Chronic obstructive pulmonary disease (COPD) is a leading cause of morbidity and mortality worldwide, resulting in an economic and social burden. Its prevalence, morbidity, and mortality vary across countries and different groups within countries1. The World Health Organization (WHO) predicted that COPD will become the third leading cause of death worldwide by 20302. Persistent respiratory symptoms and airflow limitation characterize COPD. The nature of the disease is that patients’ lung function will gradually decrease over time; even with the best available care, other unwanted symptoms, including fatigue, dyspnea, depression, anxiety, and insomnia, may occur as the disease progresses. Symptom-based palliative care treatment is required to maintain a good quality of life among these patients1.

Palliative care is an approach to improving patients’ quality of life for those who face problems associated with a life-threatening illness and their families. It prevents and relieves suffering through early identification, valid assessment, and treatment of pain and other physical, psychosocial, or spiritual problems3,4. Early palliative care in cancer patients can improve quality of life and reduce unnecessary hospitalizations and healthcare utilization3,5. WHO reported that only 14% of patients who need palliative care worldwide could access the services3. According to WHO, diseases requiring palliative care for adults are Alzheimer’s and other dementias, cancer, cardiovascular diseases, cirrhosis of the liver, chronic obstructive pulmonary diseases, diabetes, HIV/AIDS, kidney failure, multiple sclerosis, Parkinson’s disease, rheumatoid arthritis, and drug-resistant tuberculosis3,6. Even though terminal symptoms, especially dyspnea, fatigue, and restlessness, are presented in both COPD and cancer patients, COPD patients have less accessibility to palliative care compared with cancer patients4,7-12.

The global strategy for diagnosing, managing, and preventing chronic obstructive pulmonary disease, GOLD 2018 Report suggested physicians be aware of palliative care approaches for symptoms in their practice1. The guideline also suggested that physicians discuss their views of resuscitation, advance directives, and place of death with patients and their families1. However, the appropriate time to introduce palliative care to COPD patients is unclear. Currently, no tools can be used to predict how long COPD patients will survive13,14. Many physicians may find it hard, especially when discussing end-of-life decisions4,14-16. The current evidence supports that good palliative symptom management helps increase patients’ quality of life13,15. Vulnerability in the health system, such as lack of resources, low health literacy, and no supporting policy, may make the patients unaware of palliative care. Moreover, without palliative care services, the patients may undergo unnecessary life-prolonging aggressive treatments or interventions10,15,17.

This study aims to measure the prevalence of palliative care awareness in COPD patients and explore factors associated with patients’ palliative care perceptions.



A newly developed paperless questionnaire to collect demographic data, disease severity, previous experience with treatment, and questions on palliative care perception was developed. Three experts evaluated content validity: one pulmonologist and two palliative care specialists. After generating the questionnaire, a pilot study was done on 30 COPD patients in a general hospital for reliability testing, which resulted in Cronbach’s α=0.78–0.82.

Disease severity was evaluated using the COPD assessment test (CAT) score and modified Medical Research Council dyspnea scale (mMRC) score. Participants’ latest percent predicted forced expiratory volume in one second (% predicted FEV1) were the only spirometric data collected from medical records and used as one of the severity assessment tools.

CAT is a questionnaire developed by COPD experts from many countries, consisting of 8 questions, including patients’ symptoms and quality of life. Each question has a rating scale ranging 0–5. Patients with scores ≥10 should manage their symptoms. Cronbach’s α for the Thai version’s reliability was 0.85318

The mMRC dyspnea scale score ranging 0–4, is used to measure the patient’s dyspnea, where 0 suggests no breathlessness except during vigorous exercise, and 4 corresponds to breathlessness on minimal exertion. A score ≥2 indicates that the patient suffers from dyspnea19. The percent predicted forced expiratory volume in one second (% predicted FEV1) is used as COPD severity evaluation. The severity is classified into four levels regarding airflow limitation severity: mild (% predicted FEV1 ≥80), moderate (50≤ % predicted FEV1 <80), severe (30≤ % predicted FEV1 <50), and very severe (% predicted FEV1 <30).

Previous treatments mentioned the history of emergency department visits or hospitalization due to COPD exacerbation. In this part, we also asked about the experience of using respiratory support, including home oxygen therapy, ventilatory support, and non-invasive mechanical ventilation. Moreover, in this part, we supplemented patients’ disease perception questions, asking them if they knew they were now suffering from an incurable disease.

We used the following question to measure participants’ perception of palliative care: ‘Have you ever heard the term palliative care?’. Then, the question: ‘Do you know that palliative care is a part of COPD care?’ was used to measure COPD patients’ palliative care awareness.

The question: ‘How do you score your palliative care understanding?’ with a score of 1 indicating no palliative care knowledge, and 10 high level of knowledge.

Participants’ knowledge about the advanced care plan, ‘living will’, and end-of-life issues were evaluated directly. However, we used the simulated scenarios to explore participants’ end-of-life decisions regarding life-prolonging procedures, including cardiopulmonary resuscitation (CPR), endotracheal intubation, and ventilatory support, to prevent any discomfort caused by posting direct questions about end-of-life.

Study design and patient recruitment

Our study design was a cross-sectional study conducted on 280 COPD patients visiting outpatient pulmonology clinics in 2 tertiary hospitals, one university hospital, and another regional hospital, from 1 July to 28 December 2018.

The inclusion criteria for patients to participate were: aged ≥40 years, had been diagnosed with COPD by internists or pulmonologists for at least one year, and had their pulmonary function tests done no more than five years. Information and consent forms were given to every patient who met the criteria. Patients with cognitive impairment were excluded from the study.

Ethical considerations

The content of the questionnaire contained death and dying issues. Therefore, we carefully developed the interview script using gentle words but could keep the content we aimed to access. We also thoroughly trained our research assistant to use appropriate words and comforting body language throughout the interview. If any patients felt uncomfortable, a distress protocol would be activated. The protocol included stopping the interview process, and the primary investigator would be promptly notified to give immediate support. The patient would be escorted to a safe and quiet area. The primary investigator would primarily evaluate the patients and then be given appropriate counseling or referral to a psychiatrist depending on the severity of the distress. In addition, all participants were informed that they had the right to stop the interview process whenever they felt uncomfortable. However, the distress protocol was not activated throughout the study.

This study was approved by the Human Research Ethics Committee of Prince of Songkla University, Hatyai, Thailand (Approval number: 61-022-9-4). Written informed consent forms were signed before each interview.

Statistical analysis

The sample size was calculated using the one-sample proportion equation, according to a prevalence of 13% of COPD patients who knew that palliative care was part of COPD care9. The margin of error was set at 0.03–0.05, resulting in the appropriate sample size of 154–354. Therefore, according to the sample size calculation, this study included 280 participants, which was adequate according to the sample size calculation.

Statistical analysis was performed by Program R Version 3.4.2. Continuous data are presented using mean and standard deviation, or median with interquartile range (IQR), depending on the distribution. Percentages were used to describe qualitative data. Associated factors to palliative care perceptions were assessed by either Wilcoxon’s or Fisher’s exact test. A value of p<0.05 was considered statistically significant. A backward stepwise based on the Akaike information criterion (AIC) was used for variable selection for multivariate logistic regression.


Of 280 participants, the proportions of participants receiving services at university hospitals to center hospitals were 77.1% and 22.9%, respectively. Male participants were predominant (91.1%). The average age of the group was 73 years. Most participants had an education level lower than the diploma level (86%). Most of the participants were Buddhists (82.1%). More than half of the participants (54.3%) were unemployed, and 63.8% of the participant reported no comorbidity. Almost three-quarters (73.2%) reported their financial status as low-income. There were 12% of current smokers in the study. The median of % predicted FEV1, CAT, and mMRC scores were 62, 11, and 2, respectively (Table 1).

Table 1

Baseline characteristics of participants (N=280)

Place of study
University hospital21677.1
Regional hospital6422.9
Age (years), median (IQR)73 (65–78)
Education level
Lower than primary school176.1
Primary school16960.4
Secondary school5720.4
Bachelor’s degree or higher238.2
General worker113.9
Government officer113.9
Retired government officer4817.1
Worker in agriculture3813.6
Cancer (all types and stages)186.5
Heart disease (all types)5720.4
Diabetes mellitus3612.9
HIV infection31.1
Chronic kidney disease currently on hemodialysis10.4
Parkinson’s disease20.7
Drug-resistant tuberculosis10.4
None of the above17863.8
Economic status
Middle or higher7526.8
Smoking status
Current smoker3412.1
Disease severityMedian (IQR)
% predicted FEV162 (47–80)
CAT score11 (6–16)
mMRC score2 (1–3)
Number of years since diagnosis3 (2–5)
Number of days in the longest period of hospitalization due to exacerbation3 (0–7)
Number of hospital admissions within the recent year due to exacerbation0 (0–1)
Number of emergency room visits within one year due to exacerbation0 (0–1)

Participants’ disease perception and palliative care awareness are shown in Table 2. In all, 11.1% of the participants did not know they had COPD. Almost half of them (42.3%) did not realize that COPD was incurable. Three-fifths (59.3%) had been hospitalized due to disease exacerbation. Approximately one-third (30.4%) experienced endotracheal intubation, and 20% experienced non-invasive ventilator support (Bi-PAP).

Table 2

Participants’ disease perceptions, previous COPD treatments, palliative care decisions, and end-of-life decisions (N=280)

Knowing that they had COPD24988.9
Knowing that COPD is an incurable disease16157.7
Previous treatments
Having been admitted due to COPD exacerbation16659.3
Having visited an emergency room due to COPD exacerbation13849.5
Having experienced endotracheal intubation8530.4
Having experienced non-invasive ventilator support (Bi-PAP)5620.0
Having home oxygen279.6
Palliative related contents
Knowing the term palliative care7627.1
Knowing that palliative care is part of COPD care6322.6
Knowing the term CPR13949.6
Knowing endotracheal intubation17963.9
Knowing non-invasive ventilator support (Bi-PAP)8028.7
Having an advanced care plan7025.0
Knowing the term ‘living will’3412.1
Feeling worried while talking about palliative care4215.0
Having discussed with healthcare providers about palliative care and/or advanced care plan issues196.8
Self-evaluating palliative care knowledge score range 1–10, median (IQR) 1 (1–4)
End-of-life decisions
Cardiopulmonary resuscitation (CPR)
Do not resuscitate20272.1
Endotracheal intubation
Do not intubate22680.7
Non-invasive ventilator support (Bi-PAP)
Do not use Bi-PAP17963.9

The prevalence of participants who knew the term palliative care was 27.1%, and 22.6% knew that palliative care was part of COPD care. The average self-assessment of palliative care knowledge was 1 out of 10. Only 19 participants (6.8%) had discussed palliative care or advanced care plan issues with their doctors or healthcare providers (Table 2).

The exploratory analysis showed that factors significantly associated with COPD patients’ palliative care perception (knowing the term palliative care) were a younger age (p=0.003), knowing the term endotracheal intubation (p=0.027), knowing the term CPR (p<0.001), having advanced care plan (p=0.003), knowing the term ‘living will’ (p<0.001) and end-of-life decision of using non-invasive ventilator support (p=0.047) (Table 3).

Table 3

Participants’ palliative care perceptions compared with various factors*

FactorsHave perception n (%)Do not have perception n (%)p
Male69 (90.8)186 (91.2)
Female7 (9.2)18 (8.8)
Age (years), median (IQR)70 (61–76.2)74 (67–79)0.003b
% predicted FEV1, median (IQR)62.5 (44.8–79)62 (47–80)0.660b
CAT score, median (IQR)9.5 (5–15)11.5 (7–16)0.160b
mMRC score, median (IQR)1 (1–3)2 (1–3)0.129b
Knowing that they had COPD1.000a
Yes68 (89.5)181 (88.7)
No8 (10.5)23 (11.3)
Knowing about endotracheal intubation0.027a
Yes57 (75)122 (59.8)
No19 (25)82 (40.2)
Having experienced endotracheal intubation0.478a
Yes26 (34.2)59 (28.9)
No50 (65.8)145 (71.1)
Having home oxygen0.323a
Yes10 (13.2)17 (8.3)
No66 (86.8)187 (91.7)
Knowing the term CPR<0.001a
Yes52 (68.4)87 (42.6)
No24 (31.6)117 (57.4)
Having an advanced care plan0.003a
Yes29 (38.2)41 (20.1)
No47 (61.8)163 (79.9)
Knowing the term ‘living will’<0.001c
Yes20 (26.3)14 (6.9)
No56 (73.7)190 (93.1)
Having ‘living will’1.000c
Yes1 (1.3)3 (1.5)
No75 (98.7)201 (98.5)
Having discussed with healthcare providers about palliative care and/or advanced care plan issues0.211a
Yes8 (10.5)11 (5.4)
No68 (89.5)193 (94.6)
Feeling worried while talking about palliative care1.000a
Yes11 (14.5)31 (15.2)
No65 (85.5)173 (84.4)
End-of-life decisions
Cardiopulmonary resuscitation0.616a
Do not resuscitate57 (75)145 (71.1)
Endotracheal intubation0.957a
Do not intubate62 (81.6)164 (80.4)
Non-invasive ventilator support (Bi-PAP)0.047a
Do not use Bi-PAP41 (53.9)138 (67.6)

* Measured by answering the question: ‘Have you ever heard the term palliative care?’.

a Chi-squared test.

b Wilcoxon rank sum test.

c Fisher’s exact test.

The multivariate logistic regression analysis showed only two factors associated with palliative care perception: the end-of-life decision to use non-invasive ventilator support (AOR=0.51; 95% CI: 0.29–0.91) and having an advance care plan (AOR=2.54; 95% CI: 1.39–4.66) (Table 4). Factors associated with palliative care awareness (knowing that palliative care is part of COPD care) among COPD patients were younger age (p=0.014), lower mMRC score (p=0.01), knowing the term palliative care (p<0.001), knowing the term CPR (p=0.004), having an advanced care plan (p=0.049), knowing the term ‘living will’ (p<0.001) and having a discussion with healthcare providers on palliative care issues (p=0.037) (Table 5). The multivariate logistic regression analysis showed that a lower mMRC score was the only factor associated with palliative care awareness (AOR=0.67; 95% CI: 0.51–0.87) (Table 6).

Table 4

Multivariate logistic regression analysis of factors associated with palliative care perception

FactorAOR95% CIp
Knowing that they have COPD0.800.33–2.070.625
Having been admitted due to exacerbation0.940.49–1.820.865
Having experienced endotracheal intubation1.270.65–2.470.478
End-of-life decision of using Bi-PAP0.510.29–0.910.023
Having home oxygen1.820.65–4.990.246
Having advance care plan2.541.39–4.660.003
% predicted FEV11.000.99–1.010.747
CAT score0.980.91–1.060.679

[i] AOR: adjusted odds ratio.

Table 5

Participants’ palliative care awareness compared with various factors*

FactorsHave palliative care awareness n (%)Do not have palliative care awareness n (%)p
Male58 (92.1)196 (90.7)
Female5 (7.9)20 (9.3)
Age (years), median (IQR)70 (60.5–76.5)73 (67–79)0.014b
% predicted FEV1, median (IQR)65 (49.5–79.5)61 (46–79.2)0.677b
CAT score, median (IQR)9 (5–14.5)12 (7–16.2)0.058b
mMRC score, median (IQR)1 (1–2)2 (1–3)0.010b
Knowing that they had COPD0.820a
Yes57 (90.5)191 (88.4)
No6 (9.5)25 (11.6)
Knowing about endotracheal intubation0.114a
Yes46 (73)132 (61.1)
No17 (27)84 (38.9)
Having experienced endotracheal intubation0.180a
Yes24 (38.1)61 (28.2)
No39 (61.9)155 (71.8)
Having home oxygen0.845a
Yes7 (11.1)20 (9.3)
No56 (88.9)196 (90.7)
Knowing the term palliative care<0.001a
Yes52 (82.5)23 (10.6)
No11 (17.5)193 (89.4)
Knowing the term CPR0.004a
Yes42 (66.7)97 (44.9)
No21 (33.3)119 (55.1)
Having an advanced care plan0.049a
Yes22 (34.9)47 (21.8)
No41 (65.1)169 (78.2)
Knowing the term ‘living will’<0.001a
Yes18 (28.6)16 (7.4)
No45 (71.4)200 (92.6)
Having ‘living will’0.221a
Yes2 (3.2)2 (0.9)
No61 (96.8)214 (99.1)
Feeling worried while talking about palliative care1.000a
Yes9 (14.3)33. (15.3)
No54 (85.7)183 (84.7)
Having discussed with healthcare providers about palliative care and/or advanced care plan issues0.037c
Yes8 (12.7)10 (4.6)
No55 (87.3)206 (95.4)
End-of-life decisions
Cardiopulmonary resuscitation0.321a
Do not resuscitate49 (77.8)152 (70.4)
Endotracheal intubation0.119a
Do not intubate46 (73)179 (82.9)
Non-invasive ventilator support (Bi-PAP)0.784a
Do not use Bi-PAP39 (61.9)140 (64.8)

a Chi-squared test.

b Wilcoxon rank sum test.

c Fisher’s exact test.

* Measured by answering the queston: ‘Do you know that palliative care is part of COPD care?’.

Table 6

Multivariate logistic regression analysis of factors associated with palliative care awareness

FactorsAOR95% CIp
Sex: Male2.130.65–10.010.262
mMRC score0.670.51–0.870.004
Longest admission due to disease exacerbation1.031.00–1.070.052
Having ‘living will’8.870.74–217.960.097

[i] AOR: adjusted odds ratio.


Approximately one-quarter of COPD patients knew the term palliative care and knew that palliative care was part of COPD care (27.1% and 22.6%, respectively). This number is higher than in the study of Fahim et al.9 in 2013 from the UK, whose prevalence of COPD patients who knew palliative care as a choice of COPD care was 13%. However, palliative care perception in our setting was still lower than the study conducted on the Taiwanese healthy elderly population in 2012, which reported 57% recognized the term hospice and palliative care20.

In this study, the lower mMRC (less dyspnea) score was the only factor associated with palliative care awareness or knowing that palliative care is part of COPD care, in contrast with previous studies which reported that palliative care awareness was associated with higher disease severity9. Moreover, other factors such as age, sex, income, and previous treatments had no association with palliative care perception or palliative care awareness, according to multivariate logistic regression analysis in our study. In contrast, some studies reported that those factors were associated with palliative care knowledge15,20-22. The explanation for this phenomenon could be that the prevalence of COPD patients with disease perception was still low. In our study, 11.1% of participants did not know they had COPD. Only 57.7% of the participants knew that COPD was incurable, indicating disease unawareness. According to previous studies, the other explanation for this result was that as much as 60% of the participants received only primary education, which could affect their disease awareness and health literacy levels, palliative care perception, and palliative care awareness21,23. Therefore, health education should be provided more to increase health literacy levels among those patients. However, palliative care is a relatively new concept among the Thai population. There is still a large gap to fill to provide better care to COPD patients.

Our study revealed that 15% of the participants felt worried while discussing palliative care issues, which was unsurprising. In Asian culture, some people believe that talking about death can bring bad luck or accelerate their death. Hence, many people avoid discussing palliative and end-of-life care as they contain death issues24, which could be considered a roadblock in initiating palliative care discussions, especially in the non-terminal stage of disease among Thai and other Asian populations.

We also found a lack of palliative care communication between patients and healthcare providers. Only 6.8% of participants in our study reported ever discussing with healthcare providers about palliative care issues, while the study from Taiwan and the USA reported 11.7% of general population and 14.6% of COPD patients had heard about palliative care from their healthcare providers20,25. This phenomenon occurs widely as current practices do not facilitate conversations about palliative care between COPD patients and clinicians14. However, initiating the discussion of palliative care issues with COPD patients could be difficult, as there is no formal standardized guidance of when and how to start26.

In Thailand, the concept of palliative care was introduced to the Thai healthcare system in the late 1990s. The endorsement of terminally ill patients’ rights to refuse medical interventions or to discontinue any life-prolonging, by writing a ‘living will’ or advanced care plan was first available in 200727 and has been included in the national health policy since 201428. However, the resources have been scarce, making accessing palliative care services difficult27,28. Like many places, cancer patients were always the priority of palliative care. Patients who have other chronic or incurable diseases remain the minorities7,10,11,27,28.

Strengths and limitations

To our knowledge, this is the first study on palliative care awareness among COPD patients in Thailand. However, there were some limitations in this study. First, this study was done in outpatient clinics. Thus, the patients were healthier than hospitalized or homebound patients with more severe disease stages, which could affect their palliative care perception and awareness. Second, we included the latest pulmonary function tests done in the past five years as one of the inclusion criteria. However, the re-evaluation of pulmonary function tests (PFTs) was not done at the same frequency in each patient, limiting further evaluation, especially survival prediction using the BODE and ADO indices, essential tools to predict survival among COPD patients29-31. Moreover, in the study, there were some patients whose latest %FEV1 was >80, which could be explained by the fact that the PFTs we had collected were not the ‘initial’ PFTs but the latest ones. Therefore, those numbers should have increased if the patients were treated correctly, resulting in %FEV1> 8032. Finally, this study is a cross-sectional study. Thus, a limitation was the ability to conclude the causal effects, even though the multivariate regression analysis was performed.


Our study results showed that the prevalence of palliative care awareness among COPD patients and the perception of palliative care was low. Moreover, communication between patients and healthcare providers about palliative care issues was still lacking. For this reason, both COPD patients and healthcare providers should receive adequate information on palliative care to create an adequate perception of palliative care and be aware that palliative care is a part of COPD care. Moreover, according to this study, the health literacy level among COPD patients was low. Therefore, patient education should be provided to all COPD patients as early as the diagnosis is made. Further research about this issue should include inpatients or homebound patients and focus on the intervention to enhance palliative care awareness among patients and caregivers of COPD and other non-cancer patients.