Page 147 - ADULT-ONSET ASTHMA PREDICTORS OF CLINICAL COURSE AND SEVERITY

PREDICTORS OF FREQUENT EXACERBATIONS IN (EX)SMOKING AND NEVER SMOKING ADULTS WITH SEVERE ASTHMA
ASSESSMENTS
Clinical parameters - Systema c medical history was taken with regard to asthma symptoms, including asthma-speci c ques onnaires (Asthma Control Ques onnaire (ACQ)16 and Asthma Quality of Life Ques onnaire (AQLQ)),17 medica on use and asthma related healthcare consump on in the previous year (number of courses oral cor costeroids). Co-morbidi es with possible in uence on asthma symptoms were recorded; gastro-esophageal re ux disease, chronic rhinosinusi s (based on symptoms and combined with sinus CT-scan or nasal endoscopy if available).
Func onal parameters - The following lung func on measurements were performed according to interna onal standards: spirometry (prebronchodilator and postbronchodilator FEV1 and forced vital capacity (FVC)),18 single breath carbon monoxide di using capacity of the lung (TLCOc/VA),19 sta c lung volumes by bodyplethysmography (total lung capacity (TLC) and residual volume (RV))20 and airway hyperresponsiveness to methacholine (provoca ve concentra on causing a 20% drop in FEV1 (PC20)).21
In ammatory markers - Frac on of exhaled nitric oxide (FeNO) was measured with a portable rapidresponse chemoluminescent analyzer (NIOX system, Aerocrine, Sweden).22 Venous blood was collected and di eren al white blood cell counts, total and speci c IgE to common allergens (ImmunoCAP, Thermo Scien c, Uppsala, Sweden) measurements were performed. Atopy was de ned as one or more speci c IgE levels above 0.35 kU/L. Sputum induc on and processing was done according to interna onally accepted standards as described previously.23 Results for di erent sputum cell types are presented as percentage of total non-squamous cell count.
STATISTICAL ANALYSIS
First, characteris cs of (ex)smokers and never smokers were compared. Second, (ex)smoking pa ents with and without frequent exacerba ons and never smoking pa ents with and without frequent exacerba ons were compared. Comparisons were made by either student T-test, Mann-Whitney U test or chi square, whenever appropriate.
In order to iden fy variables poten ally associated with frequent exacerba ons, all variables with a p-value <0.10 in the comparison between pa ents with and without frequent exacerba ons in the (ex)smoker or never smoker group were used in a univariate logis c regression analysis. A er that, variables with a p-value <0.05 in the univariate logis c regression analysis were used in the mul variate logis c regression analysis. The nal mul variate models were created by stepwise selec on.
145