如何正确运用t检验——两几何均值比较一般差异性t检验及SAS实现

本文目的是介绍几何均数以及两组近似对数正态分布数据几何均数的一般差异性t检验及SAS实现。首先对几何均数的概念、适用条件和计算公式进行介绍;并且将几何均数与统计学中常用的算术均数的特点进行比较;其次,呈现两组药物浓度数据,该数据近似服从对数正态分布,符合几何均数的适用条件,使用SAS的TTEST过程对两组几何均数的一般差异性进行t检验,并对相关结果进行解释和比较;最后,讨论几何均数使用时的注意事项。     

Gaining weight through retirement? Results from the SHARE survey

This paper estimates the causal impact of retirement on the Body Mass Index (BMI) of adults aged 50–69 years old, on the probability of being either overweight or obese and on the probability of being obese. Based on the 2004, 2006 and 2010–2011 waves of the Survey of Health, Ageing and Retirement in Europe (SHARE), our identification strategy exploits variation in European Early Retirement Ages (ERAs) and stepwise increases in ERAs in Austria and Italy between 2004 and 2011 to examine an exogenous shock to retirement behavior. Our results show that retirement induced by discontinuous incentives in early retirement schemes causes a 12-percentage point increase in the probability of being obese among men within a two- to four-year period. We find that the impact of retirement is highly non-linear and mostly affects the right-hand side of the male BMI distribution. Additional results show that this pattern is driven by men retiring from strenuous jobs and by those who were already at risk of obesity. In contrast, no significant results are found among women.     

Another look at returns to birthweight

We revisit the causal effect of birthweight. Because variation in birthweight in developed countries primarily stems from variation in gestational age rather than intrauterine growth restriction, we depart from the widely-used twin fixed-effects estimator and employ an instrumental variable – the diagnosis of placenta previa, which provides exogenous variation in gestation length. We find protective effects of additional birthweight against infant mortality and health capital loss, such as cerebral palsy, but in contrast to sibling and twin studies, no strong evidence for non-health long-run outcomes, such as test scores. We also find that short-run birthweight effects have diminished significantly over the decades.     

Post-contrast 3D T1-weighted TSE MR sequences (SPACE,CUBE, VISTA/BRAINVIEW,isoFSE, 3D MVOX): Technical aspects and clinical applications

Post-contrast three-dimensional T1-weighted imaging of the brain is widely used for a broad range of vascular, inflammatory or tumoral diseases. The variable flip angle 3D TSE sequence is now available from several manufacturers (CUBE, General Electric; SPACE, Siemens; VISTA/BRAINVIEW, Philips; isoFSE, Itachi; 3D MVOX, Canon). Compared to gradient-echo (GRE) techniques, 3D TSE offers the advantages of useful image contrasts and reduction of artifacts from static field inhomogeneity. However, the respective role of 3D TSE and GRE MR sequences remains to be elucidated, particularly in the setting of post-contrast imaging. The purpose of this review was (1) to describe the technical aspects of 3D TSE sequences, (2) to illustrate the main clinical applications of the post-contrast 3D T1-w TSE sequence through clinical cases, (3) to discuss the respective role of post-contrast 3D TSE and GRE imaging in the field of neuroimaging.     

Alcances de la aleatorización mendeliana para el control de confusores no observables en epidemiología

The Mendelian randomization is an epidemiologic method proposed to control for spurious associations in observational studies. These associations are commonly caused by confusion derived from social, environmental, and behavioral factors, which can be difficult to measure. Mendelian randomization is based on the selection of genetic variants that are used as instrumental variables that influence exposure patterns or are associated with an intermediate phenotype of the disease. The present work aims to discuss how to select the appropriate genetic variants as instrumental variables and to present methodological tools to deal with the limitations of this epidemiological method. The use of instrumental variables for modifiable exposures has the potential to mitigate the effects of common limitations, such as confusion, when robust genetic variants are chosen as instrumental variables.     

基于Adaptive-Lasso Logistics模型的方剂学学习动机影响因素分析

目的讨论Adaptive lasso-logistic回归模型在方剂学学习动机的影响因素中的应用。方法基于成都市某大学在校2017级学生进行分层抽样,综合运用方差分析及lasso-logistic回归模型,利用R软件分析学生在方剂学学习过程中关于学习动机的影响因素。结果模型共纳入662名学生,通过交叉验证法,可知Adaptive lasso-logistic回归模型中纳入的变量为:外驱动机、学习信念、考试焦虑、性别、专业、科别。从时间成本和模型简洁度来看,Adaptive-lasso logistic模型优于全变量logistic模型和ridge logistic模型。结论Adaptive-Lasso Logistics回归模型在压缩模型、变量筛选以及预测准确率上更有效,有助于了解学生的学习动机,为教学工作者后续教学安排提供一定的参考意见。     

Mortality reductions from marginal increases in public spending on health

There is limited empirical evidence of the nature of any relationship between health spending and health outcomes in Australia. We address this by estimating the elasticity of health outcomes with respect to public healthcare spending using an instrumental variable (IV) approach to account for endogeneity of healthcare spending to health outcomes. Results suggest that, based on the conditional mean, a 1% increase in public health spending was associated with a 2.2% (p < 0.05) reduction in the number of standardised Years of Life Lost (YLL). Sensitivity analyses and robustness checks supported this conclusion. Further exploration using IV quantile regression indicated that marginal returns on public health spending were significantly greater for areas with poorer health outcomes compared to areas with better health outcomes. On average, marginal increases in public health spending reduce YLL, but areas with poorer health outcomes have the greatest potential to benefit from the same marginal increase in public health spending compared to areas with better health outcomes. Understanding the relationship between health spending and outcomes and how this differs according to baseline health outcomes can help meet dual policy objectives to improve the productivity of the healthcare system and reduce inequity.     

Results of a comparative study analyzing octogenarians with renal cell carcinoma in a competing risk analysis with patients in the seventh decade of life

ObjectivesTo analyze clinicopathological features and survival of surgically treated patients with renal cell carcinoma (RCC)≥80 years of age in comparison with patients between the ages of 60 and 70 years.Materials and methodsThe data for 2,516 patients with a median follow-up of 57 months were retrieved from a multinational database (Collaborative Research on Renal Neoplasms Association [CORONA]), including data for 6,234 consecutive patients with RCC after radical or partial nephrectomy. Comparative analysis of clinicopathological features of 241 octogenarians (3.9% of the database) and 2,275 reference patients between the ages of 60 and 70 years (36.5%) was performed. Multivariable regression analysis adjusted for competing risks was applied to identify the effect of advanced age on cancer-specific mortality (CSM) and other-cause mortality (OCM). Furthermore, instrumental variable analysis was employed to reduce residual confounding by unmeasured parameters.ResultsSignificantly more women were present (50% vs. 40%, P = 0.004), and significantly less often nephron-sparing surgery was performed in octogenarians compared with the reference group (11% vs. 20%, P<0.001). Although median tumor size and stages did not significantly defer, older patients less often had advanced or metastatic disease (N+/M1) (4.6% vs. 9.6%, P = 0.009). On multivariable analysis, higher CSM (hazard ratio = 1.48, P = 0.042) and OCM rates (hazard ratio = 4.32, P<0.001) were detectable in octogenarians (c-indices = 0.85 and 0.72, respectively). Integration of the variable age group in multivariable models significantly increased the predictive accuracy regarding OCM (6%, P<0.001), but not for CSM. Limitations are based on the retrospective study design.ConclusionsOctogenarian patients with RCC significantly differ in clinical features and display significantly higher CSM and OCM rates in comparison with their younger counterparts.