The economic burden of chronic obstructive pulmonary disease from 2004 to 2013

Abstract

Objectives:

This study examines the epidemiology and economic impact of chronic obstructive pulmonary disease (COPD) at a nationwide level in South Korea.

Methods:

This retrospective analysis used the societal cost-of-illness framework, consisting of direct medical costs, direct non-medical costs, and indirect costs. In order to analyze the societal costs of patients with COPD, this study used a data mining and a macro-costing method on data from a South Korean national-level health survey and a national health insurance claims database from 2004–2013.

Results:

The total societal cost of COPD in 2013 was estimated to be $439.9 million for 1,419,914 patients. The direct medical cost for COPD was $214.3 million, which included a hospitalization cost of $96.3 million, an outpatient cost of $76.4 million, and a pharmaceutical cost of $41.6 million. The direct non-medical cost was estimated at $43.5 million. The indirect overall cost associated with the morbidity and mortality of COPD was $182.2 million in 2013.

Conclusions:

This study showed that COPD has a major effect on healthcare costs, particularly direct medical costs. Thus, appropriate long-term interventions are recommended to lower the economic burden of COPD in South Korea.

Keywords::

• COPD
• Societal cost
• Burden of illness
• Time cost

Introduction

Chronic obstructive pulmonary disease (COPD) is an irreversible respiratory disease that occurs because of long-term smoking or air pollution. Given that the morbidity and mortality of COPD has continually increased, the burden of COPD has gradually increased as well1. The overall worldwide prevalence of COPD was reported to be 1–20%, which is a relatively high level2–4. Unlike other diseases that have shown decreased mortality rates for the past 30 years, the COPD mortality has reportedly increased by 163%5, accounting for 3.8% of all deaths in high-income countries6. For this reason, a high societal cost is incurred because of COPD.

In South Korea, COPD morbidity reportedly stands at 8.8% in people older than 19 years of age and 13.4% in people over 40 years of age, which is high, although similar to other countries7. Furthermore, COPD was either the sixth or seventh most common cause of death in South Korea from 2006–2010, depending on the year8. The medical costs for COPD was reportedly $87.0 million in 2003 and $91.8 million in 2008, which were relatively high, compared with the costs incurred because of other diseases9^,10.

Given that COPD is an irreversible disease that occurs because of risk factors, the morbidity and mortality rates of COPD are expected to increase continually as the population ages under the conditions of severe environmental pollution, especially in Asia11. For this reason, the societal costs of COPD are expected to increase gradually as well. Research, thus far, has focused primarily on the prevalence or epidemiological aspects of COPD, or predictors of COPD-related healthcare utilization1^,7^,12. In this study, we comprehensively investigated the economic burden of COPD over 10 years by analyzing its direct medical costs, direct non-medical costs, and indirect costs at a nationwide level in South Korea.

Methods

Design

This research was designed as a retrospective analysis using national-level health survey results, a macroeconomic statistics database, and a health insurance claims database from the National Health Insurance (NHI) Corporation, all of which are related to the entire South Korean population (51 million people).

The NHI claims database included details on the overall performance of the NHI program and healthcare utilization by the entire population of South Korea, such as healthcare expenditures as well as financial outlays and reimbursements. It also provided integrated enrollment and medical and prescription data that were representative of the South Korean population in terms of both age and gender (Figure 1).

Figure 1. Process of extracting COPD data from the national health insurance database. ICD, International Classification of Disease.

In this study data was used from 2004–2013 from the aforementioned database to investigate the societal costs of COPD. Because source databases of the government and public institutions were used, there was no requirement mandated for approval of this study by an institutional review board or ethics committee.

Subjects and subject disease

Data on the study subjects with COPD were extracted from a nationwide South Korean database to estimate the societal cost according to the definition of COPD. Asthma, characterized by a reversible airflow obstruction, was excluded, while chronic bronchitis and emphysema were included13^,14. The subject diseases in this study included the codes of J41, J42, J43, and J44, as listed in the International Statistical Classification of Disease and Related Health Problems 10th Revision (ICD-10). The codes related to asthma, such as J45 and J46, were excluded from this study.

Estimating the societal costs of COPD

This study used the societal cost of illness framework15. The societal costs of COPD is composed of its direct and indirect costs. The direct cost includes direct medical costs and direct non-medical costs. The indirect cost includes productivity loss because of morbidity and the value of lost future income because of early death.

The direct medical cost consists of hospitalization, outpatient care, and the pharmaceutical costs. The direct medical cost for COPD constitutes NHI payments and out-of-pocket payments. The direct medical costs paid by NHI were obtained from the NHI database.

A data mining process was applied in this study, as shown in Figure 1. First, the ICD-10 codes for COPD were identified and entered into the database.

Second, using the ICD-10 codes for COPD (J41, J42, J43, J44), medical utilization data were extracted, such as the number of patients, hospital visits for outpatient, and inpatient days. Moreover, the NHI payment and the pharmaceutical costs for the codes of COPD were extracted, with a subsequent calculation of costs by gender, age, and treatment type. Third, out-of-pocket payments were estimated using a national health expenditure survey that was conducted by the NHI16. Finally, the total direct medical cost was derived by computing the sum of NHI payments, the out-of-pocket payments, and the pharmaceutical costs.

The direct non-medical cost for COPD treatment included transportation costs and care-giving costs. To calculate the total transportation costs for this study, the one-way transportation cost, as listed in the Korea National Health and Nutrition Examination Survey (KNHANES), was adjusted annually by the transportation price index and was then multiplied by the number of hospital visits17. To calculate the care-giving costs, the average care-giving cost from the KNHANES in 2007–2009 was adjusted by the medical service price index and then multiplied by the number of hospitalizations and the paid caregiver rate18. (1) where i = 0, 1, … , n (age), j = 1, 2 (gender); α, β = out-of-pocket payment rate for hospitalization and outpatient care, respectively; E_ij, OE_ij = medical cost for hospitalization and outpatient care, respectively; O_ij = total number of hospital visits; M = transportation costs; N_ij = days of hospitalization; and I = caregiver cost per day.

The indirect cost included productivity loss because of morbidity and the values of lost future income because of early death. For the cost of productivity loss because of morbidity, non-working days were calculated by multiplying 1/3 by the number of hospitalization days and outpatient care18. We then multiplied the non-working days by the labor force participation rate, the employment rate, and the average daily income19. The value of lost future income was estimated through a gross loss output approach, using the expected income to be earned during an expected lifetime without an early death. For the discount rate, the expected future income increased in proportion to the rate of wage increase. Considering that a 5% discount rate was similar to a wage increase rate of 4.8–5.4% for the past 4–5 years, the discount rate was assumed to be 0%20. The personal life-span income at each age was calculated using the monthly wages by gender and age from the basic statistics compiled by the South Korean Ministry of Employment and Labor21. The number of deaths attributable to COPD was calculated using the raw data of death causes from the South Korean National Statistical Office, as shown in equation (2(2) )8. (2) where = average expected income in the year, t = age at death; τ = life year; F_ij = number of deaths; p_ij = labor force participation rate; e_ij = employment rate; and y_ij = average income (per day).

Results

The number of patients in South Korea with COPD steadily increased until 2012, and then decreased slightly in 2013 (Figure 2). During a recent 10-year period, the number of COPD patients increased by 33.1%, from 1,067,199 patients in 2004 to 1,419,914 patients in 2013. The proportion of male patients was ∼51%, and a similar trend was observed during those 10 years. In terms of age, patients aged 60 years or older accounted for the majority of the total number of COPD patients. Specifically, patients aged 60–69 years and 70–79 years accounted for 22.4% and 20.3% of the total in 2013, respectively (Table 1).

Figure 2. Trends in patients and societal cost of COPD for 10 years.

Table 1. Epidemiology of COPD in South Korea from 2004–2013.

	2004	2009	2013
Total number of patients with COPD	1,067,199	1,305,540	1,419,914
% of population in South Korea	2.2	2.6	2.8
Gender
Male	546,794 (51.2%)	669,893 (51.3%)	724,990 (51.1%)
Female	520,405 (48.8%)	635,647 (48.7%)	694,924 (48.9%)
Age
<40 years	264,130 (24.7%)	275,902 (21.1%)	319,717 (22.5%)
40–49 years	139,426 (13.1%)	141,919 (10.9%)	172,993 (12.2%)
50–59 years	155,177 (14.5%)	199,788 (15.3%)	217,010 (15.3%)
60–69 years	246,598 (23.1%)	287,970 (22.1%)	318,298 (22.4%)
70–79 years	194,810 (18.3%)	290,257 (22.2%)	288,479 (20.3%)
≥80 years	67,058 (6.3%)	109,704 (8.4%)	103,417 (7.3%)
% of over 60 years	47.6	52.7	50.0
Total number of deaths from COPD	5250	4964	5079
Mortality (per 100,000)	10.8	10.0	10.1
% of deaths from respiratory diseases	37.0	29.2	22.6
Gender
Male	3575 (68.1%)	3371 (67.9%)	3412 (67.2%)
Female	1675 (31.9%)	1593 (32.1%)	1667 (32.8%)
Age
<65 years	646 (12.3%)	547 (11.0%)	458 (9.0%)
≥65 years	4604 (87.7%)	4417 (89.0%)	4621 (91.0%)

The number of patients related to COPD was on the decline over a 10-year period, from 5250 patients in 2004 to 5079 in 2013, which was the lowest since 2009 (4964 deaths). This was ∼10.0 patients per 100,000 people in 2009, accounting for 29.2% of the deaths caused by respiratory diseases. Of all deaths, men accounted for 67.2–68.1% during those 10 years, which was approximately twice as high as the rate for women. Patients, aged 65 years or older, accounted for more than 87% of the total deaths from 2004–2013.

The total number of days spent visiting medical institutions for COPD examinations and treatment increased in both males and females for the 10 years of this study (Table 2). The days of hospital visits for outpatients were 5.8–6.8-times higher than that for inpatients. The average days of hospital visits per patient was ∼2.8 days, showing the highest level in males and over 65 year olds.

Table 2. Distribution of healthcare utilization by group in South Korea.

	2004	2009	2013
Total days of hospital visits	2,900,830	3,652,333	3,856,059
Average days of hospital visits per capita	2.72	2.80	2.72
Type
Inpatients	372,480	539,237	569,316
Outpatients	2,528,350	3,113,096	3,286,744
Outpatients/Inpatients	6.8	5.8	5.8
Gender (days of hospital visits per capita)
Male	1,549,528 (2.83)	1,929,283 (2.88)	2,036,898 (2.81)
Female	1,351,302 (2.60)	1,723,050 (2.71)	1,819,161 (2.62)
Age (days of hospital visits per capita)
<65 years	987,515 (1.77)	1,113,080.0 (1.80)	1,175,167 (1.66)
≥65 years	1,913,315 (3.76)	2,539,253.0 (3.69)	2,680,892 (3.77)

The total societal cost of COPD increased by 1.85-times during the 10 years of this study, from $238.0 million in 2004 and $439.9 million in 2013 (Table 3). The total societal cost per person ranged $223.0 in 2004 to $309.8 in 2013, as the total societal cost was divided by the number of COPD patients. Of the societal cost, the direct cost accounted for 58.6–61.0%, which is higher than the indirect cost. Moreover, the medical cost was the majority of the direct cost.

Table 3. Societal cost of COPD in South Korea from 2004–2013.

	2004	2009	2013
Total societal cost ($ million)	238.0 (100.0%)	380.7 (100.0%)	439.9 (100.0%)
Societal cost per capita ($)	223.0	291.6	309.8
Direct cost ($ million)	140.2 (58.9%)	232.2 (61.0%)	257.7 (58.6%)
Medical cost	117.1 (49.2%)	194 (51.0%)	214.3 (48.7%)
Inpatient	45.7 (19.2%)	76.6 (20.1%)	96.3 (21.9%)
% of Tertiary hospitals	15.9	15.4	12.1
% of General hospitals	53.5	48.6	45.9
% of Hospitals	25.7	26.8	27.5
% of Long-term care hospitals	1.3	7.1	12.9
% of Clinics	3.6	2.1	1.5
Outpatient	44.2 (18.6%)	72.9 (19.1%)	76.4 (17.4%)
% of Tertiary hospitals	9.7	13.3	11.8
% of General hospitals	21.5	25.0	27.4
% of Hospitals	8.6	8.3	8.2
% of Long-term care hospitals	0.1	0.4	1.1
% of Clinics	60.0	53.0	51.6
Pharmaceutical	27.2 (11.4%)	44.5 (11.7%)	41.6 (9.4%)
Non-medical cost	23.1 (9.7%)	38.2 (10.0%)	43.5 (9.9%)
Transportation	19.8 (8.3%)	32.2 (8.5%)	36.6 (8.3%)
Caregiving	3.3 (1.4%)	6.0 (1.6%)	6.8 (1.6%)
Indirect cost ($ million)	97.8 (41.1%)	148.5 (39.0%)	182.2 (41.4%)
Productivity loss	83.7 (35.2%)	126.7 (33.3%)	160.9 (36.6%)
Loss from early death	14.1 (5.9%)	21.8 (5.7%)	21.3 (4.8%)

Note: All costs are measured at current prices.

Of the direct medical costs for COPD, the cost for inpatients amounted to $45.7–$96.3 million, whereas it was $44.2–$76.4 million for outpatients and $27.2–$41.6 million for pharmaceutical costs over the 10-year period. The portions of the inpatient, outpatient, and pharmaceutical costs were stable over the 10-year period of this study. In terms of the medical institution, for the cases of inpatient care, the costs from general hospitals were a big part of the total medical cost for inpatients, whereas the outpatient cost from clinics accounted for the majority of the total medical cost, which was attributed to primary care.

Among the direct non-medical costs for COPD, the transportation cost amounted for ∼$19.8 million, which was 8.3% of the total societal cost in 2004, and rose to $36.6 million in 2013. The care-giving cost was estimated at $3.3 million in 2004; and increased ∼2-fold over the 10 years to $6.8 million in 2013.

The indirect cost was estimated at $97.8 million in 2004, and $182.2 million in 2013, which included lost wages (production loss) of $83.7 million in 2004 and $160.9 million in 2013; and the future lost income ranged from $14.1 million in 2004 to $21.3 million in 2013.

Discussion

This study investigated healthcare utilization and the societal cost of patients with COPD, from a large-scale health insurance dataset representing medical claims and healthcare utilization by the entire population of South Korea. Using the database of the health insurance statistical yearbook and the national health insurance database, to calculate the number of patients, it was found that there were 1,419,914 patients diagnosed with COPD in the entire South Korean population of 51 million in 2013. This figure means 27.3 patients per 1000 persons. However, this value was much lower, compared with the estimated prevalence of COPD in other counties, which ranged from 11.4–26.1%; and it was also lower than the 7.8–8.8% prevalence of COPD in previous large-scale survey studies about the South Korean COPD morbidity based on a spirometry test1^,4^,7.

The lower value in this study may indicate that the majority of COPD patients did not make hospital visits to receive proper treatments or examinations. Or, it may be a factor of the characteristics of COPD patients in the structure of the healthcare in South Korea. This assumption could be supported by the study result of Jung et al.12, in which only 25.4% of patients with airway obstruction later visited medical institutions in South Korea for respiratory symptoms related to airway obstruction. In the study by Jung et al.12, patients with COPD, but with mild severity levels, made fewer hospital visits compared with patients with moderate-to-severe COPD. Patients with COPD, especially low severity cases, were less likely to see a doctor for examinations or treatment; thus, the NHI data system didn’t detect the majority of existing undiagnosed or untreated patients.

Another assumption was that patients with COPD seek healthcare for issues other than COPD or respiratory symptoms. Patients with COPD may make non-COPD-related hospital visits, as they may have a comorbidity and are likely to be treated for the COPD symptoms in the course of treatment for other diseases, such as cardiovascular diseases. Because of the characteristics of COPD as a chronic disease, the elderly, aged 70 years or older, showed higher morbidity than the other age groups in this study, which were consistent with the results of most studies in other countries22–24. As of the end of 2008, the elderly aged 70 years or older, representing 6.7% of the overall population, accounted for 22.8% of total health expenditures in South Korea10. Thus, as the elderly population increases because of the increased average life-expectancy, the COPD morbidity and the resultant societal cost are also expected to continually increase in South Korea.

In addition, the low prevalence rate was attributed to the difference of the study subjects. We only included the ICD codes of J41, J42, J43, and J44 in accordance with the medical environment in South Korea, which differed from that of J40–J44 in other studies25–29. The ICD codes used in this study affected the research results, such as gender distribution and age distribution. In the case of bronchitis, such as J41 and J42, there were slightly more women patients than men. Age distribution also showed a similar pattern to gender distribution, according to the ICD codes. Consequently, for a more accurate estimate, further research is necessary on the study subjects related to COPD.

The COPD mortality rate in South Korea was found to be ∼10.0 per 100,000 persons in this study, which was lower than the mortality rates of 12.0–130.5 (per 100,000 persons) of other countries except for Japan (4.4 persons per 100,000 persons)30^,31. Of the total South Korean deaths, death from COPD accounted for 2.0–2.2% of all deaths, which was lower compared with 3.8% for high-income countries, and 4.9% for low-income countries6. However, according to a previous study that utilized the same method used here, the mortality of South Korean COPD patients was 25 persons per 100,000 persons, which was approximately twice as high as the figure in this study32. The difference in the mortality rates was attributable to the fact that the previous study used a broader definition of COPD, whereas this study limited the scope of the COPD definition to only those as classified by the ICD-10 codes of J41, J42, J43, and J44.

The days of hospital visits for examinations or treatment increased continually during the 10 years. This result might reflect that the severity of COPD, for patients who visited medical institutions and underwent treatment, become gradually more serious. A previous study reported that the presence of respiratory symptoms and a higher severity level of airway obstruction were predictive of COPD-related hospital visits by a patient with COPD12. Synthesizing that finding with the result of this study, further research on the societal cost of COPD, based on severity, is required to investigate changes in the economic burden based on status.

The societal cost of COPD increased continually from 2004 to 2013, which was ∼0.09% of the total cost of diseases and ∼62% of the societal cost of respiratory diseases in South Korea18. Compared with the result of a study reporting that COPD accounted for 0.7–1.2% of the total national healthcare expenditure amounts in European countries, the proportion of the societal cost of COPD was shown to be relatively low in South Korea, considering the total national cost of diseases33. It was also shown to be lower than the $23.9 billion figure in the US in 1999, in terms of both the absolute level and the contrasting relative GDP value25.

Concerning the societal cost, the direct medical costs increased from $117.1 million to $214.3 million, of which the inpatient, outpatient, and pharmaceutical costs accounted for 21.9%, 17.4%, and 9.4% of the total, respectively, as of 2013. Compared with the results of other studies, the current study showed 12–18-fold lower costs for inpatient care and 5–15-fold higher costs for the pharmaceutical cost26^,27. This figure was also lower in terms of the contrasting relative GDP value26–29. This difference in the direct medical cost was also attributable to differences in healthcare systems and to the severity levels of COPD patients among the various countries. The main payment system for healthcare in South Korea is based upon a fee-for-service (FFS). Furthermore, all healthcare costs are determined by the FFS system. Moreover, South Korea has a single payer health insurance system. Under a single payer system, all the population in South Korea is covered (in theory) for all medically necessary services, and the government organizes healthcare financing. In South Korea, the costs for surgery, special diagnostic examination, and inpatient service are relatively high, whereas the costs for outpatient care is at a relatively low level. The healthcare cost per capita of respiratory disease in South Korea was ∼$100 in 2013, which was a lower level compared with other diseases10. That cost was attributed to the characteristics of respiratory disease, of which the use of outpatient care was higher than that of surgery and special diagnostic examination. In addition, in case of pharmaceutical costs, there is a possibility that we partially included the cost related to not only COPD but also other diseases, because we extracted all the data using the pertinent ICD codes.

In medical cost, we included not only the costs of medical institutions, but also that for community healthcare. The community healthcare in South Korea consisted of (1) hospital-based home health service, (2) public health center-based home health service for vulnerable people, and (3) long-term healthcare service. As the home healthcare service provided by both hospitals and public health centers were covered by the NHI, we extracted the costs of (1) and (2) from the NHI database. However, we could not include the long-term care cost related to COPD. In 2013, the long-term healthcare cost related to dyspnea was estimated at $14 million. Because of insufficient information about the causes of dyspnea, we excluded the long-term care cost in order to avoid over-estimation. Thus, when we included part of long-term care cost related to dyspnea, the medical cost is expected to increase.

The care-giving cost was included only with respect to paid care-giving. Costs are also incurred for family caring; however, this could not be included because of insufficient data on family caregivers. Further study is required to obtain more accurate caregiving costs.

The proportion of the indirect cost decreased from 41.1% in 2004 to 38.2% in 2011, and then increased slightly in 2013 (41.4%). Unlike the direct medical cost, the share of the indirect cost over the total societal cost decreased overall, despite the fact that the number of COPD patients has been increasing. This may be attributable to the fact that the number of patients aged 50 years or younger decreased annually, despite the increased total number of COPD patients. Thus, the indirect cost is unlikely to significantly increase, like the direct medical cost, even though the number of COPD patients is expected to increase continually in the future.

This study has its limitations. It is likely that the societal cost in this study is under-estimated, such as emergency services, long-term care service, lost leisure time, and the training of new workers to replace early retirees because of disease, as well as intangible costs (pain and emotional anxiety). If the aforementioned limitations are addressed in a further study, then more accurate measurements of the societal cost could be realized.

Conclusion

COPD is expected to become one of three major causes of death by 2030, according to the World Health Organization. Because of the increasing COPD morbidity and its characteristics as an irreversible respiratory disease, the societal cost of COPD is expected to increase continually. The number of COPD cases in South Korea, calculated from the NHI database, was shown to be relatively low compared with the worldwide COPD prevalence rate. This may imply that there are many patients who have undiagnosed COPD or who have not received appropriate treatment because of the slow progression of COPD or insignificant early symptoms, or both, as well as other factors. Furthermore, according to this study, the COPD morbidity and its societal cost have gradually increased according. For proper resolution of the problem about the continually increasing societal cost of COPD, an appropriate method to control the cost should be explored based on the results of this study. In addition, there should be further studies that reliably ascertain the societal cost of COPD and the characteristics of undiagnosed or untreated patients.

Transparency

Declaration of funding

Funding for this research has been provided by Novartis Pharma AG, Hallym University, and the South Korean Ministry of Health and Welfare.

Declaration of interest

The data collection process of this study was supported by a grant from the Healthcare Technology R&D Project of the Ministry of Health and Welfare of the Republic of Korea (A040153). Moreover, the work of writing and submitting of the report was supported by Hallym University (H21050081). All the authors declare there have been no involvements that might raise the question of bias in the work reported or in the conclusions, implications, or opinions stated. JME peer reviewers on this manuscript have no relevant financial or other relationships to disclose.