ESTIMATION OF STANDARD ERROR OF REGRESSION EFFECTS IN LATENT REGRESSION MODELS USING BINDER'S LINEARIZATION

Background: Heterotopic ossification (HO) commonly occurs after total hip arthroplasty (THA) and can adversely impact clinical outcomes. The purpose of this study is to propose a more reliable HO grading method that is better predictive of patient-reported outcomes (PROs) after THA than the Brooker classification. Methods: 513 THAs (62 ± 10 years old) were reviewed. The incidence and grade of HO was evaluated using the Brooker grading system and a simplified biplanar classification system (grade 1: ⩾1 cm between bone on both anteroposterior and lateral views, grade 2: <1 cm between bone on either view). The modified Harris Hip Score (mHHS), Forgotten Joint Score (FJS), and visual analogue scale (VAS) for pain were collected at minimum of 2 years after surgery and were compared between HO grades using multiple regression models. Results: The incidence of HO varied by Brooker grade (grade 1, 23.4%; grade 2, 22.4%; grade 3, 7.2%; grade 4, 0%) and biplanar grade (grade 1: 45.6%; grade 2: 7.4%). The biplanar classification demonstrated higher interobserver reliability than the Brooker classification (κ = 0.95 and 0.91, respectively). Brooker grade 3 HO decreased the mHHS by 6.5 (standard error: 2.7) but did not have a significant effect on FJS or VAS. Biplanar grade 2 HO decreased the mHHS by 9.9 (standard error: 2.7), the FJS by 12.9 (standard error: 4.51) and increased the VAS pain score by 0.81 (standard error: 0.35). The Cox test was used to compare the fit of regression models and determined the biplanar classification was a significantly better predictor than the Brooker classification ( p < 0.001). Conclusions: Biplanar grade 2 HO had a significant negative influence on PROs. Contrary to previous literature, these results show clinical significance of non-bridging HO. Compared with the Brooker classification, the biplanar classification has greater interobserver reliability and is more predictive of outcomes after THA.

Download Full-text

Physiological Aldosterone Concentrations Are Associated with Alterations of Lipid Metabolism: Observations from the General Population

International Journal of Endocrinology ◽

10.1155/2018/4128174 ◽

2018 ◽

Vol 2018 ◽

pp. 1-6 ◽

Cited By ~ 6

Author(s):

M. Hannich ◽

H. Wallaschofski ◽

M. Nauck ◽

M. Reincke ◽

C. Adolf ◽

...

Keyword(s):

Lipid Metabolism ◽

General Population ◽

Linear Regression ◽

Standard Error ◽

Regression Models ◽

Adult Population ◽

Density Lipoprotein ◽

Lipoprotein Cholesterol ◽

Inverse Association ◽

Linear Regression Models

Objective. Aldosterone and high-density lipoprotein cholesterol (HDL-C) are involved in many pathophysiological processes that contribute to the development of cardiovascular diseases. Previously, associations between the concentrations of aldosterone and certain components of the lipid metabolism in the peripheral circulation were suggested, but data from the general population is sparse. We therefore aimed to assess the associations between aldosterone and HDL-C, low-density lipoprotein cholesterol (LDL-C), total cholesterol, triglycerides, or non-HDL-C in the general adult population. Methods. Data from 793 men and 938 women aged 25–85 years who participated in the first follow-up of the Study of Health in Pomerania were obtained. The associations of aldosterone with serum lipid concentrations were assessed in multivariable linear regression models adjusted for sex, age, body mass index (BMI), estimated glomerular filtration rate (eGFR), and HbA1c. Results. The linear regression models showed statistically significant positive associations of aldosterone with LDL-C (β-coefficient = 0.022, standard error = 0.010, p=0.03) and non-HDL-C (β-coefficient = 0.023, standard error = 0.009, p=0.01) as well as an inverse association of aldosterone with HDL-C (β-coefficient = −0.022, standard error = 0.011, p=0.04). Conclusions. The present data show that plasma aldosterone is positively associated with LDL-C and non-HDL-C and inversely associated with HDL-C in the general population. Our data thus suggests that aldosterone concentrations within the physiological range may be related to alterations of lipid metabolism.

Download Full-text

PARAMETER RECOVERY AND SUBPOPULATION PROFICIENCY ESTIMATION IN HIERARCHICAL LATENT REGRESSION MODELS

ETS Research Report Series ◽

10.1002/j.2333-8504.2007.tb02069.x ◽

2007 ◽

Vol 2007 (2) ◽

pp. i-39 ◽

Cited By ~ 3

Author(s):

Deping Li ◽

Andreas Oranje ◽

Yanlin Jiang

Keyword(s):

Regression Models ◽

Parameter Recovery ◽

Proficiency Estimation ◽

Latent Regression

Download Full-text

A Convenient Method of Decomposing the Gini Index by Population Subgroups

Journal of Official Statistics ◽

10.2478/jos-2014-0005 ◽

2014 ◽

Vol 30 (1) ◽

pp. 91-105 ◽

Cited By ~ 5

Author(s):

Tomson Ogwang

Keyword(s):

Convenient Method ◽

Standard Error ◽

Regression Models ◽

Gini Index ◽

Artificial Data ◽

Estimated Parameters ◽

Subgroup Decomposition

Abstract We propose a convenient method of estimating the within-group, between-group, and interaction components of the overall traditional Gini index from the estimated parameters of underlying “trick regression models” involving known forms of heteroscedasticity related to income. Two illustrative examples involving both real and artificial data are provided. The issue of appropriate standard error of the subgroup decomposition is also discussed.

Download Full-text

Evaluation of standard error of forecast of runoff volume using linear regression models

Canadian Journal of Civil Engineering ◽

10.1139/l02-055 ◽

2002 ◽

Vol 29 (5) ◽

pp. 635-640 ◽

Cited By ~ 4

Author(s):

Wuben Luo ◽

Eric Weiss

Keyword(s):

Linear Regression ◽

Standard Error ◽

Regression Models ◽

Linear Regression Models ◽

Modeling Error ◽

Reservoir Operations ◽

Microsoft Excel ◽

Runoff Volume ◽

Modeling Errors ◽

Significant Underestimation

Optimizing reservoir operations requires forecasts of seasonal inflow and a good understanding of the associated uncertainties. When forecasting seasonal runoff volume to a reservoir using a linear regression model, hydrologic forecasters typically use the standard error of residuals as the standard error of forecast to give water managers a sense of uncertainties in the forecast. However, this practice accounts for only the random error and ignores the modeling error in the volume forecast, resulting in underestimation of the standard error of the forecast. The underestimation can become significant in extreme runoff years for which reservoir operations tend to be most critical. This paper presents the algorithm for calculating the standard error of forecast, which takes into consideration both random and modeling errors. A simple way of calculating the standard error of forecast using built-in functions in Microsoft Excel is described. An example is used to demonstrate the potentially significant underestimation of the true error of a forecast if modeling error is ignored.Key words: standard error of forecast, residuals, runoff volume forecast, regression analysis.

Download Full-text