNA signature predicting postoperative survival and the value of adjuvant chemotherapy for CRLM remains elusive. Using miRNA sequencing and the LASSO model, we constructed an miRNA signature comprising four exosomes. The signature showed a good predictive performance for patient outcome and the advantage of adjuvant chemotherapy after hepatectomy in two institutions’ training and validation cohorts. In addition, we found that the four miRNAs could target signaling molecules playing crucial roles in colorectal cancer metastasis, vesicle-related processing, and T cell activation. Furthermore, the exosomal miRNA score also increased with the decreasing Immunoscore. We believe that our signature can predict the prognosis and guide adjuvant chemotherapy decisions after liver metastasectomy in CRLM patients, further improving the predictive performance of the current CRLM predictive model system. Abstract Background: The clinical risk score (CRS) for prediction and treatment decision in colorectal liver metastasis (CRLM) is important, but imprecise. Exosomal miRNAs play critical roles in CRLM-related biological behavior. However, an exosomal miRNA score system for predicting posthepatectomy survival and the adjuvant chemotherapy benefit of CRLM remains elusive. Methods: miRNA sequencing was used to identify differentially expressed miRNAs, and the LASSO model was used to select miRNAs to construct the intent model. The predictive performance of the model was evaluated by the area under the ROC curve (AUC) in the training, internal validation, and external validation cohorts. Results: Sixteen differentially expressed exosomal miRNAs were identified, and four miRNAs were selected for model construction. Our model performed well in predicting prognosis with five-year AUCs of 0.70 (95% CI: 0.59–0.81), 0.70 (0.61–0.81), and 0.72 (057–0.86) in the training, internal, and external validation cohorts, respectively. miRNA classifier high-risk patients had better survival benefit from adjuvant chemotherapy regardless of CRS. All four miRNAs target signaling molecules play crucial roles in colorectal cancer metastasis, vesicle-related processing, and T cell activation. It also negatively correlated with the liver metastasis Immunoscore. Conclusion: We developed a circulating exosomal miRNA signature that can predict the prognosis and guide adjuvant chemotherapy decisions after hepatectomy in CRLM.

BACKGROUND AND AIM Endoscopic ultrasound (EUS) is the most accurate diagnostic modality for polypoid lesions of the gallbladder (GB), but is limited by subjective interpretation. Deep learning-based artificial intelligence (AI) algorithms are under development. We evaluated the diagnostic performance of AI in differentiating polypoid lesions using EUS images. METHODS The diagnostic performance of the EUS-AI system with ResNet50 architecture was evaluated via three processes: training, internal validation, and testing using an AI development cohort of 1039 EUS images (836 GB polyps and 203 gallstones). The diagnostic performance was verified using an external validation cohort of 83 patients and compared with the performance of EUS endoscopists. RESULTS In the AI development cohort, we developed an EUS-AI algorithm and evaluated the diagnostic performance of the EUS-AI including sensitivity, specificity, positive predictive value, negative predictive value, and accuracy. For the differential diagnosis of neoplastic and non-neoplastic GB polyps, these values for EUS-AI were 57.9%, 96.5%, 77.8%, 91.6%, and 89.8%, respectively. In the external validation cohort, we compared diagnostic performances between EUS-AI and endoscopists. For the differential diagnosis of neoplastic and non-neoplastic GB polyps, the sensitivity and specificity were 33.3% and 96.1% for EUS-AI; they were 74.2% and 44.9%, respectively, for the endoscopists. Besides, the accuracy of the EUS-AI was between the accuracies of mid-level (66.7%) and expert EUS endoscopists (77.5%). CONCLUSIONS This newly developed EUS-AI system showed favorable performance for the diagnosis of neoplastic GB polyps, with a performance comparable to that of EUS endoscopists.

Background The incidence of primary intestinal diffuse large B-cell lymphoma (PI-DLBCL) is much lower than primary gastric DLBCL, and large-scale analyses on the clinical characteristics, molecular features, therapeutic strategies, and risk stratification have been seldomly performed in PI-DLBCL. Methods To assess prognostic model development, 107 PI-DLBCL patients diagnosed before 2014 were studied for prognosis factors including different primary involved sites and treatment strategies. For internal validation, a non-random split sample set with 77 PI-DLBCL patients after 2014 was included for validation of the prognosis factors. Results Patients with an ileocecal lesion presented with better survival time than those with non-ileocecal sites, with surgical resection significantly influencing the prognosis. Non-ileocecal patients who underwent surgery with lymphadenectomy had superior overall survival (OS) and progression-free survival (PFS) compared to those receiving surgery without lymphadenectomy or those not receiving (without) surgery. For ileocecal patients, surgery with or without lymphadenectomy resulted in better OS and PFS than those without surgery. For biomarker analysis, only BCL-2 >50% or Ki67 >80% on tumor cells indicated poor clinical outcome. In multivariate analysis, age, Eastern Cooperative Oncology Group (ECOG) score, and site of origin were independent prognostic factors for inferior OS in PI-DLBCL. A prognosis model was set up based on age, ECOG score, and site of origin, and validated well. Conclusions The prognosis in patients with PI-DLBCL with ileocecal involvement showed was better than those with non-ileocecal involvement. Surgical strategy can impact the clinical outcome of PI-DLBCL patients.

Background: Surgery is increasingly accepted as an adjunctive approach to treat multidrug-resistant tuberculosis (MDR-TB) or extensively drug-resistant tuberculosis (XDR-TB). However, a model that includes all factors to predict the risk of postoperative complications is lacking.Methods: We developed a prediction model based on 138 patients who had undergone surgery as treatment for drug-resistant tuberculosis (DR-TB) after 24 months. Clinical features on the lesion type (L), treatment history (T), physiologic status of the body (B), and surgical approach (S) were evaluated. Multivariable logistic regression analysis was conducted by clinical features selected in the least absolute shrinkage and selection operator (LASSO) to build a nomogram. The discrimination, calibration, and clinical usefulness of the nomogram were assessed using the C-Index, calibration plots, and decision curves. Internal validation was assessed using bootstrapping.Results: The nomogram contained the features L, B, T, cavitary, recurrent chest infection (RCI) and MDR-TB/XDR-TB. The model displayed good discrimination with a C-Index of 0.879 (95% CI: 0.799–0.967). A high C-Index of 0.824 was achieved in the interval validation. Decision-curve analysis showed that the nomogram was clinically useful if intervention was decided at the non-adherence possibility threshold of 4%.Conclusion: Our novel nomogram could be used conveniently to predict postoperative complication risk in DR-TB patients.

Malignant mesothelioma (MM) is an aggressive cancer primarily diagnosed on the basis of histological criteria1. The 2015 World Health Organization classification subdivides mesothelioma tumors into three histological types: epithelioid, biphasic and sarcomatoid MM. MM is a highly complex and heterogeneous disease, rendering its diagnosis and histological typing difficult and leading to suboptimal patient care and decisions regarding treatment modalities2. Here we have developed a new approach-based on deep convolutional neural networks-called MesoNet to accurately predict the overall survival of mesothelioma patients from whole-slide digitized images, without any pathologist-provided locally annotated regions. We validated MesoNet on both an internal validation cohort from the French MESOBANK and an independent cohort from The Cancer Genome Atlas (TCGA). We also demonstrated that the model was more accurate in predicting patient survival than using current pathology practices. Furthermore, unlike classical black-box deep learning methods, MesoNet identified regions contributing to patient outcome prediction. Strikingly, we found that these regions are mainly located in the stroma and are histological features associated with inflammation, cellular diversity and vacuolization. These findings suggest that deep learning models can identify new features predictive of patient survival and potentially lead to new biomarker discoveries.

Purpose In recent years, there has been an increase in the incidence of small renal masses (SRMs) and nephrectomy was the standard management of this disease in the past. Currently, the use of active surveillance has been recommended as an alternative option in the case of some patients with SRMs due to its heterogenicity. However, limited studies focused on the regarding risk stratification. Therefore, in the current study, we developed a nomogram for the purpose of predicting the presence of high-grade SRMs on the basis of the patient information provided (clinical information, hematological indicators, and CT imaging data). Patients and methods A total of 329 patients (consisting of development and validation cohort) who had undergone nephrectomy for SRMs between January 2013 and May 2016 retrospectively were recruited for the present study. All preoperative information, including clinical predictors, hematological indicators, and CT predictors, were obtained. Lasso regression model was used for data dimension reduction and feature selection. Multivariable logistic regression analysis was applied for the establishment of the predicting model. The performance of the nomogram was assessed with respect to its calibration and discrimination properties and externally validated. Results The predictors used in the assessment of the nomogram included tumor size, CT tumor contour, CT necrosis, CT tumor exophytic properties, and CT collecting system oppression. Based on these parameters, the nomogram was evaluated to have an effective discrimination and calibration ability, and the C-index was found to be 0.883 after internal validation and 0.887 following external validation. Conclusion Based on the aforementioned findings, it can be concluded that CT imaging-based preoperative nomogram is an effective predictor of SRMs and hence can be used in the preoperative evaluation of SRMs, due to its calibration and discrimination abilities.

Background Increased attention has been focused on cancer immunity gene signature. However, the threshold of immune scores to predict disease‐free survival (DFS) and overall survival (OS) in breast cancer has not yet been defined. This study aimed to explore the association of immune scores with prognosis and build a clinical nomogram to predict the survival of patients with breast cancer. Methods A total of 986 subjects were analyzed, and clinicopathological characteristics and immune scores were obtained from the TCGA database. Cox proportional hazards regression model was used to estimate the adjusted hazard ratios (HRs). Based on results of multivariate analysis, nomograms were built. The models were subjected to bootstrap internal validation. The predictive accuracy and discriminative ability were measured by concordance index (C‐index) and the calibration curve. Results The patients were divided into three subgroups according to their immune scores. We found that compared with patients with low immune scores, those with intermediate and high immune scores had significantly improved DFS (HR and 95% confidence interval [CI]: 0.439 [0.242‐0.799], 0.541 [0.343‐0.855], respectively), whereas only intermediate immune scores significantly indicated better OS (HR and 95% CI: 0.385 [0.163‐0.910]). The C‐index for DFS and OS prediction was 0.723 (95% CI, 0.661‐0.785) and 0.800 (95% CI, 0.724‐0.877), respectively. The calibration curves for probability of 3‐ and 5‐year DFS showed significant agreement between nomogram predictions and the actual observations. Conclusions High and/or intermediate immune scores are significantly correlated with better DFS and OS in patients with breast cancer. Moreover, the nomograms for predicting prognosis may help to estimate the survival of patients.

This study aimed to develop and validate nomograms to predict overall survival (OS) and cancer-specific survival (CSS) in patients with prostate cancer. Clinical data of patients with mPCa between 2010 and 2014 were retrieved retrospectively, and randomized into training (2/3) and validation sets (1/3). Nomograms were built with potential risk factors based on COX regression analysis. Accuracy was validated using the discrimination and calibration curve for the training and validation groups, respectively. 6659 mPCa patients were collected and enrolled, including 4440 in the training set and 2219 in the validation set. Multivariate analysis showed that age, marital status, PSA, biopsy Gleason score, T stage, and bone metastasis were independent risk factors for both OS and CSS. The concordance index (C-index) of OS was 0.735 (95% CI 0.722–0.748) for the internal validation and 0.735 (95% CI 0.717–0.753) for the external validation. For CSS, it was 0.734 (95% CI 0.721–0.747) and 0.742 (95% CI 0.723–0.761), respectively. The nomograms for predicting OS and CSS displayed better discrimination power in both training and validation sets. Moreover, a favorable consistency between the predicted and actual survival probabilities was demonstrated using calibration curves. The nomograms showed good performances for predicting OS and CSS in patients with prostate cancer. It might be a convenient individualized predictive tool for prognosis in clinical practice.

Purpose: To analyze pre-transplantation and early postoperative factors affecting post-transplantation urine output and develop a predictive nomogram. Patients and Methods: Retrospective analysis of non-preemptive first transplanted adult patients between 2001-2016. The outcomes were hourly diuresis in mL/Kg in the 1st (UO1) and 8th (UO8) postoperative days (POD). Predictors for both UO1 and UO8 were cold ischemia time (CIT), patient and donor age and sex, HLA I and II compatibility, pre-transplantation duration of renal replacement therapy (RRT), cause of ESRD (ESRD) and immunosuppressive regimen. UO8 predictors also included UO1, 1st/0th POD plasma creatinine concentration ratio (Cr1/0), and occurrence of acute cellular rejection (AR). Multivariable linear regression was employed to produce nomograms for UO1 and UO8. Results: Four hundred and seventy-three patients were included, mostly deceased donor kidneys’ recipients (361, 70.4%). CIT inversely correlated with UO1 and UO8 (Spearman's p=-0.43 and −0.37). CR1/0 inversely correlated with UO8 (p=-0.47). On multivariable analysis UO1 was mainly influenced by CIT, with additional influences of donor age and sex, HLA II matching and ESRD. UO1 was the strongest predictor of UO8, with significant influences of AR and ESRD. Conclusions: The predominant influence of CIT on UO1 rapidly wanes and is replaced by indicators of functional recovery (mainly UO1) and allograft's immunologic acceptance (AR absence). Mean absolute errors for nomograms were 0.08 mL/Kg h (UO1) and 0.05 mL/Kg h (UO8).

Background Triple negative breast cancer (TNBC) is an aggressive and heterogeneous disease. Nomograms predicting outcomes of TNBC are needed for risk management. Methods Nomograms were based on an analysis of 296 non-metastatic TNBC patients treated at Sun Yat-sen Memorial Hospital from 2002 to 2014. The end points were disease-free survival (DFS) and overall survival (OS). Predictive accuracy and discriminative ability were evaluated by concordance index (C-index), area under the curve (AUC) and calibration curve, and compared with the American Joint Committee on Cancer (AJCC) staging system, PREDICT and CancerMath. Models were subjected to bootstrap internal validation and external validation using independent cohorts of 191 patients from the second Xiangya Hospital and Peking University Shenzhen Hospital between 2007 and 2012. Results On multivariable analysis of training cohort, independent prognostic factors were stromal tumor-infiltrating lymphocytes (TILs), tumor size, node status, and Ki67 index, which were then selected into the nomograms. The calibration curves for probability of DFS and OS showed optimal agreement between nomogram prediction and actual observation. The C-index of nomograms was significantly higher than that of the seventh and eighth AJCC staging system for predicting DFS (training: 0.743 vs 0.666 (P = 0.003) and 0.664 (P = 0.024); validation: 0.784 vs 0.632 (P = 0.02) and 0.607 (P = 0.002)) and OS (training: 0.791 vs 0.683 (P = 0.004) and 0.677 (P