Can physicians’ judgments of futility be accepted by patients?: A comparative survey of Japanese physicians and laypeople

Study design

We performed a cross-sectional study using an anonymous questionnaire survey. All participants were asked identical questions. The present study was approved by the ethics committee at Kumamoto University Graduate School of Medical Science in July 2010.

Participants

To sample physician-respondents, we contacted 42 physicians who practice at the forefront of Japanese healthcare settings one-by-one. All consented and distributed eleven-page questionnaires to co-workers (physicians) at their facilities. A total of 500 questionnaires were distributed at 53 medical facilities of 11 prefectures in Japan. Meanwhile, we commissioned a reputable market researcher (Cross Marketing Inc. http://www.cross-m.co.jp/) to sample layperson-respondents. The inclusion criteria were: non-healthcare worker, Japanese nationality, living in Japan, able to understand Japanese, and 20–69 years old. Participants were chosen from 1.35 million registered panelists, who accessed and answered the questionnaire via the Internet. All questionnaires were sent in August 2010.

Questionnaire

Statistical analysis

The 5-point Likert scale answers were divided into two categories: YES (“I definitely think so” and “I think so”) and NO (“I am unsure,” “I somewhat don’t think so,” and “I don’t think so”). (We assumed that physician-respondents who answered “I am unsure” would not recommend the medical intervention in question in the vignettes and that layperson-respondents who answered it would not consent to the treatment. Therefore we categorized the response “I am unsure” as “NO”.) Differences in proportions among independent categorical variables were tested using the chi-square test and numerical variables were compared using the unpaired t-test. A logistic regression model was used when univariate analysis revealed statistically significant relations between independent variables (age, sex, religious belief, and experience of feeling inappropriate medical treatment) and respondents’ answers. A P-value of less than 0.05 was considered significant. All statistical analyses were conducted using SPSS (version 18.0).

Study participants

Attitudes toward treatments in hypothetical vignettes

In all vignettes, laypeople were significantly more likely to answer in the affirmative than physicians for providing treatments in question (Table 1). The logistic regression model revealed that age was the only significant independent factor in all cases (p = 0.000, 0.000 and 0.016, respectively). Younger respondents tended to support providing the treatments in question.

Factors for judging a certain treatment futile

Respondents chose three items from the list (Table 2). Items regarding the patient’s preference for the treatment (1, 3) and “Likelihood of recovery or cure due to the treatment” (15) were stressed in both groups. “Therapeutic effect for QOL (quality of life) of the patient” (9), “Physical harm (side effect or complication) of the treatment” (16), “Theoretical medical appropriateness of providing the treatment” (17), and “Patient being close to death” (7) were stressed by the physicians. On the contrary, “Request for the treatment from the patient’s family” (2) and “Psychological impact of providing the treatment on the patient” (12) were stressed by the laypeople.

Quantitative futility

One hundred sixty-six (41.4%) physicians and 380 (33.5%) laypeople did not identify a threshold that indicated quantitative futility. The thresholds chosen varied widely in both groups: 0–80% in physicians and 0–95% in laypeople (Figure 1). Median and mode were both 10% in both groups. Among the respondents who identified a threshold, 79 physicians (33.6%) and 314 laypeople (41.5%) chose thresholds under 10%. Seventeen physicians (7.2%) and 125 laypeople (16.5%) identified a threshold of 0%.

Provision of futile treatment

Among 401 physician-respondents, 88.3 percent had experienced providing futile treatment: 159 (39.7%) several times a year, 95 (23.7%) several times a month, 66 (16.5%) several times a week, and 34 (8.5%) almost every day. Request for the treatment from the patient or patient’s family (1, 2), to satisfy the patient (7), lack of public standards about judging futility or forgoing treatment (13, 14), avoidance of legal issues (16), and communication issues (5, 6) were main reasons for providing treatments that physician-respondents judged futile (Table 3).

The gap in the importance of medical information

“Healthcare worker’s inadequate explanation about futility of the treatment” (5) and “Patient’s insufficient understanding about futility of the treatment” (6) were main reasons physician-respondents gave for providing treatments judged futile (Table 3). Communication between healthcare workers and patients is a principal dimension of medical futility [2, 3, 6, 8, 12]. Its improvement, including the provision of medical information, will contribute to the resolution of the issue. Physicians are in an optimal position to evaluate the empirical and physiological aspects of futility. Four aspects of medical information gaps between physicians and patients in Japan have been discussed: knowledge of medicine, outlook after intervention, experience with similar cases, and views on the event (patients are subjective, whereas physicians are objective) [16]. By eliminating anxiety, fear and doubts of patients, and helping them make rational judgments, healthcare workers should be able to address these gaps. However, even if these gaps are removed, there is doubt about agreement between both sides.

In Vignette 1 of our questionnaire, physician-respondents stressed the professional opinion of the doctors more than laypeople (77.5% versus 44.6%, P = 0.000, Table 1). Also, physicians placed more emphasis on “Physical harm (side effect or complication) of the treatment” (16) and “Theoretical medical appropriateness of providing the treatment” (17) as factors for judging a certain treatment futile (Table 2). Because adverse effects of treatments may cause the death of debilitated patients, physicians may be nervous about providing such treatments. The principle of non-maleficence may allow physicians to forego them [4]. With the development of evidence-based medicine, physicians are urged to practice not only based on theory or rationale but more importantly based on actual physiological effectiveness, and are reluctant to deviate from them. The findings of our survey suggest that such professional standards may not be accepted by patients and their families. Even if patients have sufficient information because of efforts by healthcare workers, judgments may differ if there are gaps in degree of importance of this medical information.

The involvement of the patient’s family in medical decision making

“Request for the treatment from the patient’s family” (2) was considered to be an important factor for judging futility by laypeople rather than physicians (Table 2). The involvement of family members of Japanese patients in medical decision-making has been illustrated in previous literature. In a recent survey, Japanese patients themselves wished for their family’s involvement in medical decision-making and patients’ family members wished to take part [17]. Collectivism (individuals consistently conceptualized as part of a larger group and expected to subordinate personal goals to those of the group) and Confucianism (in which harmony and obedience are viewed as virtues) are underlying characteristics of the family’s strong influence on medical decision-making in Japan [9]. A triadic relationship is emphasized in this model of decision making, or rather, patients may sometimes withhold their own interests and follow their family members’ opinion, even when it differs from their own. Furthermore, in a public survey in 2008, attitudes of laypeople toward life-prolonging treatments differed between their own end-of-life conditions and their family members’: such treatments were wished for family members more than for themselves [10]. In another survey regarding percutaneous endoscopic gastrostomy for severely demented elderly (which was raised as a futile treatment by some participants in our preliminary survey [11]), maintaining the family’s peace of mind was a trigger for the procedure [18]. Sentiments and thoughts of family members often act in the direction of life prolongation of the end-of-life patient. Japanese physicians must be careful of the patient’s family’s wishes and consider the psychological benefits generated by the treatments. The family’s strong involvement in medical decision-making and their request for treatment, which was valued by laypeople in this questionnaire (Table 2), potentially become triggers for providing treatments that bring the patient little quantitative or qualitative benefit, and are sometimes harmful. Physicians who sincerely adopt a patient-centered approach must be reluctant to provide such treatments.

Quality of life and life-prolongation

“Therapeutic effect for quality of life (QOL) of the patient” (9) was stressed as a factor for judging a certain treatment futile by physicians (Table 2). Physicians’ priority of QOL seems to also be a reason for the differences between groups (Table 1). Some guidelines regarding end-of-life care have been established in Japan in the last few years that include mention of “meaningless” or “non-beneficial” life-prolonging treatments. Though medical futility is not explained in these guidelines, it is supposed to be one factor for their establishment. The guidelines established by the Japan Medical Association mentions that QOL should be respected and is one basis for forgoing life-prolonging treatment, as well as wishes of the patient and human dignity [14]. The interpretation is that forgoing active treatments for end-of-life patients based on improvement or maintenance of the patient’s QOL is more beneficial than continuation of treatments that may potentially harm the patient. Including the spread of palliative care in recent years, it is supposed that many physicians have come to include QOL, which justifies forgoing treatments judged futile.

However, the opposite form of end-of-life care, which continues active treatments to the last breath of the patient, has also been employed in Japanese healthcare settings. It is thought that this form is based on Shintoism, which governs Japanese spirituality and established the Japanese view of death. In Shintoism, death is a curse. Abhorrence of death still resides in the mind of the Japanese and facilitates reluctance towards any termination of life [19]. Some participants in a survey of family members of deceased cancer patients wished for continuation of active treatments, and the authors discussed that the QOL of the patient included the continuation of such treatment as long as possible [20]. In our survey, physicians emphasized patient conditions of brain death and severe dementia (in which the QOL of the patients are thought to be objectively low) significantly more than laypeople (Table 1). Asai et al., in their analysis of Japanese public opinion regarding brain death and the definition of death in the revised Organ Transplant Act, discussed that some Japanese think that medical doctors have a duty to make an all-out effort to save a person’s life, regardless of the condition, and that patients in irreversible comatose states should be valued and offered medical care unstintingly [21]. Even if the patient cannot be aware of the benefit from medical treatments, there are likely to be some people who affirm the provision of active or life-prolonging treatments for such patients. Forgoing treatments based on consideration of the patient’s QOL may be unacceptable for proponents of such perspectives. It is supposed that even today conventional views about end-of-life, such as the value of prolonging life and abhorrence of death steadfastly remain. Healthcare workers’ excessive emphasis on QOL in end-of-life care conflicts potentially leads to disagreement with patients holding such views. Also, psychological impact on patients and their families (12, 13 in Table 3) was emphasized by laypeople in the third section of the questionnaire. This is interpreted as the treatments themselves providing significant psychological benefits for people who value providing life-prolonging or aggressive treatments as much as possible, no matter what the effect on the body and QOL of the patient is. Such benefit, which is not substantive but collateral, may be very important for the patient and their family. Forgoing such treatment may cause the patient and their family to feel abandoned [6].

Arbitrariness of quantitative futility

In our survey, both groups chose “Likelihood of recovery or cure due to the treatment” (15) as a factor for judging a certain treatment futile (Table 2). This indicates that the quantitative aspect of futility (numerical probability that an act will produce the desired physiological effect) is important for judging futility. The recent rise of evidence-based medicine (EBM) may be very useful for judging futility [8]. In a survey conducted by McCrary and colleagues that asked US physicians about quantitative assessments of futility in 1991, there was a lack of agreement regarding the threshold that defines futility, [22] similar to the results of our survey in which the threshold showed wide variation in both groups (Figure 1). Additionally, 41% of physicians and 34% of laypeople did not identify a threshold. Many respondents commented that quantifying was very difficult and that the threshold was case-dependent (data not shown). These findings demonstrate the arbitrariness of the quantitative futility concept. The decision that treatment is quantitatively futile involves value choices [7]. On the other hand, very low thresholds (0–1%) were identified more frequently by laypeople than by physicians (27.0% versus 17.9%, Figure 1). This is consistent with situations in which patients who believe in miraculous successes request provision of treatments that healthcare workers are reluctant to provide (based on their futility judgment).

In contrast, what about cases of patients who do not prefer very low thresholds of quantitative futility? In such cases, should healthcare workers follow the professional demand to continue treatment as long as there are chances of recovery or to make full effort even if no chance exists? If they follow it, they may be forced to provide treatments that the patients do not want. Establishing a consistent threshold of quantitative futility may cause stress for both physicians and patients. McCrary and colleagues discussed that physicians’ unilateral determination of futility based on quantitative assessments raises the potential abuse of discretion and casts doubt on the justice of such a decision-making process because of its lack of agreement [22]. In addition, we found that patients are also unable to reach a consensus on the numerical fixation of quantitative futility because of its arbitrariness.

Physicians’ experiences of futile treatment in daily clinical settings

In our survey, 88% of physicians reported that they had experienced providing treatments that they judged futile. This indicates that futility, which has been seldom discussed in Japan, occurs in daily clinical settings and has not been adequately addressed. The lack of a system that provides guidance on forgoing treatments and avoiding legal repercussions was the main reason for providing such treatments (Table 3). These results are consistent with reports from other countries [12]. Although medical futility is often portrayed as a disagreement between physicians and patients, our results suggest that mutual agreement between the two parties is insufficient. In Japan, there were some cases in which removing mechanical ventilation from end-stage patients was considered to equate with breaking the law, regardless of mutual agreement. There has been turmoil about withdrawal of life-prolonging treatment that may result in inappropriate accusations against well-meaning doctors, put unnecessary burdens on dying patients and their families, mislead the mass media and public, and trouble police and prosecutors [23]. As expected, communication was another major reason for providing futile treatments (Table 3). Taking the results at face value, an improved decision-making process and preparation of guidelines and laws regarding end-of-life care may resolve the futility issue. This survey was not designed to cover measures for resolving the futility issue, but to demonstrate differences in value judgments between physicians and laypeople. The differences indicate that professional standards of healthcare workers cannot be applied to patients in some situations. A large interventional survey discussed that enhancing opportunities for communication is inadequate, and that social commitment and proactive, forceful measures might be needed [24]. Guidelines and laws addressing medical futility may help healthcare workers, but what measures can obtain mutual agreement on treatment goals and moreover, a social consensus?

Limitations

There may be a social desirability bias because the questionnaire asked about forgoing of treatments, which could potentially cause patients’ deaths and medical conflicts. Though all respondents responded anonymously, some might not have answered with their true intent.

Sampling bias must also be considered. We sampled physician-respondents, who answered paper-and-pencil questionnaires, via personal networks and we did not adjust their distributions (age, sex, practice term, specialty and working facility). Results of this study cannot represent the overall awareness of Japanese physicians on medical futility. On the other hand, laypeople were sampled and completed the questionnaire through a web-based method. In Japan, where the Internet penetration rate was 78.2% in 2010, internet surveys have headed the list of marketing research methods since 2005. Although this method has some advantages, including improved data quality, rapid response and a lower likelihood for social desirability bias, there is a coverage bias. Attitudes of Internet non-users and laypeople over 70 years old were not reflected in this survey. Also, this method potentially yields different results compared with off-line methods [25]. Though the questionnaire answered by both groups had the same content, it can’t strictly be said that both groups were compared under the same conditions.

In this survey, the completion rate of the laypeople-respondents who answered via the Internet questionnaire was 7%. Response rates for web-based questionnaires are generally lower than for postal questionnaires [25]. If participants are sampled randomly from a large number of potential responders, a low response rate is not necessarily a problem [26]. Our consigned research company (Cross Marketing Inc.) sampled randomly from 1.35 million registered panelists. The fact that there were panelists who failed to review the questionnaire within the delivery term and the difficulty of understanding the content of the questionnaire were considered reasons for the low completion rate. The survey was designed so that the number of deliveries is set based on the delivery term and estimated number of responses (1134 responses in about 2 days for the present survey). A substantial numbers of e-mails were delivered due to the fact that many panelists fail to connect to the Internet within the specified term. McGee pointed out technical difficulties of studies assessing patients’ reactions to futility: outcome variables are difficult to define and questionnaires are difficult to validate [27]. Though we used vignettes for clarification of the survey theme, added brief explanations of medical terminologies and asked closed-ended questions, all laypeople-respondents might not have understood completely. In fact, some respondents commented as such. However, 886 (78.1%) of laypeople-respondents freely commented about their actual experiences including end-of-life care, life-prolonging treatments, decision-making, care for the elderly, health expenditures, and so on. This may suggest that many respondents answered through their deliberation.

On the other hand, many layperson-panelists who failed to complete the questionnaire might also have had difficulty in understanding medical information. Because less educated people may tend to have unrealistic expectations to medicine in actual clinical setting, the low completion rate of this survey may suggest that the current results could have underestimated the difference of attitudes between laypeople and physicians towards the issue of medical futility.