Risk Prediction Of Outcome Research Articles

Evaluating individualized treatment effect predictions: A model-based perspective on discrimination and calibration assessment.

In recent years, there has been a growing interest in the prediction of individualized treatment effects. While there is a rapidly growing literature on the development of such models, there is little literature on the evaluation of their performance. In this paper, we aim to facilitate the validation of prediction models for individualized treatment effects. The estimands of interest are defined based on the potential outcomes framework, which facilitates a comparison of existing and novel measures. In particular, we examine existing measures of discrimination for benefit (variations of the c-for-benefit), and propose model-based extensions to the treatment effect setting for discrimination and calibration metrics that have a strong basis in outcome risk prediction. The main focus is on randomized trial data with binary endpoints and on models that provide individualized treatment effect predictions and potential outcome predictions. We use simulated data to provide insight into the characteristics of the examined discrimination and calibration statistics under consideration, and further illustrate all methods in a trial of acute ischemic stroke treatment. The results show that the proposed model-based statistics had the best characteristics in terms of bias and accuracy. While resampling methods adjusted for the optimism of performance estimates in the development data, they had a high variance across replications that limited their accuracy. Therefore, individualized treatment effect models are best validated in independent data. To aid implementation, a software implementation of the proposed methods was made available in R.

Scanning Electron Microscopy of Glucose-Induced Cataract Treated with Garcinia mangostana Linn Pericarp Extract

The elderly is susceptible to infection due to immunosenescence causing high cases of infection with significant deaths due to Coronavirus Disease 2019 (COVID-19) pandemic. Many studies have examined inflammatory parameters as predictors of the outcome of COVID-19 cases, but there has been no comparison of the values of these parameters between the elderly and adult groups. This study aims to explain the relationship between inflammatory parameter values in the elderly group compared to the adult group with the outcome of confirmed COVID-19 patients treated at Ulin Hospital Banjarmasin. An analytical observational cross-sectional design study with the samples was patients with examined Neutrophil-Lymphocyte Ratio (NLR), Platelet-Lymphocyte Ratio (PLR), Erythrocyte Sedimentation Rate (ESR), and C-reactive protein (CRP) when admitted to the COVID-19 isolation room at Ulin Banjarmasin Hospital and was divided into elderly and adult groups. A bivariate correlation test was conducted to assess the relationship between each parameter and continued by finding the cut-off value using the Receiver Operating Characteristic (ROC) curve. Odds Ratio (OR) is sought to find out how much the outcome risk prediction is based on the cut-off value. The relative coefficients of NLR, PLR, CRP, and ESR in the elderly group of 1147 subjects were 0.304; 0.119; 0.378; and 0.071. While the cut-off value of NLR, PLR, CRP, and ESR for the elderly group is 6.77; 257.5; 95.65, and 53.4. Outcomes prediction of the confirmed COVID-19 patients in the elderly group from the most significant are CRP, NLR, and PLR. Meanwhile, ESRs are statistically insignificant. The results showed that there are differences in inflammatory parameter values between the elderly and adult groups to predict the outcome of COVID-19 cases.

Comparison of Inflammation Parameters in the Elderly and Adults to the Outcomes of Confirmed Covid-19 Patients

The elderly is susceptible to infection due to immunosenescence causing high cases of infection with significant deaths due to Coronavirus Disease 2019 (COVID-19) pandemic. Many studies have examined inflammatory parameters as predictors of the outcome of COVID-19 cases, but there has been no comparison of the values of these parameters between the elderly and adult groups. This study aims to explain the relationship between inflammatory parameter values in the elderly group compared to the adult group with the outcome of confirmed COVID-19 patients treated at Ulin Hospital Banjarmasin. An analytical observational cross-sectional design study with the samples was patients with examined Neutrophil-Lymphocyte Ratio (NLR), Platelet-Lymphocyte Ratio (PLR), Erythrocyte Sedimentation Rate (ESR), and C-reactive protein (CRP) when admitted to the COVID-19 isolation room at Ulin Banjarmasin Hospital and was divided into elderly and adult groups. A bivariate correlation test was conducted to assess the relationship between each parameter and continued by finding the cut-off value using the Receiver Operating Characteristic (ROC) curve. Odds Ratio (OR) is sought to find out how much the outcome risk prediction is based on the cut-off value. The relative coefficients of NLR, PLR, CRP, and ESR in the elderly group of 1147 subjects were 0.304; 0.119; 0.378; and 0.071. While the cut-off value of NLR, PLR, CRP, and ESR for the elderly group is 6.77; 257.5; 95.65, and 53.4. Outcomes prediction of the confirmed COVID-19 patients in the elderly group from the most significant are CRP, NLR, and PLR. Meanwhile, ESRs are statistically insignificant. The results showed that there are differences in inflammatory parameter values between the elderly and adult groups to predict the outcome of COVID-19 cases.

Machine learning risk prediction model for acute coronary syndrome and death from use of non-steroidal anti-inflammatory drugs in administrative data

Our aim was to investigate the usefulness of machine learning approaches on linked administrative health data at the population level in predicting older patients’ one-year risk of acute coronary syndrome and death following the use of non-steroidal anti-inflammatory drugs (NSAIDs). Patients from a Western Australian cardiovascular population who were supplied with NSAIDs between 1 Jan 2003 and 31 Dec 2004 were identified from Pharmaceutical Benefits Scheme data. Comorbidities from linked hospital admissions data and medication history were inputs. Admissions for acute coronary syndrome or death within one year from the first supply date were outputs. Machine learning classification methods were used to build models to predict ACS and death. Model performance was measured by the area under the receiver operating characteristic curve (AUC-ROC), sensitivity and specificity. There were 68,889 patients in the NSAIDs cohort with mean age 76 years and 54% were female. 1882 patients were admitted for acute coronary syndrome and 5405 patients died within one year after their first supply of NSAIDs. The multi-layer neural network, gradient boosting machine and support vector machine were applied to build various classification models. The gradient boosting machine achieved the best performance with an average AUC-ROC of 0.72 predicting ACS and 0.84 predicting death. Machine learning models applied to linked administrative data can potentially improve adverse outcome risk prediction. Further investigation of additional data and approaches are required to improve the performance for adverse outcome risk prediction.

Can we predict patient outcome before extracorporeal membrane oxygenation for refractory cardiac arrest?

BackgroundRefractory cardiac arrest resistant to conventional cardiopulmonary resuscitation (C-CPR) has a poor outcome. Although previous reports showed that extracorporeal cardiopulmonary resuscitation (E-CPR) can improve the clinical outcome, there are no clinically applicable predictors of patient outcome that can be used prior to the implementation of E-CPR. We aimed to evaluate the use of clinical factors in patients with refractory cardiac arrest undergoing E-CPR to predict patient outcome in our institution.MethodsThis is a single-center retrospective study. We report 112 patients presenting with refractory cardiac arrest resistant to C-CPR between January 2012 and November 2017. All patients received E-CPR for continued life support when a cardiogenic etiology was presumed. Clinical factors associated with patient outcome were analyzed. Significant pre-ECMO clinical factors were extracted to build a patient outcome risk prediction model.ResultsThe overall survival rate at discharge was 40.2, and 30.4% of patients were discharged with good neurologic function. The six-month survival rate after hospital discharge was 36.6, and 25.9% of patients had good neurologic function 6 months after discharge. We stratified the patients into low-risk (n = 38), medium-risk (n = 47), and high-risk groups (n = 27) according to the TLR score (low-flow Time, cardiac arrest Location, and initial cardiac arrest Rhythm) that we derived from pre-ECMO clinical parameters. Compared with the medium-risk and high-risk groups, the low-risk group had better survival at discharge (65.8% vs. 42.6% vs. 0%, p < 0.0001) and at 6 months (60.5% vs. 38.3% vs. 0%, p = 0.0001). The low-risk group also had a better neurologic outcome at discharge (50% vs. 31.9% vs. 0%, p = 0.0001) and 6 months after discharge (44.7% vs. 25.5% vs. 0%, p = 0.0003) than the medium-risk and high-risk groups.ConclusionsPatients with refractory cardiac arrest receiving E-CPR can be stratified by pre-ECMO clinical factors to predict the clinical outcome. Larger-scale studies are required to validate our observations.

Методика прогнозирования клинических исходов при помощи логистического регрессионного анализа в статистической программе IBM SPSS Statistics 20.0 (часть 1)

We presented a description of the organization of the method of constructing a mathematical model of clinical outcome risk prediction on the identified medical and social factors. Purpose: description of binary logistic regression analysis methodology in medical research. Research methods. Logistics regression is a form of multiple regression whose general purpose is to analyze the relationship between several independent variables and a dependent variable. For binary logistic regression analysis application, the following conditions must be met: the resulting variable must assume two values (favorable clinical outcome/adverse clinical outcome), the independent variables can be both categorical and interval, and must be independent of each other. Results of a research. On the basis of the studied medical and social factors, a mathematical model of the prediction of the probability of developing primary disability among persons who have suffered brain injuries has been developed. The obtained regression equation contains four variables: the patient's age, the nature of the brain injury, the level of formation, the severity of the brain injury. The mathematical model predicts the onset of an adverse clinical outcome with a probability of 87.6 %. The obtained results allow calculating the predicted probability of the studied clinical outcome with the permissible proportion of sensitivity and specificity. Conclusion. The presented technique is of practical importance in terms of improving the knowledge of the beginner researcher in the field of building mathematical models for predicting the probability of the event being studied.

Tim-4 expression increases in ischemic stroke patients and is associated with poor outcome

T cell immunoglobin and mucin domain (Tim)-4 on monocytes is involved in immune regulation. Here, we investigated Tim-4 expression on circulating monocytes and in plasma of ischemic stroke. Tim-4 expression was significantly increased on day 2 and day 5 after stroke. Furthermore, stroke severity was positively correlated with Tim-4 expression on monocytes or in plasma. Increased Tim-4 expression was related to stroke associated with infection (SAI) on day 2. Up-regulated Tim-4 expression on monocytes or in plasma on day 2 was a risk predictor of outcome. Our findings suggest that Tim-4 can act as a prognostic biomarker of ischemic stroke.

Abstract 064: Improving Aortic Valve Replacement Risk Prediction: The Role of Geographically Derived Socioeconomic Factors

Objective: Low socioeconomic status (SES) is associated with worse surgical outcomes. Current risk adjustment models for aortic valve replacement (AVR) surgery do not include (SES), and therefore centers that treat large numbers of low-SES patients may be disadvantaged in hospital outcomes comparisons. This study evaluates whether inclusion of SES improves AVR risk prediction models. Methods: All patients undergoing isolated aortic valve replacement (AVR) at a single institution from 2005-2015 were evaluated. We estimated patients’ SES using census-tract-level data, which are more precise than ZIP-code-level data. We excluded patients (~5%) with addresses that could not be geolocated to census tracts. SES covariates were available for 95% of the study population. SES measures included mean rates of unemployment, poverty, household income, home value, educational attainment, and housing density. The risk scores for mortality, complications and increased length of stay were generated using models published by the Society for Thoracic Surgeons. Univariate models were fitted for each SE covariate with a cut-off of p &lt;0.2 for inclusion in the multivariable models for (a) mortality, (b) any complication, and (c) prolonged length of stay (PLOS) in addition to the expected risk. We evaluated the incremental value of SES covariates using area under the curve (AUC). Results: Amongst the 1,386 patients undergoing AVR included in the study, the overall mortality was 2.8%, any complication rate was 15.1% and PLOS was 9.7%. In univariate models, higher education quartile was associated with decreased mortality (OR 0.96, p = 0.04) and complications (OR 0.97, p &lt;0.01). Poverty was associated with increased length of stay (LOS) (OR 1.02, p =0.02). In the multivariable models, the inclusion of SES covariates increased the area under the curve (AUC) for mortality (0.735 to 0.750, p=0.14), for any complications (0.663 to 0.680, p&lt;0.01), and for PLOS (0.749 to 0.751, p=.12) Conclusions: The inclusion of census-tract-level socioeconomic factors into the STS risk predication models is new and shows potential to improve risk prediction for outcomes following AVR, particularly for predictions of any complications following AVR. With the possibility of reimbursement and institutional ranking based on these outcomes, this study represents an improvement in risk predication model even when limited to census tracts and a single institution’s experience.

Composite risk scores and composite endpoints in the risk prediction of outcomes in anticoagulated patients with atrial fibrillation

Several validated risk stratification schemes for prediction of ischaemic stroke (IS)/thromboembolism (TE) and major bleeding are available for patients with non-valvular atrial fibrillation (NVAF). On the basis for multiple common risk factors for IS/TE and bleeding, it has been suggested that composite risk prediction scores may be more practical and user-friendly than separate scores for bleeding and IS/TE. In a long-term prospective hospital registry of anticoagulated patients with newly diagnosed AF, we compared the predictive value of existing risk prediction scores as well as composite risk scores, and also compared these risk scoring systems using composite endpoints. Endpoint 1 was the simple composite of IS and major bleeds. Endpoint 2 was based on a composite of IS plus intracerebral haemorrhage (ICH). Endpoint 3 was based on weighted coefficients for IS/TE and ICH. Endpoint 4 was a composite of stroke, cardiovascular death, TE and major bleeding. The incremental predictive value of these scores over CHADS2 (as reference) for composite endpoints was assessed using c-statistic, net reclassification improvement (NRI) and integrated discrimination improvement (IDI). Of 8,962 eligible individuals, 3,607 (40.2%) had NVAF and were on OAC at baseline. There were no statistically significant differences between the c-statistics of the various risk scores, compared with the CHADS2 score, regardless of the endpoint. For the various risk scores and various endpoints, NRI and IDI did not show significant improvement (≥1%), compared with the CHADS2 score. In conclusion, composite risk scores did not significantly improve risk prediction of endpoints in patients with NVAF, regardless of how endpoints were defined. This would support individualised prediction of IS/TE and bleeding separately using different separate risk prediction tools, and not the use of composite scores or endpoints for everyday 'real world' clinical practice, to guide decisions on thromboprophylaxis.

Risk Prediction of Outcome following Carotid Endarterectomy

The quoted combined mortality and morbidity following carotid endarterectomy is about 5–7%. In an attempt to identify a subgroup of high risk patients, a review has been undertaken of 404 carotid endarterectomies performed between January 1985 and March 1994. The perioperative mortality rate was 2%, with 3.4% of patients experiencing transient neurological deficits and 4% permanent strokes. Multiple logistic regression analysis was used to estimate the influence on outcome of age, gender, indication for surgery, bilateral internal carotid artery disease, hypertension and smoking. No significant explanators were identified. Copyright © 1996 The International Society for Cardiovasular Surgery.

Risk prediction of outcome following carotid endarterectomy

The quoted combined mortality and morbidity following carotid endarterectomy is about 5–7%. In an attempt to identify a subgroup of high risk patients, a review has been undertaken of 404 carotid endarterectomies performed between January 1985 and March 1994. The perioperative mortality rate was 2%, with 3.4% of patients experiencing transient neurological deficits and 4% permanent strokes. Multiple logistic regression analysis was used to estimate the influence on outcome of age, gender, indication for surgery, bilateral internal carotid artery disease, hypertension and smoking. No significant explanators were identified.