Search

Simpson’s paradox (SP) is a statistical phenomenon where the association between two variables reverses, disappears, or emerges, after conditioning on a third variable. It has been proposed (by, e.g., Judea Pearl) that SP should be analyzed using the framework of graphical causal models (i.e., causal DAGs) in which SP is diagnosed as a symptom of confounding bias. This paper contends that this confounding-based analysis cannot fully capture SP: there are cases of SP that cannot be explained away in terms of confounding. Previous works have argued that some cases of SP do not require causal analysis at all. Despite being a logically valid counterexample, we argue that this type of cases poses only a limited challenge to Pearl’s analysis of SP. In our view, a more powerful challenge to Pearl comes from cases of SP that do require causal analysis but can arise without confounding. We demonstrate with examples that accidental associations due to genetic drift, the use of inappropriate aggregate variables as causes, and interactions between units (i.e., inter-unit causation) can all give rise to SP of this type. The discussion is also extended to the amalgamation paradox (of which SP is a special form) which can occur due to the use of non-collapsible association measures, in the absence of confounding. [ABSTRACT FROM AUTHOR]

Background: There is substantial evidence to support the use of several methods for preventing deep-vein thrombosis (DVT) following intracerebral hemorrhage (ICH). However, the extent to which these measures are implemented in clinical practice and the factors influencing patients' receipt of preventive measures remain unclear. Therefore, we aimed to evaluate the rate of the early implementation of DVT prophylaxis and the factors associated with its success in patients with ICH. Methods: This study enrolled 49,950 patients with spontaneous ICH from the Chinese Stroke Center Alliance (CSCA) between August 2015 and July 2019. Early DVT prophylaxis implementation was defined as an intervention occurring within 48 h after admission. Univariate and multivariate logistic regression analyses were conducted to identify the rate and factors associated with the implementation of early prophylaxis for DVT in patients with ICH. Results: Among the 49,950 ICH patients, the rate of early DVT prophylaxis implementation was 49.9%, the rate of early mobilization implementation was 29.49%, and that of pharmacological prophylaxis was 2.02%. Factors associated with an increased likelihood of early DVT prophylaxis being administered in the multivariable model included receiving early rehabilitation therapy (odds ratio [OR], 2.531); admission to stroke unit (OR 2.231); admission to intensive care unit (OR 1.975); being located in central (OR 1.879) or eastern regions (OR 1.529); having a history of chronic obstructive pulmonary disease (OR 1.292), ischemic stroke (OR 1.245), coronary heart disease or myocardial infarction (OR 1.2); taking antihypertensive drugs (OR 1.136); and having a higher Glasgow Coma Scale (GCS) score (OR 1.045). Conversely, being male (OR 0.936), being hospitalized in tertiary hospitals (OR 0.778), and having a previous intracranial hemorrhage (OR 0.733) were associated with a lower likelihood of early DVT prophylaxis being administered in patients with ICH. Conclusions: The implementation rate of early DVT prophylaxis among Chinese patients with ICH was subpar, with pharmacological prophylaxis showing the lowest prevalence. Various controllable factors exerted an impact on the implementation of early DVT prophylaxis in this population. [ABSTRACT FROM AUTHOR]

We consider estimation of a functional parameter of a realistically modeled data distribution based on observing independent and identically distributed observations. The highly adaptive lasso estimator of the functional parameter is defined as the minimizer of the empirical risk over a class of cadlag functions with finite sectional variation norm, where the functional parameter is parametrized in terms of such a class of functions. In this article we establish that this HAL estimator yields an asymptotically efficient estimator of any smooth feature of the functional parameter under a global undersmoothing condition. It is formally shown that the L1-restriction in HAL does not obstruct it from solving the score equations along paths that do not enforce this condition. Therefore, from an asymptotic point of view, the only reason for undersmoothing is that the true target function might not be complex so that the HAL-fit leaves out key basis functions that are needed to span the desired efficient influence curve of the smooth target parameter. Nonetheless, in practice undersmoothing appears to be beneficial and a simple targeted method is proposed and practically verified to perform well. We demonstrate our general result HAL-estimator of a treatment-specific mean and of the integrated square density. We also present simulations for these two examples confirming the theory. [ABSTRACT FROM AUTHOR]

Objectives: To identify independent factors for head and neck cancer (HNC) patients with carotid blowout syndrome (CBS) and construct a nomogram to predict risk of CBS preoperatively based on computed tomography angiography (CTA) imaging. Subject and Methods: From January 2010 to July 2020, 73 HNC patients who had surgery in hospitalization and underwent CTA examination for head and neck region were included in this study. Vascular alterations and the relationship between carotid artery (CA) and tumor were evaluated in CTA. Clinical and CTA imaging features were distinguished by logistic regression analysis and used to perform receiver operating curve analysis. Nomogram was created to predict risk of CBS and assessed by concordance index (C‐index) and calibration curve. Results: Three independent risk factors were identified, including radical neck dissection, CA surrounded by tumor, and CA invaded by tumor without clear boundary. Area under curve of the combination of 3 variables was 0.836 (95% CI, 0.72–0.952, p < 0.001). The C‐index of nomogram was 0.84 (95% CI, 0.73–0.94), and the calibration plot showed a good fitting between prediction and observation. Conclusions: We established a useful nomogram based on CTA imaging, which showed a satisfied efficacy for evaluating risk of CBS in HNC patients preoperatively. [ABSTRACT FROM AUTHOR]

The purpose of this study was to investigate the current status of physical restraint of patients in the neurosurgical intensive care unit (NSICU) and analyze the factors influencing this measure using a cross-sectional study design. A total of 312 patients from four tertiary hospitals in NSICU were investigated in Beijing, China. The rate of physical restraint of patients in the NSICU was 42.9%. In 41.8% of cases, nurses performed physical restraint based on experience, and 45.5% of patients had physical restraint-related nursing records. Binary logistic regression analyses revealed that physical restraint was associated with delirium, mild-to-moderate disturbance of consciousness, history of extubation, surgery, and use of sedatives within 24 hour. Analysis of related factors can provide a reference for nurses and managers to improve physical restraint strategies. [ABSTRACT FROM AUTHOR]

The Highly-Adaptive least absolute shrinkage and selection operator (LASSO) Targeted Minimum Loss Estimator (HAL-TMLE) is an efficient plug-in estimator of a pathwise differentiable parameter in a statistical model that at minimal (and possibly only) assumes that the sectional variation norm of the true nuisance functions (i.e., relevant part of data distribution) are finite. It relies on an initial estimator (HAL-MLE) of the nuisance functions by minimizing the empirical risk over the parameter space under the constraint that the sectional variation norm of the candidate functions are bounded by a constant, where this constant can be selected with cross-validation. In this article we establish that the nonparametric bootstrap for the HAL-TMLE, fixing the value of the sectional variation norm at a value larger or equal than the cross-validation selector, provides a consistent method for estimating the normal limit distribution of the HAL-TMLE. In order to optimize the finite sample coverage of the nonparametric bootstrap confidence intervals, we propose a selection method for this sectional variation norm that is based on running the nonparametric bootstrap for all values of the sectional variation norm larger than the one selected by cross-validation, and subsequently determining a value at which the width of the resulting confidence intervals reaches a plateau. We demonstrate our method for 1) nonparametric estimation of the average treatment effect when observing a covariate vector, binary treatment, and outcome, and for 2) nonparametric estimation of the integral of the square of the multivariate density of the data distribution. In addition, we also present simulation results for these two examples demonstrating the excellent finite sample coverage of bootstrap-based confidence intervals. [ABSTRACT FROM AUTHOR]

Researchers in observational survival analysis are interested in not only estimating survival curve nonparametrically but also having statistical inference for the parameter. We consider right‐censored failure time data where we observe n independent and identically distributed observations of a vector random variable consisting of baseline covariates, a binary treatment at baseline, a survival time subject to right censoring, and the censoring indicator. We assume the baseline covariates are allowed to affect the treatment and censoring so that an estimator that ignores covariate information would be inconsistent. The goal is to use these data to estimate the counterfactual average survival curve of the population if all subjects are assigned the same treatment at baseline. Existing observational survival analysis methods do not result in monotone survival curve estimators, which is undesirable and may lose efficiency by not constraining the shape of the estimator using the prior knowledge of the estimand. In this paper, we present a one‐step Targeted Maximum Likelihood Estimator (TMLE) for estimating the counterfactual average survival curve. We show that this new TMLE can be executed via recursion in small local updates. We demonstrate the finite sample performance of this one‐step TMLE in simulations and an application to a monoclonal gammopathy data. [ABSTRACT FROM AUTHOR]

The members of the Snail superfamily of zinc-finger transcription factors, including Snai1 and Snai2, are involved in essential biological processes, such as epithelial-mesenchymal transition (EMT). Although Snai1 has been investigated in a number of cancers, our knowledge on Snai2 and its role(s) in squamous cell carcinoma of oral tongue (SCCOT) is limited. In this study, we confirmed the previous observation that over-expression of Snai2 is a frequent event in SCCOT. We further demonstrated that Snai2 over-expression is associated with lymph node metastasis in two independent SCCOT patient cohorts (total n = 129). Statistical analysis revealed that Snai2 over-expression was correlated with reduced overall survival. Furthermore, over-expression of Snai2 was correlated with reduced E-cadherin expression and enhanced Vimentin expression, suggesting a functional role of Snai2 in EMT. These observations were confirmed in vitro, in which knockdown of Snai2 induced a switch from a mesenchymal-like morphology to an epithelial-like morphology in SCCOT cell lines, and suppressed the cell invasion and migration. In contrast, ectopic transfection of Snai2 led to enhanced cell invasion and migration. Furthermore, Snai2 knockdown attenuated TGFβ1-induced EMT in SCCOT cell lines. Taken together, these data suggest that Snai2 plays major roles in EMT and the progression of SCCOT and may serve as a therapeutic target for patients at risk of metastasis. [ABSTRACT FROM AUTHOR]