Rasch analysis of the Hospital Anxiety andDepression Scale (HADS) in patients with major depression / Análisis de Rasch de la Escala Hospitalaria de Ansiedad y Depresión (Hospital Anxiety and Depression Scale, HADS) en pacientes con depresión mayor

La Hospital Anxiety and Depression Scale (HADS) [Escala Hospitalaria de Ansiedad y Depresión] es una herramienta ampliamente utilizada para detección de la depresión y la ansiedad en pacientes con enfermedades médicas. Este estudio tuvo como objetivo explorar las propiedades psicométricas de la HADS en la detección de pacientes con depresión mayor utilizando el método de la teoría de respuesta al ítem. Un total de 460 pacientes con depresión mayor completaron el HADS. Se utilizó el análisis de Rasch para examinar la unidimensionalidad, el ajuste de los ítems, la dependencia local, la confiabilidad, el orden de las categorías, el funcionamiento diferencial de los ítems (DIF) y la focalización. La HADS mostró una construcción bidimensional. Todos los ítems se ajustaban al modelo de Rasch. Tres pares de ítems mostraron una dependencia local menor pero desconsiderada. Ambas subescalas tuvieron una confiabilidad aceptable. Ninguno de los ítems mostró categorías desordenadas o DIF. Todos los ítems estaban bien dirigidos y los participantes con niveles altos y bajos de angustia fueron menos objetivo que aquellos con niveles moderados de angustia. Finalmente, se generó una tabla de conversión para transformar las puntuaciones brutas en medidas de intervalo. El HADS demostró propiedades psicométricas adecuadas para evaluar la depresión y la ansiedad en pacientes con depresión mayor. Fue más apropiado para evaluar niveles de angustia moderados que altos o bajos. La tabla de conversión se puede utilizar para una medición más precisa. Estos resultados pueden allanar el camino para métodos eficientes y sensibles para analizar la respuesta a los síntomas de depresión en la investigación y en la práctica clínica.(AU)

The Hospital Anxiety and Depression Scale (HADS) is a widely used screening tool for depression and anxiety in patients with medical ill-nesses. This study aimed to explore the psychometric properties of the HADS in screening for patients with major depression using item response theory method.A total of 460 patients with major depression completed the HADS. Rasch analyses were used to examine unidimensionality, item fit, local dependency, reliability, ordering of categories, differential item functioning (DIF)and targeting. The HADS showed a two-dimensional construct.All items fit the Rasch model.Three pairs of items showed mi-norbut inconsiderate local dependency.Both subscales had acceptable re-liability.None of the items displayed disordered categoriesor DIF.All items werewelltargeted, and participants with high and low levels of dis-tress were less targeted than those with moderate levels of distress.Finally, a conversion table to transform the raw scores into interval measures was generated. The HADS demonstrated adequate psychometric properties in assessing depression and anxiety in patients with major depression. It was more appropriate for assessing moderate than high or low levels of dis-tress.The conversion table can be used for more precise measurement.These results may pave the way for efficient and sensitive methods of ana-lyzing depression symptom response in research and in clinical practice.(AU)

Presbyopic correction use and its impact on quality of vision symptoms.

AIM: To assess real-world adoption of presbyopic correction and its impact on quality of vision. METHOD: The use of visual corrections by 529 sequential patients (aged 36 years to 85 years, 50.4% female) attending 4 optometric practices in diverse areas across London were surveyed by interview and completed the quality of vision (QoV) questionnaire to evaluate visual symptoms. RESULTS: Over half of the population (54.7%) managed without glasses at least some of the time, while between 30 and 40% wore distance, reading and progressive spectacles with those using Progressive Addition Lenses wearing them over 80% of the time, while those wearing reading spectacles only approximately 25% of the time. Age, sex and driving frequency had no effect of QoV (p>0.05), whereas the distance of the task significantly impacted QoV (p<0.01). In all QoV metrics, regardless of the far, intermediate or near blur assessment, QoV was rated higher by patients whose main tasks were far focused (n=231, 43.9%), than those who principally conduct intermediate tasks (n=165, 31.4%) and worse still for those whose main tasks were near (n=130, 24.7%), regardless of the form of correction. CONCLUSION: Majority of tasks are in the distance and these had a higher QoV than intermediate tasks with near focused tasks being even worse. It is important to discuss with patients the principal distance of the tasks they generally perform and the forms of presbyopic correction used from the outset.

[Rasch Model in the Validation of the Paediatric Quality of Life Inventory™ 4.0 (PedsQL 4.0™) in Colombian Children and Adolescents]. / Validación por modelo de Rasch del Cuestionario de Calidad de Vida (PedsQL 4.0(®)) en niños y adolescentes colombianos.

INTRODUCTION: The aim of this study was to validate the PedsQL 4.0™ in Colombian children and adolescents using the Rasch model. The Paediatric Quality of Life Inventory (PedsQL 4.0™) has demonstrated to be a reliable and sensitive measurement to changes in health status, as well as being quick and easy to use. METHODS: Validation study of measurement tools. The PedsQL 4.0™ was applied to a convenience sample of 375 children and adolescents between 5 and 17 years old and 500 caregivers of children between 2 and 18 years old in five Colombian cities. The psychometric properties were analysed according to the Rasch model, including adjustment, separation, and differential item functioning (DIF). RESULTS: The Rasch model provided adequate fits to data. The social dimension, for both versions, had greater difficulty than the physical health dimension. Internal consistency for the items was observed, while for individuals, the values of reliability and separation were lower than that established. The DIF occurred in very few variables, especially when comparing cities. The characteristic curves for the items presented disordered thresholds. CONCLUSIONS: The items had adequate internal consistency. Analysis showed adequate individual separation, but disordered thresholds were found in the response categories. No DIF was observed by sex or disease, but it is noteworthy that the DIF occurred between cities.

Validación por modelo de Rasch del Cuestionario de Calidad de Vida (PedsQL 4. 0®) en niños y adolescentes colombianos / Rasch Model in the Validation of the Paediatric Quality of Life Inventory TM 4. 0 (PedsQL 4.0TM) in Colombian Children and Adolescents

Introducción: La escala PedsQL 4.0® se ha probado confiable y sensible a los cambios en el estado de salud, y es de rápida y fácil aplicación. El propósito de este estudio es validar la PedsQL 4.0®a partir de la aplicación del modelo de Rasch en niños y adolescentes colombianos de ambos sexos. Métodos: Se llevó a cabo un estudio observacional de validación de instrumentos de medición. Se realizó un muestreo no probabilístico por conveniencia, conformado por 375 sujetos de 5 a 18 años de edad y 500 cuidadores de menores de 2 a 18 años en cinco ciudades colombianas. Las propiedades psicométricas de la escala se analizaron según el modelo de Rasch, entre ellas el ajuste, la separación y el funcionamiento diferencial del ítem. Resultados: Se encontró un adecuado ajuste de los datos al modelo de Rasch. La dimensión social presentó mayor dificultad que la dimensión de salud física en las dos versiones. Se observó consistencia interna de los ítems, mientras que para las personas, los valores de confiabilidad y separación estuvieron por debajo de lo establecido. El funcionamiento diferencial del ítem ocurrió en unas pocas variables, en especial, al comparar por ciudad. Las curvas características de los ítems presentaron umbrales desordenados. Conclusiones: Los ítems presentaron una consistencia adecuada; el análisis por personas no mostró una separación adecuada; sin embargo, se encontraron umbrales desordenados en las categorías de respuesta. No se presentó funcionamiento diferencial del ítem por sexo o por enfermedad, pero llama la atención que el sí ocurriera entre ciudades.

Introduction: The aim of this study was to validate the PedsQL 4.0TM in Colombian children and adolescents using the Rasch model. The Paediatric Quality of Life Inventory (PedsQL 4.0TM) has demonstrated to be a reliable and sensitive measurement to changes in health status, as well as being quick and easy to use. Methods: Validation study of measurement tools. The PedsQL 4.0TMwas applied to a convenience sample of 375 children and adolescents between 5 and 17 years old and 500 caregivers of children between 2 and 18 years old in five Colombian cities. The psychometric properties were analysed according to the Rasch model, including adjustment, separation, and differential item functioning (DIF). Results: The Rasch model provided adequate fits to data. The social dimension, for both versions, had greater difficulty than the physical health dimension. Internal consistency for the items was observed, while for individuals, the values of reliability and separation were lower than that established. The DIF occurred in very few variables, especially when comparing cities. The characteristic curves for the items presented disordered thresholds. Conclusions: The items had adequate internal consistency. Analysis showed adequate individual separation, but disordered thresholds were found in the response categories. No DIF was observed by sex or disease, but it is noteworthy that the DIF occurred between cities

Análisis de la confiabilidad de los resultados de la Prueba de Diagnóstico Matemática en la Universidad Nacional de Costa Rica utilizando el modelo de Rasch

El objetivo del presente estudio es evidenciar cómo la implementación de la teoría de respuesta a los ítems, en particular el modelo de Rasch, ha logrado mejorar los índices de confiabilidad de la prueba de diagnóstico matemático en la Universidad Nacional de Costa Rica, pasando de un alfa de .51 en el 2010 a uno de .78 en el 2012. El análisis psicométrico permitió brindar recomendación acerca de la construcción de la prueba que redundaron en mejores indicadores y elaboración de ítems para medir los diferentes niveles de habilidad que es el propósito fundamental de una prueba diagnóstica.

Summary. The goal of this study is to show how the implementation of the theory of item response, particularly the Rasch model has improved the reliability indices math test diagnosis at the National University of Costa Rica, from an alpha of 0.51 in 2010 to one of 0, 78 in 2012. The psychometric analysis allowed providing recommendation regarding the construction of the test which resulted in better indicators and preparation of items to measure the different levels of skill that is the fundamental purpose of a diagnostic test.

Funcionamento Diferencial do Item no Alcohol Use Disorders Identification Test / Differential Item Functioning in the Alcohol Use Disorders Identification Test / Funcionamiento Diferencial del Ítem en el Alcohol Use Disorders Identification Test

O objetivo do presente estudo foi investigar a presença de Funcionamento Diferencial do Item (DIF) para o sexo nos itens de uma popular medida de rastreio para a dependência do álcool, o Alcohol Use Disorders Identification Test (AUDIT). Os participantes foram 254 estudantes universitários. As análises de DIF foram realizadas a partir do modelo Partial Credit. Três itens não apresentaram ajuste adequado ao modelo, enquanto outro item apresentou um viés de grande magnitude para o sexo feminino. Especificamente, o itemAUDIT-6 mostrou ser, de fato, um aspecto do beber com maior severidade para indivíduos do sexo feminino do que para aqueles dosexo masculino. Recomenda-se cautela na utilização de escores brutos a partir dos 10 itens originais do instrumento e a exclusão doitem AUDIT-6 ou sua reformulação para a aplicação em indivíduos do sexo feminino.

The aim of the present study was to investigate Differential Item Functioning (DIF) for sex in a popular screening measure of alcohol dependence, the Alcohol Use Disorders Identification Test (AUDIT). Participants were 254 undergraduate students. Raschanalysis-based DIF procedures were employed to evaluate the possibility of sex bias in the AUDIT items. Results showed three items with inadequate fit to the unidimensional Partial Credit model and one item with a large bias against women. Specifically, item AUDIT-6 showed a larger difficulty estimate for women, suggesting a more severe aspect of drinking behavior for these individuals. Recommendations are made with regard to using raw scores based on the sum of the 10 AUDIT items. Dropping out or reformulating the AUDIT-6 item to accommodate sex differences and to reduce errors of individual ability (severity) estimates is strongly advised.

El objetivo del presente estudio fue investigar la presencia de Funcionamiento Diferencial del Ítem (DIF) para el sexo en los ítems de una popular medida de rastreo para la dependencia del alcohol, el Alcohol Use Disorders Identification Test (AUDIT). Participaron 254 estudiantes universitarios. Los análisis de DIF fueron realizados a partir del modelo Partial Credit. Tres ítems no presentaron ajusteade cuado al modelo, mientras otro ítem mostró una tendencia de gran magnitud para el sexo femenino. Específicamente, el ítem AUDIT-6 mostró ser, de hecho, un aspecto del beber con mayor severidad para individuos del sexo femenino que para aquellos del sexo masculino. Se recomienda precaución en el uso de escores brutos a partir de los 10 ítems originales del instrumento y la exclusióndel ítem AUDIT-6 o su reformulación para la aplicación en individuos del sexo femenino.

Funcionamento Diferencial do Item no Alcohol Use Disorders Identification Test / Differential Item Functioning in the Alcohol Use Disorders Identification Test / Funcionamiento Diferencial del Ítem en el Alcohol Use Disorders Identification Test

O objetivo do presente estudo foi investigar a presença de Funcionamento Diferencial do Item (DIF) para o sexo nos itens de uma popular medida de rastreio para a dependência do álcool, o Alcohol Use Disorders Identification Test (AUDIT). Os participantes foram 254 estudantes universitários. As análises de DIF foram realizadas a partir do modelo Partial Credit. Três itens não apresentaram ajuste adequado ao modelo, enquanto outro item apresentou um viés de grande magnitude para o sexo feminino. Especificamente, o itemAUDIT-6 mostrou ser, de fato, um aspecto do beber com maior severidade para indivíduos do sexo feminino do que para aqueles dosexo masculino. Recomenda-se cautela na utilização de escores brutos a partir dos 10 itens originais do instrumento e a exclusão doitem AUDIT-6 ou sua reformulação para a aplicação em indivíduos do sexo feminino.(AU)

The aim of the present study was to investigate Differential Item Functioning (DIF) for sex in a popular screening measure of alcohol dependence, the Alcohol Use Disorders Identification Test (AUDIT). Participants were 254 undergraduate students. Raschanalysis-based DIF procedures were employed to evaluate the possibility of sex bias in the AUDIT items. Results showed three items with inadequate fit to the unidimensional Partial Credit model and one item with a large bias against women. Specifically, item AUDIT-6 showed a larger difficulty estimate for women, suggesting a more severe aspect of drinking behavior for these individuals. Recommendations are made with regard to using raw scores based on the sum of the 10 AUDIT items. Dropping out or reformulating the AUDIT-6 item to accommodate sex differences and to reduce errors of individual ability (severity) estimates is strongly advised.(AU)

El objetivo del presente estudio fue investigar la presencia de Funcionamiento Diferencial del Ítem (DIF) para el sexo en los ítems de una popular medida de rastreo para la dependencia del alcohol, el Alcohol Use Disorders Identification Test (AUDIT). Participaron 254 estudiantes universitarios. Los análisis de DIF fueron realizados a partir del modelo Partial Credit. Tres ítems no presentaron ajusteade cuado al modelo, mientras otro ítem mostró una tendencia de gran magnitud para el sexo femenino. Específicamente, el ítem AUDIT-6 mostró ser, de hecho, un aspecto del beber con mayor severidad para individuos del sexo femenino que para aquellos del sexo masculino. Se recomienda precaución en el uso de escores brutos a partir de los 10 ítems originales del instrumento y la exclusióndel ítem AUDIT-6 o su reformulación para la aplicación en individuos del sexo femenino.(AU)