Mortality In High-risk Group Research Articles

Evaluation of neutrophil gelatinase-associated lipocalin (NGAL), hypoxia-induced factor-1 alpha (HIF-1α) and apelin 13 levels as new potential biomarkers for pulmonary thromboembolism: A prospective clinical study

AimThe objective was to evaluate the serum levels of neutrophil gelatinase-associated lipocalin (NGAL), hypoxia-induced factor-1 alpha (HIF-1α), and apelin 13 in patients with acute pulmonary thromboembolism (PE) and to investigate their diagnostic and prognostic role in PE patients with different mortality risk groups. Material and methodsThis study was conducted in a tertiary referral center and included 124 subjects with 94 cases of PE and 30 cases of healthy control group. All subjects were 18 years of age or older. The diagnosis of PE was done with computed tomography angiography of the thorax. After the diagnosis of acute PE, the serum levels of neutrophil gelatinase-associated lipocalin (NGAL), hypoxia-induced factor-1 alpha (HIF-1α), and apelin 13 levels were measured with a commercial enzyme-linked immunosorbent assay (ELISA) kit. ResultsThe median and IQR (interquartile range) age of patients and control groups were 68 (56–76) and 61.5 (56–67) years, respectively. The majority of patients with PE had risk factors (97.88 %), and only two (2.12 %) had no known risk factors. HIF-1 alpha level was found to be higher in the patient group than in the control group (p = 0.03). At the same time, the HIF-1 alpha level was found to be higher in the high mortality risk group than in the control group, low mortality risk group and intermediate-low mortality risk group (p = 0.000, 0.011, 0.002, respectively). While there was no significant difference in NGAL level between the patient group and the control group, a significant difference was observed between the mortality groups. NGAL level was found to be higher in the high mortality risk group than the control group, low mortality risk group, and medium-low mortality risk group (p = 0.001, 0.000, 0.010, respectively). Apelin 13 levels did not differ significantly in all groups. ConclusionHIF-1 alpha is a promising biomarker in distinguishing between patients and control groups and in identifying those with high mortality risk in the patient group. At the same time, NGAL can be used as a successful biomarker in determining the group with high mortality risk in cases of PE.

Which medicare advantage enrollees are at highest one-year mortality risk?

BackgroundWhile a number of tools exist to predict mortality among older adults, less research has described the characteristics of Medicare Advantage (MA) enrollees at higher risk for 1 year mortality. ObjectivesTo describe the characteristics of MA enrollees at higher mortality risk using patient survey data. Research DesignRetrospective cohort. SubjectsMA enrollees completing the 2019 MA Consumer Assessment of Healthcare Providers and Systems (CAHPS) Survey. MeasuresLinked demographic, health, and mortality data from a sample of MA enrollees were used to predict 1-year mortality risk and describe enrollee characteristics across levels of predicted mortality risk. ResultsThe mortality model had a 0.80 c-statistic. Mortality risks were skewed: 6 % of enrollees had a ≥ 10 % 1-year mortality risk, while 45 % of enrollees had 1 % to < 5 % 1-year mortality risk. Among the high-risk (≥10 %) group, 47 % were age 85+ versus 12 % among those with mortality risk <5 %. 79 % were in fair or poor self-rated health versus 29 % among those with mortality risk of <5 %. 71 % reported needing urgent care in the prior 6 months versus 40 % among those with a mortality risk of 1 to<5 %. ConclusionsRelatively few older adults enrolled in MA are at high 1-year mortality risk. Nonetheless, MA enrollees over age 85, in fair or poor health, or with recent urgent care needs are far more likely to be in a high mortality risk group.

Survival prediction for stage I-IIIA non-small cell lung cancer using deep learning

Background and purposeThe aim of this study was to develop and evaluate a prediction model for 2-year overall survival (OS) in stage I-IIIA non-small cell lung cancer (NSCLC) patients who received definitive radiotherapy by considering clinical variables and image features from pre-treatment CT-scans. Materials and methodsNSCLC patients who received stereotactic radiotherapy were prospectively collected at the UMCG and split into a training and a hold out test set including 189 and 81 patients, respectively. External validation was performed on 228 NSCLC patients who were treated with radiation or concurrent chemoradiation at the Maastro clinic (Lung1 dataset). A hybrid model that integrated both image and clinical features was implemented using deep learning. Image features were learned from cubic patches containing lung tumours extracted from pre-treatment CT scans. Relevant clinical variables were selected by univariable and multivariable analyses. ResultsMultivariable analysis showed that age and clinical stage were significant prognostic clinical factors for 2-year OS. Using these two clinical variables in combination with image features from pre-treatment CT scans, the hybrid model achieved a median AUC of 0.76 [95 % CI: 0.65–0.86] and 0.64 [95 % CI: 0.58–0.70] on the complete UMCG and Maastro test sets, respectively. The Kaplan-Meier survival curves showed significant separation between low and high mortality risk groups on these two test sets (log-rank test: p-value < 0.001, p-value = 0.012, respectively) ConclusionWe demonstrated that a hybrid model could achieve reasonable performance by utilizing both clinical and image features for 2-year OS prediction. Such a model has the potential to identify patients with high mortality risk and guide clinical decision making.

The Efficacy of Pretreatment 18F-FDG PET-CT-Based Deep Learning Network Structure to Predict Survival in Nasopharyngeal Carcinoma.

Previous studies have shown that the 5-year survival rates of patients with nasopharyngeal carcinoma (NPC) were still not ideal despite great improvement in NPC treatments. To achieve individualized treatment of NPC, we have been looking for novel models to predict the prognosis of patients with NPC. The objective of this study was to use a novel deep learning network structural model to predict the prognosis of patients with NPC and to compare it with the traditional PET-CT model combining metabolic parameters and clinical factors. A total of 173 patients were admitted to 2 institutions between July 2014 and April 2020 for the retrospective study; each received a PET-CT scan before treatment. The least absolute shrinkage and selection operator (LASSO) was employed to select some features, including SUVpeak-P, T3, age, stage II, MTV-P, N1, stage III and pathological type, which were associated with overall survival (OS) of patients. We constructed 2 survival prediction models: an improved optimized adaptive multimodal task (a 3D Coordinate Attention Convolutional Autoencoder and an uncertainty-based jointly Optimizing Cox Model, CACA-UOCM for short) and a clinical model. The predictive power of these models was assessed using the Harrell Consistency Index (C index). Overall survival of patients with NPC was compared by Kaplan-Meier and Log-rank tests. The results showed that CACA-UOCM model could estimate OS (C index, 0.779 for training, 0.774 for validation, and 0.819 for testing) and divide patients into low and high mortality risk groups, which were significantly associated with OS (P < .001). However, the C-index of the model based only on clinical variables was only 0.42. The deep learning network model based on 18F-FDG PET/CT can serve as a reliable and powerful predictive tool for NPC and provide therapeutic strategies for individual treatment.

Spontaneous Bacterial Peritonitis in Decompensated Liver Cirrhosis—A Literature Review

Background: Spontaneous bacterial peritonitis (SBP) is defined as a bacterial infection of the ascitic fluid without a surgically treatable intra-abdominal infection source. SBP is a common, severe complication in cirrhosis patients with ascites, and if left untreated, in-hospital mortality may exceed 90%. However, the incidence of SBP has been lowered to approx. 20% through early diagnosis and antibiotic therapy. Clinical awareness, prompt diagnosis, and immediate treatment are advised when caring for these patients to reduce mortality and morbidity. Aim: To discuss important issues comprising types of SBP, pathogenesis, bacteriology, including the emergence of multidrug-resistant (MDR) microorganisms, prompt diagnosis, risk factors, prognosis, treatment strategies, as well as recurrence prevention through antibiotic prophylaxis until liver transplantation and future trends in treating and preventing SBP in detail. Methods: This article is a literature review and appraisal of guidelines, randomized controlled trials, meta-analyses, and other review articles found on PubMed from between 1977 and 2022. Results: There are three types of SBP. Bacterial translocation from GI tract is the most common source of SBP. Therefore, two thirds of SBP cases were caused by Gram-negative bacilli, of which Escherichia coli is the most frequently isolated pathogen. However, a trend of Gram-positive cocci associated SBP has been demonstrated in recent years, possibly related to more invasive procedures and long-term quinolone prophylaxis. A diagnostic paracentesis should be performed in all patients with cirrhosis and ascites who require emergency room care or hospitalization, who demonstrate or report consistent signs/symptoms in order to confirm evidence of SBP. Distinguishing SBP from secondary bacterial peritonitis is essential because the conditions require different therapeutic strategies. The standard treatment for SBP is prompt broad-spectrum antibiotic administration and should be tailored according to community-acquired SBP, healthcare-associated or nosocomial SBP infections and local resistance profile. Albumin supplementation, especially in patients with renal impairment, is also beneficial. Selective intestinal decontamination is associated with a reduced risk of bacterial infection and mortality in high-risk group. Conclusions: The standard treatment for SBP is prompt broad-spectrum antibiotic administration and should be tailored according to community-acquired SBP, healthcare-associated or nosocomial SBP infections and local resistance profile. Since the one-year overall mortality rates for SBP range from 53.9 to 78%, liver transplantation should be seriously considered for SBP survivors who are good candidates for transplantation. Further development of non-antibiotic strategies based on pathogenic mechanisms are also urgently needed.

Association of Systemic Inflammation and Overall Survival in Elderly Patients with Cancer Cachexia - Results from a Multicenter Study.

BackgroundSystemic inflammation and cachexia are associated with adverse clinical outcomes in elderly patients with cancer. The survival outcomes of elderly patients with cancer cachexia (EPCC) with high inflammation and a high risk of mortality are unknown. This study aimed to investigate the impact of high inflammation on the prognosis of EPCC patients with high mortality.Patients and MethodsThis multicenter cohort study included 746 EPCC (age >65 years) with a mean age of 72.00 ± 5.24 years, of whom 489 (65.5%) were male. The cut-off value for the inflammation index was obtained using the optimal survival curve. The different inflammatory indicators were assessed using the concordance index (C-index), decision curve analysis (DCA), and prognostic receiver operating characteristic (ROC). The high mortality risk group of EPCC was defined by the 2011 Fearon Cancer Diagnostic Consensus. EPCC were divided into the high-risk group, which satisfies three diagnostic criteria, and a low-risk group, which satisfies only one or two diagnostic criteria.ResultsThe C-index, DCA, and prognostic ROC indicated the superiority of advanced lung cancer inflammation index (ALI) compared with other indicators, including neutrophil–lymphocyte ratio (NLR), prognostic nutritional index (PNI), systemic immune-inflammation index (SII), and platelet–lymphocyte ratio (PLR). Whether ALI was used as a continuous or a categorical variable, ALI had a better prognostic value in EPCC compared with other inflammatory indicators. In particular, patients with low ALI (<25.03) had a worse overall survival (OS) than patients with high ALI (≥25.03) (P < 0.001, HR [95% CI] = 2.092 [1.590–2.751]). The combination effect analysis showed that the risk of mortality of the patients in the low-ALI and high-risk groups was 3.095-fold higher than that of patients in the high-ALI and low-risk groups.ConclusionThe prognostic and discriminative value of the inflammatory indicator ALI was better than that of NLR, PNI, SII, and PLR in EPCC. The high-risk group of EPCC with a low ALI would increase the death risk of OS.

Hepatic Hydrothorax-An Independent Decompensating Event Associated with Long-Term Mortality in Patients with Cirrhosis.

Background: Hepatic hydrothorax (HH) is an understudied complication of decompensated cirrhosis. We aimed to evaluate the long-term prognosis of patients with HH by comparing them with a matched non-HH group. Methods: This retrospective study included 763 consecutive patients hospitalized for decompensated cirrhosis and ascites. Ninety-seven patients with HH were matched for survival analysis with non-HH patients based on liver disease severity. Results: The prevalence of HH was 13.1%. Patients with HH had significantly worse overall liver function. Upon matching, patients with HH had a lower long-term survival (15.4% vs. 30.9% at 5 years) with a mean overall survival of 22.2 ± 2.2 months for the HH group vs. 27.1 ± 2.6 months for the non-HH group (Log Rank–0.05). On multivariate survival analysis using Cox regression, the MELD-Na score, ALBI grade, hepato-renal syndrome, and grade III ascites had a significant impact on mortality in patients with HH. In patients with HH, a MELD-Na score ≥ 16, ALBI grade III, hepato-renal syndrome, or severe ascites delineated high-mortality risk groups. Conclusions: HH is consistently associated with more advanced liver disease. Patients with HH have worse long-term survival, their prognosis being closely intertwined with overlapping decompensating events.

NIH Bayesian Score As a Stratification Tool in Sickle Cell Disease - Results from a Single Center Cohort in Brazil

▪Background: Adults with SCD are notoriously difficult to classify into prognostic index of mortality. Our objective was to test the ability of the NIH Bayesian severity score to predict the risk of death in a cohort of patients with SCD followed at a Sickle Cell Outpatient Clinic of the Clinics Hospital, University of Sao Paulo Medical School. Methods: Since 1992 we have been following adult patients with SCD in our outpatient clinic. Data from patients with a follow-up of at least 2 years was collected. When patients completed 2 years of follow-up, data for applying the NIH Bayesian score, available online (http://bios.ugr.es/dss-calculator/) were acquired. The calculator generates a score ranging from 0 to 1, and patients were followed-up until 2016. Patients with more than three missing information were excluded. Survival analysis was performed by means of Kaplan-Meier curves, divided according to quartiles of the different score levels; the log rank test was used for curve comparison. Results: We analyzed 547 records and 219 were excluded due to missing data, leaving 328 patients for analysis. Mean follow-up period was 6.8 years (ranging from 2 to 25 years). We divided the patients into SS/Sβ0-thalassemia (243 patients) and SC patients (85 patients). During this period, 64 patients died (53 in the SS/Sβ0 group, 11 in the SC group). Patients in the group of the higher score (> 0.72) had higher mortality (p= 0.009; Figure 1) . Among SS/Sβ0 patients that died during follow-up, 58% had a score > 0.72, and among SC patients, only 18.2% had this score. Discussion and conclusions: This is the first external validation of the NIH Bayesian score with mortality evaluation. We were able to identify a high mortality risk group (score> 0.72), especially after 5 years of follow-up. In our study, the performance of the score was better to predict risk of death in the SS/Sβ0 group than in the SC group. This study has limitations due to its retrospective nature and the missing information. There is a need for prospective studies to reinforce these findings, but the NIH Bayesian score model for evaluation of SCD severity may help in treatment decision-making, including the indication of bone marrow transplantation especially in SS/Sβ0-thalassemia patients. [Display omitted] DisclosuresNo relevant conflicts of interest to declare.

Outcomes of Sepsis and Septic Shock in Cancer Patients: Focus on Lactate.

The number of oncological patients (OP) admitted to intensive care units (ICU) for sepsis/septic shock has dramatically increased in recent years. The definition of septic shock has been modified, adding hyperlactatemia as a severity biomarker for mortality. However, it remains poorly reported in septic OP. We performed a retrospective analysis from a prospective database of sepsis/septic shock patients admitted to our ICU between September 2017 and September 2019 and followed until day 90. We identified 251 patients and 31.9% had active oncological comorbidity, mainly solid tumor (81.3%). Septic shock criteria were met for 112 (44.6%). Hyperlactatemia was observed in 136 (54.2%) patients and this was associated with a lower survival rate. Overall 90-day mortality was 15.1%. In OP vs. non-OP, hyperlactatemia was more frequent (65% vs. 49.1%, p = 0.013) and associated with lower survival (65.4% vs. 85.7%, p = 0.046). In OP, poor performance status was also associated with lower survival (HR 7.029 [1.998–24.731], p = 0.002) In an adjusted analysis, cancer was associated with lower 90-day survival (HR 2.690 [1.402–5.160], p = 0.003). In conclusion, septic OP remains a high mortality risk group in whom lactate levels and performance status could help with better risk stratification.

Coronavirus disease 2019: a comprehensive review and meta-analysis on cardiovascular biomarkers.

Preventive cardiology has an important role to play in the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic. The SARS-CoV-2 pandemic has been observed to have a greater mortality impact on subgroups of people in the population who are deemed to be at higher medical disease risk. Individuals with cardiovascular disorders are one such COVID-19-associated high-mortality risk group. Evidence is accumulating that COVID-19 infection may worsen an individual's future cardiovascular health, and, preinfection/postinfection cardiovascular evaluation may be warranted to determine if progressive cardiovascular damage has occurred because of COVID-19 infection. In this study, we conducted a systematic review and meta-analysis, focusing on the association between COVID-19 severity and cardiac-specific biomarkers, including N-terminal pro-B-type natriuretic peptide (NT-proBNP), troponin T (TnT)/troponin I (TnI), lactate dehydrogenase (LDH), creatine kinase, and creatine kinase isoenzyme (CK-MB). TnT had the highest odds ratio or OR (11.83) indicating the greatest association with COVID-19 severity, followed by NT-proBNP (7.57), TnI (6.32), LDH (4.79), D-dimer (4.10), creatine kinase (3.43), and CK-MB (3.35). All of the biomarkers studied were significantly correlated with COVID-19 severity including severe symptoms, ICU care, and mortality (P < 0.0001, except P < 0.01 for CK-MB). COVID-19 infection results in short-term and long-term disease risk that may involve adverse cardiovascular health issues including heart failure. Cardiac-specific biomarkers appear to identify a subset of COVID-19 patients who have the highest risk of an adverse medical outcome. Preventive cardiology has an important role to play in the COVID-19 pandemic.The risk/benefit analysis of maintaining or eliminating the use of the angiotensin receptor blockers (ARB) and angiotensin-converting enzyme inhibitor (ACE-I) medications deserves further investigation.

Prognostic model for patients with advanced cancer using a combination of routine blood test values.

The purpose of this study was to develop a simple prognostic model based on objective indicators alone, i.e., routine blood test data, without using any subjective variables such as patient's symptoms and physician's prediction. The subjects of this retrospective study were patients at the palliative care unit of Tohoku University Hospital, Japan. Eligible patients were over 20years old and had advanced cancer (n = 225). The model for predicting survival was developed based on Cox proportional hazards regression models for univariable and multivariable analyses of 20 items selected from routine blood test data. All the analyses were performed according to the TRIPOD statement ( https://www.tripod-statement.org/ ). The univariable and multivariable regression analyses identified total bilirubin, creatinine, urea/creatinine ratio, aspartate aminotransferase, albumin, total leukocyte count, differential lymphocyte count, and platelet/lymphocyte ratio as significant risk factors for mortality. Based on the hazard ratios, the area under the curve for the new risk model was 0.87 for accuracy, 0.83 for sensitivity, and 0.74 for specificity. Diagnostic accuracy was higher than provided by the Palliative Prognostic Score and the Palliative Prognostic Index. The Kaplan-Meier analysis demonstrated a survival significance of classifying patients according to their score into low-, medium-, and high-mortality risk groups having median survival times of 67days, 34days, and 11days, respectively (p < 0.001). We developed a simple and accurate prognostic model for predicting the survival of patients with advanced cancer based on routine blood test values alone that may be useful for appropriate advanced care planning in a palliative care setting.

DNA methylation biomarker selected by an ensemble machine learning approach predicts mortality risk in an HIV-positive veteran population

ABSTRACT Background: With the improved life expectancy of people living with HIV (PLWH), identifying vulnerable subpopulations at high mortality risk is important. Evidences showed that DNA methylation (DNAm) is associated with mortality in non-HIV populations. Here, we established a panel of DNAm biomarkers that can predict mortality risk among PLWH. Methods: 1,081 HIV-positive participants from the Veterans Ageing Cohort Study (VACS) were divided into training (N = 460), validation (N = 114), and testing (N = 507) sets. VACS index was used as a measure of mortality risk among PLWH. Model training and fine-tuning were conducted using the ensemble method in the training and validation sets and prediction performance was assessed in the testing set. The survival analysis comparing the predicted high and low mortality risk groups and the Gene Ontology enrichment analysis of the predictive CpG sites were performed. Results: We selected a panel of 393 CpGs for the ensemble prediction model that showed excellent performance in predicting high mortality risk with an auROC of 0.809 (95%CI: 0.767,0.851) and a balanced accuracy of 0.653 (95%CI: 0.611, 0.693) in the testing set. The high mortality risk group was significantly associated with 10-year mortality (hazard ratio = 1.79, p = 4E-05) compared with low risk group. These 393 CpGs were located in 280 genes enriched in immune and inflammation response pathways. Conclusions: We identified a panel of DNAm features associated with mortality risk in PLWH. These DNAm features may serve as predictive biomarkers for mortality risk among PLWH. Abbreviations: AUC: Area Under Curve; CI: Confidence interval; DMR: differentially methylated region; DNA: Deoxyribonucleic acid; DNAm: DNA methylation; DAVID: Database for Annotation, Visualization, and Integrated Discovery; EWA: epigenome-wide association; FDR: False discovery rate; FWER: Family-wise error rate; GLMNET: elastic-net-regularized generalized linear models; GO: Gene ontology; HIV: Human immunodeficiency virus; HM450K: Human Methylation 450 K BeadChip; k-NN: k-nearest neighbours; NK: Natural killer; PC: Principal component; PLWH: people living with HIV; QC: Quality control; SVM: Support Vector Machines; VACS: Veterans Ageing Cohort Study; XGBoost: Extreme Gradient Boosting Tree

An 11-lncRNA risk scoring model predicts prognosis of lung squamous cell carcinoma.

The study aims to construct a multi-gene risk scoring model that can be used to predict the prognosis of patients with lung squamous cell carcinoma (LUSC). RNA-seq data from 494 LUSC tumor samples and 49 peripheral lung tissue samples were obtained from TCGA_LUSC database. Differential analysis was conducted using edgeR to screen differentially expressed lncRNAs (DElncRNAs) between LUSC and normal samples. Univariate Cox regression analysis was used to screen lncRNAs that were significantly correlated with LUSC prognosis. LASSO regression model was built to reduce complexity. The LUSC prognostic model based on lncRNAs was established by multivariate Cox regression analysis, which was evaluated by ROC curves and survival analysis. ROC and Kaplan-Meier survival curves of each lncRNA in the model were plotted and compared. 2085 DElncRNAs were identified. Combined with univariate Cox regression analysis, 342 prognosis-related genes were screened. After LASSO regression analysis, 11 lncRNAs tightly related to LUSC prognosis were identified and a risk scoring model was constructed. ROC curve analysis proved the good performance of the model. The Kaplan-Meier survival curve showed that the mortality in high-risk group was significantly higher. The survival analysis results of each lncRNA were also consistent with the prediction in Cox regression. Our results suggested that the 11-lncRNA risk scoring model may provide a new insight for predicting prognosis of LUSC patients.

Deep Learning Predicts Lung Cancer Treatment Response from Serial Medical Imaging.

Tumors are continuously evolving biological systems, and medical imaging is uniquely positioned to monitor changes throughout treatment. Although qualitatively tracking lesions over space and time may be trivial, the development of clinically relevant, automated radiomics methods that incorporate serial imaging data is far more challenging. In this study, we evaluated deep learning networks for predicting clinical outcomes through analyzing time series CT images of patients with locally advanced non-small cell lung cancer (NSCLC).Experimental Design: Dataset A consists of 179 patients with stage III NSCLC treated with definitive chemoradiation, with pretreatment and posttreatment CT images at 1, 3, and 6 months follow-up (581 scans). Models were developed using transfer learning of convolutional neural networks (CNN) with recurrent neural networks (RNN), using single seed-point tumor localization. Pathologic response validation was performed on dataset B, comprising 89 patients with NSCLC treated with chemoradiation and surgery (178 scans). Deep learning models using time series scans were significantly predictive of survival and cancer-specific outcomes (progression, distant metastases, and local-regional recurrence). Model performance was enhanced with each additional follow-up scan into the CNN model (e.g., 2-year overall survival: AUC = 0.74, P < 0.05). The models stratified patients into low and high mortality risk groups, which were significantly associated with overall survival [HR = 6.16; 95% confidence interval (CI), 2.17-17.44; P < 0.001]. The model also significantly predicted pathologic response in dataset B (P = 0.016). We demonstrate that deep learning can integrate imaging scans at multiple timepoints to improve clinical outcome predictions. AI-based noninvasive radiomics biomarkers can have a significant impact in the clinic given their low cost and minimal requirements for human input.

Is syncope a predictor of mortality in acute pulmonary embolism?

Whether syncope as a presenting symptom independently classifies acute pulmonary embolism (APE) into a high mortality risk group remains a matter of controversy.We retrospectively included all consecutive patients admitted to our clinic with APE from January 2014 to December 2016.Our sample consisted of 76 patients with a mean age of 69 ±13.6 years, 64.5% female. 14.3% presented with syncope at admission. In-hospital mortality was 20.8%. Patients with syncope were more likely to require inotropic support (OR = 5.2, 95 % 1.17-23.70, p=0.03) due to the association of cardiogenic shock (OR= 15.95% CI 3.02-74.32, p < 0.001) and systolic blood pressure below 90 mmHg (OR=5.52, 95% CI 1.24-24.47, p=0.03). Patients with syncope had a higher PESI score (150.9 ± 51.1 vs 99.9 ± 30.1, p < 0.001) and a greater in-hospital mortality (OR= 4.5, 95% CI 1.14-17.62, p=0.03). However, multivariate logistic regression equations did not identify syncope as an independent predictor of mortality.In our sample, syncope did not independently reclassify the patient in a higher mortality group, but due to the association with hemodynamic instability, which remains the primary tool in therapeutic decision-making.

Deep learning for lung cancer prognostication: A retrospective multi-cohort radiomics study.

BackgroundNon-small-cell lung cancer (NSCLC) patients often demonstrate varying clinical courses and outcomes, even within the same tumor stage. This study explores deep learning applications in medical imaging allowing for the automated quantification of radiographic characteristics and potentially improving patient stratification.Methods and findingsWe performed an integrative analysis on 7 independent datasets across 5 institutions totaling 1,194 NSCLC patients (age median = 68.3 years [range 32.5–93.3], survival median = 1.7 years [range 0.0–11.7]). Using external validation in computed tomography (CT) data, we identified prognostic signatures using a 3D convolutional neural network (CNN) for patients treated with radiotherapy (n = 771, age median = 68.0 years [range 32.5–93.3], survival median = 1.3 years [range 0.0–11.7]). We then employed a transfer learning approach to achieve the same for surgery patients (n = 391, age median = 69.1 years [range 37.2–88.0], survival median = 3.1 years [range 0.0–8.8]). We found that the CNN predictions were significantly associated with 2-year overall survival from the start of respective treatment for radiotherapy (area under the receiver operating characteristic curve [AUC] = 0.70 [95% CI 0.63–0.78], p < 0.001) and surgery (AUC = 0.71 [95% CI 0.60–0.82], p < 0.001) patients. The CNN was also able to significantly stratify patients into low and high mortality risk groups in both the radiotherapy (p < 0.001) and surgery (p = 0.03) datasets. Additionally, the CNN was found to significantly outperform random forest models built on clinical parameters—including age, sex, and tumor node metastasis stage—as well as demonstrate high robustness against test–retest (intraclass correlation coefficient = 0.91) and inter-reader (Spearman’s rank-order correlation = 0.88) variations. To gain a better understanding of the characteristics captured by the CNN, we identified regions with the most contribution towards predictions and highlighted the importance of tumor-surrounding tissue in patient stratification. We also present preliminary findings on the biological basis of the captured phenotypes as being linked to cell cycle and transcriptional processes. Limitations include the retrospective nature of this study as well as the opaque black box nature of deep learning networks.ConclusionsOur results provide evidence that deep learning networks may be used for mortality risk stratification based on standard-of-care CT images from NSCLC patients. This evidence motivates future research into better deciphering the clinical and biological basis of deep learning networks as well as validation in prospective data.

Diagnosed depression and sociodemographic factors as predictors of mortality in patients with dementia.

Potentially modifiable risk factors for developing dementia have been identified. However, risk factors for increased mortality in patients with diagnosed dementia are not well understood. Identifying factors that influence prognosis would help clinicians plan care and address unmet needs.AimsTo investigate diagnosed depression and sociodemographic factors as predictors of mortality in patients with dementia in UK secondary clinical care services. We conducted a cohort study of patients with a dementia diagnosis in an electronic health records database in a UK National Health Service mental health trust. In 3374 patients with 10 856 person-years of follow-up, comorbid depression was not associated with mortality (adjusted hazard ratio 0.94; 95% CI 0.71-1.24). Single patients had higher mortality than those who were married (adjusted hazard ratio 1.25; 95% CI 1.03-1.50). Patients of Asian ethnicity had lower mortality rates than White British patients (adjusted hazard ratio 0.50; 95% CI 0.34-0.73). Clinically diagnosed depression does not increase mortality in patients with dementia. Patients who are single are a potential high-mortality risk group. Lower mortality rates in Asian patients with dementia that have been reported in the USA also apply in the UK.Declaration of interestNone.

Running does not increase symptoms or structural progression in people with knee osteoarthritis: data from the osteoarthritis initiative.

Higher levels of moderate to vigorous physical activity improve all-cause mortality and cardiovascular events. However, the effect of running, a moderate to vigorous activity, in those with knee osteoarthritis (OA), a common arthritis that occurs with aging, a high-risk group for mortality and cardiovascular events, is unclear. Therefore, we aimed to evaluate the association of self-selected running on OA symptom and structure progression in people with knee OA. This nested cohort study within the Osteoarthritis Initiative (OAI) (2004-2014) included those at least 50years old with OA in at least one knee. Runners were defined using a self-administered questionnaire at the 96-month visit. At baseline and 48-months, symptoms were assessed and radiographs were scored for Kellgren-Lawrence (KL) grade (2-4) and medial Joint Space Narrowing (JSN) score (0-3). We evaluated the association of self-selected running with outcomes: KL worsening, medial JSN worsening, new knee pain, and improved knee pain over 48months, adjusting for baseline age, sex, body mass index (BMI), KL score, contralateral KL score, contralateral knee pain, and injury. If data were not available at the 48-month visit, then they were imputed from the 36-month visit. One thousand two hundred three participants had a mean age of 63.2 (7.9) years, BMI of 29.5 (4.6) kg/m2, 45.3% male, and 11.5% runners. Data from 8% of participants required imputation. Adjusted odds ratios for KL grade worsening and new frequent knee pain were 0.9 (0.6-1.3) and 0.9 (0.6-1.6) respectively. Adjusted odds ratio for frequent knee pain resolution was 1.7 (1.0-2.8). Among individuals 50years old and older with knee OA, self-selected running is associated with improved knee pain and not with worsening knee pain or radiographically defined structural progression. Therefore, self-selected running, which is likely influenced by knee symptoms and may result in lower intensity and shorter duration sessions of exercise, need not be discouraged in people with knee OA.

Severe Fetal Abnormality and Outcomes of Continued Pregnancies: A French Multicenter Retrospective Study.

Objectives To describe a population choosing to continue with their pregnancy despite being eligible to receive a medical termination of pregnancy (TOP). Methods Nine-year retrospective study of data (01/01/2006 to 31/12/2014) from three French prenatal diagnostic centers describing the perinatal outcomes of these pregnancies. Pregnancies were classified according to etiology and severity of its fetal pathology. Several perinatal parameters were described: maternal characteristics, parental prenatal choices and information on the pregnancy and neonatal outcomes. These parameters were classified in function of the severity of fetal pathology according to the classification proposed by Dommergues et al. (Prenatal Diagnosis 30(6):531-539, 2010) Results Overall, 155 pregnancies were continued; 140 have been included in our study. Pregnancy outcomes consisted of four TOPs (2.9%); 20 in utero deaths (14.3%); 110 live births (78.6%) of which 55.4% were still alive at 2 years old as the most recent information; and 6 (4.2%) with unknown outcomes. In 27 cases, perinatal palliative care was requested (an increase of 37% over 9 years). 36.4% of cases were classified as having a high mortality risk; 19.3% with a severe handicap risk; 11.4% with a risk of isolated intellectual disability; and 32.9% with an uncertain prognosis. The parental decisions to choose perinatal palliative care were significantly higher within the high mortality risk group as compared to other severity groups (p < 0.001); this group also had a significantly higher mortality (p < 0.001), with a survival rate of 26.3%. Conclusion Over the study period, in France, there was an increase in continued pregnancies, despite a diagnosis of severe fetal pathology in France. Therefore, it is essential that perinatal professionals are provided with a palliative care framework and training in their approach for this population which is heterogeneous in terms of etiology.

Acute Kidney Injury After Craniotomy Is Associated With Increased Mortality: A Cohort Study.

Acute kidney injury (AKI) is a serious postoperative complication. To determine whether AKI in patients after craniotomy is associated with heightened 30-day mortality. We performed a 2-center, retrospective cohort study of 1656 craniotomy patients who received critical care between 1998 and 2011. The exposure of interest was AKI defined as meeting RIFLE (Risk, Injury, Failure, Loss of Kidney Function, and End-stage Kidney Disease) class risk, injury, and failure criteria, and the primary outcome was 30-day mortality. Adjusted odds ratios were estimated by multivariable logistic regression models with inclusion of covariate terms thought to plausibly interact with both AKI and mortality. Additionally, mortality in craniotomy patients with AKI was analyzed with a risk-adjusted Cox proportional hazards regression model and propensity score matching as a sensitivity analysis. The incidences of RIFLE class risk, injury, and failure were 5.7%, 2.9%, and 1.3%, respectively. The odds of 30-day mortality in patients with RIFLE class risk, injury, or failure fully adjusted were 2.79 (95% confidence interval [CI], 1.76-4.42), 7.65 (95% CI, 4.16-14.07), and 14.41 (95% CI, 5.51-37.64), respectively. Patients with AKI experienced a significantly higher risk of death during follow-up; hazard ratio, 1.82 (95% CI, 1.34-2.46), 3.37 (95% CI, 2.36-4.81), and 5.06 (95% CI, 2.99-8.58), respectively, fully adjusted. In a cohort of propensity score-matched patients, RIFLE class remained a significant predictor of 30-day mortality. Craniotomy patients who suffer postoperative AKI are among a high-risk group for mortality. The severity of AKI after craniotomy is predictive of 30-day mortality. AKI, acute kidney injuryAPACHE II, Acute Physiology and Chronic Health Evaluation IICI, confidence intervalCPT, Current Procedural TerminologyICD-9-CM, International Classification of Diseases, Ninth Revision, Clinical ModificationRIFLE, risk, injury, failure, loss of kidney function, and end-stage kidney diseaseRPDR, Research Patient Data Registry.