Rate of severe exacerbations, healthcare resource utilisation and clinical outcomes in patients with COPD in low-income and middle-income countries: results from the EXACOS International Study

Abstract

Introduction

The EXAcerbations of Chronic obstructive lung disease (COPD) and their OutcomeS (EXACOS) International Study aimed to quantify the rate of severe exacerbations and examine healthcare resource utilisation (HCRU) and clinical outcomes in patients with COPD from low-income and middle-income countries.

Methods

EXACOS International was an observational, cross-sectional study with retrospective data collection from medical records for a period of up to 5 years. Data were collected from 12 countries: Argentina, Brazil, Chile, Colombia, Costa Rica, Dominican Republic, Guatemala, Hong Kong, Mexico, Panama, Russia and Taiwan. The study population comprised patients ≥40 years of age with COPD. Outcomes/variables included the prevalence of severe exacerbations, the annual rate of severe exacerbations and time between severe exacerbations; change in lung function over time (measured by the forced expiratory volume in 1 s (FEV₁)); peripheral blood eosinophil counts (BECs) and the prevalence of comorbidities; treatment patterns; and HCRU.

Results

In total, 1702 patients were included in the study. The study population had a mean age of 69.7 years, with 69.4% males, and a mean body mass index of 26.4 kg/m². The mean annual prevalence of severe exacerbations was 20.1%, and 48.4% of patients experienced ≥1 severe exacerbation during the 5-year study period. As the number of severe exacerbations increased, the interval between successive exacerbations decreased. A statistically significant decrease in mean (SD) FEV₁ from baseline to post-baseline was observed in patients with ≥1 severe exacerbation (1.23 (0.51) to 1.13 (0.52) L; p=0.0000). Mean BEC was 0.198 x10⁹ cells/L, with 64.7% of patients having a BEC ≥0.1 x10⁹ cells/L and 21.3% having a BEC ≥0.3 x10⁹ cells/L. The most common comorbidity was hypertension (58.3%). An increasing number of severe exacerbations per year was associated with greater HCRU.

Discussion

The findings presented here indicate that effective treatment strategies to prevent severe exacerbations in patients with COPD remain a significant unmet need in low-income and middle-income countries.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Severe exacerbations, defined as exacerbations requiring hospitalisation, are associated with higher mortality and elevated healthcare resource utilisation (HCRU) compared with moderate exacerbations.

• Although 90% of global chronic obstructive lung disease (COPD)-related deaths occur in low-income and middle-income countries with limited healthcare resources, there are little published data on the rate of severe exacerbations, or HCRU and clinical outcomes associated with severe exacerbations in patients with COPD in these countries.

WHAT THIS STUDY ADDS

• EXACOS International is the first study to report comprehensive data on severe exacerbation rates, HCRU and clinical outcomes (including blood eosinophil count (BEC), comorbidities and prescribed treatments) in patients with COPD from low-income and middle-income countries.

• The overall mean prevalence of severe exacerbations was 20.1%, and 48.4% of patients experienced ≥1 severe exacerbation during the 5-year study period. As the number of severe exacerbations increased, the interval between successive events decreased. A statistically significant decrease in mean forced expiratory volume in 1 s from baseline to post-baseline was observed in the study population. An increasing number of severe exacerbations per year was associated with greater HCRU.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• Our results indicate that effective treatment strategies to prevent severe exacerbations and to improve overall care for patients with COPD remain a significant unmet need in low-income and middle-income countries.

Introduction

Chronic obstructive pulmonary disease (COPD) is a major global cause of morbidity and mortality.¹ According to the WHO, COPD is the third leading cause of death worldwide.² Epidemiological studies indicate that COPD is strongly associated with a history of exposure to tobacco smoke; however, a substantial proportion of patients with COPD (25–45%) report that they have never smoked.³ Other risk factors for the development of COPD, which are particularly relevant in developing countries, include biomass smoke exposure,^4–6 poor lung growth and development,^7–11 inadequately treated childhood respiratory infections,¹² occupational air pollution,¹³ chronic childhood asthma¹¹ and post-tuberculosis lung disease.¹³

Exacerbations are clinically significant events as they accelerate lung function decline,^14–17 negatively impact patient quality of life,^{18 19} and are associated with increased hospital admissions and mortality.²⁰ Studies have shown that a history of exacerbations is the most important risk factor for future exacerbations.^21–24 In the natural history of COPD, exacerbations become more frequent and more severe as the severity of the underlying disease increases.²⁴ Indeed, two large healthcare database studies from the UK have demonstrated that the risk of mortality increases with both the frequency and severity of exacerbations. Thus, in a study of 99 574 patients with COPD from the UK Clinical Practice Research Datalink (CPRD) database (mean follow-up: 4.9 years), increasing frequency of exacerbations was associated with an increased risk of death in a graduated fashion, ranging from HR (95% CI) 1.10 (1.03 to 1.18) for 2 moderate exacerbations to HR 1.57 (1.45 to 1.70) for ≥5 moderate exacerbations compared with no exacerbation at baseline. Severe exacerbations were associated with an even higher risk of death (HR 1.79; 1.65 to 1.94).²² Results from EXAcerbations of COPD and their OutcomeS (EXACOS)-UK, an observational cohort study of 340 515 patients that used the CPRD and Hospital Episode Statistics database, demonstrated that increasing frequency and severity of exacerbations were associated with increasing risk of future exacerbations, all-cause mortality and COPD-related mortality.²³

Other studies have shown that severe exacerbations, defined as exacerbations requiring hospitalisation, are associated with higher mortality,^25–27 and elevated healthcare resource utilisation (HCRU) and costs.²⁸ Using a large, population-based inception cohort of 73 106 patients hospitalised for the first time for COPD, Suissa et al presented data indicating that severe exacerbations recur progressively sooner after each subsequent severe event and that the mortality rate increases with every new severe exacerbation.²⁹

Despite the fact that approximately 90% of global COPD-related deaths occur in low-income and middle-income countries,² there are little published data on the burden of severe exacerbations in patients with COPD in these countries. Therefore, the EXACOS International Study was instigated to quantify the association between the rate of severe exacerbations and HCRU and clinical outcomes in patients with COPD in 12 countries with limited healthcare resources, including countries from North, Central and Latin America, Russia, the Caribbean and Asia.

Methods

Study design and data sources

EXACOS International was an observational, cross-sectional study with retrospective data collection from medical records for a period of up to 5 years prior to the study visit. There was one study visit, during which data were captured in electronic case report forms (eCRFs) by a doctor or trained nurse. As there was only one study visit, no prospective data collection took place. Data were collected from 65 sites within 12 countries: Argentina, Brazil, Chile, Colombia, Costa Rica, Dominican Republic, Guatemala, Hong Kong, Mexico, Panama, Russia and Taiwan (online supplemental figure 1). The study period was the 5-year period prior to the index date (the date of signed informed patient consent).

Study population

Patients who met the following inclusion criteria were eligible for enrolment in the study: patients with COPD treated by a pulmonologist, with an investigator-confirmed diagnosis of COPD for ≥5 years prior to the index date; age ≥40 years; smokers or ex-smokers (quit smoking no longer than 15 years before the study visit); COPD-related data stored in medical records for ≥5 years, including spirometry, modified Medical Research Council (mMRC) Dyspnoea Scale³⁰ and medication data; and a signed informed consent form at the time of the study visit indicating that the patient understood the purpose of the study and the procedures required for the study, and was willing to participate in the study. A history of asthma was not included in the exclusion criteria as the intention was to study a broad patient population with COPD. Patients with a concomitant diagnosis of bronchiectasis, sarcoidosis or interstitial lung diseases were excluded.

Exposures and outcomes

A severe exacerbation was defined as hospitalisation (with or without an intensive care unit (ICU) visit) or an emergency department (ED) visit as a result of worsening COPD symptoms. Key outcomes included the number of severe exacerbation events per year (0, 1, 2, 3 and >3); the proportion of patients who experienced severe exacerbations per year; the time interval between severe exacerbation events; and change in lung function over time, as measured by the forced expiratory volume in 1 s (FEV₁; evaluated in patients with baseline (date of first spirometry data assessment) and post-baseline (date of second spirometry data assessment) FEV₁ measurements ≥6 months apart (FEV₁ population)). The prevalence of comorbidities, including cardiovascular diseases (CVDs) and other conditions, was assessed. Peripheral blood eosinophil counts (BECs) were measured at the study visit. The HCRU burden was assessed by the proportion of patients with hospitalisation, ED visits and ICU visits, and the number and duration of visits. Respiratory maintenance and reliever treatment medications were evaluated during the 12-month period prior to the study visit. Oral corticosteroids (OCS), injected systemic corticosteroids (SCS) and antibiotic prescriptions were evaluated during the 5-year study period.

Statistical methods

Categorical variables were summarised descriptively with the number and percentage of patients in each category. Continuous variables were summarised descriptively with the median (range) or mean (SD). The number of severe exacerbation events, the proportion of patients who experienced severe exacerbations and the time interval between severe exacerbation events were stratified by 0, 1, 2, 3 and >3 severe exacerbations per year; these groups were mutually exclusive. The statistical significance of the mean change in FEV₁ from baseline to post-baseline was analysed using the paired t-test. For evaluation of the HCRU burden, hospitalisation, ED visits and ICU visits were stratified by 0, 1, 2, 3 and >3 severe exacerbations per year. The analyses of the statistical significance of HCRU rates in patients with 1, 2, 3 and >3 severe exacerbations vs 0 severe exacerbation were performed using the Fisher’s exact test or the χ² test. Statistical analysis was performed using SAS (V.9.4 or higher) software (SAS Institute).

Patient and public involvement

Neither patients nor members of the public were involved in the design, conduct, reporting or dissemination plans of the study.

Results

Study population

Overall, 1755 patients with COPD signed the informed consent form and were eligible for inclusion in the study. Of these patients, 47 failed screening and 6 did not complete at least 5 years following a COPD diagnosis. Therefore, a total of 1702 patients from a total of 12 countries met the inclusion criteria and were included in the study population (online supplemental figure 1).

The demographic and clinical characteristics of the study population are shown in table 1 and online supplemental table 1. The mean age of the total population was 69.7 years, with 69.4% being male. The mean body mass index was 26.4 kg/m², and 57.9% of patients were overweight or obese. Overall, 49.0% of patients reported full healthcare reimbursement, with 33.1% having no healthcare reimbursement. Altogether, 17.0% of patients reported university and/or postgraduate education, and 28.1% of patients were current smokers. The distribution of severity of airway obstruction in the study population was: 41.1% moderate COPD (Global Initiative for Chronic Obstructive Lung Disease (GOLD) 2), 33.5% severe COPD (GOLD 3) and 12.5% very severe COPD (GOLD 4); while 55.9% of patients presented with moderate-to-severe dyspnoea (mMRC 2–3) and 8.6% with very severe dyspnoea (mMRC 4).

Table 1.

Demographic and clinical characteristics: total population and by country*

	Total (n=1702)	Arg (n=147)	Bra (n=294)	Chi (n=136)	Col (n=112)	Mex (n=179)	Rus (n=326)	Tai (n=300)
Age, years, mean (SD)	69.7 (9.1)	66.3 (8.5)	69.0 (8.1)	71.6 (7.4)	72.1 (7.9)	73.0 (8.1)	64.8 (9.2)	72.0 (8.8)
Male, n (%)	1181 (69.4)	67 (45.6)	123 (41.8)	75 (55.1)	73 (65.2)	119 (66.5)	284 (87.1)	295 (98.3)
BMI, kg/m2, mean (SD)	26.4 (5.5)	27.5 (6.3)	26.2 (5.4)	28.0 (5.9)	24.7 (4.3)	27.1 (6.1)	27.9 (4.6)	24.5 (4.0)
Weight category, n (%)
Normal weight	627 (36.8)	48 (32.7)	102 (34.7)	40 (29.4)	50 (44.6)	65 (36.3)	77 (23.6)	159 (53.0)
Underweight	90 (5.3)	8 (5.4)	18 (6.1)	4 (2.9)	9 (8.0)	6 (3.4)	8 (2.5)	14 (4.7)
Overweight	602 (35.4)	46 (31.3)	111 (37.8)	49 (36.0)	38 (33.9)	62 (34.6)	138 (42.3)	106 (35.3)
Obese	383 (22.5)	45 (30.6)	63 (21.4)	43 (31.6)	15 (13.4)	46 (25.7)	103 (31.6)	21 (7.0)
Healthcare insurance status, n (%)
No reimbursement	564 (33.1)	57 (38.8)	81 (27.6)	46 (33.8)	17 (15.2)	105 (58.7)	147 (45.1)	0 (0.0)
Partial reimbursement	304 (17.9)	54 (36.7)	19 (6.5)	53 (39.0)	36 (32.1)	2 (1.1)	81 (24.9)	0 (0.0)
Full reimbursement	834 (49.0)	36 (24.5)	194 (66.0)	37 (27.2)	59 (52.7)	72 (40.2)	98 (30.1)	300 (100.0)
Education category, n (%)
Primary and/or secondary school	758 (44.5)	77 (52.4)	150 (51.0)	82 (60.3)	23 (20.5)	61 (34.1)	35 (10.7)	184 (61.3)
High school	417 (24.5)	37 (25.2)	66 (22.4)	17 (12.5)	34 (30.4)	12 (6.7)	162 (49.7)	71 (23.7)
University and/or postgraduate	290 (17.0)	28 (19.0)	33 (11.2)	15 (11.0)	25 (22.3)	34 (19.0)	95 (29.1)	30 (10.0)
Smoking status, n (%)
Ex-smoker	1223 (71.9)	100 (68.0)	255 (86.7)	91 (66.9)	101 (90.2)	130 (72.6)	160 (49.1)	213 (71.0)
Current smoker	479 (28.1)	47 (32.0)	39 (13.3)	45 (33.1)	11 (9.8)	49 (27.4)	166 (50.9)	87 (29.0)
COPD GOLD grades, n (%)†	n=1686	n=146	n=289	n=134	n=112	n=172	n=326	n=300
GOLD 1: FEV1% predicted ≥80%	203 (12.0)	12 (8.2)	13 (4.5)	18 (13.6)	10 (8.9)	22 (12.8)	17 (5.2)	64 (21.3)
GOLD 2: FEV1% predicted 50–79%	699 (41.5)	64 (43.8)	114 (39.4)	62 (46.3)	56 (50.0)	57 (33.1)	137 (42.0)	125 (41.7)
GOLD 3: FEV1% predicted 30–49%	571 (33.9)	53 (36.3)	116 (40.1)	41 (30.6)	34 (30.4)	64 (37.2)	135 (41.4)	84 (28.0)
GOLD 4: FEV1% predicted <30%	213 (12.6)	17 (11.6)	46 (15.9)	13 (9.7)	12 (10.7)	29 (16.9)	37 (11.3)	27 (9.0)
mMRC Scale, n (%)‡
Grade 0	122 (7.2)	8 (5.4)	40 (13.6)	5 (3.7)	9 (8.0)	3 (1.7)	14 (4.3)	28 (9.3)
Grade 1	483 (28.4)	37 (25.2)	91 (31.0)	47 (34.6)	44 (39.3)	38 (21.2)	43 (13.2)	129 (43.0)
Grade 2	543 (31.9)	54 (36.7)	59 (20.1)	35 (25.7)	30 (26.8)	62 (34.6)	182 (55.8)	71 (23.7)
Grade 3	408 (24.0)	36 (24.5)	69 (23.5)	39 (28.7)	19 (17.0)	51 (28.5)	76 (23.3)	55 (18.3)
Grade 4	146 (8.6)	12 (8.2)	35 (11.9)	10 (7.4)	10 (8.9)	25 (14.0)	11 (3.4)	17 (5.7)

mMRC Dyspnoea Scale grading: grade 0: ‘I only get breathless with strenuous exercise’; grade 1: ‘I get short of breath when hurrying on the level or walking up a slight hill’; grade 2: ‘I walk slower than people of the same age on the level because of breathlessness, or I have to stop for breath when walking on my own pace on the level’; grade 3: ‘I stop for breath after walking about 100 m or after a few minutes on the level’; grade 4: ’I am too breathless to leave the house or I am breathless when dressing or undressing’.

The entries in this column were bolded since these represent data for the total poulation.

*Data presented for countries with ≥100 patients (Arg, Bra, Chi, Col, Mex, Rus, Tai); please see online supplemental table 1 for full data from 12 countries (including Cos, Dom, Gua, HK and Pan).

†COPD GOLD grade measurements of airflow limitation¹ were based on the latest available FEV₁ measurement.

‡The severity of dyspnoea at the study visit was assessed with the mMRC Dyspnoea Scale (used with the permission of the MRC; https://www.ukri.org/councils/mrc/facilities-and-resources/find-an-mrc-facility-or-resource/mrc-dyspnoea-scale/).

Arg, Argentina; BMI, body mass index; Bra, Brazil; Chi, Chile; Col, Colombia; COPD, chronic obstructive pulmonary disease; Cos, Costa Rica; Dom, Dominican Republic; FEV₁, forced expiratory volume in 1 s; GOLD, Global Initiative for Chronic Obstructive Lung Disease; Gua, Guatemala; HK, Hong Kong; Mex, Mexico; mMRC, modified Medical Research Council; Pan, Panama; Rus, Russia; Tai, Taiwan.

Blood eosinophil counts

The overall mean BEC was 0.198 x10⁹ cells/L, with 64.7% of patients having a BEC ≥0.1 x10⁹ cells/L and 21.3% having a BEC ≥0.3 x10⁹ cells/L (online supplemental table 2).

Comorbidities

Overall, 76.2% of patients had one or more comorbid conditions. Russia (84.4%) and Brazil (84.0%) reported the highest prevalence of comorbidities (online supplemental figure 2A). The most common comorbidity was hypertension (58.3%), followed by diabetes (20.4%); hypertension and diabetes together were observed in 13.3% of patients. Acute myocardial infarction was reported in 7.0% of patients (online supplemental figure 2B). Considering the comorbidity data by country, Russia had the highest prevalence of hypertension (75.8%), while the highest prevalence of diabetes was reported by Taiwan (25.0%). Regarding the prevalence of other comorbidities in the total population, depression or anxiety was reported in 16.5% of patients, and gastro-oesophageal reflux disease (GORD) was reported in 16.3% of patients (online supplemental figure 2C).

Respiratory inhaler medication during the 12-month period prior to the study visit

Overall, commonly prescribed inhaler devices included: short-acting β₂-agonist (SABA) (45.5%), inhaled corticosteroid (ICS)+long-acting β₂-agonist (LABA) combination (single inhaler) (41.7%), long-acting muscarinic antagonist (LAMA) (37.9%) and LABA+LAMA combination (single inhaler) (36.5%). Regarding triple therapy, 16.3% of patients were on an ICS+LABA+LAMA single inhaler (online supplemental table 3), and 14.4% were on multiple-inhaler triple therapy (online supplemental table 4). Considering the inhaler data according to country, Hong Kong had the highest rate of SABA prescribing (100%), Brazil had the highest rate of ICS+LABA prescribing (75.9%) (excluding Guatemala because of its small sample size (n=5)), Costa Rica had the highest rate of LAMA prescribing (76.2%), and Hong Kong the highest rates of ICS (46.1%) and LABA+LAMA prescribing (61.8%). The highest rate of OCS maintenance therapy prescribing was observed in Costa Rica (14.3%) (online supplemental table 3). The usage of other inhaler regimens of interest (total population) is shown in online supplemental table 4.

Annual rate of severe exacerbations during the 5-year study period

The overall mean annual prevalence of severe exacerbations (based on ≥1 severe exacerbation) over the 5-year study period was 20.1% (figure 1A). Additionally, 48.4% of patients experienced at least one severe exacerbation during the 5-year study period. Russia (76.7%) reported the highest rates of severe exacerbations, while Costa Rica (23.8%) and Mexico (26.3%) reported the lowest rates of severe exacerbations (excluding Guatemala because of its small sample size (n=5)) (figure 1B). Patients experienced a total of 2926 severe exacerbations during the 5-year study period, equating to an overall mean annual rate per patient of 0.34.

Figure 1.

(A) Annual prevalence of severe exacerbations by number of events over the 5-year study period: total population. (B) Prevalence of ≥1 severe exacerbation over 5 years: total population and by country.

Time interval between severe exacerbations

The overall median interval between ≥2 severe exacerbations (n=289) was 479 days (approximately 16 months) (online supplemental figure 3). As the number of severe exacerbations increased, the interval between successive events decreased, suggesting that severe exacerbation frequency was associated with an increased risk of future exacerbations (online supplemental figure 3).

Prescribed COPD exacerbation treatments during the 5-year study period

Overall, 46.7% of patients received OCS bursts, 25.3% received SCS and 56.2% received antibiotics. Patients received a median of two OCS, SCS and antibiotics courses each, with treatments lasting for a median of 7, 5 and 7 days, respectively (table 2).

Table 2.

Prescribed COPD exacerbation treatments during the 5-year study period: total population and by country

	Prescribed treatment	Patients who received treatment, n (%)	Number of treatment courses, median (range)	Treatment duration, days, median (range)
Total (n=1702)	OCS burst	795 (46.7)	2 (1–49)	7 (1–189)
	SCS	431 (25.3)	2 (1–27)	5 (1–120)
	Antibiotics	957 (56.2)	2 (1–50)	7 (1–60)
Argentina (n=147)	OCS burst	75 (51.0)	1 (1–15)	5 (1–22)
	SCS	32 (21.7)	1 (1–5)	3 (1–17)
	Antibiotics	94 (63.9)	1 (1–12)	7 (1–14)
Brazil (n=294)	OCS burst	160 (54.4)	3 (1–18)	5.5 (3–60)
	SCS	53 (18.0)	1 (1–10)	5 (1–60)
	Antibiotics	186 (63.3)	2 (1–16)	7 (1–60)
Costa Rica (n=21)	OCS burst	9 (42.9)	1 (1–3)	6 (5–7)
	SCS	3 (14.3)	1 (1–1)	5 (1–5)
	Antibiotics	7 (33.3)	1 (1–2)	7 (7–7)
Chile (n=136)	OCS burst	62 (45.5)	2 (1–14)	5 (1–21)
	SCS	40 (29.4)	1 (1–12)	3 (1–30)
	Antibiotics	72 (52.9)	2 (1–14)	7 (1–15)
Colombia (n=112)	OCS burst	53 (47.3)	2 (1–17)	5 (1–30)
	SCS	22 (19.6)	1 (1–8)	1 (1–21)
	Antibiotics	43 (38.4)	1 (1–11)	7 (1–15)
Dominican Republic (n=58)	OCS burst	46 (79.3)	2 (1–6)	5 (4–10)
	SCS	2 (3.4)	1 (1–1)	61.5 (3–120)
	Antibiotics	39 (67.2)	2 (1–11)	7 (5–55)
Guatemala (n=5)	OCS burst	0 (0.0)	0	0
	SCS	3 (60.0)	2 (1–2)	5 (5–8)
	Antibiotics	3 (60.0)	3 (2–3)	5 (5–7)
Hong Kong (n=89)	OCS burst	62 (69.7)	4.5 (1–49)	7 (2–12)
	SCS	29 (32.6)	3 (1–15)	5 (2–12)
	Antibiotics	66 (74.2)	6 (1–50)	7 (1–21)
Mexico (n=179)	OCS burst	69 (38.5)	1 (1–8)	5 (2–180)
	SCS	21 (11.7)	1 (1–2)	2 (1–7)
	Antibiotics	86 (48.0)	2 (1–9)	7 (1–21)
Panama (n=35)	OCS burst	7 (20.0)	1 (1–2)	5 (1–7)
	SCS	6 (17.1)	1 (1–3)	1 (1–1)
	Antibiotics	8 (22.9)	2 (1–3)	7 (2–10)
Russia (n=326)	OCS burst	132 (40.5)	2 (1–9)	10 (3–60)
	SCS	158 (48.5)	2 (1–20)	5 (1–30)
	Antibiotics	247 (75.8)	2 (1–10)	8 (1–24)
Taiwan (n=300)	OCS burst	120 (40.0)	2 (1–24)	7 (1–189)
	SCS	62 (20.7)	1 (1–27)	2.5 (1–16)
	Antibiotics	106 (35.3)	2 (1–20)	7 (1–24)

COPD, chronic obstructive pulmonary disease; OCS, oral corticosteroids; SCS, systemic corticosteroids.

Change in lung function (FEV₁) during the 5-year study period

Of the total study population with COPD (n=1702), 1686 patients underwent spirometry, of whom 817 patients (48.5%) had ≥1 severe exacerbation. The mean (SD) FEV₁ among these patients was 1.35 (0.69) L. Of the patients with ≥1 severe exacerbation, 402 (49.2%) had two FEV₁ measurements at least 6 months apart (FEV₁ population). A statistically significant decrease in mean (SD) FEV₁ from baseline to post-baseline was observed both in patients with 0 severe exacerbation (n=529) (1.54 (0.58) to 1.44 (0.61) L—equating to a 6.5% decrease in mean FEV₁; p=0.0000) and in patients with ≥1 severe exacerbation (n=402) (1.23 (0.51) to 1.13 (0.52) L—an 8.1% decrease; p=0.0000), indicating a generalised decline in lung function over time in the study population that was independent of the number of severe exacerbations. An additional notable finding was that baseline mean (SD) FEV₁ decreased as the number of severe exacerbations increased: from 1.54 (0.58) L in patients with 0 exacerbation to 1.07 (0.46) L in patients with >3 exacerbations—a mean decrease of 30.5% (table 3).

Table 3.

Change in lung function (FEV₁) during the 5-year study period by number of severe exacerbation events: FEV₁ population*

Severe exacerbations	Baseline		Post-baseline		Change from baseline
	Patients with spirometry tests, n	FEV1, L, mean (SD)	Patients with spirometry tests, n	FEV1, L, mean (SD)	FEV1, L, mean (SD)	P value†
0 event	529	1.54 (0.58)	529	1.44 (0.61)	−0.10 (0.31)	0.00000
1 event	155	1.38 (0.53)	155	1.25 (0.50)	−0.13 (0.32)	0.00000
2 events	74	1.24 (0.50)	74	1.17 (0.52)	−0.07 (0.24)	0.01183
3 events	43	1.20 (0.46)	43	1.10 (0.57)	−0.10 (0.34)	0.05001
>3 events	130	1.07 (0.46)	130	0.97 (0.48)	−0.10 (0.28)	0.00006
≥1 severe event	402	1.23 (0.51)	402	1.13 (0.52)	−0.11 (0.30)	0.0000

*Patients with two FEV₁ measurements taken at least 6 months apart.

†P value calculated using the paired t-test.

FEV₁, forced expiratory volume in 1 s.

Healthcare utilisation during the 5-year study period

In general, an increasing number of severe exacerbations per year was associated with greater HCRU (figure 2). Considering patients with >3 severe exacerbations, nearly all of them (97.2%) required an ED visit, 84.2% required hospitalisation and 11.9% required an ICU visit (figure 2).

Figure 2.

Healthcare resource utilisation during the 5-year study period by number of severe exacerbation events: total population. *P<0.00001 vs 0 severe exacerbation. ED, emergency department; ICU, intensive care unit.

For ED visits and hospitalisation, HCRU increased incrementally as the number of severe exacerbations increased. However, for ICU visits, HCRU was similar (6.0–7.3%) from 1–3 severe exacerbations, but then increased for patients with >3 severe exacerbations (11.9%). This suggested that patients with >3 severe exacerbations had more severe disease that required treatment in the ICU (figure 2). Here, it must be noted that the small number of cases with zero exacerbation who required an ED visit (n=17) or hospitalisation (n=23) was classified by the investigator as not caused by a severe exacerbation event. The mean (SD) duration of hospitalisation and ICU visits for patients with ≥1 severe exacerbation was 10.2 (7.8) days and 8.5 (8.9) days, respectively.

Discussion

Here, we have presented the results of EXACOS International, an observational cohort study aimed at quantifying the rate of severe exacerbations and the association between severe exacerbations, HCRU and clinical outcomes in patients with COPD in 12 low-income and middle-income countries, including Russia, and countries from North, Central and Latin America, the Caribbean and Asia. Overall, we found that almost half (46.0%) of patients were classified with severe or very severe (GOLD 3 or 4) COPD. This is comparable with the prevalence of GOLD 3–4 of 39% reported by Soriano et al, in a pooled analysis of 15 882 patients with COPD from 22 cohorts in seven high-income countries (Denmark, France, Norway, Spain, Switzerland, the UK and the USA).³¹ Over half (55.9%) of patients experienced moderate-to-severe (grades 2–3) dyspnoea, while 8.6% had very severe (grade 4) dyspnoea. It is notable that these figures are somewhat higher than the 40.4% of moderate-to-severe dyspnoea ((unmodified) MRC grades 3–4) and 3.2% very severe (grade 5) dyspnoea reported by Müllerová et al, in a cohort of patients with COPD identified from the UK CPRD.³² This discrepancy suggests that a higher proportion of patients from countries included in this study have moderate-to-severe and very severe dyspnoea than their counterparts in high-income countries. We suggest that the reasons for this are likely to be related to the challenges associated with delivering cost-effective prevention, diagnosis and management of COPD in resource-limited countries.³³

Studies indicate that approximately one-third of patients with COPD have sputum eosinophilia, indicating eosinophilic inflammation of the airways.^34–36 Although the eosinophil threshold(s) associated with clinical relevance in patients with COPD is currently subject to debate, a number of studies have shown that BEC holds potential in identifying patients who are more likely to respond to ICS.^37–41 Indeed, GOLD management recommendations indicate that patients with BEC ≥0.1 x10⁹ cells/L with uncontrolled COPD (≥2 moderate exacerbations/≥1 hospitalisation) and patients with BEC ≥0.3 x10⁹ cells/L should escalate their treatment from LABA+LAMA to LABA+LAMA+ICS triple therapy.¹ Here, it is important to note that 35% of patients enrolled in the current study did not have BEC taken at the study visit, indicating that BEC measurements are not standard clinical practice in the countries that participated in the study. In the current study, the mean BEC of our COPD cohort was 0.198 x10⁹ cells/L, with 21.3% having a BEC ≥300 cells/µL. These figures are consistent with those of previous studies, which reported a median BEC of 0.15–0.3 x10⁹ cells/L,^42–44 and a 20% prevalence of patients with a BEC of ≥0.3 x10⁹ cells/L.⁴⁵

A wide range of comorbidities, including obesity,^46–49 CVD (congestive heart failure, coronary artery disease, ischaemic stroke and arrhythmias), hypertension, diabetes,^{25 50–56} GORD^{57 58} and mental health disorders,^{55 59 60} have been commonly reported in patients with COPD. With regard to the mental health disorder prevalence data reported here, we should note that the prevalence of anxiety and depression was based solely on medical record data and was not confirmed with patient questionnaires.

CVD is particularly common in patients with COPD and is associated with an increased risk of hospitalisation and mortality,^61–64 accounting for approximately one-third of deaths in COPD.^65–68 In the present study, the most common comorbidity was hypertension (58.3%), followed by diabetes (20.4%). The high prevalence of cardiovascular comorbidities in the current study is consistent with previous studies demonstrating that the risk of experiencing an acute cardiovascular event is significantly increased following severe exacerbations compared with non-exacerbation periods.^69–72 A recent population-based cohort study of Danish patients with COPD showed that the odds of a cardiovascular event were 1.5 times higher in patients with a moderate exacerbation compared with more than 6 times higher in those with a severe exacerbation versus patients with no exacerbation.⁷³ Several hypotheses have been proposed to explain the association between COPD exacerbations and CVD. The onset of exacerbations is associated with elevated levels of systemic inflammatory markers, such as fibrinogen and interleukin-6.^74–76 These markers represent potent prothrombotic stimuli that have been linked to cardiovascular events.^{74 75 77} Another idea is that the systemic inflammation of COPD, combined with vessel wall abnormalities caused by sustained or intermittent hypoxia, may lead to oxidative stress, endothelial dysfunction and arterial stiffness, which, in turn, are linked with CVD.⁷⁸ Alter et al suggested that the acute worsening of airflow limitation that occurs during an exacerbation may lead to increased left ventricular wall stress and coincident heart failure.⁷⁹

GOLD 2024 management recommendations suggest that for patients with COPD with few symptoms (mMRC 0–1) and low risk of exacerbations, LABA or LAMA is preferred over SABA, except for patients with only very occasional dyspnoea. Patients taking LABA or LAMA who experience exacerbations should escalate therapy to LABA+LAMA, with further escalation to LABA+LAMA+ICS for frequent or severe exacerbations.¹ We found that the most commonly prescribed respiratory maintenance and reliever treatment was SABA (45.5%), which was presumably used as rescue medication. The second most commonly prescribed treatment was ICS+LABA (41.7%), even though GOLD 2024 no longer endorses the use of this combination therapy.¹ Interestingly, ICS+LABA is one of the most commonly prescribed medications for COPD in Latin American countries,⁸⁰ likely due to ICS+LABA being one of the prescribed medications that are provided free of charge to patients via government funding.⁸¹ Accordingly, we observed a generally high rate of ICS+LABA prescribing in Latin American countries, particularly in Brazil (75.9%). Brazil also had a high rate of LAMA monotherapy prescribing (69.4%). This is because patients taking ICS+LABA only need to buy a LAMA inhaler to enable escalation to ICS+LABA+LAMA. Thus, our interpretation of the Brazil-specific prescription data is that the majority of patients in Brazil using ICS+LABA+LAMA triple therapy use two different inhalers (ICS+LABA inhaler and a LAMA inhaler). Similarly, the high rate of ICS and LABA+LAMA prescribing in Hong Kong is likely indicative of patients taking triple therapy via multiple inhalers rather than via a single triple device. Our findings suggest that in Brazil and Hong Kong, many more patients are using triple therapy via multiple inhalers than indicated by the number of patients prescribed triple therapy via a single device. The problem with this trend is that multiple-inhaler triple therapy is associated with lower medication adherence and persistence, and worse clinical outcomes compared with single-inhaler triple therapy.^82–84 In contrast, the high prescribing rates of single-inhaler triple therapy observed in Taiwan (33.3%) are related to the fact that all inhaler therapies, including triple therapies, are fully reimbursed under Taiwan’s health insurance system.⁸⁵ Overall, healthcare system-related factors, including access to medicines, as well as economic and cultural factors, are the main drivers of prescribing patterns in low-income and middle-income countries.

The use of OCS/SCS and antibiotics is associated with the treatment of moderate exacerbations,^{24 86} and therefore the observed frequent use of these medications is indicative of the additional burden of moderate exacerbations in this population with COPD. We note that the use of oral medications, such as OCS and theophylline, is high in Asia.⁸⁷ A possible reason for this includes patient preferences for oral medication rather than inhalers (cultural factors); therefore, physicians in Asia tend to prescribe oral medications as add-on therapy to inhalers. This is reflected in the results of the current study: theophylline use was highest in Taiwan (55.0%) and OCS use was high in both Taiwan (11.3%) and Hong Kong (13.5%). Additionally, in Taiwan, when patients are hospitalised or admitted to the ED because of severe exacerbations, maintenance OCS is often prescribed in line with patient preferences.⁸⁷

Overall, almost half (48.4%) of the population experienced at least one severe exacerbation during the 5-year study period, suggesting that severe exacerbations remain a significant clinical burden in low-income and middle-income countries. The overall mean annual prevalence of severe exacerbations over the 5-year study period was 20.1%, indicating that one in five patients experienced at least one severe exacerbation every year. The mean annual rate of severe exacerbations per patient was calculated to be 0.34. This figure is in close agreement with the rate of 0.33 severe exacerbation per patient-year reported by Sadatsafavi et al, a 1-year randomised trial of the use of azithromycin for the prevention of COPD exacerbations.⁸⁸ Oshagbemi et al reported severe exacerbation incidence rates of 11–43 per 1000 patient-years in a UK population-based study using the CPRD database.⁸⁹ Other studies have estimated the annual rate of moderate-to-severe exacerbations to range widely from 0.34 to 3.0 exacerbations per patient.⁹⁰ We observed that Russia (76.7%) had the highest rates of severe exacerbations. This finding could be related to the long Russian winter, as several studies have shown that patients with COPD experience more frequent and more severe exacerbations in the winter months.^{86 91–94}

Our data showed that, as the number of severe exacerbations increased, the interval between successive events decreased, suggesting that severe exacerbation frequency was associated with an increased risk of future exacerbations. These findings are consistent with the results of previous studies showing that a history of exacerbations is a major risk factor for future exacerbations.^{22 23 29} Even a single COPD exacerbation may result in a significant increase in the rate of decline of lung function,¹⁵ which may not always fully recover to pre-exacerbation levels.¹⁶ Therefore, preventing the first severe exacerbation should be a critical therapeutic goal in the management of COPD.^95–97

It should be noted that the present study was conducted during the COVID-19 pandemic period (years 2–3 equate to 2020–2021). The modification of healthcare services, including a general reduction in services and an increase in remote consultations⁹⁸ to reduce the risk of viral transmission, may have reduced hospital admissions for severe exacerbations. In a study conducted by Alqahtani et al, it was observed that there was a 50% reduction in admissions for COPD exacerbations during the COVID-19 pandemic period compared with pre-pandemic times, which was likely associated with a reduction in respiratory viral infections that trigger exacerbations.⁹⁹ In accordance with these results, we observed a 38% reduction in the annual prevalence of severe exacerbations from pre-pandemic year 1 (2019; 27.3%) to year 2 (2020; 16.8%).

A statistically significant decrease in mean FEV₁ of approximately 100 mL (range: 70–130 mL) from baseline to post-baseline was observed across all patient subgroups, indicating a generalised decline in lung function over time that is independent of the number of severe exacerbations experienced by patients in this population with COPD. It is possible that the 6-month baseline to post-baseline period was too short a time to capture significant differences in lung function decline over time across patient subgroups. However, we observed a substantial decrease in baseline lung function as the number of severe exacerbations increased, indicating that the percentage decrease in lung function was higher in the frequent severe exacerbator group (>3 events, 9.3%) than in the zero severe exacerbator group (6.5%). These findings corroborate a number of previous studies demonstrating that exacerbations accelerate lung function decline, an effect which increases for patients with severe exacerbations.^14–17

COPD imposes a substantial burden on healthcare systems worldwide and is associated with significant economic costs. The main direct healthcare costs associated with COPD are due to hospitalisation, primarily as a result of exacerbations and the medications used to treat exacerbations, and are correlated with exacerbation severity.^{28 100–102} In line with these findings, we have shown in the present study that an increasing number of severe exacerbations was associated with greater HCRU, as defined by hospitalisation, ED visits and ICU visits. The Continuing to Confront COPD International Patient Survey reported that the annual direct medical costs associated with COPD in Mexico, Brazil and Russia (three countries included in the current study) were $1003, $555 and $742 per patient, respectively, considerably lower than the US costs of $9981 per patient. In Mexico, general practitioner visits (30.0%), specialist visits (20.0%) and home oxygen (16.0%) were the main drivers of direct costs; whereas in Brazil and Russia, the main contributors to direct costs were inpatient hospitalisation, indicative of severe exacerbations (22.0–27.0%), medications (24.0%) and home oxygen (28.0–33.0%).¹⁰³

Taken together, the results presented here demonstrate that effective prevention of severe exacerbations remains a significant unmet need in COPD. An increased understanding of the key factors that lead to severe exacerbations, coupled with optimisation of individual patient treatment in order to reduce the risk of severe exacerbations, are crucial therapeutic goals for patients with COPD.⁹⁶ This is particularly true for resource-limited, low-income and middle-income countries, who are lagging behind their Western counterparts in terms of access to affordable inhaled medicines and in the provision of effective diagnosis and care for patients with COPD.^104–108

Results from this study should be interpreted in light of several limitations. A history of asthma was not included in the exclusion criteria to enable an evaluation of a broad patient population with COPD. However, this may have created an upward bias in the measured rate of severe exacerbations as the risk of having a hospitalised exacerbation has been shown to be higher in patients with COPD and comorbid asthma compared with COPD alone.¹⁰⁹ Bearing in mind that COPD underdiagnosis is a major problem in primary care worldwide,^{110 111} the patient population studied in the present study may not be fully representative of the broader patient population with COPD. We used the classification of GOLD grades 1–4 that was current at the time the data were collected. An alternative possibility was to use the GOLD 2011 ABCD classification system—A: few symptoms, better lung function; B: more symptoms, better lung function; C: few symptoms, poor lung function; D: more symptoms, poor lung function.¹¹² However, this system has since evolved into the GOLD ABE assessment tool, comprising three risk groups. Group A: low risk (0–1 moderate exacerbations per year, not requiring hospitalisation) and fewer symptoms (mMRC 0–1 or COPD assessment test (CAT) <10); group B: low risk (0–1 moderate exacerbation per year, not requiring hospitalisation) and more symptoms (mMRC ≥2 or CAT ≥10); and group E: high risk (≥2 moderate exacerbations per year or ≥1 requiring hospitalisation) and any level of symptoms.¹ This was an observational study, and, as such, is subject to the limitations inherent to this type of study.¹¹³ ED visits in developing countries might not always translate into a severe exacerbation but rather reflect ease of access to a healthcare facility ED, especially in national healthcare systems where primary care is inefficient and patients have a low quality of care by international standards, such as in Russia.¹¹⁴ In these situations, defining all exacerbation-related ED visits as ‘severe exacerbations’ could result in an overestimation of the rate of severe exacerbations. A further limitation was that no patients with exposure to biomass smoke were included in the study. These patients present similar characteristics in acute exacerbations as those exposed to tobacco smoke.⁶ Thus, it was not possible to present a fully comprehensive representation of the burden of severe exacerbations in low-income and middle-income countries. Additionally, as a retrospective data study, it relied on the accuracy/completeness of the data that were captured in the eCRFs, and may be subject to recall bias. The observed heterogeneity of data is likely related to differences between the participating countries, in terms of their socioeconomic structures, healthcare systems and access to medications. Data on prescribed treatments and BEC were only available during the 12-month period prior to the study visit. Finally, the focus of the study was on patients who remained alive after severe exacerbation events; therefore, no data on patient mortality are presented. Nevertheless, a key strength of this study is that the EXACOS International Study quantified the rate of severe exacerbations over a 5-year period in a relatively large sample of patients with COPD (n=1702) from 12 low-income and middle-income countries. Moreover, to the best of our knowledge, this is the first study to report data on robust clinical outcomes, including severe exacerbation rates and lung function, as well as clinical characteristics (GOLD classification and mMRC Dyspnoea Scale, BEC, comorbidities and prescribed treatments) in a low-income and middle-income population with COPD. Additionally, as the study was conducted during the COVID-19 pandemic, it has the potential to provide valuable insights into the potential impact of the pandemic on serious exacerbations and HCRU.

Conclusions

The EXACOS International Study is the first study to report comprehensive data on severe exacerbation rates, HCRU and clinical outcomes (including BEC, comorbidities and prescribed treatments) in patients with COPD from low-income and middle-income countries with resource-limited healthcare systems. We found that nearly half of the population with COPD had at least one severe exacerbation over the 5-year study period. The observed frequent use of OCS/SCS and antibiotics is indicative of the additional burden of moderate exacerbations in this population with COPD. These results indicate that both moderate and severe exacerbations remain a significant clinical burden in patients with COPD in low-income and middle-income countries. Overall, it is clear that an increased understanding of the key causative factors that lead to severe exacerbations, coupled with optimisation of individual patient treatment in order to prevent severe exacerbations, are crucial therapeutic goals for patients with COPD in low-income and middle-income countries.

Data availability statement

Data are available upon reasonable request. Data underlying the findings described in this manuscript may be obtained in accordance with AstraZeneca’s data-sharing policy described at: https://astrazenecagrouptrials.pharmacm.com/ST/Submission/Disclosure. Data for studies directly listed on Vivli can be requested through Vivli at www.vivli.org. Data for studies not listed on Vivli could be requested through Vivli at https://vivli.org/members/enquiries-about-studies-not-listed-on-the-vivli-platform/. The AstraZeneca Vivli member page is also available, outlining further details: https://vivli.org/ourmember/astrazeneca/.

Ethics statements

Patient consent for publication

Not applicable.

Ethics approval

This study involves human participants and was approved by the Ethics Committee/Institutional Review Board/Independent Ethics Committee (study code: D5980R00042) in each country. Participants gave informed consent to participate in the study before taking part.