The strategy for predicting future exacerbation of asthma using a combination of the Asthma Control Test and lung function test

• Published in: The Journal of asthma Vol. 46; no. 7; p. 677
• Main Authors: Sato, Ryuji, Tomita, Katsuyuki, Sano, Hiroyuki, Ichihashi, Hideo, Yamagata, Shigeyoshi, Sano, Akiko, Yamagata, Toshiyuki, Miyara, Takayuki, Iwanaga, Takashi, Muraki, Masato, Tohda, Yuji
• Format: Journal Article
• Language: English
• Published: England 01.01.2009
• Subjects: Aged; Algorithms; Asthma - diagnosis; Asthma - physiopathology; Breath Tests; Disease Progression; Female; Forced Expiratory Volume - physiology; Humans; Male; Middle Aged; Models, Statistical; Nitric Oxide - metabolism; Predictive Value of Tests; Respiratory Function Tests; Retrospective Studies; ROC Curve; Sensitivity and Specificity; Surveys and Questionnaires
• ISSN: 1532-4303
• DOI: 10.1080/02770900902972160

Various factors have been reported to be useful for predicting future exacerbations. This study was intended to determine a usefulness of a combination of a patient-based questionnaire, such as the Asthma Control Test (ACT) score with objective assessments, such as forced expiratory volume in 1 second (FEV(1)) and/or exhaled nitric oxide (FE(NO)), for predicting future exacerbations in adult asthmatics. We therefore enrolled 78 subjects with mild to moderate asthma, who were clinically stable for 3 months who all had been regularly receiving inhaled steroid treatment. All subjects underwent a routine assessment of asthma control including the ACT score, spirometry, and FE(NO), and then were followed up until a severe exacerbation occurred. The predictors of an increased risk of severe exacerbation were identified and validated using decision trees based on a classification and regression tree (CART) analysis. The properties of the developed models were the evaluated with the area under the ROC curve (AUC) (95% confidence interval [CI]). The CART analysis automatically selected the variables and cut-off points, the ACT score <or=23 and FEV(1) <or= 91.8%, with the greatest capacity for discriminating future exacerbations within one year or not. When the probability was calculated by the likelihood ratio of a positive test (LP), the ACT score <or=23 was identified with a 60.3% probability, calculated by 1.82 of LP, whereas the combined ACT score <or=23 and the percentage of predicted FEV(1) <or= 91.8% were identified with an 85.0% probability, calculated by an LP score of 5.43, for predicting future exacerbation. These results demonstrated that combining the ACT score and percentage of predicted FEV(1), but not FE(NO,) can sufficiently stratify the risk for future exacerbations within one year.