Search

BackgroundThe Centers for Medicare & Medicaid Services require that dialysis patients' health-related quality of life be assessed annually. The primary instrument used for this purpose is the Kidney Disease Quality of Life 36-Item Short-Form Survey (KDQOL-36), which includes the SF-12 as its generic core and 3 kidney disease-targeted scales: Burden of Kidney Disease, Symptoms and Problems of Kidney Disease, and Effects of Kidney Disease. Despite its broad use, there has been limited evaluation of KDQOL-36's psychometric properties.Study designSecondary analyses of data collected by the Medical Education Institute to evaluate the reliability and factor structure of the KDQOL-36 scales.Settings & participantsKDQOL-36 responses from 70,786 dialysis patients in 1,381 US dialysis facilities that permitted data analysis were collected from June 1, 2015, through May 31, 2016, as part of routine clinical assessment.Measurements & outcomesWe assessed the KDQOL-36 scales' internal consistency reliability and dialysis facility-level reliability using coefficient alpha and 1-way analysis of variance. We evaluated the KDQOL-36's factor structure using item-to-total scale correlations and confirmatory factor analysis. Construct validity was examined using correlations between SF-12 and KDQOL-36 scales and "known groups" analyses.ResultsEach of the KDQOL-36's kidney disease-targeted scales had acceptable internal consistency reliability (α=0.83-0.85) and facility-level reliability (r=0.75-0.83). Item-scale correlations and a confirmatory factor analysis model evidenced the KDQOL-36's original factor structure. Construct validity was supported by large correlations between the SF-12 Physical Component Summary and Mental Component Summary (r=0.40-0.52) and the KDQOL-36 scale scores, as well as significant differences on the scale scores between patients receiving different types of dialysis, diabetic and nondiabetic patients, and patients who were employed full-time versus not.LimitationsUse of secondary data from a clinical registry.ConclusionsThe study provides support for the reliability and construct validity of the KDQOL-36 scales for assessment of health-related quality of life among dialysis patients.

Asymptotically optimal correlation structure methods with binary data can break down in small samples. A new correlation structure methodology based on a recently developed odds-ratio (OR) approximation to the tetrachoric correlation coefficient is proposed as an alternative to the LPB approach proposed by Lee et al. (1995). Unweighted least squares (ULS) estimation with robust standard errors and generalized least squares (GLS) estimation methods were compared. Confidence intervals and tests for individual model parameters exhibited the best performance using the OR approach with ULS estimation. The goodness-of-fit chi-square test exhibited the best Type I error control using the LPB approach with ULS estimation.

The current study focuses on the relationships among a trauma history, a substance use history, chronic homelessness, and the mediating role of recent emotional distress in predicting drug treatment participation among adult homeless people. We explored the predictors of participation in substance abuse treatment because enrolling and retaining clients in substance abuse treatment programs is always a challenge particularly among homeless people. Participants were 853 homeless adults from Los Angeles, California. Using structural equation models, findings indicated that trauma history, substance use history and chronicity of homelessness were associated, and were significant predictors of greater recent emotional distress. The most notable result was that recent emotional distress predicted less participation in current substance abuse treatment (both formal and self-help) whereas a substance use history alone predicted significantly more participation in treatment. Implications concerning treatment engagement and difficulties in obtaining appropriate dual-diagnosis services for homeless mentally distressed individuals are discussed.

Asymptotically optimal correlation structure methods with binary data can break down in small samples. A new correlation structure methodology based on a recently developed odds-ratio (OR) approximation to the tetrachoric correlation coefficient is proposed as an alternative to the LPB approach proposed by Lee et al. (1995). Unweighted least squares (ULS) estimation with robust standard errors and generalized least squares (GLS) estimation methods were compared. Confidence intervals and tests for individual model parameters exhibited the best performance using the OR approach with ULS estimation. The goodness-of-fit chi-square test exhibited the best Type I error control using the LPB approach with ULS estimation.

Studies among normative samples generally demonstrate a positive impact of marriage on health behaviors and other related attitudes. In this study, we examine the impact of marriage on HIV/AIDS risk behaviors and attitudes among impoverished, highly stressed, homeless couples, many with severe substance abuse problems. A multilevel analysis of 368 high-risk sexually intimate married and unmarried heterosexual couples assessed individual and couple-level effects on social support, substance use problems, HIV/AIDS knowledge, perceived HIV/AIDS risk, needle-sharing, condom use, multiple sex partners, and HIV/AIDS testing. More variance was explained in the protective and risk variables by couple-level latent variable predictors than by individual latent variable predictors, although some gender effects were found (e.g., more alcohol problems among men). The couple-level variable of marriage predicted lower perceived risk, less deviant social support, and fewer sex partners but predicted more needle-sharing.

Content is related to other aspects of writing, but exactly how they are related has remained unclear or has not received sufficient critical attention. Consequently, in most writing assessments, content has been treated as just one among several relatively distinct but equal elements. However, in this study, the authors have quantified these relationships, and as a result can suggest, although this may seem controversial to many readers, that writing assessments may henceforth be simplified to assessments of content in cases where analysis of separate skills is not required. The data are stories written by children in English and scored by human readers or the Coh-Metrix. Three models of content were evaluated by using structural equation modeling. The model in which content is influenced either directly or indirectly by five writing elements--namely, vocabulary diversity, text length, coherence, originality, and grammar--provided a reasonable explanation of the data. The model shows that content can be defined as the sum of these five elements, which together explain content to a substantial extent. Since the substantial portion of content is the combined result of all these elements, we can simply assess content and know that, through their direct and indirect effects, content also represents all five textual elements.

Although coefficient alpha is the most widely used measure of internal consistency, it does not optimally describe the unidimensional internal consistency of a composite. Coefficients based on a one-factor model have been suggested as improved estimators of internal consistency reliability. When the 1-factor model does not fit the data, however, the meaning of such a coefficient is unclear. A new identification condition for factor analytic models is proposed that assures the composite can be modeled with only one common factor. The associated 1-factor reliability is the maximal internal consistency coefficient for a unit-weighted composite. The coefficient also describes k-factor reliability, the greatest lower bound to reliability, and reliability for any composite from a latent variable model with additive errors.

This article describes a Monte Carlo study of 2 methods for treating incomplete non-normal data. Skewed, kurtotic data sets conforming to a single structured model, but varying in sample size, percentage of data missing, and missing-data mechanism, were produced. An asymptotically distribution-free available-case (ADFAC) method and structured-model expectation-maximization (EM) with non-normality corrections were applied to these data sets, and the 2 methods were then compared in terms of bias in parameter estimates, bias in standard-error estimates, efficiency of parameter estimates, and model chi-squares. The results favored the non-normality corrected EM over the ADFAC method in almost all respects, the only important exception involving bias in standard-error estimates with large samples.

The vast majority of structural equation models contain no mean structure, that is, the population means are estimated at the sample means and are then eliminated from modeling consideration. Generalized least squares methods are proposed to estimate potential mean structure parameters and to evaluate whether the given model can be successfully augmented with a mean structure. A simulation evaluates the performance of some alternative tests. A method that takes variability due to estimation of covariance structure parameters into account in the mean structure estimator, as well as in the weight matrix of the generalized least squares function, performs best. In small samples, the F-test and Yuan-Bentler adjusted X2 test perform best.

Following up and correcting prior work by Teel, Bearden, and Sharma (1986) in this journal, a general approach to variance explained in latent dependent variables of nonrecursive linear structural equation models is given. A new method of its estimation, easily implemented in EQS or LISREL, is described and illustrated.