FormalPara Key Points for Decision Makers

Symptomatic chronic obstructive pulmonary disease (COPD) leads to impairment in the health state of patients.

Symptomatic COPD was associated with high economic burden across all countries. It influenced both direct and indirect costs associated with symptoms, dyspnoea severity, and duration of disease.

The health-related quality of life of patients with symptomatic COPD is, in general, low and influenced by multifactorial causality.

1 Introduction

Chronic obstructive pulmonary disease (COPD) is characterised by varying degrees of airflow obstruction that reflect the severity of symptoms such as breathlessness, chronic cough, and wheeze [1]. It is the leading cause of morbidity and mortality across all countries with varying financial status [2] and thus imposes a challenging problem to the healthcare system [3]. Although the Global Initiative for Chronic Obstructive Lung Disease (GOLD) recommends routine use of the modified Medical Research Council (mMRC) and COPD Assessment Test (CAT) to assess and treat COPD symptoms, symptomatic COPD has not been defined adequately in the published literature [1]. Increases in the number and severity of general symptoms and in the disease severity of COPD are associated with a humanistic and economic burden [410]. Furthermore, various comorbid conditions such as cardiac disorders, diabetes, obesity, osteoporosis, lung cancer, pneumonia, and psychological disorders may further impact the disease process [1113]. Taken together, COPD has a high-level impact on health-related quality of life (HRQoL) and healthcare resource utilisation [14].

Patients with COPD usually suffer from a marked deterioration in HRQoL [15]. This impact on HRQoL can be measured in terms of general health status, mental health, functional status, fatigue, and quality of sleep [16, 17]. The findings of these HRQoL instruments, which provide non-invasive patient-centered monitoring, reflect the humanistic burden of COPD [14]. Similarly, there exists a strong correlation between disease severity and economic burden [18]. The main focus of COPD treatment has been on the reduction and prevention of exacerbations [19], which are costly to treat and exert a negative impact on the HRQoL of the affected individuals [20]. Thus, patients with no or very few exacerbations also incur significant costs through repeated general practitioner (GP) visits and associated healthcare contact for their symptoms, in addition to a compromised HRQoL, indicating that symptomatic COPD is associated with high economic and humanistic burdens [10].

The objective of this systematic review was to identify and evaluate humanistic (HRQoL) and economic burdens among symptomatic patients with COPD.

2 Methods

2.1 Study Design

This systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [21]. It evaluated published evidence of humanistic burden of symptomatic COPD, as measured by HRQoL indices. The economic burden was assessed from cost-of-illness and resource use studies in patients with symptomatic COPD.

2.2 Information Sources

A comprehensive search of Excerpta Medica Database (EMBASE®), Medical Literature Analysis and Retrieval System Online (MEDLINE®), MEDLINE® In-Process, Cochrane Central Trials Register, Cochrane National Health Service Economic Evaluation Database (NHS EED), EconLit® , and Literatura Latino Americana em Ciências da Saúde (LILACS®) was performed from 2003 to 31 May 2013. In addition to database searching, abstracts from conference proceedings from the American Thoracic Society (ATS), American College of Chest Physicians (ACCP), European Respiratory Society (ERS), and International Society for Pharmacoeconomics and Outcomes Research (ISPOR) were hand searched from 2010 to 2013. A comprehensive search strategy was designed to retrieve relevant clinical data from published literature. An example of the search strategy is presented in detail in electronic supplementary Table S1.

The searches were restricted to include studies published in the English language, with the exception of LILACS®, which was not restricted to English articles. Relevant data from all included studies were extracted by a single reviewer, using a pre-defined extraction grid, which was subsequently validated by an independent reviewer. Thereafter, studies meeting the eligibility criteria were critically appraised using the Downs and Black checklist [22] for the humanistic burden, and the Drummond and Jefferson checklist for the economic burden [23].

2.3 Eligibility Criteria

The patient population of interest comprised adults of any race or gender with symptomatic COPD. Considerable variability in the definition of symptomatic COPD suggests that it is poorly defined in the clinical literature [1]. Therefore, in line with the GOLD guidelines, patients with one or more symptoms such as cough, breathlessness, sputum production, wheeze, and chest tightness were considered symptomatic [24]. Based on these recommendations, symptoms were also considered for evaluation in situations where they were based on physicians’ discretion only. Furthermore, the burden of symptomatic COPD patients in early stages might be higher than for non-symptomatic patients in slightly advanced disease, making quantification of this disease burden even more challenging. The GOLD guidelines have defined dyspnoea as a sense of increased effort to breathe, heaviness, air hunger, or gasping. In addition to the above mentioned criteria, the GOLD guidelines have recommended the mMRC scale and CAT for assessing COPD symptoms. An mMRC score ≥2 (breathlessness) and a CAT score ≥10 (high level of symptoms) represents patients with high symptoms [1]. Thus, patients with an mMRC score ≥2 and a CAT score ≥10 were also included in the review.

The humanistic burden was restricted to include cohort/longitudinal, case–control, cross–sectional, single-arm, and utility studies along with analyses of hospital records/databases and surveys. The outcomes of interest were patient-reported outcomes/HRQoL measures. Randomized controlled trials (RCTs) were excluded as they primarily focus on the assessment of efficacy and safety on a highly selective patient population rather than burden observed in clinical practice. Studies evaluating the impact of pharmacological or non-pharmacological interventions on improvement/deterioration of HRQoL were also excluded.

The economic burden was restricted to include cost studies/surveys/analysis and resource use studies in adult patients with symptomatic COPD, irrespective of the type of treatment investigated. As economic models present data for costs or cost effectiveness related to specific interventions under investigation, economic evaluations from modeling studies were excluded. The outcomes of interest were cost (direct, indirect, or total) and resource use associated with symptomatic COPD care.

2.4 Humanistic Assessments

The humanistic burden of illness (e.g. COPD) is characterised by the impact of illness on health and consequences of treatment, including impact on HRQoL or the relationship between various HRQoL scales and symptomatic COPD, loss of productivity and absenteeism, and frequency of clinical events (i.e. exacerbations, severity of symptoms). The complete list of HRQoL scales included in this review is available in electronic supplementary Table S2. Moreover, the relationship between patients’ demographic and comorbid characteristics (age, gender, body mass index [BMI], smoking status, comorbid conditions, etc.) and their humanistic burden was also evaluated.

2.5 Cost Assessments

The economic burden was characterised by the estimation of cost-of-illness (direct, indirect, and total) required for treating the COPD cohort of interest. These included costs associated with symptoms (severity and frequency) and dyspnoea care (based on the MRC dyspnoea scale score), and costs related to comorbidity.

3 Results

3.1 Study Characteristics

A total of 32 studies reporting humanistic burden and 74 economic studies specifically reporting cost and resource use involving patients with symptomatic COPD were included. Six studies reported both humanistic and economic impact of COPD. Figure 1 shows the study selection process, in line with PRISMA guidelines. The characteristics of patients and studies included in the humanistic burden review are presented in Table 1. The impact of varying levels of symptoms on the humanistic burden in patients with symptomatic COPD is presented in electronic supplementary Table S3.

Fig. 1
figure 1

Study selection process, in line with the PRISMA guidelines. PRISMA Preferred Reporting Items for Systematic Reviews and Meta-Analyses, NOI Not of Interest, SGA Subgroup Analysis

Table 1 Characteristics of patients and studies included in the humanistic burden review

3.2 Humanistic Burden

3.2.1 Impact of Demographic Factors on Humanistic Burden of Symptomatic Chronic Obstructive Pulmonary Disease

Mollaoglu and colleagues observed that with advancing age (age groups 65–69, 70–74, and ≥75 years), elderly patients with symptomatic COPD demonstrated a trend towards increasing fatigue, disability, and overall decrease in energy (p > 0.05). Moreover, females were associated with higher fatigue and low energy levels compared with males, which was attributed to traditional female responsibilities within a family; however, the association was not statistically significant. Married patients had significantly increased fatigue compared with their single counterparts (p = 0.02). Furthermore, lower education level and employment (housewives and full-time employees) were associated with lower overall HRQoL (p > 0.05) [25].

3.2.2 Impact of Respiratory Symptoms and Dyspnoea

An increase in the severity of respiratory symptoms was associated with a corresponding decline in the HRQoL of patients [4, 5, 2528].

Antoniu and colleagues evaluated HRQoL by categorising patients on the basis of symptom levels (mild, moderate, and severe). Patients experiencing severe sputum and cough had significantly higher fatigue and disability levels and significantly lower energy levels compared with patients with mild symptoms [28]. In contrast, a study by Al Moamary and colleagues revealed no significant differences in HRQoL in patients experiencing cough and wheezing evaluated using CAT and Chronic Respiratory Questionnaire–Self-Administered Standardized (CRQ-SAS) scores [4].

Based on the AQ20 score, patients reporting frequent cough, breathlessness during domestic work, and chest problems experienced limitation in full enjoyment of their life, whereas patients who did not present any respiratory symptoms scored the best possible result (score 0) [29]. Another study by Bellas and colleagues evaluated the relationship between cough and HRQoL in patients with COPD by using the CAT and the Leicester Cough Questionnaire (LCQ) as a cough-specific HRQoL measure. It was observed that increasing cough severity was associated with greater impairments in HRQoL [5].

This was also reflected in a significantly high mean of poor self-rated sleep quality [Pittsburgh Sleep Quality Index (PSQI)] in patients with a higher frequency of nocturnal wheezing, worrying, and uncontrolled thoughts. Other specific respiratory symptoms, not significant but associated with poor sleep quality, included cough, dyspnoea, chest pain, heartburn, and palpitations [30].

The association of dyspnoea with humanistic burden of symptomatic COPD was assessed in six studies. An increase in MRC scores from 0 to 4 was associated with worsening of HRQoL [4, 25, 27, 3133]. Punekar and colleagues demonstrated a greater impairment in HRQoL with higher levels of dyspnoea, such that EuroQol (EQ-5D) utilities ranged from 0.88 to 0.17 in the primary care physician (PCP) cohort and 0.88 to 0.29 in the respiratory specialist cohort, with MRC score increasing from 0 to 4 [33]. Furthermore, among elderly patients, an increase in the severity of dyspnoea from mild to severe was associated with an increase in fatigue and disability and decrease in energy levels [25]. In a cross-sectional survey by Al Moamary and colleagues, patients with shortness of breath had significantly lower HRQoL compared with patients with no symptoms of shortness of breath on using the CAT score in contrast to the CRQ-SAS score, where association was negative but statistically non-significant [4]. Furthermore, using Short Form-12 (SF-12), the questionnaire-based ‘symptom of grade 2 dyspnoea’ reported the lowest mean Physical Component Summary (PCS), whereas dyspnoea attacks had the lowest mean Mental Component Summary (MCS). Both scores decreased significantly with increasing number of symptoms; however, this was more evident for physical than mental scores. The larger impact of dyspnoea on these scores could be attributed to other disease conditions such as heart disease [27].

3.2.3 Impact of Time to Occurrence of Symptoms

Patients with symptomatic COPD experiencing daytime as well as night-time symptoms had lower HRQoL compared with patients with either or none of the symptoms [33, 34]. Multivariate analysis demonstrated significant impact of both symptoms on HRQoL after controlling for age, comorbidities, country, and treatment setting measured using the EQ-5D [33]. Sullivan and colleagues showed that compared with patients without symptoms, those with both or either (night-time/early morning) symptoms reported worse HRQoL and dyspnoea, as indicated by significantly lower SF-12 PCS and MCS mean scores (p < 0.01), as well as higher CAT and mMRC mean scores (p < 0.01) [34].

3.2.4 Impact of Comorbidities

Life-event stress is associated with depressive symptoms and worse HRQoL in elderly individuals with COPD. It was observed that life-event stress was associated with more depressive symptoms and worse QoL in individuals with symptomatic COPD [35]. Carvounis and colleagues also showed that the mean Geriatric Depression Scale (GDS) score for patients >65 years was low (7.8 ± 0.9), indicating mild depression. These patients also presented with a higher risk of exacerbations [36].

3.3 Economic Burden

Results from the published studies reporting cost and healthcare resource use were variable and could be attributed to a number of influencing parameters (country, patient characteristics, disease severity, and the cost/resource item). Characteristics of patients and studies included in the economic review are presented in Table 2. The economic burden of symptomatic COPD is presented in electronic supplementary Tables S4 and S5.

Table 2 Characteristics of studies included in the economic review

3.3.1 Resource Use and Direct Costs

Two global surveys provided comprehensive evidence on the economic burden of COPD across countries. Symptomatic COPD was associated with high economic burden across all countries studied in these two surveys [37, 38]. The Confronting COPD survey was a large-scale international survey conducted across eight countries in Europe and North America (between August 2000 and January 2001). The annual direct costs of COPD for the year 2002 ranged from US$522 in France to US$4119 in the US. The composition of direct costs differed between countries, with more than 50 % of direct costs resulting from inpatient admissions in Canada, Italy, Spain, the UK, and the US, and almost 50 % due to regularly prescribed medicines in The Netherlands [38]. In the US, the mean direct annual per patient costs of PCP visits, specialist visits, and emergency room (ER) visits for the year 2002 were US$225.19, US$150.20, and US$289.10, respectively (data evaluated between August 2000 and January 2001). Patients in the US also reported the maximum specialist visits, with a mean of 1.8 visits per patient in the past year [39]. The costs were lowest in France, with oxygen therapy accounting for 25 % of the total direct costs [38, 39].

Another large survey (the BREATHE study, conducted between June 2010 and December 2011) was conducted in 2012 in 11 countries of the Middle East and North Africa. Physician consultations were the most frequently utilised resource, followed by ER visits and hospitalisations. Resource use was significantly higher in subjects with more symptoms (CAT score ≥10) compared with those with fewer symptoms (p < 0.0001) [37]. Murtagh and colleagues reported that using a random sample of the Northern Ireland population (for the year 2000), the mean direct healthcare cost for each COPD patient was estimated at £171.69 (US$309) per annum, as per the year 2000 estimates [40].

Dalal and colleagues evaluated the direct healthcare costs (data evaluated between 1 January 2005 and 30 September 2008) associated with COPD care during 2008 in a hospital setting among COPD patients aged ≥40 years. The mean costs for emergency department (ED) visits were US$647 (SD US$445), simple admissions were US$7,242 (SD US$7,987), and complex admissions (intensive care unit [ICU; general, surgical, or medical] or intubation) were US$20,757 (SD US$41,370). Complex admissions accounted for 20.9 % of the total costs for COPD management in a hospital setting [41].

3.3.2 Impact of Demographic Characteristics

An association between age of patients with COPD and increase in healthcare resource utilisation was observed, whereas gender was not considered as an independent predictor of healthcare resource utilisation [4245]. Readmission at least twice in the first year of follow-up was observed among 9 % of total patients with COPD studied and was related to increasing age and male gender [46]. However, Berkius and colleagues observed that although the majority of readmissions were observed in women, female gender was not an independent determinant of readmission [odds ratio (OR) 0.97; 95 % confidence interval (CI) 0.75–1.25]. Age was observed to be independently associated with readmissions, with older patients demonstrating less chance of being readmitted (OR 0.68; 95 % CI 0.59–0.79 per 10-year increase in age) [42]. Furthermore, with an increase in years of schooling, a fall in direct medical costs associated with COPD was observed [40].

3.3.3 Impact of Respiratory Symptoms and Dyspnoea

In multiple studies across several countries evaluating data from the Confronting COPD survey, costs were higher in patients with severe dyspnoea [MRC dyspnoea scale (dyspnoea score 5)] compared with those with mild (dyspnoea score 0–2) and moderate (dyspnoea score 3–4) dyspnoea [9, 39, 4751]. In the UK, costs ranged from €590 (direct cost €395; indirect cost €195) in patients with mild dyspnoea, to €4,670 (direct cost €3,036; indirect cost €1,634) in those with severe dyspnoea in the Confronting COPD survey [9]. In Italy, the corresponding costs ranged from €441 in patients with mild dyspnoea to €6,365 in those with severe dyspnoea in the Confronting COPD survey [48].

The annual costs per patient increased with an increase in dyspnoea severity (according to the mMRC criteria) and duration of the disease (higher cost in the second year). However, it was observed that with the exclusion of medication costs, the costs 2 years post-diagnosis decreased compared with the costs 1 year post-diagnosis (£2,499.85 and £3,149.53, respectively). This could be attributed to better symptom control and disease management strategies in the primary care setting [52]. Gabriel and colleagues reported that COPD patients with chronic symptoms of airflow obstruction were associated with a higher rate of hospital service utilisation (1000/year) (19.6, 95 % CI 8.8–43.7 vs. 3.3, 95 % CI 0.8–13.2) compared with those without chronic symptoms [53].

3.3.4 Impact of Comorbid Conditions

Across countries, the annual societal costs associated with symptomatic COPD were higher among patients with comorbidities compared with those without any comorbidity [9, 39, 41, 4751]. The rate of hospital admissions, ER visits, and physician consultations was higher among patients with comorbidities [37]. Perera and colleagues [54] reported that comorbidities were associated with significantly higher hospitalisation costs in the US (p < 0.001).

3.3.5 Indirect Cost

Very few studies have examined the impact of symptomatic COPD on indirect costs, despite the fact that these are quite substantial. In the US, productivity loss was significantly higher in patients with both night-time and early morning symptoms compared with those without symptoms [55]. The Confronting COPD survey estimated the average indirect costs among US participants at 27 % of the total societal costs [39]. In Italy, costs due to absenteeism from work decreased significantly over a period of time due to the substantial decline in the number of exacerbations experienced by patients with COPD [56]. These observations were in line with those of an analysis based on an RCT in patients with COPD in which the increased cost of treatment was offset by indirect cost savings as a result of a decrease in the frequency of exacerbations [57].

4 Discussion

The current systematic literature review was undertaken to establish evidence on humanistic burden (based on HRQoL parameters) and economic burden in patients with symptomatic COPD.

4.1 Humanistic Burden

The results of humanistic burden showed variability in the inclusion criteria based on disease severity evaluated by using spirometry (mild, moderate, severe, and very severe) at the time of diagnosis. The evaluated studies included patients aged ≥40 years, with few studies particularly focussing on geriatric patients (>65 years of age) [25, 35, 5860]. Patients with symptomatic COPD were associated with impaired HRQoL. These findings are in line with the findings from Jones and colleagues who reported impaired HRQoL in patients with a CAT score ≥10 [61].

Demographic factors such as age and gender, and other factors such as marital status, education level, or employment status, were found to have an impact on the humanistic burden of the disease. An association between advancing age and deteriorating HRQoL was reported in a majority of the studies (n = 17). In addition, duration of illness was the disease-driven factor, while severity of dyspnoea and time of symptoms (daytime or night-time symptoms) were disease-specific factors affecting the humanistic burden of the disease [4, 30, 33, 36, 62]. These findings are in line with the findings of the systematic review conducted by Tsiligianni and colleagues [63], where gender, BMI, smoking, symptoms, comorbidity, depression, anxiety, and exacerbations were reported as important parameters impacting the HRQoL of patients with COPD. Therefore, treatment strategies should also influence modifiable demographic factors to improve HRQoL in the COPD patient population.

Disease-specific symptoms such as dyspnoea considerably influenced the health of patients with COPD [31, 32, 63]. A longer duration of illness, longer hospital stay, and higher number of hospitalisations were other important factors influencing HRQoL in patients with symptomatic COPD [25, 28, 63, 64]. HRQoL also deteriorated with an increase in general respiratory symptoms and their severity. The symptoms evaluated in these studies included breathlessness or shortness of breath while walking, cough, sputum, wheeze, fatigue, and disability [46, 8, 2527, 29, 63]. Therefore, a symptom-oriented approach involving treatment strategies aimed at severity of dyspnoea and exacerbations may improve HRQoL of COPD patients.

4.2 Economic Burden

Respiratory diseases are major causes of mortality and morbidity, and consume a considerable amount of healthcare resources. Among chronic respiratory diseases, COPD is associated with more healthcare consumption than other chronic diseases, and accounts for around five times as many hospital admissions in the UK compared with asthma [65]. A majority of included studies utilised reimbursement/claims/other databases as the sources of cost and resource use data. Variation in terms of country of study, year, currency, analysis perspective, disease severity, and cost/resource item under evaluation was observed in the published economic burden studies, making it difficult to compare across different health economies. Overall, both modifiable and non-modifiable demographic risk factors such as low education status, increasing age, and female gender were associated with greater annual societal costs. Higher societal costs were incurred for patients with increasing comorbidities, disease severity, frequency and severity of exacerbation, and symptoms. These factors directly influenced treatments administered, equipment used, labour, diagnosis, hospital admissions, PCP visits, specialist visits, ER visits, etc., thus, increasing the total direct costs across all the studies. Furthermore, indirect costs of absenteeism from work also suggest that a cost-effective treatment should focus on all aspects of the response to the disease per se, its consequence, and its progression.

4.3 Limitations

This systematic review is a qualitative evaluation of available evidence in symptomatic COPD; therefore, it does not provide a complete picture of COPD burden. Although better and reliable associations of various factors with symptomatic COPD can be achieved through quantitative analysis of HRQoL data, it was not conducted for the current review. Vague definitions of symptomatic COPD across studies made evidence generation a challenging task. In many studies, generic questionnaires [EQ-5D, SF-12, SF-36, and Functional Assessment of Chronic Illness Therapy–Fatigue (FACIT–F)] were used to measure HRQoL, which are less sensitive when compared with disease-specific tools for COPD (Clinical COPD Questionnaire (CCQ), St. George’s Respiratory Questionnaire (SGRQ), Sino-Nasal Outcome Test-22 (SNOT-22), CAT, CRQ-SAS). Also, a majority of the studies (>50 %) recruited patients with symptomatic COPD from an outpatient setting, mostly primary care centres; therefore, they may not represent the COPD population at large. The HRQoL component of the review focused on studies published in English only. Non–English language studies meeting the inclusion criteria of the review were included only from the LILACS® database. Nearly 30 % of the included studies on humanistic and economic burden were published as conference abstracts. Very few studies presented evidence on indirect costs associated with symptomatic COPD, although indirect costs are a major contributor to the overall economic burden of symptomatic COPD. This could be attributed to the fact that symptoms act as triggers that generate indirect costs followed by direct costs from clinical evaluation and treatment. Therefore, a better understanding of symptoms and symptom control should be the primary focus of future research.

5 Conclusions

HRQoL among patients with symptomatic COPD, as well as management of the disease, were associated with a substantial burden, especially in the developed nations of Europe and North America. This is particularly relevant because the burden of symptomatic COPD patients in the early stages might be higher than for non-symptomatic patients in slightly advanced disease. Therefore, this review attempted to better understand and quantify this difficult-to-define symptomatic COPD. HRQoL in patients with symptomatic COPD was influenced by various factors, of which dyspnoea was the most important determinant. In terms of economic burden, after excluding increase in disease severity and exacerbations, the presence of comorbidities, increasing age, and female gender were the critical parameters associated with an increase in the societal burden of symptomatic COPD. Future COPD cost-of-illness studies should take into account all types of costs incurred by patients, including indirect costs, to determine the most appropriate estimation of economic burden of the disease. In addition, early diagnosis and better treatment and disease management strategies should be undertaken to further curtail the economic burden of symptomatic COPD on patients, especially in the developed nations of the world.