Análisis factorial exploratorio de una escala de gravedad de obstrucción bronquial evaluada en menores de tres años, Chile, 2015 / Exploratory factor analysis of a score for grading the severity of bronchial obstruction evaluated in children younger than 3 years old, Chile, 2015

RESUMEN

Introducción. La escala de Tal modificada es utilizada en Chile para determinar la gravedad de la obstrucción bronquial. Existen informes de su validez; sin embargo, carece de antecedentes del análisis de su estructura como escala para cuantificar de forma significativa la gravedad de la obstrucción bronquial. Objetivo. Identificar el patrón de relación entre los ítems de la escala y la dimensión gravedad de la obstrucción bronquial en una muestra de menores de tres años que consultan en dos centros de salud de la provincia de Concepción, Chile. Método. Análisis factorial exploratorio. Selección no aleatoria y voluntaria de participantes, menores de 36 meses, agosto-noviembre de 2015. Evaluados de manera independiente por médico y kinesiólogo, en tiempo seguido. Se aplican medidas de tendencia central, análisis factorial exploratorio, por subgrupo de menores y mayores de 6 meses, y consistencia interna. Resultados. 204 mediciones a 102 menores. Con una varianza explicada de 39%, los ítems se dirigen a dos conceptos diferentes, un factor que agrupa frecuencia respiratoria, sibilancia y retracción, y otro que solo lo determina cianosis, con alfa de Cronbach de 0, 5. Al extraer cianosis del análisis, muestra matriz monofactorial con 38% de varianza total explicada y consistencia interna de 0, 62. Conclusiones. Por medio de sus ítems, la estructura de la escala no se ajusta adecuadamente para esta muestra. Cianosis se muestra como un ítem disociado en la estructura. Al ajustar el modelo a solo tres ítems, la estabilidad del alfa como medida de consistencia interna aumenta, pero a un valor cuestionable, lo que, sin duda, variará al ser aplicada en otra muestra.

ABSTRACT

Introduction. The modified Tal's score has been used in Chile to determine the severity of bronchial obstruction. Its validity has been reported already; however, its structure as a scale to significantly quantify the severity of bronchial obstruction has not been analyzed. Objective. To identify the relationship pattern among the score's items and the severity of bronchial obstruction in a sample of children younger than 3 years seen in two health care centers in the province of Concepción, Chile. Method. Exploratory factor analysis. Nonrandom, voluntary selection of participants youngerthan36months, August-November 2015. Participants were assessed independently and subsequently by a physician and a physical therapist. The following measurements were applied central tendency, exploratory factor analysis, subgroups of participants younger and older than 6 months, and internal consistency. Results. A total of 204 measurements were done in 102 children. With a 39% explained variance, items incline towards to two different concepts one factor groups respiratory rate, wheezing, and retractions, and the other determines only cyanosis, with a Cronbach's alpha of 0.5. Once cyanosis is removed from the analysis, the original one-factor matrix sample shows a 38% total explained variance and a 0.62 internal consistency. Conclusions. Based on its items, the score's structure does not fit adequately to this sample. Cyanosis appears as a dissociated item within the structure. Once the model is adjusted to only 3 items, the stability of Chronbach's alpha as a measure of internal consistency increases but at an objectionable value, which will undoubtedly vary once it is applied to a different sample.