Retrospective Evaluation of Bayesian Risk Models of LVAD Mortality at a Single Implant Center.
Ontology highlight
ABSTRACT:
SUBMITTER: Lohmueller LC
PROVIDER: S-EPMC6176112 | biostudies-literature | 2018
REPOSITORIES: biostudies-literature
Publications
Retrospective Evaluation of Bayesian Risk Models of LVAD Mortality at a Single Implant Center.
Lohmueller Lisa C LC Kanwar Manreet K MK Bailey Stephen S Murali Srinivas S Antaki James F JF
Frontiers in medicine 20181002
Use of a left ventricular assist device (LVAD) can benefit patients with end stage heart failure, but only with careful patient selection. In this study, previously derived Bayesian network models for predicting LVAD patient mortality at 1, 3, and 12 months post-implant were evaluated on retrospective data from a single implant center. The models performed well at all three time points, with a receiver operating characteristic area under the curve (ROC AUC) of 78, 76, and 75%, respectively. This ...[more]
Similar Datasets
Project description:Modern predictive models require large amounts of data for training and evaluation, absence of which may result in models that are specific to certain locations, populations in them and clinical practices. Yet, best practices for clinical risk prediction models have not yet considered such challenges to generalizability. Here we ask whether population- and group-level performance of mortality prediction models vary significantly when applied to hospitals or geographies different from the ones in which they are developed. Further, what characteristics of the datasets explain the performance variation? In this multi-center cross-sectional study, we analyzed electronic health records from 179 hospitals across the US with 70,126 hospitalizations from 2014 to 2015. Generalization gap, defined as difference between model performance metrics across hospitals, is computed for area under the receiver operating characteristic curve (AUC) and calibration slope. To assess model performance by the race variable, we report differences in false negative rates across groups. Data were also analyzed using a causal discovery algorithm "Fast Causal Inference" that infers paths of causal influence while identifying potential influences associated with unmeasured variables. When transferring models across hospitals, AUC at the test hospital ranged from 0.777 to 0.832 (1st-3rd quartile or IQR; median 0.801); calibration slope from 0.725 to 0.983 (IQR; median 0.853); and disparity in false negative rates from 0.046 to 0.168 (IQR; median 0.092). Distribution of all variable types (demography, vitals, and labs) differed significantly across hospitals and regions. The race variable also mediated differences in the relationship between clinical variables and mortality, by hospital/region. In conclusion, group-level performance should be assessed during generalizability checks to identify potential harms to the groups. Moreover, for developing methods to improve model performance in new environments, a better understanding and documentation of provenance of data and health processes are needed to identify and mitigate sources of variation.
| S-EPMC9931319 | biostudies-literature
Project description:Existing risk assessment tools for patient selection for left ventricular assist devices (LVADs) such as the Destination Therapy Risk Score and HeartMate II Risk Score (HMRS) have limited predictive ability. This study aims to overcome the limitations of traditional statistical methods by performing the first application of Bayesian analysis to the comprehensive Interagency Registry for Mechanically Assisted Circulatory Support dataset and comparing it to HMRS. We retrospectively analyzed 8,050 continuous flow LVAD patients and 226 preimplant variables. We then derived Bayesian models for mortality at each of five time end-points postimplant (30 days, 90 days, 6 month, 1 year, and 2 years), achieving accuracies of 95%, 90%, 90%, 83%, and 78%, Kappa values of 0.43, 0.37, 0.37, 0.45, and 0.43, and area under the receiver operator characteristic (ROC) of 91%, 82%, 82%, 80%, and 81%, respectively. This was in comparison to the HMRS with an ROC of 57% and 60% at 90 days and 1 year, respectively. Preimplant interventions, such as dialysis, ECMO, and ventilators were major contributing risk markers. Bayesian models have the ability to reliably represent the complex causal relations of multiple variables on clinical outcomes. Their potential to develop a reliable risk stratification tool for use in clinical decision making on LVAD patients encourages further investigation.
| S-EPMC4414734 | biostudies-literature
Project description: Not available
| S-EPMC10701355 | biostudies-literature
Project description:(1) Background: There is a marked proportion of spondylodiscitis patients who die during the early stage of the disease despite the applied therapy. This study investigates this early mortality and explores the associated risk factors. (2) Methods: We conducted a retrospective analysis of spondylodiscitis patients treated at our Level I spine center between 1 January 2018 and 31 December 2022. (3) Results: Among 430 patients, 32 (7.4%) died during their hospital stay, with a median time of 28.5 days (range: 2.0-84.0 days). Six of these patients (18.75%) did not undergo surgery due to dire clinical conditions or death prior to scheduled surgery. Identified causes of in-hospital death included multiorgan failure (n = 15), acute bone marrow failure (2), cardiac failure (4), liver failure (2), acute respiratory failure (2), acute renal failure (1), and concomitant oncological disease (1). In a simple logistic regression analysis, advanced age (p = 0.0006), diabetes mellitus (p = 0.0002), previous steroid medication (p = 0.0279), Charlson Comorbidity Index (p < 0.0001), and GFR level at admission (p = 0.0008) were significant risk factors for in-hospital death. In a multiple logistic regression analysis, advanced age (p = 0.0038), diabetes mellitus (p = 0.0002), and previous steroid medication (p = 0.0281) remained significant. (4) Conclusions: Despite immediate treatment, a subset of spondylodiscitis patients experience early mortality. Particular attention should be given to elderly patients and those with diabetes or a history of steroid medication, as they face an elevated risk of a rapidly progressing and fatal disease.
| S-EPMC10707421 | biostudies-literature
Project description:BackgroundSevere influenza, especially influenza pneumonia, causes large numbers of deaths each year. Some patients who develop severe influenza have no known risk factors. In this study we investigated risk factors for mortality of patients with influenza A-related pneumonia who have different basic conditions. We also evaluated the power of pneumonia severity assessment tools in Chinese patients hospitalized with influenza A-related pneumonia. Together, these results could provide a basis for a screening method that has improved ability for the early identification of critical patients who will have poor prognoses in clinical practice.MethodsThis single-center, retrospective case-control study included 152 adult patients with severe influenza over six influenza seasons. Data for diagnoses and demographics, as well clinical data, laboratory findings, treatment methods, 30-day and 60-day outcomes of the patients were collected. Patients who had any of the risk factors for severe influenza were included in the high-risk group, and those that had no known risk factors were included in the low-risk group.ResultsThe PSI, CURB-65 and PIRO-CAP tools all underestimated the mortality rate of patients hospitalized with influenza A-related pneumonia, and this underestimate was more pronounced for low-risk patients. D-dimer (Odds ratio [OR] = 1.052, 95% confidence interval [CI] 1.001-1.106, p = 0.045) and direct bilirubin (OR = 1.143, 95%CI 1.049-1.246, p = 0.002) were independent risk factors for mortality of patients with influenza A-related pneumonia. When used in combination with ferritin and D-dimer, the area under receiver operator characteristic curve (AUCROC) was 0.851 (95%CI 0.780-0.922, p < 0.001), 0.840 (95%CI 0.763-0.916, p < 0.001) and 0.829 (95%CI 0.748-0.911, p < 0.001) for PSI, CURB-65 and PIRO-CAP, respectively, which was higher than that obtained using PSI, CURB-65 and PIRO-CAP alone.ConclusionsThe findings demonstrate that currently used community-acquired pneumonia (CAP) scoring systems could underestimate the risk of influenza A-related pneumonia mortality. D-dimer was shown to be an independent risk factor of mortality for influenza A-related pneumonia in hospitalized patients, and a combination of D-dimer with ferritin could improve the predictive value of PSI, CURB-65 and PIRO-CAP for adverse prognoses of patients with influenza A-related pneumonia.
| S-EPMC9841622 | biostudies-literature
Project description:Tricuspid regurgitation (TR) is a common finding in patients with end stage heart failure referred for implantation of left ventricular assist devices. While functional TR frequently resolves after left ventricular unloading, patients with residual and progressive TR demonstrate increased rates of RV dysfunction and poor survival. Criteria for intervention on the tricuspid valve have focused on the degree of tricuspid annular dilatation and the severity of tricuspid regurgitant volume. The surgical decision making regarding intervention on the tricuspid valve remains obscure and historical cohort data cannot distinguish cause from effect.
| S-EPMC9720263 | biostudies-literature
Project description:BackgroundCardiac resynchronization therapy (CRT) reduces morbidity and mortality among individuals with dyssynchronous systolic heart failure (HF). However, patient outcomes vary, with some at higher risk than others for HF progression and death.ObjectiveTo develop a risk prediction score incorporating variables associated with mortality, left ventricular assist device (LVAD) implant, or heart transplant in recipients of a primary prevention cardiac resynchronization therapy-defibrillator (CRT-D).MethodsWe followed 305 CRT-D patients from the Prospective Observational Study of Implantable Cardioverter-Defibrillators for the composite outcome of all-cause mortality, LVAD implant, or heart transplant soon after device implantation. Serum biomarkers and electrocardiographic and clinical variables were collected at implant. Multivariable analysis using the Cox proportional hazards model with stepwise selection method was used to fit the final model.ResultsAmong 305 patients, 53 experienced the composite endpoint. In multivariable analysis, 5 independent predictors ("HF-CRT") were identified: high-sensitivity C-reactive protein >9.42 ng/L (HR = 2.5 [1.4, 4.5]), New York Heart Association functional class III/IV (HR = 2.3 [1.2, 4.5]), creatinine >1.2 mg/dL (HR = 2.7 [1.4, 5.1]), red blood cell count <4.3 × 10(6)/μL (HR = 2.4 [1.3, 4.7]), and cardiac troponin T >28 ng/L (HR = 2.7 [1.4, 5.2]). One point was attributed to each predictor and 3 score categories were identified. Patients with scores 0-1, 2-3, and 4-5 had a 3-year cumulative event-free survival of 96.8%, 79.7%, and 35.2%, respectively (log-rank, P < .001).ConclusionA simple score combining clinical and readily available biomarker data can risk-stratify CRT patients for HF progression and death. These findings may help identify patients who are in need of closer monitoring or early application of more aggressive circulatory support.
| S-EPMC4656051 | biostudies-literature
Project description:BackgroundPatients infected with severe acute respiratory syndrome coronavirus (SARS-CoV-2) can develop severe illness necessitating intensive care admission. Critically ill patients are susceptible for the development of secondary bacterial infections. Due to a combination of virus- and drug-induced immunosuppression, critically ill patients with corona virus disease 2019 (COVID-19) may even have a higher risk of developing a secondary infection. These secondary infections can aggravate the severity of illness and increase the risk of death. Further research on secondary infections in COVID-19 patients is essential. Therefore, the objective of this study was to investigate the incidence and associated risk factors of secondary bacterial infections and to identify the most common groups of pathogens in critically ill COVID-19 patients.MethodsThis mono-center, retrospective observational cohort study was performed at the intensive care unit (ICU) of the Jessa Hospital, Hasselt, Belgium. All adult COVID-19 patients admitted to the ICU from 13th March 2020 until 17th October 2020, were eligible for inclusion in the study. Data from the resulting 116 patients were prospectively entered into a customized database. The resulting database was retrospectively reviewed to investigate three types of secondary bacterial infections (secondary pneumonia, bloodstream infections of unknown origin, catheter-related sepsis).ResultsOf 94 included patients, 68% acquired at least one of the studied secondary bacterial infections during their ICU stay. Almost two thirds of patients (65.96%, n = 62) acquired a secondary pneumonia, whereas 29.79% (n = 28) acquired a bacteremia of unknown origin and a smaller proportion of patients (14.89%, n = 14) acquired a catheter-related sepsis. Male gender (P = 0.05), diabetes mellitus (P = 0.03) and the cumulative dose of corticosteroids (P = 0.004) were associated with increased risk of secondary bacterial infection. The most common pathogens detected in the cultures of patients with secondary pneumonia were Gram-negative bacilli. Bacteremia of unknown origin and catheter-related sepsis were mostly caused by Gram-positive cocci.ConclusionThis study confirms that the incidence of secondary bacterial infections is very high in critically ill COVID-19 patients. These patients are at highest risk of developing secondary pneumonia. Male gender, a history of diabetes mellitus and the administration of corticosteroids were associated with increased risk of secondary bacterial infection.
| S-EPMC8890021 | biostudies-literature
Project description:BackgroundValve replacement surgeries holds risks of morbidity and mortality.Materials and methodsThe study cohort included 346 patients who underwent different types of valve surgery, excluding redo and Bentall operations. All operations were performed through a median sternotomy using cardiopulmonary bypass.ResultsMean patient age was 51.6 ± 16.1 years, and 51% were male. Approximately 21% had diabetes, and 44.6% were hypertensive. Aortic valve replacement (AVR) was performed in 125 patients (37%), mitral valve replacement (MVR) in 95 (28%), combined AVR and MVR in 42 (13%), AVR plus coronary artery bypass grafting (CABG) in 19 (6%), and MVR plus CABG in 32 (10%). Operative mortality was 5.8% (n = 20). In the bivariate-level analysis, older age, operation type, hypertension, emergency surgery, use of a biological valve in the aortic or mitral position, pump time greater than 120 min, and aortic clamp time greater than 60 min were significant predictors of 30-day mortality. Use of medications stratified by duration (less than or more than a month) was also shown to be a predictor of mortality. Use of angiotensin-converting enzyme inhibitors, digoxin, beta-blockers, statins, and loop diuretics was associated with mortality. Older age, emergency/salvage surgery, use of beta-blockers for less than 1 month preoperatively, and use of a biological valve in the aortic position were significant and independent predictors of 30-day mortality.ConclusionAge, emergency valve surgery, use of a biological valve, use of beta-blockers for less than 1 month before surgery, type of surgery, EF<35%, pump time, and cross clamp time were all found to be independent predictors of mortality in patients undergoing valve surgery. Further prospective multicenter studies may be needed to provide a comprehensive assessment of mortality in patients undergoing valve surgery in Jordan.
| S-EPMC7851328 | biostudies-literature
Project description:ObjectivesWe investigated the impact of anemia based on admission hemoglobin (Hb) level as a prognostic risk factor for severe outcomes in hospitalized patients with coronavirus disease 2019 (COVID-19).MethodsA single-center, retrospective cohort study was conducted from a random sample of 733 adult patients (age ≥ 18 years) obtained from a total of 4356 laboratory confirmed SARS-CoV-2 cases who presented to the Emergency Department of Montefiore Medical Center between March-June 2020. The primary outcome was a composite endpoint of in-hospital severe outcomes of COVID-19. A secondary outcome was in-hospital all-cause mortality.ResultsAmong the 733 patients included in our final analysis, 438 patients (59.8%) presented with anemia. 105 patients (14.3%) had mild, and 333 patients (45.5%) had moderate-severe anemia. Overall, 437 patients (59.6%) had a composite endpoint of severe outcomes. On-admission anemia was an independent risk factor for all-cause mortality, (Odds Ratio 1.52, 95% CI [1.01-2.30], p = 0.046) but not for composite severe outcomes. However, moderate-severe anemia (Hb < 11 g/dL) on admission was independently associated with both severe outcomes (OR1.53, 95% CI [1.05-2.23], p = 0.028) and mortality (OR 1.67, 95% CI [1.09-2.56], p = 0.019) during hospitalization.ConclusionAnemia on admission was independently associated with increased odds of all-cause mortality in patients hospitalized with COVID-19. Furthermore, moderate-severe anemia (Hb <11 g/dL) was an independent risk factor for severe COVID-19 outcomes. Moving forward, COVID-19 patient management and risk stratification may benefit from addressing anemia on admission.
| S-EPMC8007204 | biostudies-literature