Hernando Torres-Zevallos1, Alfredo G Guerreros2, Juan C Bazo-Alvarez3, German F Alvarado1, Karinna Vazquez4, Jonathan M Feldman5, Thomas Ritz6. 1. Facultad de Salud Pública y Administración "Carlos Vidal Layseca," Universidad Peruana Cayetano Heredia, Lima, Peru. 2. Clinica Internacional, Lima, Peru. 3. Instituto de Investigación, Universidad Católica Los Angeles de Chimbote, Chimbote, Peru. 4. Memorial Rehabilitation Institute, Hollywood, FL, USA. 5. Albert Einstein College of Medicine/Montefiore Medical Center, Ferkauf Graduate School of Psychology/Yeshiva University, Bronx, NY, USA. 6. Southern Methodist University, Dallas, TX, USA.
Abstract
Background: Perceptions of asthma triggers provide important guidance for patients' disease management. A psychometrically valid instrument, the Asthma Trigger Inventory (ATI), is available in English and German language versions, however, a version in Spanish as major world language has been missing.Method: A Spanish-language version of the ATI was evaluated in 339 adult patients with asthma, 223 of these in Peru and 107 in the USA. Principal Component analysis (PCA) with Varimax rotation was used to identify coherent trigger domains across and within samples. Resulting subscales were evaluated for internal consistency. Results: PCA suggested differences in factor structures between sites. Whereas the USA sample largely replicated original factors for animal allergens, pollen allergens, physical activity, air pollution/irritants, infections, and psychology, the initial analysis of the Peru sample suggested substantial overlap of air pollution/irritant, infection, and allergen items. Subsequent analysis of an expanded research form of the ATI for the Peru site culminated in extraction of five factors related to psychology, climate/temperature, combined pollen and animal allergens, physical activity, and infection. Internal consistencies were in an acceptable to excellent range (α = 0.74 to 0.94). Additional free trigger responses confirmed the importance of climate variables for patients in Peru. Psychological triggers were reported by 26% (Peru) and 31% (USA) of patients. Conclusion: Coherent trigger domains are readily identified and measured reliably by a Spanish-language ATI version. However, factor structures vary between samples from different Hispanic/Latino cultural and geographic domains. Culturally adapted versions of this instrument are therefore required for Hispanic/Latino population studies.
Background: Perceptions of asthma triggers provide important guidance for patients' disease management. A psychometrically valid instrument, the Asthma Trigger Inventory (ATI), is available in English and German language versions, however, a version in Spanish as major world language has been missing.Method: A Spanish-language version of the ATI was evaluated in 339 adult patients with asthma, 223 of these in Peru and 107 in the USA. Principal Component analysis (PCA) with Varimax rotation was used to identify coherent trigger domains across and within samples. Resulting subscales were evaluated for internal consistency. Results: PCA suggested differences in factor structures between sites. Whereas the USA sample largely replicated original factors for animal allergens, pollen allergens, physical activity, air pollution/irritants, infections, and psychology, the initial analysis of the Peru sample suggested substantial overlap of air pollution/irritant, infection, and allergen items. Subsequent analysis of an expanded research form of the ATI for the Peru site culminated in extraction of five factors related to psychology, climate/temperature, combined pollen and animal allergens, physical activity, and infection. Internal consistencies were in an acceptable to excellent range (α = 0.74 to 0.94). Additional free trigger responses confirmed the importance of climate variables for patients in Peru. Psychological triggers were reported by 26% (Peru) and 31% (USA) of patients. Conclusion: Coherent trigger domains are readily identified and measured reliably by a Spanish-language ATI version. However, factor structures vary between samples from different Hispanic/Latino cultural and geographic domains. Culturally adapted versions of this instrument are therefore required for Hispanic/Latino population studies.