VALIDATION OF SERBIAN VERSION OF CHRONIC OBSTRUCTIVE PULMONARY DISEASE ASSESSMENT TEST

Background/Aim. The Chronic obstructive pulmonary disease (COPD) Assessment Test (CAT) is a simple and reliable tool designed to measure overall COPD related health status and complement physician assessment in routine clinical practice. Objective of this study was to evaluate the validity of the Serbian version of CAT. Methods. Study included 140 outpatients in the stable COPD, recruited from two centres: Clinic for Pulmonology, Clinical Center of Serbia, Belgrade, and Institute for Pulmonary Diseases of Vojvodina, Sremska Kamenica. All patients completed pulmonary function testing ? spirometry, the CAT and the modified Medical Research Council (mMRC) dyspnea scale at baseline visit. The CAT test-retest reliability was tested in 20 patients by the same investigator (physician). Results. We demonstrated that Serbian version of CAT had high internal consistency with Cronbach?s alpha 0.88. Test-retest analysis showed good correlation between CAT scores in two time points (Spearman?s ? = 0.681, p < 0.01). In our study the CAT correlated moderately to mMRC scale (? = +0.57), weakly to FEV1 (? -0.214), was positively related to number of exacerbations, but did not showed exact regularity with change in the Global Initiative for Chronic Obstructive lung disease (GOLD) stage. Conclusion. The Serbian version of CAT is a reliable, simple and easy-to-use tool that can be used in everyday clinical practice to assess the health status of COPD patients in Serbia.


INTRODUCTION
The Global initiative for Chronic Obstructive Lung Disease (GOLD) strategy document defined COPD as a preventable and treatable disease, with persistent respiratory symptoms and airflow limitation [1].Smoking is the main risk factor for disease development along with environmental exposures to biomass fuels and air pollution.Also, individual predisposing factors such as genetic abnormalities, abnormal lung development, and accelerated aging contribute to COPD developing.A course of the disease is often progressive and associated with significant comorbidities which increase its morbidity and mortality.
COPD is one of the most common diseases with a global prevalence of 11,7% [2].
COPD is one of the fastest growing causes of death and it is expected to be the 3rd cause of mortality by 2020 worldwide [1].The disease may cause disability and the quality of life is one of the treatment goals.Health status assessment in COPD patients is a routine in clinical research studies with comprehensive but time-consuming tools such as St.
The COPD Assessment Test (CAT) is a simple and reliable tool designed to measure overall COPD related health status and complement physician assessment in routine clinical practice [8].It is a short, self-administrated, eight-item questionnaire that includes symptoms, limitation of daily activity, sleep quality and energy, providing a single score.It Patients are further classified according to the airflow limitation severity based on postbronchodilator FEV 1 in patients with FEV 1 /FVC < 0.70 to GOLD stages: mild /GOLD 1 (FEV 1 ≥80%), moderate/GOLD 2 (FEV 1 50-79%), severe/GOLD 3 (FEV 1 30-49%), and very severe/GOLD 4 (FEV1<30%).CAT was administered to all the patients at the baseline visit by the same investigator (physician).CAT was again administered to twenty patients by the same investigator, at the second visit, 14 days after the first one.

Questionnaire
The CAT is a disease-specific questionnaire assessing health status in individuals with COPD.COPD Assessment Test and CAT logo is a trade mark of GlaxoSmithKline group of companies.The CAT can be freely used.For this study we used already available Serbian translation of CAT written in consistent and understandable language, so there was no need for back-translation analysis.CAT consists of the following eight items, each formatted as a minimum and maximum score of 0 to 5, respectively: cough, phlegm, chest tightness, and breathlessness going up hills/stairs, activity limitations at home, confidence leaving home, sleep, and energy.Individual item scores are summarized to provide a total CAT score that can range from 0 (floor) to 40 (ceiling).

Statistical analysis
Internal consistency of CAT questionnaire was tested by Cronbach's α coefficient analysis.
Correlation analysis between CAT score and mMRC dyspnea scale was performed, so as correlation between CAT score and pulmonary function measures (FEV 1 (L, %), FVC (L, %), FEV 1 /FVC).Test-retest analysis obtained as correlation analysis between CAT scores in two time points, performed by the same investigators (physicians).Differences between continuous variables was tested by using Student t test, ANOVA for variables with normal distribution, or Mann-Whitney U test or Kruskal-Wallis nonparametric ANOVA for parameters which distribution deviated from normal Gaussian distribution pattern.
Differences in frequency of the categorical variables were tested with Chi-square test.
Correlation analysis performed by using Spearman's nonparametric correlation methods.
All differences were set at 0, 05 alpha.

RESULTS
From May 2017 to January 2018, 140 patients with COPD completed the CAT questionnaire and mMRC dyspnea scale, and 20 patients completed CAT test in two time points, performed by the same investigator (physician).General characteristic of subjects (mean age 64.4±9.3 years; 84 men and 56 women) are summarized in Table 1.The majority of subject were ex-smokers (n=82, 60%), and the rest were active smokers (n=58, 41%).There were no subjects who never smoked.The mean body mass index (BMI) was 26.6±5.4kg/m 2 .Average FEV 1 (%) was 47.6±19.1 which indicated moderate to severe airflow limitation.Average mMRC score was 1.93±1.11and average CAT was 19.5±8.9, which implied that patients had more symptoms and more pronounced breathlessness (Table 1).The Cronbach's α was 0.887 for CAT test.Neither Cronbach's α item deleted value wasn't larger than basic value of 0.887, so we concluded that all questions are consistent with the questionnaire topic.We found significant positive correlation between CAT score and mMRC score (ρ=+0.570,P<0.001).CAT score showed weak but significant negative correlation with FVC (L): ρ= -0.274, P<0.01) and FEV 1 (L): (-0.214,P<0.05), but did not correlate with any other pulmonary function measure.Test-retest analysis showed good correlation between CAT scores in two time points (Spearman's ρ =0.681, P<0.01).
Next, patients were classified according to GOLD stage (I-IV).In GOLD stage groups we compared different general, anthropometric, clinical, pulmonary function parameters, exacerbation status and maintenance therapy, results are summarized in Table 2.
There was no difference in distribution of ex-smokers and smokers, and cumulative smoking status expressed in pack/years among GOLD I-IV groups.Patients with longer disease duration tended to be in higher GOLD stage groups, but there was no significant difference.Average COPD duration was from 4 years in GOLD I group to 8, 5 years in GOLD IV group.Patients with higher GOLD stage had significantly more acute exacerbation episodes (GOLD III and IV groups compared to II and I), and also higher number of exacerbations requiring hospitalization (especially GOLD IV group compared to other three groups).Regarding maintenance therapy higher percent of patients used LAMA and ICS/LABA in higher GOLD stage groups.MMRC breathlessness score, was highest in GOLD IV group compared to other three GOLD groups.On the contrary CAT total score didn't show exact regularity with GOLD stage change (Table 2).
Relation between CAT score and presence of comorbidities, all and cardiovascular, and number of exacerbation/year is summarized in Table 3.
There was no significant difference in CAT score between subjects with or without investigated comorbidities.On the contrary, increase in exacerbation number was related consistently with higher CAT score (P<0.01),so patients with 2 or more exacerbations during one year had significantly higher CAT score (Table 3).We also compared CAT scores between BMI subgroups, but did not find any regularity, neither statistical significance.

DISCUSSION
The study demonstrated good internal consistency and reliability of CAT score in a population of COPD patients in Serbia.CAT correlated moderately with the mMRC scale, did not differ significantly across spirometric GOLD stages and was higher in patients who experienced frequent exacerbations.
CAT is a short (8-item), self-administered questionnaire developed by Jones et al for the purpose of measuring health status of patients with COPD.It was derived from the data from three international observational prospective studies including 1 503 COPD patients from Belgium, France, Germany, the Netherlands, Spain and the USA following rigorous methodological approach and was subsequently validated in the subgroup of patients from the USA.This study showed excellent consistency of the questionnaire with the Cronbach's alpha of 0.88 and a good reliability [8].Soon after it was developed CAT was incorporated into GOLD guidelines as a part of a multidimensional assessment of COPD patients.Currently CAT is the preferred method for symptom assessment over traditionally used unidimensional mMRC scale that measures only breathlessness, as it is more comprehensive (GOLD).Also, CAT demonstrated a very good correlation with St George's Respiratory Questionnaire (SGRQ) that is commonly used to access the impact of COPD on health status in clinical trials but is too complex for use in a busy every day practice [8].
Since CAT questionnaire was developed and validated in English language it is possible that cultural, social, and linguistic differences may affect its performance in other populations.Hence, after its publication in 2009, CAT has been translated and validated in various countries including Japan [11], Indonesia, Korea, Vietnam [12], Thailand [13], Brazil [14], Turkey [15], Iran [16] and Arabic speaking countries [17].To our knowledge our study is the first that validated the use of CAT in Serbian language.
We demonstrated that Serbian version of CAT has high internal consistency with Cronbach's alpha 0.88 that is identical to original version of CAT [8] and comparable to other validation studies in which Cronbach's alpha ranged from 0.73 to 0.98 thus exhibiting high item correlation [18,16].The demographic characteristics of our study population were similar to derivation cohort [8] with 60% men, mean age 64 years, but our patient had more severe airway obstruction (mean FEV1 47.6±19.1% predicted compared to 52.3±18.9%predicted in US and 57.8±19.9%predicted in EU cohort).The mean values of CAT score was 19.5±8.9 indicating that patients had high symptom burden and pronounced breathlessness which is comparable to CAT scores in studies from Belgium, France, Germany, US, Portugal and Asian population [12,14,19].
Test-retest reproducibility measured at two time points in our study was good (Spearman's ρ =0.681) and consistent with other validation studies.When compared to other important functional and physiological variables CAT correlated moderately to mMRC scale (ρ=+0.57)and weakly to FEV1 (ρ -0.214).This is in line with previous studies in which the correlation between CAT and mMRC scale (ρ=0.29-0.61)and FEV1 (ρ=-0.56--0.23)was found to be moderate at best [18].Although in our study more severe COPD patients (group 3 and 4) had higher CAT scores the difference between the COPD groups was not significant.By contrast a cross sectional European study [19] showed a constant increase of CAT score across COPD stages with 3 points difference between the classes.
In our study CAT was positively related to number of exacerbations.Frequent exacerbators (≥2 exacerbations) had higher CAT scores compared to non exacerbators (20.0±8.3 vs. 12.0±6.4).Previous studies have also demonstrated that infrequent exacerbators have lower values of CAT compared to patients with ≥2 exacerbations [20; 21].In addition, CAT values are shown to be higher in patients experiencing exacerbation compared to stable patients with a mean difference of 4.7 units between the groups [22; 19, 8].
We found no difference in CAT scores in patients with and without comorbidities.This is consistent with study by Kwon in Asian population [12] but differs from large European study where presence of 3 or more comorbidities was associated with higher CAT scores [19].This may be due to a difference in sample size, as study by Jones included significantly higher number of patients that allowed subgroup analysis.Similarly to aforementioned studies we found no difference in CAT scores between patients with and without cardiovascular disease.Also, in contrast to previous data that showed that patients with lower BMI have higher CAT scores [12] in our study CAT did not differ across BMI groups.
Our study has several limitations.First, the study included COPD patients from two academic pulmonary hospitals that are not necessarily generalizable to all COPD patients in Serbia.Second, although number of patients is comparable to other questionnaire validation studies, our study was likely underpowered to detect between group differences which may explain the observed dissimilarities (such as non-significant differences in CAT values across GOLD stages and BMI groups) when compared to larger studies.
Nevertheless, this is a first study to validate CAT in Serbian language that confirmed its reliability and consistency.
In conclusion, the Serbian version of the CAT is an easy to administer and reliable tool that could be used in everyday clinical practice for assessment of health status in Serbian COPD patients.
is easy to complete by the patient and interpret by the clinician.The CAT was developed and validated internationally, in the Europe and United States, and has been translated into many languages worldwide [9].GOLD recognized the importance of the CAT in the multidimensional system of assessment of disease severity and selection of pharmacological treatment, as well as monitoring of the disease.METHODS Aim of this cross-sectional study was to evaluate the validity of the Serbian version of COPD Assessment Test -CAT.Study was conducted in accordance with Good Clinical Practice as outlined in the Declaration of Helsinki 2000.All necessary approvals for the trial were obtained from respective institutional review end ethical boards.Study population and design Study was independently conducted, and 140 patients were recruited from two Serbian centres Clinic for pulmonology, Clinical center of Serbia, Belgrade (98 subjects) and Institute for Pulmonary Diseases of Vojvodina Sremska Kamenica (42 subjects) from May 2017 to January 2018.They were outpatients in the stable stage of COPD, older than 40 years of age, smokers or ex-smokers.COPD was diagnosed according to Global Initiative for Chronic Obstructive Disease -GOLD criteria [1], no earlier than 6 months prior to study.Inclusion criteria was COPD stable stage and written informed consent.Exclusion criteria was: active respiratory disorder other than COPD, immunosuppression, or subjects unable to complete the questionnaires.Stable COPD is defined as no change in respiratory state in duration of 4 weeks that requires no change in therapy, or systemic steroids and/or antibiotics use.Patient characteristics included demographic information, smoking, COPD and exacerbations history, therapy and comorbidities (cardiovascular diseases -heart failure, arterial hypertension, ischemic heart disease, arrhythmia, peripheral artery diseases, osteoporosis, depression, diabetes, and gastroesophageal reflux).At the baseline visit, the patient's breathlessness was assessed using the Modified Medical Research Council (mMRC) dyspnea scale, and spirometry test was performed according to American Thoracic Society/European Respiratory Society ATS/ERS spirometry guidelines [10].

Table 3
CAT score values, comorbidities and exacerbation rates Revised on April 18, 2018.Accepted on April 20, 2018.Online First May, 2018.
aa Numbers in parenthesis means number, percent of patients P from ANOVA; aa P<0.001 vs. group without exacerbation Received on February 20, 2018.