Medication availability and economic barriers to adherence in asthma and COPD patients in low-resource settings.

Inhaled medication is essential to control asthma and COPD, but availability and proper adherence are challenges in low-middle income countries (LMIC). Data on medication availability and adherence in Central Asia are lacking. We aimed to investigate the availability of respiratory medication and the extent of financially driven non-adherence in patients with COPD and asthma in Kyrgyzstan. A cross-sectional study was conducted in two regions of Kyrgyzstan. Patients with a physician- and spirometry confirmed diagnosis of asthma and/or COPD were included. The main outcomes were (1) availability of respiratory medication in hospitals and pharmacies, assessed by a survey, and (2) medication adherence, assessed by the Test of Adherence to Inhalers (TAI). Logistic regression analyses were used to identify predictors for adherence. Of the 300 participants (COPD: 264; asthma: 36), 68.9% were buying respiratory medication out-of-pocket. Of all patients visiting the hospital, almost half reported medication not being available. In pharmacies, this was 8%. Poor adherence prevailed over intermediate and good adherence (80.7% vs. 12.0% and 7.3%, respectively). Deliberate and erratic non-adherence behavior patterns were the most frequent (89.7% and 88.0%), followed by an unconscious non-adherent behavioral pattern (31.3%). In total, 68.3% reported a financial reason as a barrier to proper adherence. Low BMI was the only factor significantly associated with good adherence. In this LMIC population, poor medication availability was common and 80% were poorly adherent. Erratic and deliberate non-adherent behaviors were the most common pattern and financial barriers play a role in over two-thirds of the population.

Introduction

Worldwide, COPD and asthma are the most prevalent chronic respiratory diseases placing a significant clinical and economic burden on patients and societies[1]. Notably, the burden of chronic respiratory diseases seems disproportionally high in low- and middle income countries (LMIC)[2,3]. Importantly, the higher the control of asthma and COPD and the lower their exacerbation rate, the lower the level of health care expenditures on these diseases[4,5].

To control asthma and COPD, effective inhaled respiratory medicines have been developed and are widely recommended in clinical guidelines[6,7]. These medicines may however not all be available in all global settings, especially in LMIC. The World Health Organization (WHO) has developed the Essential Medicines List (EML), a list of essential medicines to help countries choose medicines that improve health and lower costs[8]. A survey of 32 LMIC indicated that over 90% had national EMLs in place[9]. While national EMLs aim to support the availability of a minimum number of respiratory drugs, actual, real-world, availability in LMIC is often much lower. Indeed, in LMIC, the availability of medicines for asthma was only 30.1% and 43.1%, in private and public sectors respectively[10]. An earlier study also found that the availability of three cornerstone asthma drugs (beclomethasone, budesonide, and salbutamol) in 52 LMIC was poor across treatment centers and hospitals[11]. More recently, a Ugandan study demonstrated that the majority of asthma and COPD medicines were largely unavailable, especially in public hospitals, and were unaffordable in the private sector[12].

In addition, even when some medication may be available in LMIC, financial barriers may hamper proper adherence to inhalers in asthma and COPD patients. A significant association was found between non-adherence and poor disease outcomes (e.g., exacerbations), and greater health-economic burden (e.g., direct and indirect health care costs)[13,14]. Previous studies have shown that non-adherence can occur due to a number of reasons with cost-related reasons being an important one in high resource settings, with medicines potentially being considered not cost-effective[15]. The extent to which economic barriers play a role in medication non-adherence in LMIC is unknown. Assessing adherence to asthma/COPD medication has long been difficult to perform in LMIC settings given the lack of validated measurement methods. Especially for LMIC in the Central Asian region, limited data on medication availability and adherence has been published. Recently, the “Test of Adherence to Inhalers” (TAI) for asthma and COPD patients was developed, which potentially could provide required insights, also in LMIC. The TAI is a validated questionnaire to identify non-adherence and to assess barriers related to the use of inhalers in asthma and COPD, including financial reasons[16].

The aim of this study is to describe availability of respiratory drugs and the extent of financially driven non-adherence in patients with COPD and asthma in a LMIC, taking Central Asian country Kyrgyzstan as a case study.

Methods

Study design

This was a cross-sectional study that was conducted from June 2021 to July 2021. We used questionnaires in a representative sample of patients with asthma and COPD. Ethical approval was obtained from the ethical committee of the National Center of Cardiology and Internal Medicine (NCCIM) with protocol number 4. The study is reported according to the STROBE checklist for cross-sectional studies (Supplementary Information).

Setting

The study was conducted in two regions of Kyrgyzstan. Kyrgyzstan is a land-locked, mountainous lower-middle income country in Central Asia with, according to the Worldbank, a gross domestic product per capita of $1178 in 2020. The regions were Jalal-Abad and Chui (Bishkek). In the city of Jalal-Abad, patients were recruited in the primary health care center and at the pulmonology department of the Jalal-Abad regional hospital, and in Bishkek patients were recruited from several randomly selected primary health care centers.

Participants

We included patients with asthma and COPD fulfilling the following inclusion and exclusion criteria.

Inclusion criteria:

Age ≥18 years

Born and live in Kyrgyzstan and speaking Russian and/or Kyrgyz

Physician confirmed diagnosis of asthma/COPD

COPD: FEV1/FVC ratio <0.7

Asthma: Post-bronchodilator increase in FEV1 > 12% or 200 ml from baseline

Patient consent to participate and willing to sign the consent form

Exclusion criteria:

FEV1/FVC ratio >0.7 in COPD patients

No reversibility after bronchodilation test in asthma patients

Patients that have a disability in communication

Data sources

For this study, multiple primary (i.e., a patient questionnaire) and secondary data (i.e., clinical record data from the branches of primary health care centers) sources were used. The questionnaire was developed by the research team, piloted among a small group, and optimized before large-scale data collection. The collecting process for the patient questionnaires is described below. As part of our Data Management Plan, only anonymized data were shared between study site collaborators and the research team, limiting data to the part that was kept securely taking the participant’s privacy into consideration.

Data collecting process

Data collection was performed by the study site collaborators who were healthcare workers at the Respiratory Department of each hospital. To minimize bias in the selection of patients, every third diagnosed COPD/asthma patient from the pulmonologist registries was selected and contacted by phone by one of the study site collaborators to invite them (asking them to bring all the medication they use).

Firstly, study site collaborators introduced the meaning and target of the study to patients and then invited them to take part in the study.

Secondly, if patients agreed (and signed the informed consent that was written in the questionnaire, see Supplementary Information) the diagnosis was confirmed. In patients who did not have spirometry in their case-record, spirometry was performed for confirmation.

Afterwards, patients received the survey: patients did the survey by themselves with help from the study site collaborators if needed. In case of misunderstanding of any questions, collaborators helped to explain them.

Finally, when patients finished their survey and submitted it to the study site collaborators, collaborators checked it for detecting any missing answers or mistakes and asked them to correct it.

Data to be collected

The Test of Adherence to Inhalers (TAI) (Supplementary Information) was used to identify problems with adherence to inhalation therapy. The TAI is a validated questionnaire consisting of 10 items (with a scale of 1–5) to be completed by the patient and two questions (with a scale of 1–2) to be completed by a healthcare professional. In this study, we used the formally translated Russian version of the TAI as available from www.taitest.com. The total score is 50 for the 10-item TAI and 54 for the 12-item TAI. The TAI distinguishes good adherence (TAI-10 = 50), intermediate adherence (TAI-10 = 46–49), and poor adherence (TAI-10 < 46). In addition, the type of non-adherence can be assessed including erratic (sum of TAI questions 1–5: <25), deliberate (sum of TAI questions 6–10: <25) and unconscious non-adherence (sum of TAI questions 11–12: <4). We had particular interest in TAI question 10 that focuses on financial barriers for non-adherence[16].

Medication availability in hospitals, private clinics, and pharmacies as well as having received a straightforward technique training instruction on inhaler use was assessed by a patient survey that also included several baseline characteristics (Supplementary Information).

Outcomes

The main study outcomes were (1) availability of respiratory medication and inhaler training in the hospital, community pharmacy, and private clinic, as measured by the survey, and (2) adherence to respiratory medication, as measured with TAI questionnaire. Of the two outcomes, the availability of medication is a health system barrier for which we expected little impact of patient-related co-variables. For the other outcome, i.e., medication adherence, we hypothesized that some patient co-variates may impact this. Therefore, several predefined demographic (e.g., age, sex, education, work status), clinical (e.g., disease severity, comorbidities), pharmaceutical (daily regimen), and socio-economic (e.g. monthly income, self-buying of medication) predictors for non-adherence were collected based on literature and the authors’ experience (Supplementary Information).

Statistical methods and sample size

Descriptive statistics (mean, median, SD, minimum, maximum) and the Chi-square test were used for the comparison of categorical variables. Univariable and multivariable regression analyses were used to assess associations between predictors and the outcomes (i.e. proper medication adherence, defined as a 10-item TAI score >45). Anticipating to a degree of non-adherence of 50%[17], a regression model with 15 potential predicting variables for non-adherence and the estimated requirement of 10 events of non-adherence per variable in logistic regression analysis[18], we aimed for the inclusion of 300 patients. Variables with p < 0.25 in univariable regression analyses were included in multivariable regression analyses[19]. Statistical analysis was performed with Statistical Package for the Social Sciences (SPSS, Chicago, IL, USA) (version 26.0 for Windows). Statistical significance was set at p < 0.05.

Table 1

Sociodemographic and clinical characteristics of asthma and/or COPD participants (n = 300).

	All (n = 300)	COPD group (n = 264)	Asthma group (n = 36)
Age, y
Mean (SD)	58.5 (11.8)	59.8 (10.8)	49.1 (14.5)
Median (minimum; maximum)	60 (19; 100)	61 (19; 100)	49 (21; 77)
Sex
Male, No. (%)	171 (57.0)	157 (59.5)	14 (38.9)
Female, No. (%)	129 (43.0)	107 (40.5)	22 (61.1)
BMI, kg/m2
Mean (SD)	28.2 (6.2)	28 (5.9)	29.4 (7.9)
Median (minimum; maximum)	27.7 (16.2; 93.3)	27.6 (18; 93.3)	29 (16.2; 59.3)
Monthly income, USD$
Mean (SD)	128.9 (113.2)	127.2 (109.6)	142 (138.5)
Median (minimum; maximum)	98.5 (0; 944.3)	94.4 (0; 944.3)	118 (0; 802.6)
Missing value	6	5	1
Education
Primary/secondary, No. (%)	84 (28.2)	73 (27.9)	11 (30.6)
Professional, No. (%)	127 (42.3)	110 (42.0)	17 (47.2)
University, No. (%)	87 (29.2)	79 (30.2)	8 (22.2)
Missing value	2	2	–
Working status
Working, No. (%)	126 (42.0)	110 (41.7)	16 (44.4)
Unemployed, No. (%)	49 (16.3)	39 (14.8)	10 (27.8)
Retired, No. (%)	125 (41.7)	115 (43.6)	10 (27.8)
Having insurance, No. (%)	297 (99.0)	261 (98.9)	36 (100.0)
Missing value	3	3	–
Smoking status
Current smoker, No. (%)	50 (16.7)	46 (17.4)	4 (11.1)
Ex-smoker, No. (%)	94 (31.3)	85 (32.2)	9 (25.0)
Never smoker, No. (%)	156 (52.0)	133 (50.4)	23 (63.9)
Biomass using for heating/cooking, No. (%)
Yes	145 (48.3)	131 (49.6)	14 (38.9)
No	155 (51.7)	133 (50.4)	22 (61.1)
Disease duration, years
Mean (SD)	9.8 (6.4)	9.7 (6.5)	10.6 (5.9)
Median (minimum; maximum)	10 (1; 40)	10 (1; 40)	10 (1; 20)
Pulmonary function tests
FEV1, % predicted
Mean (SD)	53.5 (14.9)	53.4 (15.4)	53.7 (10.9)
Median (minimum; maximum)	55 (18; 83)	55 (18; 83)	55 (34; 72)
FEV1/FVC ratio, %
Mean (SD)	59 (11.0)	58.2 (8.5)	65.6 (21.3)
Median (minimum; maximum)	60 (0; 83.4)	60 (30.5; 83.4)	72 (0; 83)
Reversibility, %
Mean (SD)	–	–	23.2 (10.0)
Median (minimum; maximum)	–	–	18.5 (9; 48)
COPD, GOLD grade, No. (%)
GOLD 1	–	3 (1.0)	–
GOLD 2	–	160 (60.6)	–
GOLD 3	–	79 (29.9)	–
GOLD 4	–	22 (8.3)	–
ACO, No. (%)	11 (3.7)		–
Comorbidities
Cardiovascular diseases, No. (%)	144 (48.0)	135 (51.1)	9 (25.0)
Allergic rhinitis, No. (%)	49 (16.3)	27 (10.2)	22 (61.1)
Bronchiectasis, No. (%)	16 (5.3)	15 (5.7)	1 (2.8)
Diabetes, No. (%)	30 (10.0)	26 (9.8)	4 (11.1)
Depression, No (%)	7 (2.3)	6 (2.3)	1 (2.8)
Co-medications
0 co-medication, No (%)	160 (53.3)	142 (53.8)	18 (50.0)
1 co-medication, No (%)	35 (11.7)	23 (8.7)	12 (33.3)
2 co-medications, No (%)	44 (14.7)	42 (15.9)	2 (5.6)
≥3 co-medications, No (%)	61 (20.3)	57 (21.6)	4 (11.1)
Type of prescribed medication
SABA	112 (37.3)	88 (33.3)	24 (66.7)
SAMA	171 (57.2)	167 (63.5)	4 (11.1)
SABA/SAMA	34 (11.3)	28 (10.6)	6 (16.7)
LAMA	3 (1.0)	3 (1.1)	0 (0.0)
LABA	20 (6.7)	18 (6.8)	2 (5.6)
ICS/LABA	38 (12.7)	31 (11.7)	7 (19.4)
ICS	34 (11.3)	19 (7.2)	15 (41.7)
Xanthines	19 (6.3)	14 (5.3)	5 (13.9)
Mucolytics	15 (5.0)	13 (4.9)	2 (5.6)
Prednisolone	16 (5.3)	10 (3.8)	6 (16.7)
Antibiotics	23 (7.7)	19 (7.2)	4 (11.1)
Number of times taking respiratory medication per day
1 time, No (%)	11 (3.7)	8 (3.0)	3 (8.3)
2 times, No (%)	92 (30.8)	74 (28.1)	18 (50.0)
3 times, No (%)	56 (18.7)	56 (21.3)	–
4 times, No (%)	116 (38.8)	111 (42.2)	5 (13.9)
5 times, No (%)	24 (8.0)	14 (5.3)	10 (27.8)
Missing value	1	1	–
Buying respiratory medication, No (%)
Myself	206 (68.9)	184 (70.0)	22 (61.1)
Partly covered by health insurance	84 (28.1)	72 (27.4)	12 (33.3)
Fully covered by health insurance	9 (3.0)	7 (2.7)	2 (5.6)
Missing value	1	1	–
Co-payment for the drugs, mean (SD)
Formal, USD$	8.3 (8.2)	7.9 (8.0)	10.5 (9.0)
Informal, USD$	2.1 (4.8)	1.6 (3.5)	5.3 (9.5)

Current rate of USD$ to KGS = 84.72 som (July, 2021).

SD standard deviation, ACO asthma, COPD overlap, FEV1 forced expiratory volume in 1 s; FVC forced vital capacity, SABA short-acting ß2-agonists, SAMA short-acting muscarinic antagonists, LAMA long-acting muscarinic antagonists, LABA long-acting ß2-agonists, ICS inhaled corticosteroids.

Table 2

Availability of respiratory medication and inhaler technique education.

Variable	All (n = 300)	COPD group (n = 264)	Asthma group (n = 36)
Availability in the hospital
Yes	95 (31.7)	83 (31.4)	12 (33.3)
No	86 (28.7)	66 (25.0)	20 (55.6)
Did not visit this one	119 (39.7)	115 (43.6)	4 (11.1)
Availability in the private clinic
Yes	33 (11.0)	26 (9.8)	7 (19.4)
No	79 (26.3)	62 (23.5)	17 (47.2)
Did not visit this one	188 (62.7)	176 (66.7)	12 (33.3)
Availability in the pharmacy
Yes	265 (88.7)	234 (88.6)	31 (86.1)
No	24 (8)	19 (7.2)	5 (13.0)
Did not visit this one	11 (3.7)	11 (4.2)	–
Trained how to use inhalers
By doctor	270 (90.0)	242 (91.7)	28 (77.8)
By nurse	7 (2.3)	5 (1.7)	2 (5.6)
By pharmacist	19 (6.3)	14 (5.3)	5 (13.9)
Never	4 (1.3)	3 (1.1)	1 (2.8)

Data as frequencies and percentages in parenthesis.

Table 3

Adherence to inhalers, adherence level and non-adherence behavior patterns in asthma and/or COPD patients.

Variable	All (n = 300)	COPD group (n = 264)	Asthma group (n = 36)
Adherence level (10-item TAI)
Good	22 (7.3)	22 (8.3)	0 (0.0)
Intermediate	36 (12.0)	28 (10.6)	8 (22.2)
Poor	242 (80.7)	214 (81.1)	28 (77.8)
Non-adherence behavior (12-item TAI)
Erratic	264 (88.0)	231 (87.5)	33 (91.7)
Deliberate	269 (89.7)	235 (89.0)	34 (94.4)
Unconscious	94 (31.3)	80 (30.3)	14 (38.9)

Data as frequencies and percentages in parenthesis.

TAI test of adherence to inhalers.

Table 4

Univariate and multivariate association with adherence in asthma and COPD patients (n = 300).

Variable	Univariate OR	p-value	Multivariate OR	p-value*
Age
≤50 years	1.463 (95% CI 0.760–2.815)
>50 years	Reference group	0.255
Sex
Male	1.420 (95% CI 0.785–2.568)	0.246	1.235 (95% CI 0.656–2.327)	0.513
Female	Reference group
BMI
Low BMI	1.777 (95% CI 0.988–3.198)	0.055	1.893 (95% CI 1.015–3.530)	0.045*
High BMI	Reference group
Monthly income
Low income	1.484 (95% CI 0.827–2.661)	0.186	1.825 (95% CI 0.598–5.576)	0.291
High income	Reference group
Education
Primary/secondary	1.044 (95% CI 0.484–2.252)	0.912
Professional	1.088 (95% CI 0.542–2.183)	0.813
University	Reference group
Working status
Working	0.887 (95% CI 0.462–1.701)	0.718
Unemployed	1.774 (95% CI 0.824–3.821)	0.143	1.962 (95% CI 0.850–4.528)	0.114
Retired	Reference group
Smoking status
Current smoker	1.003 (95% CI 0.438–2.300)	0.993
Ex-smoker	1.315 (95% CI 0.697–2.481)	0.398
Never smoker	Reference group
Biomass using for heating/cooking
Yes	1.298 (95% CI 0.729–2.313)	0.376
No	Reference group
Disease duration
>10 years	Reference group
<10 years	1.415 (95% CI 0.779–2.571)	0.255
Pulmonary function tests
FEV1, % Predicted
<52	Reference group
>52	1.104 (95% CI 0.618–1.974)	0.738
FEV1/FVC ratio, %
<59	1.058 (95% CI 0.593–1.887)	0.850
>59	Reference group
COPD, GOLD grade
GOLD 1–2	Reference group
GOLD 3	1.081 (95% CI 0.551–2.122)	0.820
GOLD 4	0.997 (95% CI 0.464–2.140)	0.993
Comorbidities
Cardiovascular diseases	0.854 (95% CI 0.480–1.518)	0.590
Allergic rhinitis	2.134 (95% CI 1.070–4.257)	0.031*	1.884 (95% CI 0.874–4.060)	0.106
Diabetes	0.819 (95% CI 0.299–2.239)	0.697
Buying respiratory medication
Myself	0.568 (95% CI 0.314–1.028)	0.062	0.641 (95% CI 0.340–1.208)	0.169
Partly/fully covered by health insurance	Reference group
Previous inhaler education
By doctor	7.761 (95% CI 1.035–58.197)	0.046*	6.868 (95% CI 0.895–52.691)	0.064
By nurse/pharmacist/Never	Reference group
Number of times taking medication per day
1–2 times	Reference group
>2 times	0.578 (95% CI 0.322–1.036)	0.066	0.665 (95% CI 0.352–1.255)	0.208

BMI body mass index, FEV forced expiratory volume in 1 s, COPD chronic obstructive pulmonary disease, GOLD The Global Initiative for Obstructive Lung Disease.

*p < 0.05; High BMI is >27.7 (median), Low BMI is <27.7 (median); High monthly income is >98.5 USD$ (median), Low monthly income is <98.5 USD$ (median). Missing variables: Monthly income –6; Education –2; Buying respiratory medication –1; Number of times taking medication per day –1.