Prospective observational study validating the German version of the Control of Allergic Rhinitis and Asthma Test (CARAT10)

Abstract

The Control of Allergic Rhinitis and Asthma Test (CARAT10), developed by Portuguese experts, is the only questionnaire assessing asthma control under additional consideration of a frequently concurrent allergic rhinitis (AR), providing sub-scores for both diseases. Aims of this study were the translation and validation of its German version. Asthma patients both with and without AR were included at three primary care pulmologists´ practices in Southern Germany. After translation process, patients completed the CARAT10, the Asthma Control Questionnaire (ACQ), the Asthma Control Test (ACT), and the Standardised Asthma Quality of Life Questionnaire (AQLQ(S)). Item and scale characteristics as well as measures of reliability and validity of the CARAT10 were determined. A confirmatory factor analysis was conducted to test factorial validity. Data of 213 patients were analysed, 101 (47%) of them with concurrent AR. Missing responses were minimal and unsystematic. Cronbach´s α was 0.87 for the CARAT10 total score (TS) and 0.84 for each sub-score. Spearman´s correlation coefficients for the association of the CARAT10 TS with ACQ, ACT and AQLQ(S) were moderate to high and slightly higher in patients with AR. Higher correlations were found for its lower airway sub-score than the upper airway sub-score (all around 0.8 to all around 0.3). Confirmatory factor analysis confirmed the two-factorial scale structure of the CARAT10, with a two-factor model showing a better fit to the data than a one-factor model. The German version of the CARAT10 is an acceptable, reliable and valid tool. Our results suggest a recommended use in asthma patients with AR.

Fig. 1

Graph of the confirmatory factor analysis of the two-factor-model—The 10 items of the CARAT10 measure symptoms correlating in an upper airways factor (items 1, 2, 3, 4) and symptoms (items 5, 6, 7), aspects of quality of life (items 8, 9) and use of specific medication (item 10) correlating in a lower airway factor. The broken line arrow indicates correlation between the two factors; continuous arrows indicate factor loads

Fig. 2

Graph of the confirmatory factor analysis of the one-factor-model—The 10 items of the CARAT10 measure symptoms of the upper airways (items 1, 2, 3, 4), symptoms of the lower airways (items 5, 6, 7), aspects of quality of life (items 8, 9) and use of specific medication (item 10) correlating in an one-factor-total score. The continuous arrows indicate factor loads

Fig. 3

Bland-Altman Plots comparing CARAT10 TS and ACT (upper part), CARAT10 TS and ACQ6 (middle), and between ACT and ACQ6 (lower part). All scales are transformed to a scale ranging from 0 (worst symptom load possible) to 1 (no symptom load). Black filled circles represent value pairs from patients with symptomatic AR, grey filled circles from patients with AR currently not active, and light circles from patients without AR. The solid lines indicates the mean difference between the two scales compared, the dashed lines ± 1.96 standard deviations