Study design and participants
This facility-based cross-sectional study was conducted in the medicine outpatient department of Rajshahi Medical College Hospital, a tertiary care referral hospital from November 2020 to January 2021. All the adult patients (aged ≥18 years) visiting the department with the diagnosis of asthma were the study population. Sample size was calculated from the following formula: , where, z = 1.96 for 95% confidence level, p = assumed prevalence of poor adherence to inhaler therapy, d = allowable error of assumed prevalence. Due to lack of existing evidence, we assumed the prevalence of adherence to inhaler therapy as 50% among the asthma patients of Bangladesh and the calculated sample size was 384. Assuming 5% non-response rate we approached a total of 400 patients. Patients aged ≥18 years, had diagnosed asthma and were using at least one metered dose inhaler (MDI) with or without spacer and/or dry powder inhaler (DPI) for at least one year were included in the study. Patients having asthma-COPD overlap syndrome, other obstructive lung disease, chronic debilitating conditions (e.g. carcinoma), women with pregnancy and using inhaler for less than one year were excluded. The consecutive eligible patients during the study period were recruited until the targeted number of patients was reached. After excluding the incomplete data 357 patients were included in the final analysis.
Face-to-face interview by five trained physicians after the consultation using a structured questionnaire and check lists was conducted to collect data from the patients and their medical record respectively. The questionnaire had four parts: (i) socio-demographic characteristics of the patients, (ii) information of inhalers they used and measurement of inhaler adherence, (iii) a demonstration session of their inhaler using technique to identify any critical error, and (iv) asthma control status using the Asthma Control Test (ACT) The questionnaire was prepared in English and translated to Bangla. Back translated version was compared with original version to confirmed the equivalence across the language. A consortium was made to check the consistency of the translation and was pretested among 20 asthma patients before using it.
The 10-item Test of Adherence scale (TAI) based on a five-point Likert scale, which was developed and validated by Plaza et. al. (17) and widely used in different countries (18,19) were used to assess the inhaler adherence of asthma patients. However, the scale was not previously used among Bangladeshi patients, hence it was not previously validated. Patients are considered as good, intermediate and poor adherent if they score 50, 46-49 and ≤45 respectively (17,18). In our study, poor adherence (TAI score ≤45) was considered as non-adherent.
Patients were requested to demonstrate their inhaler using technique and were scored in a checklist adapted from a previous study based on the recommendation of the American Thoracic Society according to the steps completed by the patients to identify any critical error (2).
Outcome and independent variables
Non-adherence to inhaler among asthma patients (TAI score ≤45) was the outcome variable of the present study. Independent variables were sociodemographic characteristics of the patients (age, sex, residence, educational status, family income etc.), disease profile (smoking history, comorbidity, health seeking behavior), and inhaler related information (type, number and duration of inhaler usage, perceived difficulty and critical error of inhaler using technique, and self-reported efficacy of inhaler).
All the statistical analyses were made by using STATA version 16.0. Both univariable and multivariable logistic regression models adjusted for socio-demographic and inhaler related factors were used to determine the predictors of poor adherence to inhaler of the asthma patients. The variance inflation factor (VIF) was used to detect any evidence of multicollinearity problem among independent variables. Statistical significance level was set at p-value <0.05 for 95% confidence interval (CI).
Ethical approval was obtained from the ethical review committee of Rajshahi Medical College to conduct the study; Memo no: RMC/IRB/2020-2021/121. Informed written consent was also obtained from each respondent after explaining the purpose of the study.