With recent attention focused on patient-centered care models that incorporates the needs, values, and preferences of each individual patient in his or her treatment-related decisions, there is a growing social interest in respecting the patient’s self-determination regarding meaningless life-sustaining treatment (LST) and guaranteeing a “good death” at the end of their life. This has become especially relevant in light of the coronavirus disease 2019 pandemic where medical decisions related to extension of life could be closely linked to allocation of limited medical resources. Thus, establishing a system that gives priority to the will and desire of the patient is of utmost importance.

Decisions about end-of-life LST are known to be influenced not only by personal attitude and religious propensity of the physician in charge, but also various other factors including involvement of family members, sociocultural factors, and the legal system supporting such factors [1–6]. However, results from previous studies have been based mostly on questionnaire surveys, such as awareness surveys [5–7], where there were limitations in objectively identifying the impact of key factors affecting decisions about LST, such as economic factors or disability. To present the direction of the future for relevant policies and reach international consensus on issues related to LST, it is absolutely necessary to have objective evidence pertaining to factors that influence decision-making for LST.

With the enactment of the “Act on Hospice and Palliative Care and Decisions on Life-Sustaining Treatment for Patients at the End of Life” (hereinafter “LST Decision Act”), South Korea has established a national management system for LST (Central Hospice Center: September 2017, National Agency for Management of Life-Sustaining Treatment: February 2018), which had not existed until the introduction of the LST Decision Act. In the present study, we linked statistical data from the National Agency for Management of Life-Sustaining Treatment, who collects data from the relevant population, to national health insurance data, which reflects the population’s healthcare utilization, to identify the current status of actual LST at the national level and objectively analyze and identify factors that influence treatment decisions. We believe that the findings could be used as the basis for establishing a comprehensive strategy for systematically managing the entire end-of-life treatment decision-making stage.

The present study was the first study in South Korea to identify the current status of LST implementation by analyzing data representing the total population of South Korea. The LST determination system was being applied most often to patients aged ≥ 70 years, but the percentage of cases where the patient made the determination on LST was only 33.5%. In particular, the percentage of LST being implemented based on advance statement on LST was extremely low (1.2%).

The results showed that, compared to previous studies from individual hospitals (12.0%–29.0%), the percentage of self-determination on LST had increased (33.5%) after the enactment of the new legislation [9–12]. However, the percentage of terminating LST based on the assertion of family members still remained high (66.5%). We believe that the reason for these results may have been influenced by the unique cultural background in Asia. In Asian culture, discussions about death is considered to be a taboo topic and telling a parent, for example, the bad news that he or she has cancer and making them worry is against one’s filial piety [13,14]. Because of the culture where there is a lack of openness about death-related discussions, the percentage of patients who complete an advance statement on LST tends to be lower than in Western countries [15,16] and making it difficult to understand the health care wishes of patients in relation to LST. By contrast, some countries try to respect the wishes of patients in end-of-life care as much as possible. For example, in the United States, more than one-third of adults files any type of advance statement on LST [17] and most patients (91% to 100%) who chose a do-not-resuscitate (DNR) on the Physician Orders for Life-sustaining Treatment (POLST) are allowed a natural death without attempting CPR [18]. Additionally, UK’s research data show that 77.8% of patients died in their preferred place, with hospital admission rate decreased by 45% due to the decrease in hospital visits as per the patients’ wishes [19].

Second, the environment in which decisions are made as a family rather than as an individual should be considered. A previous study on intensive care unit doctors in Asia reported that among the factors that these doctors considered to be important in making LST-related decisions, the needs of the family accounted for 77.9%; it was only surpassed by the factors patient’s intention (84.1%) and long-term quality of life [3]. It is well-established fact that the physician’s perception of the intention of the family and the preference of the patient may sometimes be different than the patient’s original intention [1,20,21]. Personal characteristics or experience of the physician could also influence LST-related decisions, which inherently presents the risk of a decision being made contrary to the patient’s intention [11]. Therefore, it is necessary to focus on advance statement on LST to ensure the patient’s intention is respected. Indeed, our research team analyzed the completion rate of advance statement on LST based on the national data, there were 136 locations throughout South Korea who were filing advance statement on LST between February 2018 and October 2019. For every 100,000-population aged ≥ 19 years, only 973.8 persons had filed an advance statement on LST. Considering that one in three U.S. adults files any type of advance statement on LST [17], long-term policies for increasing the completion rate of advance statement on LST among adults aged ≥19 years are needed in Asia, including South Korea [20–23].

Another important issue is when an LST plan was prepared. According to previous studies, the mean number of days between completion of an advance statement on LST and death is 33 days in South Korea [11,24], whereas the period between completing a DNR order and death was within one week [9,25,26]. In other words, patients were making LST-related decisions due to imminent death without necessarily having enough time to deliberate and decide how they want to manage their own death. In particular, when the condition of a patient with altered consciousness deteriorates, it becomes difficult for them to make decisions as they may have normally wanted; instead, it becomes easier for the treatment to proceed based on the decision made by his or her family. Therefore, to prevent such situations from occurring, it is necessary to create an environment that promotes discussions about one’s own intentions in advance. In Oregon, the median time the POLST is completed before death has increased significantly, from 5 weeks (2010–2011 data) to 21 weeks (2015–2016 data), which is seen as the effect of the LST plans [27].

In the analysis of factors that influence self-determination on LST by the patient, the likelihood of self-determination about death increased with increasing age, as compared to patients aged < 30 years. However, this likelihood declined for patients 60 years or older. According to cause of death data from Statistics Korea, the primary cause of death among people aged 20–29 years was suicide, followed by traffic accidents [28]. Thus, deaths among people aged 20–29 years typically involve sudden death and not disease-associated death. Therefore, people aged < 30 years likely have less opportunity to deliberate and decide on LST in advance, as compared to other age groups.

Interestingly, the present study found that the likelihood of self-determination about death was higher among male patients without any disability, who reside in a national capital region, have a higher number of comorbidities, and was hospitalized in the preceding year. In particular, patients with a disability lacked systematic discussions about determination on LST, as compared to those without a disability. Moreover, patients with a disability have been found to face concerns about not receiving proper treatment despite serious disease conditions or their own intentions not being reflected in the determination on LST [29,30]. Despite patients with a disability having a higher prevalence of chronic diseases and greater need for healthcare services than patients without a disability, such issues were not actively handled in LST-related discussions [31,32]. Even in the new law in South Korea, discussion about people with disabilities is excluded [8]. Therefore, further contemplation about measures that can enhance the right to self-determination on LST is needed.

Insurance premium level was used in the present study as an indicator of personal economic status. The low- income group who are medical aid recipients showed higher self-determination tendencies than people who pay higher insurance premiums. This may indirectly reflect the economic burden of continuing LST, which suggests the need to improve economic accessibility with respect to healthcare utilization. There is a possibility that household composition, such as the presence of children or a spouse, influences self-determination. According to statistics from Statistics Korea [33] and Social Security Information System [34], single-person households accounted for 29.9% of the total domestic population and the ratio of single-person households to social assistance beneficiaries was 69.9%. In other words, the rate of self-determination on the LST could have been higher, as there were many medical aid beneficiaries from single-person households.

In conclusion, our study identified factors that influence self-determination on LST to gain a broader understanding about patient-centered decision-making in the end-of-life process. In particular, our study linked data from the National Agency for Management of Life-Sustaining Treatment and the NHIS in identifying the status of LST implementation at the national level. However, given that the system is still new, there were limitations in determining the long-term effects of such a system.

Nevertheless, we were able to confirm increases in the percentage of self-determination due to the establishment of the system, and above all, we identified factors that could influence self-determination, such as disability, economic status, and area of residence. These factors could be used to establish the basis for future policy development to further increase opportunity for patients to provide an advance statement on LST or an LST plan. For the development of proper institutional policies, continued monitoring of the LST decision system, implementing promotion policies and continued output of national statistics are needed. In addition, future studies should investigate the economic effects of such a system by analyzing LST-related medical costs, which could help strengthen justification for self-determination on LST among patients.