Stochastic Curtailment of Questionnaires for Three-Level Classification: Shortening the CES-D for Assessing Low, Moderate, and High Risk of Depression.
ABSTRACT: In clinical assessment, efficient screeners are needed to ensure low respondent burden. In this article, Stochastic Curtailment (SC), a method for efficient computerized testing for classification into two classes for observable outcomes, was extended to three classes. In a post hoc simulation study using the item scores on the Center for Epidemiologic Studies-Depression Scale (CES-D) of a large sample, three versions of SC, SC via Empirical Proportions (SC-EP), SC via Simple Ordinal Regression (SC-SOR), and SC via Multiple Ordinal Regression (SC-MOR) were compared at both respondent burden and classification accuracy. All methods were applied under the regular item order of the CES-D and under an ordering that was optimal in terms of the predictive power of the items. Under the regular item ordering, the three methods were equally accurate, but SC-SOR and SC-MOR needed less items. Under the optimal ordering, additional gains in efficiency were found, but SC-MOR suffered from capitalization on chance substantially. It was concluded that SC-SOR is an efficient and accurate method for clinical screening. Strengths and weaknesses of the methods are discussed.
Project description:BACKGROUND:The Center for Epidemiologic Studies Depression Scale (CES-D) is a measure of depressive symptomatology which is widely used internationally. Though previous attempts were made to shorten the CES-D scale, few have attempted to develop a Computerized Adaptive Test (CAT) version for the CES-D. OBJECTIVE:The aim of this study was to provide evidence on the efficiency and accuracy of the CES-D when administered using CAT using an American sample group. METHODS:We obtained a sample of 2060 responses to the CESD-D from US participants using the myPersonality application. The average age of participants was 26 years (range 19-77). We randomly split the sample into two groups to evaluate and validate the psychometric models. We used evaluation group data (n=1018) to assess dimensionality with both confirmatory factor and Mokken analysis. We conducted further psychometric assessments using item response theory (IRT), including assessments of item and scale fit to Samejima's graded response model (GRM), local dependency and differential item functioning. We subsequently conducted two CAT simulations to evaluate the CES-D CAT using the validation group (n=1042). RESULTS:Initial CFA results indicated a poor fit to the model and Mokken analysis revealed 3 items which did not conform to the same dimension as the rest of the items. We removed the 3 items and fit the remaining 17 items to GRM. We found no evidence of differential item functioning (DIF) between age and gender groups. Estimates of the level of CES-D trait score provided by the simulated CAT algorithm and the original CES-D trait score derived from original scale were correlated highly. The second CAT simulation conducted using real participant data demonstrated higher precision at the higher levels of depression spectrum. CONCLUSIONS:Depression assessments using the CES-D CAT can be more accurate and efficient than those made using the fixed-length assessment.
Project description:BACKGROUND:Previously, we proposed a model for ordinal scale scoring in which individual thresholds for each item constitute a distribution by each item. This lead us to hypothesize that the boundary curves of each depressive symptom score in the distribution of total depressive symptom scores follow a common mathematical model, which is expressed as the product of the frequency of the total depressive symptom scores and the probability of the cumulative distribution function of each item threshold. To verify this hypothesis, we investigated the boundary curves of the distribution of total depressive symptom scores in a general population. METHODS:Data collected from 21,040 subjects who had completed the Center for Epidemiologic Studies Depression Scale (CES-D) questionnaire as part of a national Japanese survey were analyzed. The CES-D consists of 20 items (16 negative items and four positive items). The boundary curves of adjacent item scores in the distribution of total depressive symptom scores for the 16 negative items were analyzed using log-normal scales and curve fitting. RESULTS:The boundary curves of adjacent item scores for a given symptom approximated a common linear pattern on a log normal scale. Curve fitting showed that an exponential fit had a markedly higher coefficient of determination than either linear or quadratic fits. With negative affect items, the gap between the total score curve and boundary curve continuously increased with increasing total depressive symptom scores on a log-normal scale, whereas the boundary curves of positive affect items, which are not considered manifest variables of the latent trait, did not exhibit such increases in this gap. DISCUSSION:The results of the present study support the hypothesis that the boundary curves of each depressive symptom score in the distribution of total depressive symptom scores commonly follow the predicted mathematical model, which was verified to approximate an exponential mathematical pattern.
Project description:Do animals form task-specific representations, or do those representations take a general form that can be applied to qualitatively different tasks? Rhesus monkeys (Macaca mulatta) learned the ordering of stimulus lists using two different serial tasks, in order to test whether prior experience in each task could be transfered to the other, enhancing performance. The simultaneous chaining paradigm delivered rewards only after subjects responded in the correct order to all stimuli displayed on a touch sensitive video monitor. The transitive inference paradigm presented pairs of items and delivered rewards when subjects selected the item with the lower ordinal rank. After learning a list in one paradigm, subjects' knowledge of that list was tested using the other paradigm. Performance was enhanced from the very start of transfer training. Transitive inference performance was characterized by 'symbolic distance effects,' whereby the ordinal distance between stimuli in the implied list ordering was strongly predictive of the probability of a correct response. The patterns of error displayed by subjects in both tasks were best explained by a spatially coded representation of list items, regardless of which task was used to learn the list. Our analysis permits properties of this representation to be investigated without the confound of verbal reasoning.
Project description:The development of tolerance to morphine, one of the most potent analgesics, in the management of chronic pain is a significant clinical problem and its mechanisms are poorly understood. Morphine exerts its pharmacological effects via the ?-opioid receptor (MOR). Tolerance is highly connected to G-protein-coupled receptors (GPCR) phosphorylation and desensitization increase. Because morphine desensitization previously has been shown to be MOR phosphorylation- and ß-arrestin2-independent (in contrast to agonists such as fentanyl), we examined the contribution of phosphorylation of the entire C-terminus to the development of antinociceptive tolerance to the partial (morphine) and full (fentanyl) MOR agonists in vivo. In MOR knockout (MORKO) mice, we delivered via lentivirus the genes encoding the wild-type MOR (WTMOR) or a phosphorylation-deficient MOR (Cterm(-S/T)MOR) in which all of the serine and threonine residues were mutated to alanine into the ventrolateral periaqueductal grey matter (vlPAG) or lumbar spinal cord (SC), structures that are involved in nociception. We compared the analgesic ED50 in WTMOR- and Cterm(-S/T)MOR-expressing MORKO mice before and after morphine or fentanyl tolerance was induced. Morphine acute antinociception was partially restored in WTMOR- or Cterm(-S/T)MOR-transferred MORKO mice. Fentanyl acute antinociception was observed only in MORKO mice with the transgenes expressed in the SC. Morphine antinociceptive tolerance was not affected by expressing Cterm(-S/T)MOR in the vlPAG or SC of MORKO mice. Fentanyl-induced tolerance in MORKO mice expressing WTMOR or Cterm(-S/T)MOR, is greater than morphine-induced tolerance. Thus, MOR C-terminus phosphorylation does not appear to be critical for morphine tolerance in vivo.
Project description:BACKGROUND:As the worldwide prevalence of chronic illness increases so too does the demand for novel treatments to improve chronic illness care. Quantifying improvement in chronic illness care from the patient perspective relies on the use of validated patient-reported outcome measures. In this analysis we examine the psychometric and scaling properties of the Patient Assessment of Chronic Illness Care (PACIC) questionnaire for use in the United Kingdom by applying scale data to the non-parametric Mokken double monotonicity model. METHODS:Data from 1849 patients with long-term conditions in the UK who completed the 20-item PACIC were analysed using Mokken analysis. A three-stage analysis examined the questionnaire's scalability, monotonicity and item ordering. An automated item selection procedure was used to assess the factor structure of the scale. Analysis was conducted in an 'evaluation' dataset (n?=?956) and results were confirmed using an independent 'validation' (n?=?890) dataset. RESULTS:Automated item selection procedures suggested that the 20 items represented a single underlying trait representing "patient assessment of chronic illness care": this contrasts with the multiple domains originally proposed. Six items violated invariant item ordering and were removed. The final 13-item scale had no further issues in either the evaluation or validation samples, including excellent scalability (Ho?=?.50) and reliability (Rho?=?.88). CONCLUSIONS:Following some modification, the 13-items of the PACIC were successfully fitted to the non-parametric Mokken model. These items have psychometrically robust and produce a single ordinal summary score. This score will be useful for clinicians or researchers to assess the quality of chronic illness care from the patient's perspective.
Project description:The purpose of this study was to construct a computerized adaptive test (CAT) for measuring self-care performance (the CAT-SC) in children with developmental disabilities (DD) aged from 6 months to 12 years in a content-inclusive, precise, and efficient fashion.The study was divided into 3 phases: (1) item bank development, (2) item testing, and (3) a simulation study to determine the stopping rules for the administration of the CAT-SC. A total of 215 caregivers of children with DD were interviewed with the 73-item CAT-SC item bank. An item response theory model was adopted for examining the construct validity to estimate item parameters after investigation of the unidimensionality, equality of slope parameters, item fitness, and differential item functioning (DIF). In the last phase, the reliability and concurrent validity of the CAT-SC were evaluated.The final CAT-SC item bank contained 56 items. The stopping rules suggested were (a) reliability coefficient greater than 0.9 or (b) 14 items administered. The results of simulation also showed that 85% of the estimated self-care performance scores would reach a reliability higher than 0.9 with a mean test length of 8.5 items, and the mean reliability for the rest was 0.86. Administering the CAT-SC could reduce the number of items administered by 75% to 84%. In addition, self-care performances estimated by the CAT-SC and the full item bank were very similar to each other (Pearson r = 0.98).The newly developed CAT-SC can efficiently measure self-care performance in children with DD whose performances are comparable to those of TD children aged from 6 months to 12 years as precisely as the whole item bank. The item bank of the CAT-SC has good reliability and a unidimensional self-care construct, and the CAT can estimate self-care performance with less than 25% of the items in the item bank. Therefore, the CAT-SC could be useful for measuring self-care performance in children with DD in clinical and research settings.
Project description:Short versions of the Beck Hopelessness Scale have all been created according the Classical Test Theory, but the use and the application of this theory has been repeatedly criticized. In the current study, the Item Response Theory approach was employed to refine and shorten the BHS in order to build a reasonably coherent unidimensional scale whose items/symptoms can be treated as ordinal indicators of the theoretical concept of hopelessness, scaled along a single continuum. In a sample of 492 psychiatrically hospitalized, adult patients (51.2% females), predominantly with a diagnosis of Bipolar Disorder type II, the BHS was submitted to Mokken Scale Analysis. A final set of the nine best-fitting items satisfied the assumptions of local independency, monotonicity, and invariance of the item ordering. Using the ROC curve method, the IRT-based 9-item BHS showed good discriminant validity in categorizing psychiatric inpatients with high/medium suicidal risk and patients with and without suicide attempts. With high sensitivity (>.90), this newly developed scale could be used as a valid screening tool for suicidal risk assessment in psychiatric inpatients.
Project description:Production of reactive oxygen species (ROS) can be accelerated under various biotic and abiotic stresses causing lipid peroxidation, protein degradation, enzyme inactivation, and DNA damage. Superoxide reductase (SOR) is a novel antioxidant enzyme from Pyrococcus furiosus and is employed by this anaerobic hyperthermophilic archaeon for efficient detoxification of ROS. In this study, SOR was introduced into a flowering plant Cornus canadensis to enhance its heat tolerance and reduce heat induced damage. A fusion construct of the SOR gene and Green Fluorescent Protein gene (GFP) was introduced into C. canadensis using Agrobacterium-mediated transformation. Heat tolerance of the GFP-SOR expressing transgenic plants was investigated by observing morphological symptoms of heat injury and by examining changes in photosynthesis, malondialdehyde (MDA), and proline levels in the plants. Our results indicate that the expression of the P. furiosus SOR gene in the transgenic plants alleviated lipid peroxidation of cell membranes and photoinhibition of PS II, and decreased the accumulation of proline at 40°C. After a series of exposures to increasing temperatures, the SOR transgenic plants remained healthy and green whereas most of the non-transgenic plants dried up and were unable to recover. While it had previously been reported that expression of SOR in Arabidopsis enhanced heat tolerance, this is the first report of the successful demonstration of improved heat tolerance in a non-model plant resulting from the introduction of P. furiosus SOR. The study demonstrates the potential of SOR for crop improvement and that inherent limitations of plant heat tolerance can be ameliorated with P. furiosus SOR.
Project description:To establish the reliability and validity of a shortened (10-item) depression scale used among HIV-positive patients enrolled in the Drug Treatment Program in British Columbia, Canada.The 10-item CES-D (Center for Epidemiologic Studies Depression Scale) was examined among 563 participants who initiated antiretroviral therapy (ART) between August 1, 1996 and June 30, 2002. Internal consistency of the scale was measured by Cronbach's alpha. Using the original CES-D 20 as primary criteria, comparisons were made using the Kappa statistic. Predictive accuracy of CES-D 10 was assessed by calculating sensitivity, specificity, positive predictive values and negative predictive values. Factor analysis was also performed to determine if the CES-D 10 contained the same factors of positive and negative affect found in the original development of the CES-D.The correlation between the original and the shortened scale is very high (Spearman correlation coefficient =0.97 (P<0.001). Internal consistency reliability coefficients of the CES-D 10 were satisfactory (Cronbach α=0.88). The CES-D 10 showed comparable accuracy to the original CES-D 20 in classifying participants with depressive symptoms (Kappa=0.82, P<0.001). Sensitivity of CES-D 10 was 91%; specificity was 92%; and positive predictive value was 92%. Factor analysis demonstrates that CES-D 10 contains the same underlying factors of positive and negative affect found in the original development of the CES-D 20.The 10-item CES-D is a comparable tool to measure depressive symptoms among HIV-positive research participants.
Project description:BACKGROUND: Experimental designs constitute a vital component of all Stated Choice (aka discrete choice experiment) studies. However, there exists limited empirical evaluation of the statistical benefits of Stated Choice (SC) experimental designs that employ non-zero prior estimates in constructing non-orthogonal constrained designs. This paper statistically compares the performance of contrasting SC experimental designs. In so doing, the effect of respondent literacy on patterns of Attribute non-Attendance (ANA) across fractional factorial orthogonal and efficient designs is also evaluated. The study uses a 'real' SC design to model consumer choice of primary health care providers in rural north India. A total of 623 respondents were sampled across four villages in Uttar Pradesh, India. METHODS: Comparison of orthogonal and efficient SC experimental designs is based on several measures. Appropriate comparison of each design's respective efficiency measure is made using D-error results. Standardised Akaike Information Criteria are compared between designs and across recall periods. Comparisons control for stated and inferred ANA. Coefficient and standard error estimates are also compared. RESULTS: The added complexity of the efficient SC design, theorised elsewhere, is reflected in higher estimated amounts of ANA among illiterate respondents. However, controlling for ANA using stated and inferred methods consistently shows that the efficient design performs statistically better. Modelling SC data from the orthogonal and efficient design shows that model-fit of the efficient design outperform the orthogonal design when using a 14-day recall period. The performance of the orthogonal design, with respect to standardised AIC model-fit, is better when longer recall periods of 30-days, 6-months and 12-months are used. CONCLUSIONS: The effect of the efficient design's cognitive demand is apparent among literate and illiterate respondents, although, more pronounced among illiterate respondents. This study empirically confirms that relaxing the orthogonality constraint of SC experimental designs increases the information collected in choice tasks, subject to the accuracy of the non-zero priors in the design and the correct specification of a 'real' SC recall period.