Introduction

Brain death, the neurological standard that allows a declaration of death in people with irreversible loss of whole brain or brain stem (in UK) function, has been legally endorsed and routinely practiced in the US and other Western countries for half a century [1]. The brain death doctrine confers social utility by permitting the legal consequences of death, initiating bereavement practices, preventing futile care, and supplying organs for transplant [2]. Currently, 15,000–20,000 people are declared brain dead in the US each year, and approximately half of them donate organs, accounting for nearly 90 % of all deceased donors [3].

In contrast, many Asian countries have resisted incorporating brain death into medical, social, and legal realms [4]. In China, the legalization of brain death has been unsuccessfully attempted several times. In June 1986, the neurointensive care community convened in Nanjing and proposed a guideline for diagnosing brain death [5]. In May 1999, the Chinese Medical Association and the China Transplantation Development Foundation co-sponsored a conference in Wuhan during which a brain death bill was drafted [6]. Both documents were repudiated by the Ministry of Health, who started to draft its own brain death criteria for adults in 2003, the year the first confirmed brain death case in the country was reported [7, 8]. The Chinese Organ Transplant Act was finally passed in 2007, although without any mention of brain death [9]. In 2009, the Ministry of Health put together a second draft of the brain death diagnostic criteria, but the lack of legal recognition limited its use [10]. It appeared that brain death legalization in the near future was unlikely, and therefore the Chinese transplant community announced a halt to achieve it, and the focus was shifted onto donation after cardiac death in the fall of 2013 [11]. To date, there is no national policy regarding brain death, however a handful of cities allow hospitals to declare brain death according to institutional protocols. Without formal nation-wide records, the published brain death cases add up to less than 200 [12]. The 61 documented organ donations after brain death account for a negligible portion of all deceased donor organ transplants in China [12].

Cultural, philosophical, and religious barriers have been quoted to explain the lack of a legal and official brain death doctrine in China, often in overarching statements such as, “The concept of brain death is not fully accepted in China due to long-standing cultural traditions,” [12] and “Cultural resistance is a real issue.” [13] Borrowing from medical anthropology works done on Japan [14, 15], scholars have speculated that in contrast to the body-soul dichotomy and individual autonomy valued in the West, the integrated life view in Buddhist and folk beliefs, and the Confucian social structure built upon interpersonal relationships deter the Chinese people from accepting brain death [4]. However, little empirical research has been done to test this hypothesis.

While several public opinion surveys have clearly shown that the approval rate of brain death is significantly lower in China than in the West, none probed what motivations underlie the responses. For example, in large demographically representative adult samples, 60–71 % of people in the US, UK, Germany and France considered brain death an appropriate standard for human death [16], compared to 33 % in China [17]. Even among the Chinese hospital staff, only 35 % accepted brain death [18]. Moreover, 66 % of hospitalized patients in Kunming, China, believed a brain dead person to be still alive [19], as did 40 % of professionals and government officials in Guangdong [20]. As a reference, 16 % of Americans considered someone declared brain dead alive [21]. How much religious and cultural values weigh in the observed differences remains unknown.

Interestingly, a considerable fraction of the Chinese respondents, for instance 57 % in one study, compared to 6–19 % of Westerners surveyed concurrently, either have not heard of brain death or were not sure what the term meant [16, 17]. These data suggest that confusion and misunderstanding about the concept may contribute to the apparent resistance.

To better understand the factors influencing the Chinese perception of brain death, we designed the current study. We decided to target medical providers because of their privileged role in diagnosing brain death, declaring death on brain dead patients, explaining the situation to families, and directing decisions to withdraw care and harvest organs. As mentioned above, in previous surveys, Chinese medical providers had low levels of acceptance for brain death similar to the patient populations, raising concerns for an underlying cultural barrier. Support from medical professionals would be crucial for the legalization of brain death. In the West, the Harvard Report and Declaration of Sydney by the World Medical Assembly, both published on the same day in 1968 signifying the medical community’s affirmation of the neurological criteria of death, preceded the legalization movement [22]. The first law recognizing brain death, passed in Kansas in 1970, was proposed by a physician legislator [23]. Medical professionals also play a pivotal role in educating the public and promoting the clinical application of brain death following legalization, especially in developing countries like China where health literacy is relatively low and patients rely on the providers for guidance in medical decision-making [24]. By evaluating the medical providers’ knowledge, understanding, and both the theoretical and practical acceptance of the brain death doctrine, we were able to correlate their decisions with demographic and cultural backgrounds, and thus elucidate some of the ethical and social determinants for the medical practice of brain death in China. Furthermore, our work may help identify specific obstacles in the road to legalization of brain death in China.

Methods

Statement of Human Rights and Informed Consent

This study was approved by the Institutional Review Boards at Yale School of Medicine (New Haven, USA) and Central South University (Changsha, China) and performed in accordance with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Verbal informed consent was obtained from all individual participants included in the study. Written informed consent was not required for this type of study.

Study Population and Data Collection

The study population consisted of medical providers, including medical students in their clinical years, residents, attending physicians, nurses, and nurse supervisors, from three academic hospitals affiliated with Central South University Xiangya School of Medicine in Hunan, China. In China, one becomes eligible for licensure in medical practice after 5 years of undergraduate study culminating in the bachelor of medicine degree. One may continue pursuing academic degrees at the master’s and doctorate levels through additional 3 to 6 years of graduate work in clinical medicine. Accelerated 8-year combined programs for the doctor of medicine degree are also available. Xiangya School of Medicine offers all four pathways of training. Both undergraduate and graduate medical students were included in our study. The Chinese medical students carry substantial clinical responsibilities and work fulltime in direct patient care, are therefore regarded as providers. Participants were recruited from the departments of neurology, neurosurgery, neuroscience intensive care unit (ICU), medical ICU, internal medicine, psychiatry, transplant surgery, general surgery, traditional Chinese medicine, and emergency medicine. The authors (QY, YF, QC, XL) visited the hospitals and distributed an anonymous paper survey with voluntary participation. It was estimated that at least 63 each of students, physicians, and nurses were needed for discerning 10 % difference at Σ = 0.2, α = 0.05, and power = 0.8 among the three groups. A total of 500 surveys were distributed and 476 were collected (response rate 95.2 %) and this constituted the convenience sample for the study.

Survey Design

The survey was designed to examine the respondents’ demographic background, religious and cultural beliefs, life values and beliefs about death in general, the diagnosis of brain death, and medical decision-making. The brain death knowledge test had 12 questions covering the defining characteristics of brain death such as apnea, irreversibility, lack of brain stem reflexes, lack of consciousness and voluntary movement, preservation of spinal and peripheral nerve system functions, and distinction from cardiac death and the vegetative state [25]. This included the five core questions (question number 1–5 on Table 3) previously demonstrated to have strong test-retest reliability (Spearman’s ρ 0.91), internal consistency (Cronbach’s α 0.64), and ability to discriminate clinical experts (mean score 4.8 ± 1.1) from laypersons (mean score 3.0 ± 1.1) [26]. The rest of the test items were adapted from a study in Hong Kong, which examines more detailed understanding of the brain death definition [25]. Correlations between an individual item, the total score, and scores for the 5 core items were analyzed for the current study population.

In the medical decision-making section, respondents were shown two scenarios modeled after Siminoff et al., who had designed the instrument based on focus groups and validated it in a large American sample [21]. One scenario described a brain dead patient (#1), the other a patient in the persistent vegetative state (PVS, #2):

  1. #1

    A patient is on a ventilator. There is no detectable brain function. He does not open or close eyes, does not move or withdraw from painful stimuli. There is no brain activity on electroencephalogram (EEG). There is no blood flow into the brain. He still has a heartbeat and feels warm to the touch.

  2. #2

    A patient has been in a nursing home for 5 years following a severe brain injury. He is not on any life support machines. He can breathe on his own, but needs to be tube fed. He has sleep-wake cycles but no awareness. He does not respond to people or things around him, except for withdrawing to pain. There is some brain activity on EEG.

The scenarios were further differentiated into hypothetical conditions ascribing the patient’s age, insurance status, and relationship to the provider. For each scenario, the respondent was asked to decide (a) whether the patient was dead, (b) whether to withdraw supportive measures such as mechanical ventilation and enteric nutrition, and (c) whether to permit the removal of organs for transplant from the patient in the described state. Consistency of logic in the answers was analyzed by correlating the respondents’ decisions for the hypothetical patients with answers to the knowledge questions and subjective rating of importance of psychosocial factors.

Analysis

The survey responses were independently coded into a Microsoft Excel spreadsheet by two investigators and discrepancy checked by a third to ensure accuracy. A majority of the respondents answered all questions; 106 (20.8 %) left one to two questions blank; and 80 (16.8 %) had more than two blanks. All surveys were analyzed. For each analysis, only the valid answers for the particular questions were used. Statistical analysis was carried out in SPSS v.21 for Windows (IBM). Descriptive statistics were conducted for comparison between groups. Continuous variables, such as the knowledge test score, were subjected to an independent sample t-test between two groups, and ANOVA among multiple groups. Effect size was calculated with Cohen’s d for t-test and η2 for ANOVA. Categorical variables and proportions were subjected to a Chi-square test. All reported p-values were two-tailed and considered statistically significant at p ≤ 0.05. For inter-group comparisons, post-hoc power analysis was carried out to confirm sufficient sample size in each group.

Linear regression was used to discern correlation between respondent characteristics and performance on the brain death knowledge test. Logistic regression was used to discern correlation between respondent characteristics and ethical acceptance of brain death or decision to acknowledge brain death. A positive predictor had an odds ratio above 1. Variables achieving p < 0.20 in the univariate analysis were entered into the multivariate model and retained with p ≤ 0.05. Effect size of the model was evaluated by Nagelkerke’s R2 and goodness of fit was evaluated by Hosmer and Lemeshow test as well as percentage of correct prediction. The significance of each contributing variable was expressed as by the Wald statistic and odds ratios given when appropriate. The indirect effect of select factors was evaluated by mediation analysis using the PROCESS SPSS macro developed by Hayes et al. (available from http://www.processmacro.org/) [27]. Since the main outcome of our study, the operational acknowledgement of brain death, was dichotomous, a formal structured equation modeling was not suitable.

Results

Respondent Characteristics

The survey was completed by 476 medical providers practicing in three academic hospitals in Hunan, China. Our sample included 210 medical students, 156 physicians (84 residents and 72 attendings), and 110 nurses (Table 1 and Supplementary Table 1). The medical students were doing full time clinical work pursuing bachelor (39.8 %), master (20.1 %) or doctor (40.1 %) of medicine degrees. Overall, 31.6 % were male, and 56.9 % were under the age of 25 years. The medical students and nurses were young and predominantly female, whereas the physicians tended to be male and older. The physicians also completed more education than nurses (57.7 % vs. 2.7 % with graduate degrees). Only 6.8 % of the respondents have lived, worked, or studied abroad, with the United States, Australia, and United Kingdom among the most popular destinations. The specialties represented were the clinical neurosciences (neurology and neurosurgery, 27.3 %); intensive care (12.6 %); other medical areas (14.9 %), e.g. general medicine, hematology, and psychiatry; other surgical areas (12.0 %), e.g. general surgery, transplant surgery, and emergency medicine; and traditional Chinese medicine (TCM, 19.1 %). While 14.1 % admitted to having religious faith, 36.2 % reported visiting places of worship once a year or more, most commonly Buddhist temples, and 39.5 % believed that humans have a soul. The support for organ donation was wide spread at 92.8 %.

Table 1 Respondent characteristics (N = 476)
Full size table

Experiences and Beliefs

Almost all (95.2 %) of the respondents had heard of the term “brain death” prior to this study. More than half (57.8 %) reported having seen patients who were in the “brain dead state” and 16.9 % had declared such patients as “brain dead” (Table 2). Physicians and nurses were more likely to have heard of brain death and seen brain dead patients than students, and physicians were most likely to have made brain death declarations. As expected, those in the neuroscience specialties were the most experienced with brain death and TCM providers were the least experienced (Supplementary Table 2).

Table 2 Experiences and beliefs
Full size table

Most respondents acknowledged the cardiopulmonary criteria of death, recognizing the lack of heartbeat or pulse (67.2 %) and spontaneous respiration (34.2 %) as signs of death. Over half (54.9 %) also considered the lack of detectable brain activity on EEG a sign of death. A small number accepted other brain-centric signs such as the lack of consciousness or motor function. A large majority (83.8 %) felt confident about the ability of existing diagnostic tools in accurately discerning that someone is brain dead.

When asked whether they considered brain death an ethically acceptable standard to determine human death, 65.5 % answered yes. The ethical acceptance was high among physicians (78.0 %) and low among TCM providers (38.9 %). Nearly half (47.6 %) of the respondents were unsure about the legal status of brain death in China; 33.5 % correctly thought it was illegal and 18.9 % believed it was legal.

Medical Knowledge About Brain Death

We administered a 12-question test on the medical characteristics of brain death, including apnea, irreversibility, lack of brain stem reflexes, lack of consciousness, presence of heartbeat, preservation of spinal cord activity, differentiation from the vegetative state, and possible chronicity (Table 3). We included the five core questions developed and validated by Tawil et al. (Question 1–5, Table 3) [26], as well as the questions used in a previous study from Hong Kong (Questions 6–12, Table 3) [28]. The combined 12-item test demonstrated adequate internal consistency (Cronbach’s α 0.627). In our sample, the mean score was 8.50 ± 1.83 for all 12 questions and 3.76 ± 1.08 for the five core questions. The 12-point and 5-point scores had strong correlation with each other (Pearson coefficient 0.766, p < 0.001, Supplementary Table 5). Each item on the knowledge test had significant positive correlation with the overall scores, except for #8 and #9, which were also the most commonly missed questions and both related to the preservation of spinal cord, peripheral, and autonomic nervous system functions in brain death. Notably, 64.6 % correctly answered that brain death is not the same as being “vegetative”, and 68.7 % correctly answered that a person can be declared brain dead with a beating heart. Male gender, working in the neuroscience specialties, and having seen brain dead patient previously strongly correlated with higher core scores and overall performance on the knowledge test (Table 4). Older age, having had graduate education, and having lived abroad also correlated with correct answers on the core questions. Each of the factors had small to medium size effects on the overall variation in the knowledge test performance. As several of these attributes were predictive of being a physician, the scores were re-analyzed within the physician subgroup (Supplementary Table 3). Although the trend in gender, age, and education was still observed, the differences were no longer significant. The significant effect of specialty and previous encounter with the brain dead remained. Among all participants, neuroscience attending physicians (n = 35) scored the highest points, 9.55 ± 0.89 for all questions and 4.63 ± 0.70 for the five core questions.

Table 3 Responses to knowledge questions about brain death
Full size table
Table 4 Brain death knowledge score
Full size table

Medical Decision-Making on Brain Death and Persistent Vegetative State

When presented with the scenario of a brain dead patient, 50.7 % of all respondents agreed that the person was dead, 51.9 % of all would withdraw supportive measures, and 40.6 % of all would permit the removal of organs for transplant (Table 5). In contrast, only 7.4 % considered the patient in persistent vegetative state (PVS) dead, although 13.4 % would withdraw support and 11.1 % would permit organ procurement. Correctly answering that brain death was different from the vegetative state in the knowledge test was predictive of differentiating between the brain dead and PVS patients in the scenarios (different answers for death declaration: OR 1.731 95 % CI 1.444–2.075, withdrawal: OR 1.674 95 % CI 1.390–2.017, organ removal: OR 1.706 95 % CI 1.396–2.084, p < 0.001 for all three questions).

Table 5 Medical decision-making in patient scenarios
Full size table

When given an 80 year-old versus a 20 year-old brain dead patient, respondents were much more likely to acknowledge death (60.3 % vs 45.3 %, p <0.001) and withdraw support (72.8 % vs 40.0 %, p < 0.001), but less likely to permit organ removal (33.1 % vs 43.5 % p = 0.001) from the older patient. When given an uninsured patient who must pay all medical expenses out-of-pocket versus a fully insured patient who has no out-of-pocket expenses, respondents were also more likely to acknowledge death (55.0 % vs. 50.5 %, p = 0.180) and withdraw support (56.7 % vs. 49.5 %, p = 0.033) from the uninsured patient, but only the decision to withdraw was significant in a head-to-head comparison with the fully insured. In all the aforementioned four conditions, the patient was a stranger. In comparison, if the patient was a member of their own family, the respondents were more reluctant to acknowledge death (42.5 % vs 52.7 %, p = 0.003), withdraw support (40.0 % vs 54.8 %, p < 0.001), or permit procurement (34.7 % vs 42.1 %, p = 0.026). Patient being a family member was the only condition that prompted different answers for 51 respondents, suggesting that personal relationship to the patient can independently sway decisions regarding brain death.

To examine the conscious considerations involved in the decision making process for the brain dead patient, we asked the respondents to rate the importance of several psychosocial factors on a 0–5 Likert-type scale (Table 6). Legal status of brain death was rated the most important with a mean score of 4.25 ± 1.26, followed by emotional attachment between the patient and family (3.70 ± 1.52), and financial considerations (3.38 ± 1.77), medical liability and defensive medicine (3.32 ± 1.57), and educational level of the patient’s family (3.05 ± 1.76). Religion (2.59 ± 1.86) and “face”, the Chinese concept of the social image one ought to maintain for others (1.67 ± 1.60) were regarded the least important. Although there was no difference in the overall rank order of the factors, nurses give more points to emotional attachment and financial considerations than physicians. Both physicians and nurses gave medical liability and education more importance than students. The rating for religion was slightly higher in those with religious faith (3.10 ± 1.835 vs 2.51 ± 1.897, p = 0.690). The subjective ratings were consistent with how the factors played into the patient scenarios. For example, those who gave a high rating of 4 or 5 for financial considerations were more likely to make different decisions for the uninsured and the fully insured patients, especially regarding care withdrawal (OR 2.782 95 % CI 1.502–5.150, p = 0.001). On the other hand, those who gave a high rating for emotional attachment were more likely to make different decisions for the patient who was a family member, although the trend did not reach statistical significance.

Table 6 Perceived importance of psychosocial factors in medical decision-making about brain death
Full size table

Predictors of Brain Death Medical Decision-Making

The proportion of medical providers who considered the hypothetical brain dead patient dead (“operational acknowledgement of brain death”) was lower than what we expected from the knowledge test scores and the rate of acceptance for the ethical and technical validity of brain death. For instance, 74 of the 166 (44.6 %) respondents who both accepted brain death as ethical and agreed to brain-based signs of death refused to acknowledge death in one or more of the five brain dead patients; 37 (22.3 %) considered all of the brain dead patients alive. Therefore we needed to explore other factors contributing to the decisions.

Univariate analysis revealed significant correlation between the demographic, occupational, educational, intellectual, and spiritual characteristics of the providers and their application of the brain death concept to patient scenarios (Supplementary Table 4). Increasing age, education, knowledge level, and professional and life experiences, as well as working in neuroscience or intensive care specialties and valuing brain and consciousness in the definitions about life and death, all positively correlated with the likelihood of acknowledging death in the brain dead patient. On the other hand, being a nurse, working in the TCM specialty, and having beliefs about life that emphasized heartbeat and body heat correlated with decreased brain death acknowledgment. Interestingly, religious faith and beliefs in the existences of soul, world after death, or reincarnation, did not have significant effects on how the respondents saw the brain dead patient.

In the multivariate analysis shown in Table 7, ethical acceptance of brain death (OR 3.819, 95 % CI 3.172–4.598) was the single most important independent predictor of operational acknowledgement of brain death, followed by high scores on the brain death knowledge test (OR 2.438, 95 % CI 2.055–2.892), and the belief that the soul lives in the brain (OR 1.848, 95 % CI 1.417–2.410). Providers who were male, had taken care of chronically ill family members, had witnessed the end of life in a family member, and believed in brain-centric signs of death were highly likely to acknowledge brain death.

Table 7 Independent predictors of operational brain death acknowledgement
Full size table

Patient attributes independently affected the acknowledgement of brain death after accounting for provider differences. Being elderly (OR 1.835, 95 % CI 1.398–2.410) or uninsured (OR 1.238, 95 % CI 1.004–1.525) encouraged, and being a member of the provider’s own family (OR 0.663, 95 % CI 0.535–0.819) discouraged brain death acknowledgement. Overall, the multivariate model had good fit as indicated by non-significance of the Hosmer and Lemeshow test (p = 0.091) and 68.9 % correct classification (Table 7). It is estimated that our model accounted for 27.2 % of the observed variation in the operational acknowledgement of brain death (Nagelkerke R2, Table 7). Several provider attributes exerted indirect influences on the medical decisions. Beliefs about the legality of brain death predicted the acknowledgment of brain death, so did the provider’s confidence in the accuracy of any given brain death diagnosis. However, their effects became insignificant when ethical acceptance entered the model. Both legality and confidence also correlated with ethical acceptance. Indeed ethical acceptance proved to have a significant mediating effect on these two variables, suggesting they indirectly affected brain death acknowledgement through increasing ethical acceptance. Similarly, advanced degrees and neuroscience specialty indirectly promoted brain death acknowledgment through increasing knowledge scores. On the other hand, TCM specialty partially exerted an indirect effect through lowering ethical acceptance and knowledge scores. Based on our analysis of the direct and indirect effects, we proposed a network of interacting provider and patient attributes (Fig. 1).

Fig. 1
figure 1

Schematic of factors involved in medical decision-making about brain death. Both provider and patient characteristics contribute to the provider’s decision to acknowledge death in a brain dead patient. Level of comprehension about the clinical features of brain death and the ethical acceptance of the brain death doctrine are the two most significant direct promoters of the operational acknowledgment of brain death

Full size image

Discussion

In this study, we sought to gain a deeper insight into the Chinese medical providers’ attitudes toward the brain death doctrine. Combining the survey of personal and professional experiences, the knowledge assessment, and the clinical scenarios, we were able to sketch a picture of the providers’ decision-making processes regarding the doctrine and reveal the motives underlying their choices. Reports of the respondents’ values demonstrated good logical consistency with their decisions in the patient scenarios, indicating that the survey reflected their internal beliefs. Our results showed that brain death is a familiar phenomenon to the Chinese medical community. The providers felt comfortable with the ethical and theoretical basis of brain death. Despite the lack of official legal recognition, they routinely encounter brain dead patients and make a diagnosis of the condition. Although the providers demonstrated sufficient understanding of the clinical features of brain death and a high level of acceptance of the ethical and theoretical basis of the brain concept, they hesitated to acknowledge death in the hypothetical brain dead patient. Elements from the providers’ personal and professional experiences, as well as characteristics of the patient, contributed to the decisions. Remarkably, religion did not appear to play a significant role.

It was not our aim to provide an ethical justification, or otherwise, for the Brain Death doctrine per se. However, by the use of empirical methods we were able to demonstrate the complexities that enter into responding to a clinical ethical question, in the absence of generally acceptable guidelines, both legal and professional. In a country where there are legal and professional codes and guidance, the brain death doctrine is generally, but not universally, accepted and practiced. There is a compelling social need for agreement on when death occurs and when this can be declared [29]. This determination and declaration, in a modern hospital setting, relies on both biological knowledge, and social and communal acceptance [4]. This acceptance and practice of the doctrine allows for the disposition of ventilated patients with the diagnosis, and the potential for organ donation from an individual with a beating heart [4, 30]. We will not review, in depth, the ethical issues associated with the doctrine and refer the reader to a publication of the US President’s Council on Bioethics published in 2011 [31]. The issues include: the role of the brain in defining life biologically and philosophically; whether brain death is a real pathophysiological state or an artificial construct; whether it is true that brain death is equivalent to the end of life; the justice of avoiding inappropriate medical care (often called futile care); and the social utility of providing organs for transplant [4]. We acknowledge that there are valid criticisms of the brain death doctrine [32, 33]. Skeptics may say that a brain dead individual is not a corpse and to the typical person the appearance and warmth of the body is not how they envisage a biological death. However when there is irreversible brain failure to the extent that it satisfies the criteria for a declaration of brain death [2, 25] it allows for the many goods that can follow this declaration. It is practiced as a matter of reasonable public policy despite the argument that it is a socio-medico-legal contrivance [33]. But overall, in the West, the practice is viewed as operationally useful and not intuitively offensive [4, 29, 34].

Our study contributes to the existing literature in several ways. First, we systematically tested the level of knowledge on the clinical definition of brain death among medical providers in Mainland China, providing a reference point for future research. Second, we examined the connection between the level of intellectual understanding and the application of that knowledge to medical decision-making regarding brain death. Last, we tested some of the key factors previously hypothesized to affect the acceptance of brain death, including religion, body-soul relationship, beliefs about the definitions of life and death, and social values [4]. Our empirical work complements the largely theoretical discussions to date on China’s official approach to the brain death concept. Notably, we found that contrary to what existing literature suggests, the levels of understanding and acceptance of the brain death doctrine among the Chinese medical professionals were high, and their decision to apply brain death in the clinical context was independent of religious beliefs. We also unveiled several important factors that may influence the practice of brain death in addition to legality, including knowledge level, ethical acceptance, patient’s financial background, and emotional connections between the patient and provider. We believe these warrant consideration in the country’s effort to legalize and propagate brain death.

Previous surveys of Chinese public opinions about brain death suggested substantial confusion and lack of knowledge on the subject matter [17–20]. To our surprise, the level of knowledge about brain death among Chinese medical professionals was not low. The scores on the knowledge test were nearly comparable to Western samples. For example, our highest scoring subgroup, attending neurologists and neurosurgeons (score on the five core brain death questions 4.6 ± 0.7, n = 35), approached the expert level reported by Tawil et al. (4.8 ± 1.1, n = 13, p = 0.458 when compared to our sample) [26]. The scores for medical students and nurses were also similar to their Western counterparts [35–38]. About 65 % of our sample correctly differentiated between the vegetative and brain dead states, which is below the 82 % reported for American medical students but above the 48 % reported for medical students in Hong Kong [28, 35]. In our sample, advanced degrees, working in neuroscience specialties, and past encounters with presumed brain dead patients positively correlated with performance on the knowledge test, suggesting that education and clinical experience are the main sources of knowledge for the providers. However, as is the case in American medical schools, there is very little instruction about brain death in the formal curricula in China [18, 35]. The providers most likely have accumulated knowledge through self-directed learning during actual patient care, and learning from colleagues, specialty literature, and the media [35]. Instituting formal education on the topic would help increase awareness and further eliminate confusion.

Previous surveys reported relatively low levels of brain death acceptance among the Chinese, hovering around 30 % for both the general public and medical staff [17–20]. Nearly two-thirds of our sample accepted brain death as an ethical standard for determining human death. Note that we specifically phrased the question to assess the theoretical alignment between the respondent and the concept of brain death in the ethical realm. If asked broadly, acceptance could be interpreted as the willingness to apply the standard to self or others, or the ability to submit to the reality of a brain death diagnosis. We found that ethical acceptance could be swayed by beliefs about the legality of brain death in China, underscoring the influence of legalization in order for brain death to gain widespread use in China in the future [12, 39]. This sentiment was further corroborated in the respondents’ subjective ranking of legality as the most important factor to consider when making decisions about brain death. In the West, official endorsement from the medical community was paramount in initiating the legalization process, and once brain death legislations were adopted, the doctrine quickly disseminated and gained wide-spread acceptance from medical providers to become part of the routine practice [4]. Our study result also suggest that legalizing brain death on the national level could have a positive effect on the medical providers’ willingness to declare death in the brain dead patient and trigger the associated arrangements in China.

Despite the observed sizable intellectual understanding and ethical acceptance of brain death, many of the Chinese medical providers were reluctant to apply the concept. When given the description of a brain dead patient, only half of the providers were willing to admit this patient as indeed dead. Even among the respondents who agreed to the ethical validity of brain death and recognized the clinical signs of brain death, many paradoxically claimed the brain dead patient as still alive, demonstrating a gap between the concept people claim to accept and their willingness to act on their belief. It also suggests that abstract yes or no questions about brain death “acceptance” may not accurately reflect people’s intentions and behaviors. Intrigued by this observation, we analyzed a wide range of factors that may affect the respondents’ answers to the clinical scenarios. Our results proved that medical decision-making surrounding brain death is a complex process, as illustrated in Fig. 1.

Regarding provider attributes, we identified “level of knowledge on the subject” and “ethical acceptance of the brain death standard” as the two most significant direct promoters of the operational acknowledgment of brain death. Knowledge and ethical acceptance also mediated the effect of several factors from the provider’s past experiences and belief systems. For example, advanced degrees enabled the providers to acquire more sophisticated knowledge about brain death, which in turn made them more likely to acknowledge death in the brain dead patient. Believing brain death was legal in China facilitated the acceptance of brain death as ethical, which also encouraged acknowledgement.

Religion has been hypothesized as a principle driver in the different attitudes toward brain death in Eastern and Western societies [4, 14, 15]. Surprisingly, religion did not appear to play a crucial role in our sample. Neither the presence nor the type of religious devotion had significant direct effect on the providers’ responses in the clinical scenarios. Believing that the soul resided in the brain was the only spiritual factor weighing in the acknowledgement of brain death. However beliefs about the soul did not correlate with religious faith or beliefs about the world after death or reincarnation. This suggested that the providers likely arrived at the connection between the cerebral location of the soul, and the logical validity of the brain death concept, independent of their religious indoctrination. Only 14 % of the medical providers surveyed in our study endorsed a religious faith. Similarly low proportions were reported in other surveys done in Mainland China [17, 20], corroborating the secular atmosphere. Although we cannot exclude subtle effects of religious traditions in the Chinese people’s attitude toward brain death, possibly through interactions with other values in one’s belief system, previous literature might have over-emphasized the contribution of religion.

Several of the attributes we tested had direct influence on the operational acknowledgement of brain death and no indirect effect through either knowledge or ethical acceptance. Examples include past experiences of taking care of sick loved ones, witnessing the death of a family member or friend, and a personal relationship with a brain dead patient. All three attributes channel the caring of the brain dead patient into an additional emotional experience, insinuating that psychological aspects may be a third major modulating factor in medical decision-making about brain death. The involvement of emotion has also been raised in other clinical situations, for instance in a family’s decision to refuse organ donation and coping with ICU death [40, 41]. Future studies evaluating the extent and type of emotional involvement in the brain death decision process are warranted to further explore our hypothesis.

Of note, in a previous study using similar scenarios conducted in Ohio, US, the age of the patient did not change how people responded to questions about declaring death [21]. However in our Chinese sample, being elderly increased the likelihood of brain death acknowledgment and withdrawal of supportive measures. In addition, the previously untested financial capacity proved to be another important determining factor. Brain death was more likely to be acknowledged in and care withdrawn from the uninsured patient than the fully insured patient. These results may reflect the biases in the medical providers’ perception of social utility in the different patient cohorts, especially given the scarcity of medical resources in China.

In our study, one particular subgroup deviated significantly in their responses to all three parts of the survey from the rest of the study population — the TCM providers. Their performance on the knowledge test was inferior compared to providers practicing Western style allopathic medicine. Less than 30 % of TCM providers thought the brain death doctrine was ethically acceptable. They were the least confident in the ability of modern diagnostics to accurately discern brain death. They did not believe that brain death was legal in China. Finally, they were least likely to acknowledge death in the brain dead patient. The TCM curriculum differs from the more popular Western medicine in that it takes a holistic approach, emphasizes the integration of bodily functions, and employs natural, non-invasive remedies such as herbs, acupuncture, and massage [42]. Values of TCM resonate with philosophical ideas regarded as “traditionally Eastern”, such as the correspondence between various body systems, nature worship, distributed flow of vital energy, and disregard for the brain [14]. The behavior of the TCM providers in our survey showed that cultural contexts do play a role in the acceptance of the brain death doctrine, and may be a sign that parts of traditional Chinese culture are not amenable to allowing a declaration of death when “brain death” is diagnosed. Alternatively, given that the concept of brain death was formed in the West, the TCM providers simply might not have had sufficient exposure or education in the foreign import.

Our study has several limitations. We surveyed a medically oriented group. The participants were overall highly educated and sophisticated thinkers, and we were able to probe deep into their knowledge and decisions because they had the capacity to understand the specialized questions. We expect their familiarity and openness to the brain death concept to surpass those of the general population. Therefore our results cannot be generalized to describe the attitudes of the Chinese public. However given that the medical providers play key roles in a country’s journey to fully incorporate brain death — from formulating the diagnostic criteria, introducing legislations, to performing clinical assessments and educating patients and families — we believe insights into their perception and decision-making process help guide the direction of the debate. Future studies exploring the perspectives of other stakeholders, including patients, family members, organ transplant recipients, and policymakers, would further inform the status of the brain death debate in China. Our convenience sampling and available-data analysis methods may have introduced some bias in the data. Although we made an effort to survey providers at different stages of training from a variety of departments and achieved a high response rate, we cannot exclude the possibility that providers in other practice settings may think differently. Our survey used hypothetical clinical vignettes to assess the medical provider’s intended decisions. We did not investigate the providers’ actual decision-making in clinical practice. It is unclear how much of the providers’ intention would translate into action when faced with real brain dead patients.

In summary, we have shown that Chinese medical providers’ perception of brain death and willingness to apply the concept to practice depends on a multitude of personal and social characteristics. We identified ethical acceptance, clinical knowledge, and relationship to the patient among the most important factors. Beliefs of legality had an indirect effect. Religious faith did not appear to contribute significantly. Although legalization is necessary for incorporating the brain death doctrine into medical practice and the top-down endorsement would grant the medical providers the protection to make the clinical diagnosis, legalization alone may not be sufficient to dissolve the resistance against brain death. As we have shown in this empirical survey, the decisions regarding the declaration of death in a brain dead patient is a complex and emotionally laden process. Even if legally endorsed, medical providers may not feel comfortable declaring brain death due to educational, cultural, ethical, financial, and emotional constraints. Providers may require education about the clinical features of brain death to accurately recognize and diagnose the condition. Raising awareness of the cultural, ethical, and psychological factors affecting the choices surrounding brain death may empower the providers to effectively communicate the issue to patients and families. Minimizing conflicts of interest may further promote objective decision-making and provider buy-in. All of these areas need to be considered in order to move toward successful introduction of the doctrine to China.