The judgment grounds in surrogate decision-making in Japanese clinical practice: A qualitative survey

adopted a on to reveal the judgement grounds in decision-making involving critical, in

only 29.7% have the capacity to make such a decision (2). Thus, more and more patients are expected to have someone else make decisions on their behalf (surrogate decision-maker) in the future.
In the United States, surrogate decision-makers are expected to base their decisions on the substituted judgment standard; that is, by considering what the patient, if competent, would choose (3). Buchanan and Brock introduced "a hierarchy of standards" for surrogate decision-making, which include the following three standards: a patient's known wishes, substituted judgments, and the patient's best interests (4). These standards provide guidance for surrogate decision-makers in deriving judgments and have thus far been considered the 'orthodox' view of surrogate decision-making in the eld of bioethics (5).
In practice, however, the judgment grounds in surrogate decision-making are not always considered in accordance with the hierarchy of standards. This has been pointed out in Japan as well as in other countries (6). For example, a study in Japan reported that surrogate decision-makers and physicians base their decisions on their own preferences (7). These reports suggest that patient preferences or best interests might not always form the grounds for judgment in surrogate decision-making. Notably, studies in Japan have been limited in terms of sample size as well as content pertaining to surrogate decisionmaking.In recent years, activities to promote advanced care planning (ACP) have gathered momentum in Japan. This may potentially affect the ways in which judgments are made in surrogate decision-making.
In 2018, the Ministry of Health, Labour and Welfare (MHLW) revised the "Guidelines on the decision making process for end of life care." This revision re ects the reality that activities and research efforts relating to ACP (commonly practiced in the United States and European countries) have also increased in Japan. The revised guidelines highlight the importance of discussing on a regular and repeated basis what the patient's intentions are regarding medical and care strategies as well as their desired way of living, with the premise that such intentions can change as their physical and psychological conditions change. In other words, the importance of actions relating to ACP is emphasized in the revised guidelines.
Another important point discussed in the guidelines is that the patients themselves should specify a surrogate decision-maker who would presume their intentions before they become incapable of communicating their wishes (8). The MHLW also organizes workshops for consultants across Japan as part of their jurisdictional project based on "Education For Implementing End-of-Life Discussion (E-FIELD)" for medical practitioners. The goal of these workshops is to develop a consultation system involving approximately 400 medical institutions nationwide to promote decision-making that respects patient preferences (9). The guidelines and workshops primarily advocate for the use of the standards proposed by Buchanan and Brock in eliciting grounds for judgments and recommend the use of ACP.
Dissemination of these guidelines could potentially change the judgment grounds in surrogate decisionmaking in Japan.The purpose of this study is to clarify the judgment grounds on which surrogate decisions are made in Japan, in the face of a rise in surrogate decision-making in this country. In particular, this study aims to reveal judgment grounds in surrogate decision-making involving critical, liferelated choices in acute hospitals. Finally, based on our results, we consider how the expected dissemination of ACP might change the judgment grounds in surrogate decision-making. Until now, reports on surrogate decision-making in Japan have been limited to those published within the country.
We think it is meaningful to report to the international audience the judgment grounds in surrogate decision-making currently used in Japan, where the aging of society is progressing further and activities relating to ACP are expected to intensify.

Study design
The objective of this study was to analyze the process of surrogate decision-making from the perspective of surrogate decision-makers in Japan, and to clarify the judgment grounds in surrogate decision-making in clinical practice. Given that each case is highly personal in nature, we predicted that only research methods that allow for careful analysis and examination of each decision-making process would clarify the reality of each situation. Accordingly, we adopted a qualitative research method that was based on semi-structured interviews.

Eligibility criteria
Participants of this study were individuals residing in the suburbs of Tokyo recruited through a recruitment company (10). Eligible participants were surrogate decision-makers of a patient who met all of the following ve conditions: (a) hospitalized during the period spanning April 1, 2012 through March 31, 2017; (b) had no capacity to make decisions for themselves; (c) aged 65 years or older; (d) required decision-making regarding life-sustaining treatment (dialysis, arti cial respirator, tube feeding, central venous hyperalimentation); and (e) meeting with the physician regarding treatment took place with individuals other than the patient. Exclusion criteria were as follows: (a) those who did not wish to participate in this study; (b) those with di culty in Japanese oral communication; and (c) underage individuals. Based on the study protocol and informed consent form, the recruitment company created a list of participants selected from a database in accordance with the inclusion and exclusion criteria. These participants were provided with informed consent form regarding the present study, and a list of contact information for those who expressed an intention to participate was submitted to the investigator of the present study.
Interview procedure Documents describing the contents of planned interviews were distributed to participants in advance.
Interviews were conducted by researchers at the date and time speci ed by the interviewees in a conference room of the study center (Tokyo Medical Center), where participant privacy was ensured. The interviews were conducted in line with the questions listed in the interview guide (Table 1).

Ethical considerations
This study was approved by the Ethics Committee of Tohoku University School of Medicine (Approval No. 2017-1-856). At the time of the survey, participants were provided with explanations regarding the survey.
Written consent was obtained from all participants regarding their participation, that the content of interview would be recorded, and that their statements would be reported anonymously.

Analysis methods
Analyses were performed using a qualitative analysis method by referencing the KJ method (11) and the "Ueno method"(12). The above equation method is a simpli ed version of the KJ method and was developed by Chizuko Ueno, Professor Emeritus, the University of Tokyo. There are two main features. (1) The analysis of all interview data may lead to unexpected and novel results. (2) Since the result is constructed by letting the data speak, it is possible to claim with evidences. The speci c method is as follows.
First, interviews were recorded using an IC recorder, and a verbatim report of the interview data was created. The interview data were analyzed by the following procedures.
All sentences in the raw data were subject to analysis; all researchers involved in the analysis individually read all sentences carefully. Sentences or portions of sentences relating to the same content were coded. At the time of coding, no attempt was made to simplify expressions. Codes with similar content were grouped into a subcategory and given a name that represented the shared content. When creating subcategories, consideration was given to make the meaning easily understandable (i.e., just by reading the subcategory). Similar subcategories were grouped into categories and then into core categories, with increasing levels of abstraction. The researchers veri ed the content validity by repeating discussions until a consensus was reached regarding classi cation as well as coding, in order to ensure the reliability and validity of the analyses. Analyses were performed by a multidisciplinary group of researchers including 2 physicians, 1 nurse, 2 philosophers, and 1 pharmacist.
Since the codes (sentences) derived using the present analysis method were long, it was not feasible to present them all in this report. Accordingly, we excerpted important parts within each code as appropriate, omitting some parts without changing the meaning. Data were analyzed by MAXQDA Plus12 (Release 12.2.1)software.

Overview of study participants
We received a list of 228 participants who met the inclusion criteria. Of these, 15 participants were selected from among those who were available for interviews at a date and time that was convenient for both researchers and participants (i.e., interviewees). It should be noted that the number of individuals pooled by the web survey company, to whom survey requests could be sent, was not disclosed. Interviews with the 15 participants were carried out over the course of 6 non-consecutive days (up to 3 interviewees per day) in November and December 2017. Of the interviewed participants, 14 were subject to analysis after excluding one who was a family member of a patient, who supported the surrogate decision-maker but did not actually make decisions on the patient's behalf. Details on participant attributes and other background characteristics are summarized in Table 2. All surrogate decision-makers who participated in this study were family members of the respective patients.

Analysis results
Analysis results of interviews regarding the judgment grounds in surrogate decision-making are summarized in Table 3. A total of 4 "core categories", 17 [categories], 35 < subcategories>, and 55 (codes) were extracted.
Type 1: Core category "Patient preference-oriented factor" The judgment grounds rooted in patient preferences were classi ed as the Type 1 core category. This core category comprised 2 categories, 8 subcategories, and 13 codes. Representative categories/subcategories/codes are described below.
[I respected the preference of the patient] One of the subcategories of this category was < Since the patient's preferences were clear, my decisions never swayed>, which included the following code: (I had conversations with the patient in advance. We often talked about when the patient was going to die, half-jokingly. The patient also mentioned speci c matters, such as not wanting to live with the help of various machines connected to the body). In this case, the patient mentioned speci c treatment choices in prior discussions, and the surrogate decisionmaker respected those as the judgment grounds in surrogate decision-making.
[I respected the presumed intention of the patient] One of the subcategories of this category was < I made the decision, thinking what the patient would do >,which included the following code: (We as family members tried to put ourselves in the patient's place. We wondered which one of the choices my father would pick after hearing what the doctor had said, had he been able to make his own decision). This code re ected the attitudes of the surrogate decision-maker who tried to gure out what the patient's preferences might be, from the patient's perspective.
Type 2: Core category "Patient interest-oriented factor" The judgment grounds rooted in patient interests were classi ed as the Type 2 core category. This core category comprised 4 categories, 12 subcategories, and 20 codes. Representative categories/subcategories/codes of this core category are described below.
[I tried to make the decision by considering the patient's best interests] This category included the subcategory < I thought that what would be good for the patient would be to receive medical treatment and recover>, which contained the following code: (What I thought would be good for the patient was, for example, to be able to lead a normal life as before, even if it is somewhat inconvenient. I thought any decision that would allow for this would be in the best interest of the patient and was a good decision). This surrogate decision-maker thought that a treatment option that allowed for the patient to live as usual would be in line with the patient's best interests and used this as the basis for judgment in surrogate decision-making.
[I did not want to do anything cruel to the patient] This category included the subcategory < I decided against life-prolonging treatment out of pity>, which contained the following code: (To be honest, we as family members just felt sorry for the patient, whom we couldn't even recognize anymore, and since we were no longer able to have a conversation, we did not know how much the patient was understanding what we were saying -so we did not choose lifeprolonging treatment. We clearly communicated these thoughts with the doctor and made the decision). As the patient became increasingly ill, the surrogate decision-maker judged that the patient's dignity was not being preserved; this formed the basis for the judgment to tell the physician that life support was not desired.
[I made the decision based on the patient's ADL and my communications with the patient] This category included the subcategory < I thought the patient would nd it painful to live in a vegetative state>, which contained the following code: (I might come off as an ungrateful child if I say this, but my feeling was that, rather than living in a vegetative state at age 87, the patient would be better off just dying. ... Living in pain connected to numerous tubes, just lying in bed and sleeping for 1 year, or 2 years -how pitiful, I thought, if that's what it comes to). The surrogate decision-maker felt sorry for the patient living with signi cantly reduced ADL, given the patient's age. Such a thought could potentially lead to a decision that shortens the time to death of the patient. This code also re ected a sense of guilt associated with making a surrogate decision based on the family's preferences. Type 3: Core category "Family preference-oriented factor" The judgment grounds rooted in the preferences of the surrogate decision-maker, who is a family member of the patient, were classi ed as Type 3 core category. This core category comprised 5 categories, 13 subcategories, and 17 codes. Surrogate decision-makers made decisions on behalf of the patient based on their (family's) own preferences, rather than considering the patient's preferences. In some cases, the surrogate decision-maker was unaware of the patient's preferences originally, while in other cases, the surrogate decision maker was aware of the patient's preferences but chose not to consider them, prioritizing their own preferences.

[I wanted to protect my family's life and interests]
This category included the subcategory < I realistically considered the lives of family members and decided to forgo gastrostomy>, which contained the following code: (I thought 'I must look to the best interests of my father,' but realistically speaking, my younger sister, the second daughter, had small children and was running her own business. Her life would have been affected if she did not work. As the eldest daughter, I myself was also unable to leave the house for a long period of time because I was raising my children. Therefore, it was not at all realistic for me to provide home care. I shut my eyes to his pain and wishes and decided not to have him receive gastrostomy in consideration of continuing medical treatment at the hospital). While this surrogate decision-maker wished to prioritize the patient's preferences, she had to make the decision that did not go along with the patient's preferences in light of the realistic circumstances surrounding herself as well as other family members.
[I made the decision based on the thoughts of family members and people close to the patient] This category included the subcategory < The feelings of the closest family member were important>, which contained the following code: (I needed to convince my mother-in-law, who was closest to the patient. I thought that, rather than us (the son and his wife) making decisions against her will, she should make decisions that she is satis ed with, after she has organized her own thoughts. For this reason, it took a lot more time to come to a decision, and I'm afraid my father-in-law suffered for a prolonged period). This code describes a surrogate decision-making process in which the surrogate decision-maker secured the time necessary for the family to agree with the decision. However, this in turn increased the time that the patient was in pain.
[I wanted the patient to live] This category included the subcategory < When the death suddenly became a real possibility, I as a family member wanted to prolong the patient's life>, which contained the following code: (The shock was tremendous when the doctor told us that death was inevitable, as the patient's condition worsened. At that time, I honestly just thought, 'I want the patient to live, even a day longer,' and it didn't matter if gastrostomy, or anything, had to be done. It was hard for the family to have to say goodbye all of a sudden, so I wanted the patient to get better, even just a little. I was always prepared, to no small extent. But when a doctor talks about life-or-death, you can't help but think "please just help the patient"). When the death of the patient became a real possibility with worsening of the patient's health, the hope of the surrogate decision-maker to prolong the patient's life by even one day formed the basis for judgment in surrogate decision-making. Type 4: Core category "Balanced patient/family preference-oriented factor" The core category classi ed as Type 4 was the reasoning related to an attempt to balance preferences of the patient and those of the surrogate decision-maker (i.e., family). This core category comprised 1 category, 2 subcategories, and 5 codes.

[I balanced the patient's intention and the lives of family members]
This category included the subcategory < I made the decision by considering the balance between the patient's life and lives of family members>, which contained the following code: (I had mixed feelings when I had to decide about the patient's nutrition. Considering the burden on my brother and his wife who actually provided care, I wondered how my decision might affect their lives. On the other hand, I also had to think about the feelings of my father who wanted to recuperate at home. It was hard at the time of decision-making. I was particularly worried about the burden on my sister-in-law). As suggested by this code, the surrogate decision-maker made decisions while considering the patient's wish to receive home care, as well as the burden on the lives of family members who provide the care. On these grounds, the surrogate decision-maker ultimately decided on gastrostomy as a means of nutrition support, which was not in line with the patient's wish to receive home care. This decision was also made in order to reduce the burden of care.

Discussion
In this study, we identi ed four judgment grounds that lead to decision-making by surrogate decisionmakers in Japan. With respect to Type 1-3 factors, in view of the standards proposed by Buchanan and Brock for guiding surrogate choices, the Type 3 factor represents one that must be disregarded to the extent possible in the reasoning process leading to surrogate decisions. However, the results of the present study revealed just how di cult it is to eliminate this factor. In the following sections, we analyze this factor, which re ects the di culty of making surrogate decisions based on the patient's preferences and/or best interests, while considering the background of Japanese society. We also comment on the in uence of ACP that is expected to become more widely used in the future.
<Culture in which conversations about EOL rarely occur> One of the subcategories extracted from the results of the present analysis was < I did not know what was good for the patient>. This indicated a struggle that, no matter how hard the surrogate decisionmaker tries to guess the patient's preferences, there is no way of knowing what the patient would actually choose, or how (i.e., based on what values). During the interviews, surrogate decision-makers described di culties presuming the patient's preferences. In other words, it is di cult for surrogate decision-makers to see things from the patient's point of view.
One factor contributing to this di culty is a culture in Japan that does not encourage having speci c conversations about EOL (end of life). Such talk is generally considered bad luck and even taboo in some families. According to actual data, only 5.5% of Japanese citizens reportedly talk about medical treatment in EOL situations with their family or medical care personnel, and only roughly 8% put their intentions in writing beforehand (13). Thus, it is likely that the number of surrogate decision-makers who clearly recognize the patient's preferences is low.

<Changes in social circumstances>
In 1980, almost 70% of elderly people aged ≥ 65 years lived with their children. By 2015, this rate had signi cantly decreased to 39.0%. Also, the rate of double-income families, which was 49.3% in 1980, has been rising year by year, and reached 64.4% in 2015. (14)These data suggest an increase in the number of adults (i.e., offspring of elderly individuals) who are not at home all day. With respect to the degree of communication between parents and children who do not live in the same house, Japan reportedly has the lowest frequency of older individuals meeting or contacting (e.g., by phone) their non-cohabiting children, relative to the United States, Germany, and Sweden (15). Although these international comparisons are based on a small number of countries, in Japan, the frequency of communication between elderly individuals and their non-cohabiting children or other family members may be low, according to international averages.
Based on these reports, we predict that in recent decades, it has become less common for children (i.e., potential surrogate decision-makers) to share time and space with their parents (i.e., patients) on a daily basis. This suggests that children in this generation may not be able to easily understand or imagine how their parents live, or what they value in their daily living. This may become an obstacle when they need to gure out what the patient preferences are as a surrogate decision-maker. It is also possible that this situation underlies the practice of surrogate decision-making for which factors other than the patient's preferences or best interests form the judgment grounds. All 14 of the surrogate decision-makers included in this study were children of patients or the children's spouses. Although their working statuses are unclear, the rate of cohabitation was 20%. Thus, their circumstances might have made it di cult to imagine the life and values of the patient. <Time restriction in surrogate decision-making > When considering the judgment grounds in surrogate decision-making, it is likely that time restrictions have some in uence. According to a report from the United States, 48% of surrogate decision-makers had to make critical decisions about life-sustaining treatment for patients aged ≥ 65 years within 48 hours after hospitalization in acute hospitals (16). Thus, in acute hospital settings, surrogate decision-makers may be forced to make these decisions in a short period of time. If a surrogate decision-maker was to make decisions in such a short time frame on behalf of an elderly patient who developed a serious lifethreatening disease, to what extent would the surrogate decision-maker weigh the patient's preferences? There were some cases in the present study in which the family did not choose life-prolonging treatment (e.g., <I judged it realistically impossible to provide home care>). We presume that, in a setting that requires judgment regarding treatment options related to life support, it can readily be envisioned that the life of the surrogate decision-maker (family) would be affected somewhat depending on outcomes after treatment, especially when the patient's condition is unfavorable. In such situations, the surrogate decision-maker might make a hasty decision about which treatment to choose, thinking it realistically impossible to bear the burden of care, given their own life circumstances.

<Novelty of Type 4 factor>
The Type 4 factor re ects the reasoning of surrogate decision-makers who consider not only preferences of the patient but also those of family members in an effort to balance the two. Many previous reports examining the judgment grounds in surrogate decision-making only introduced one basis for judgment per case of surrogate decision-making, i.e., one that serves as the core of decision-making. On the other hand, the present study identi ed 3 types of factors that are not necessarily mutually exclusive; in fact, our ndings suggested the possibility that in actual decision-making, multiple types of elements might be considered in reasoning and deriving surrogate decisions. The present study analyzed data obtained from an interview survey pertaining to the entire process of surrogate decision-making up to the judgment stage. For this reason, multiple judgment grounds were identi ed for each case of surrogate decisionmaking. From 14 cases of surrogate decision-making subjected to analysis, 55 codes relating to the judgement grounds were extracted. We presume these codes were considered in combination in actual settings of surrogate decision-making, and perhaps in a comparative manner. Those included in the Type 4 factor were categorized separately from others intentionally, since Type 1-3 factors re ected single judgement ground, whereas Type 4 re ected the outcome of comparative weighing of multiple grounds. In the United States, where patient autonomy is valued in most cases, there have been reports that, in actuality, surrogate decision-makers derive decisions based on their own values and circumstances in some cases (17)(18). In Japan, the present report is the rst to address this issue.
<Concerns about the potential psychological di culties in surrogate decision-making as the consequence of wide use of ACP in Japan> ACP, which is expected to become more prevalent in the future, has been suggested to activate communication between physicians and surrogate decision-makers (19). While the widespread use of ACP is desirable, there are concerns that the increased use of ACP might complicate the process of considering judgment grounds in surrogate decision-making and also increase the psychological burden on surrogate decision-makers. In other words, by clearly recognizing patient preferences (more than they already do), the surrogate decision-maker may end up with more con icts in surrogate decision-making. This is because, as demonstrated by the present study, there are actual situations in which surrogate decision-making is practiced based on preferences of the surrogate decision-maker, which differ from those of the patient.
In Japan, patients rarely talk about their own preferences and values. Having advance discussions more often would allow the surrogate decision-maker to be more aware of patient preferences than they have in the past. This may help identify judgment grounds in surrogate decision-making that are rooted in patient preferences and best interests. However, this may also lead to a clearer awareness among surrogate decision-makers of the fact that their preferences may differ from those of the patient. As a result, we worry that surrogate decision-makers might become more con icted as they struggle to decide whether to prioritize patient preferences or those of their own. This is a situation that can be inferred from the current state of surrogate decision-making as revealed in the present study, i.e., decisions are made on bases rooted in the preferences of the surrogate decision-maker. We surmise that clarifying patient preferences does not necessarily result in a situation in which those preferences are smoothly prioritized, but rather causes a struggle in surrogate decision-makers who must consider the preferences of both patients and their own families. Whether such struggles are good or bad is beyond the scope of this discussion. Nonetheless, it should be noted that this struggle may complicate the process of surrogate decision-making. Rather than focusing solely on the principles of respect for patient autonomy and standards of decision-making, the judgment grounds in surrogate decision-making should be discussed while considering multiple factors including culture, social situation, and circumstances of surrogate decision-makers. Such approaches in decision making should be allowed for future surrogate decisionmakers and medical practitioners in Japan. Given that some patients think it is permissible that, in addition to their own preferences, preferences of the surrogate decision-maker may be prioritized in the surrogate decision-making process, as revealed by previous studies (20-21).

Strengths and limitations
The present study analyses were performed by a multidisciplinary group of 6 professionals including non-medical practitioners (physician, nurse, pharmacist, philosopher). Discussions were carried out among these analysts from various perspectives, making this system more favorable compared to those adopted by previous studies. In addition, in devising the analysis method, we referenced the "Ueno method" which is based on the KJ method. This allowed us to analyze the entire process of surrogate decision-making and clarify the judgment grounds. The "Ueno method" is superior to other methods in that it allows for the analysis of entire interview contents without omission.
This study also has limitations worth noting. First, since the recruitment process was outsourced to a web research company, interview respondents were limited to Internet users. In addition, participants were restricted to those living in the suburbs of Tokyo due to the location of the interview site. Potential bias also exists as the detailed characteristics of surrogate decision-makers, such as the number of years of care experience, educational background, economic status, religion, and family composition of the patient other than the surrogate decision-maker, were not available. However, as we analyzed data from the 14 participants, we achieved theoretical saturation of concepts extracted as judgment grounds in surrogate decision-making. Thus, we did not increase the sample size any further.
The second limitation is recall bias related to the timing of the interviews. The interview survey was performed within 6 months to 3 years after surrogate decision-making. Due to the time lag, interview contents might have differed from actual events. However, given that experiences of surrogate decisionmaking might be connected to grief, ethical consideration was given such that interviews were performed after a certain amount of time had passed.
Finally, there was also a limitation regarding the "Ueno method." Although this method has the analytical advantage discussed above, it has not been validated internationally. No English description is available, and no studies using this method have been reported internationally.
Despite these limitations, the present study provides novel insights into the judgment grounds in surrogate decision-making. A large-scale cross-sectional study on this topic based on the present results would help clarify the actual diversity and its frequency of grounds used for judgment in surrogate decision-making in Japan.

Conclusions
This study revealed the current status of surrogate decision-making in Japan: when making decisions about important matters related to a patient's life, surrogate decision-makers base their decisions not only on the preferences and best interests of the patient, but on their own preferences as well. Included in the preferences of surrogate decision-makers were their own views of life and death, their values, and care burden.
In Japan, ACP is likely to become more prevalent in the future. ACP will provide a valuable source of information and is bene cial in terms of respecting patient autonomy. However, given the cultural and social backgrounds in Japan, it remains unclear whether this information can be properly re ected in judgment grounds in surrogate decision-making. It would be undesirable to base judgments solely on the principle of respect for autonomy or the principles of surrogate decision-making originating from the United States. In the future, we believe that surrogate decisions-makers will be required to consider the judgment grounds from a more diverse perspective and that such attitudes should be ethically accepted.

Declarations
Ethics approval and consent to participate This study was approved by the Ethics Committee of Tohoku University Graduate School of Medicine(2017-856).

Consent for publication Not applicable
Availability of data and material The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.
Competing interests The authors declare that they have no competing interests. Authors' contributions All authors MT,KO,AE,TO,AA contribute to the analysis and thesis writing, and they have read and approved the nal manuscript.