A Composite Ranking of Risk Factors for COVID-19 Time-To-Event Data from a Turkish Cohort

Having a complete and reliable list of risk factors from routine laboratory blood test for COVID-19 disease severity and mortality is important for patient care and hospital management. It is common to use meta-analysis to combine analysis results from different studies to make it more reproducible. In this paper, we propose to run multiple analyses on the same set of data to produce a more robust list of risk factors. With our time-to-event survival data, the standard survival analysis were extended in three directions. The first is to extend from tests and corresponding p-values to machine learning and their prediction performance. The second is to extend from single-variable to multiple-variable analysis. The third is to expand from analyzing time-to-decease data with death as the event of interest to analyzing time-to-hospital-release data to treat early recover as a meaningful event as well. Our extension of the type of analyses leads to ten ranking lists. We conclude that 20 out of 30 factors are deemed to be reliably associated to faster-death or faster-recovery. Considering correlation among factors and evidenced by stepwise variable selection in random survival forest, 10~15 factors seem to be able to achieve the optimal prognosis performance. Our final list of risk factors contains calcium, white blood cell and neutrophils count, urea and creatine, d-dimer, red cell distribution widths, age, ferritin, glucose, lactate dehydrogenase, lymphocyte, basophils, anemia related factors (hemoglobin, hematocrit, mean corpuscular hemoglobin concentration), sodium, potassium, eosinophils, and aspartate aminotransferase.

Outcomes of Patients With Hematologic Malignancies Who Received Inpatient Palliative Care Consultation

PURPOSE: Palliative care (PC) plays an established role in improving outcomes in patients with solid tumors, yet these services are underutilized in hematologic malignancies (HMs). We reviewed records of hospitalized patients with active HM to determine associations between PC consultation and length of stay, intensive care unit stay, 30-day readmission, and 6-month mortality compared with those who were not seen by PC. METHODS: We reviewed all oncology admissions at our institution between 2013 and 2019 and included patients with HM actively on treatment, stratified by those seen by PC to controls not seen by PC. Groups were compared using Wilcoxon rank-sum, chi-square, and Fisher's exact tests on the basis of the type and distribution of data. Multiple logistic regression models with stepwise variable selection methods were used to find predictors of outcomes. RESULTS: Three thousand six hundred fifty-four admissions were reviewed, among which 370 unique patients with HM were included. Among these, 102 (28%) patients saw PC, whereas the remaining 268 were controls with similar comorbidities. When compared with controls, PC consultation was associated with a statistically significant reduction in 30-day readmissions (16% v 27%; P = .024), increased length of stay (11.5 v 6 days; P < .001), increased intensive care unit admission (28% v 9%; P < .001), and increased 6-month mortality (67% v 15%; P < .001). These data were confirmed in multivariable models. CONCLUSION: In this retrospective study, more than two thirds of patients with HM did not receive PC consultation despite having similar comorbidities, suggesting that inpatient PC consultation is underutilized in patients with HM, despite the potential for decreased readmission rates.

Association between TSH Values and GFR Levels in Euthyroid Cases with Metabolic Syndrome

Background. Metabolic syndrome (MetS) is associated with the risk of developing chronic kidney disease. Although the negative effects of high thyroid-stimulating hormone (TSH) values on glomerular filtration rate (GFR) levels have been known for years, the negative effects of increased TSH on GFR in euthyroid cases have been reported in recent years. This study was aimed at investigating the association between the effect of increased TSH values and estimated-GFR (eGFR) levels in euthyroid cases with MetS. Methods. For this hospital-based descriptive study, 191 MetS cases (123 females, 68 males) were evaluated. Those whose TSH was not within 0.5–4.5 uIU/mL, eGFR was <40 mL/min/1.73 m2, and/or reported any thyroid/kidney disease were excluded. Partial correlation coefficients were calculated to investigate the relationship between the eGFR values and several other numerical variables while controlling for age and BMI in addition to the adjusted gender effect. Thereafter, the multiple linear regression analysis with a stepwise variable selection approach was used to reveal the independent factors that could affect the logarithmically transformed eGFR. Results. The median age was 52 (19–65) years, the median eGFR was 94.3 (41.3–194) mL/min/1.73 m2, and the median TSH was 1.58 (0.50–4.50) uIU/mL in the whole group. Increased TSH even in the normal range was associated with eGFR after adjusting for age and body mass index (BMI), especially in females. The high age (b = −0.160, p = 0.005 ), high BMI (b = −0.134, p = 0.020 ), high TSH (b = −0.380, p < 0.01 ), and high uric acid (b = −0.348, p < 0.01 ) were found as significant predictors of the eGFR in MetS patients. Conclusion. Independent of age and BMI, elevated TSH even in the euthyroid range showed an association with the eGFR in female MetS cases who had normal kidney functions. This correlation was stronger than the correlations between the eGFR and the MetS diagnostic parameters. These findings need further studies on the issue..

POS0937 TREATMENT SURVIVAL ON BIOLOGICS IN PATIENTS WITH AXIAL SPONDYLOARTHRITIS – DATA FROM MOSCOW UNIFIED ARTHRITIS REGISTRY (MUAR)

Background:The number of new biologics in treatment of axial spondyloarthritis (axSpA) is rapidly increasing. It is important to assess timely their place in the treatment of axSpA, especially with regard to retention on therapy.Objectives:To compare retention on therapy with different biologics in patients with axSpA.Methods:We retrospectively analyzed the data of axSpA patients receiving biologics from the MUAR register. Predictors of retention on therapy were selected by forward stepwise variable selection within Cox regression proportional hazard model. These predictors were considered as confounders when comparing the risks of biologics withdrawal.Results:990 treatment episodes in 640 patients with axSpA were analyzed (non-radiographic axSpA – 4.1%, ankylosing spondylitis - 95.9%). The duration of episodes was 824±920 days. Men were 66,6%, mean age 46,4±11,4.The patients were treated with Adalimumab (ADA) (n= 252 treatment episodes), Golimumab (GOL) (n=82), Infliximab (INF) (n=167), Netakimab (NET) (n=9), Secukinumab (SEC) (n=75), Certolizumab pegol (CER) (n=66), Etanercept (ETA) (n=339).The following predictors of withdrawal risk were identified –1.The total duration of the disease2.The duration of the disease before the onset of biologic treatment3.Gender4.Family history of non-inflammatory spondylopathy (degenerative spinal disease)5.The line of biologic treatment6.The level of educationThe severity of radiographic sacroiliitis and HLA B-27 positivity were not associated with the risk of discontinuation of biologics.The identified predictors were further considered as confounders. Adjusted for confounders, ETA had the lowest treatment withdrawal risk (Figure 1). ADA, GOL, INF, SEC, CER had significantly higher risk of withdrawal compared with ETA (Table 1).Conclusion:Our analysis detected predictors associated with risk of biologics withdrawal in axSpA patients in real clinical practice. There are significant differences between biologics regarding retention on treatment.Table 1.Hazard ratio for treatment withdrawalDrugHazard ratio (Exp B)pADA1.52*0.004GOL2.95*0.000INF2.574*0.000NET3.680*0.073SEC2.133*0.005CER2.922*0.000*- withdrawal risk relative to ETAFigure 1.Picture 1. Treatment withdrawal riskDisclosure of Interests:None declared

Utilizing machine learning dimensionality reduction for risk stratification of chest pain patients in the emergency department

Background Chest pain is among the most common presenting complaints in the emergency department (ED). Swift and accurate risk stratification of chest pain patients in the ED may improve patient outcomes and reduce unnecessary costs. Traditional logistic regression with stepwise variable selection has been used to build risk prediction models for ED chest pain patients. In this study, we aimed to investigate if machine learning dimensionality reduction methods can improve performance in deriving risk stratification models. Methods A retrospective analysis was conducted on the data of patients > 20 years old who presented to the ED of Singapore General Hospital with chest pain between September 2010 and July 2015. Variables used included demographics, medical history, laboratory findings, heart rate variability (HRV), and heart rate n-variability (HRnV) parameters calculated from five to six-minute electrocardiograms (ECGs). The primary outcome was 30-day major adverse cardiac events (MACE), which included death, acute myocardial infarction, and revascularization within 30 days of ED presentation. We used eight machine learning dimensionality reduction methods and logistic regression to create different prediction models. We further excluded cardiac troponin from candidate variables and derived a separate set of models to evaluate the performance of models without using laboratory tests. Receiver operating characteristic (ROC) and calibration analysis was used to compare model performance. Results Seven hundred ninety-five patients were included in the analysis, of which 247 (31%) met the primary outcome of 30-day MACE. Patients with MACE were older and more likely to be male. All eight dimensionality reduction methods achieved comparable performance with the traditional stepwise variable selection; The multidimensional scaling algorithm performed the best with an area under the curve of 0.901. All prediction models generated in this study outperformed several existing clinical scores in ROC analysis. Conclusions Dimensionality reduction models showed marginal value in improving the prediction of 30-day MACE for ED chest pain patients. Moreover, they are black box models, making them difficult to explain and interpret in clinical practice.

Determinants of Malnutrition Among Ever-married Women in Bangladesh

Underweight and overweight problems have serious consequences on the health status of women in Bangladesh. The objective of this study is to find the important factors that may influence a woman for being underweight, overweight and obese. Multinomial logistic regression model is fitted for this purpose. The stepwise variable selection procedure is used to select covariates for the model. Information of ever-married 15,323 non-pregnant women is extracted from Bangladesh Demographic and Health Survey, 2014 data. Seven covariates (region, living place, wealth index, respondent‟s marital status, current working status, education, and current age) are selected finally for the model from the initially considered thirteen variables. The results of the study demonstrate that the women living in Sylhet region, rural area, widowed or divorced, having less education and younger age are more likely to become underweight. Conversely, the women are living in Khulna region, urban area, married, not working, having more than 10 years of schooling and age 35-49 are at higher risk of experiencing overweight or obesity. Thus, the Government of Bangladesh should take proper initiatives to improve underweight and overweight problem of women considering the findings of this study. Dhaka Univ. J. Sci. 69(1): 7-13, 2021 (January)

National Outcomes of Elective Hybrid Arch Debranching with Endograft Exclusion versus Total Arch Replacement Procedures: Analysis of the Society of Thoracic Surgeons Adult Cardiac Surgery Database

Objective Hybrid arch procedures (arch vessel debranching with thoracic endovascular aneurysm repair [TEVAR] coverage of arch pathology) have been presented as an alternative to total arch replacement (TAR). But multicenter-based analyses of these two procedures are needed to benchmark the field and establish areas of improvement. Methods The Society of Thoracic Surgeons (STS) Adult Cardiac Surgery Database from July 2014 to December 2015 was queried for elective TAR and hybrid arch procedures. Demographics and operative characteristics were compared and stepwise variable selection was used to create a risk-set used for adjustment of all multivariable models. Results A total of 1,011 patients met inclusion criteria, 884 underwent TAR, and 127 had hybrid arch procedures. TAR patients were younger (mean age: 62.7 ± 13.3 vs. 66.7 ± 11.9 years; p = 0.001) and had less peripheral vascular disease (34.0 vs. 49.6%; p < 0.001) and preoperative dialysis (1.7 vs. 4.7%; p = 0.026), but similar history of stroke (p = 0.91)/cerebrovascular disease (p = 0.52). TAR patients had more concomitant procedures (60 vs. 34.6%; p < 0.0001). TAR patients had lower mortality (6.7 vs. 12.6%; p = 0.02), stroke (6.9 vs. 15%; p = 0.002), paralysis (1.8 vs. 7.1%; p = 0.002), renal failure (4.6 vs. 8.7%; p = 0.045), and STS morbidity (34.2 vs. 42.5%; p = 0.067). Composite mortality, stroke, and paralysis were significantly lower with TAR (11.5 vs. 25.2%; p = 0.0001). After risk adjustment, analysis showed hybrid arch procedures imparted an increased odds of mortality (odds ratio [OR] = 1.91, p = 0.046), stroke (OR = 2.3, p = 0.005), and composite endpoint of stroke or mortality (OR = 2.31, p = 0.0002). Conclusion TAR remains the gold standard for elective aortic arch pathologies. Despite risk adjustment, hybrid arch procedures were associated with increased risk of mortality and stroke, advocating for careful adoption of these strategies.

Machine Learning Dimensionality Reduction Showed Marginal Performance Benefit Over Stepwise Regression for Risk Stratification of Chest Pain Patients in the Emergency Department

Background: Chest pain is among the most common presenting complaints in the emergency department (ED). Swift and accurate risk stratification of chest pain patients in the ED may improve patient outcomes and reduce unnecessary costs. Traditional logistic regression with stepwise variable selection has been used to build risk prediction models for ED chest pain patients. In this study, we aimed to investigate if machine learning dimensionality reduction methods can achieve superior performance than the stepwise approach in deriving risk stratification models. Methods: A retrospective analysis was conducted on the data of patients >20 years old who presented to the ED of Singapore General Hospital with chest pain between September 2010 and July 2015. Variables used included demographics, medical history, laboratory findings, heart rate variability (HRV), and HRnV parameters calculated from five to six-minute electrocardiograms (ECGs). The primary outcome was 30-day major adverse cardiac events (MACE), which included death, acute myocardial infarction, and revascularization. Candidate variables identified using univariable analysis were then used to generate the stepwise logistic regression model and eight machine learning dimensionality reduction prediction models. A separate set of models was derived by excluding troponin. Receiver operating characteristic (ROC) and calibration analysis was used to compare model performance.Results: 795 patients were included in the analysis, of which 247 (31%) met the primary outcome of 30-day MACE. Patients with MACE were older and more likely to be male. All eight dimensionality reduction methods marginally but non-significantly outperformed stepwise variable selection; The multidimensional scaling algorithm performed the best with an area under the curve (AUC) of 0.901. All HRnV-based models generated in this study outperformed several existing clinical scores in ROC analysis.Conclusions: HRnV-based models using stepwise logistic regression performed better than existing chest pain scores for predicting MACE, with only marginal improvements using machine learning dimensionality reduction. Moreover, traditional stepwise approach benefits from model transparency and interpretability; in comparison, machine learning dimensionality reduction models are black boxes, making them difficult to explain in clinical practice.

Machine learning dimensionality reduction for heart rate n-variability (HRnV) based risk stratification of chest pain patients in the emergency department

BackgroundChest pain is among the most common presenting complaints in the emergency department (ED). Swift and accurate risk stratification of chest pain patients in the ED may improve patient outcomes and reduce unnecessary costs. Traditional logistic regression with stepwise variable selection has been used to build risk prediction models for ED chest pain patients. In this study, we aimed to investigate if machine learning dimensionality reduction methods can achieve superior performance than the stepwise approach in deriving risk stratification models.MethodsA retrospective analysis was conducted on the data of patients >20 years old who presented to the ED of Singapore General Hospital with chest pain between September 2010 and July 2015. Variables used included demographics, medical history, laboratory findings, heart rate variability (HRV), and HRnV parameters calculated from five to six-minute electrocardiograms (ECGs). The primary outcome was 30-day major adverse cardiac events (MACE), which included death, acute myocardial infarction, and revascularization. Candidate variables identified using univariable analysis were then used to generate the stepwise logistic regression model and eight machine learning dimensionality reduction prediction models. A separate set of models was derived by excluding troponin. Receiver operating characteristic (ROC) and calibration analysis was used to compare model performance.Results795 patients were included in the analysis, of which 247 (31%) met the primary outcome of 30-day MACE. Patients with MACE were older and more likely to be male. All eight dimensionality reduction methods marginally but non-significantly outperformed stepwise variable selection; The multidimensional scaling algorithm performed the best with an area under the curve (AUC) of 0.901. All HRnV-based models generated in this study outperformed several existing clinical scores in ROC analysis.ConclusionsHRnV-based models using stepwise logistic regression performed better than existing chest pain scores for predicting MACE, with only marginal improvements using machine learning dimensionality reduction. Moreover, traditional stepwise approach benefits from model transparency and interpretability; in comparison, machine learning dimensionality reduction models are black boxes, making them difficult to explain in clinical practice.