Prevalence of Formal Accusations of Murder and Euthanasia against Physicians

Abstract

Background

Little is known about how often physicians are formally accused of hastening patient deaths while practicing palliative care.

Methods

We conducted an Internet-based survey on a random 50% sample of physician-members of a national hospice and palliative medicine society.

Results

The final sample consisted of 663 physicians (response rate 53%). Over half of the respondents had had at least one experience in the last 5 years in which a patient's family, another physician, or another health care professional had characterized palliative treatments as being euthanasia, murder, or killing. One in four stated that at least one friend or family member, or a patient had similarly characterized their treatments. Respondents rated palliative sedation and stopping artificial hydration/nutrition as treatments most likely to be misconstrued as euthanasia. Overall, 25 physicians (4%) had been formally investigated for hastening a patient's death when that had not been their intention—13 while using opiates for symptom relief and six for using medications while discontinuing mechanical ventilation. In eight (32%) cases, another member of the health care team had initiated the charges. At the time of the survey, none had been found guilty, but they reported experiencing substantial anger and worry.

Conclusions

Commonly used palliative care practices continue to be misconstrued as euthanasia or murder, despite this not being the intention of the treating physician. Further efforts are needed to explain to the health care community and the public that treatments often used to relieve patient suffering at the end of life are ethical and legal.

Introduction

Over the past decade, pundits, advocacy groups, and politicians have publically conflated illegal, active euthanasia with ethically permissible and legal end-of-life care practices. In 2005, efforts to stop artificial hydration and nutrition for Terry Schiavo, a young woman in a persistent vegetative state, resulted in a public controversy and were depicted in national media reports as murder.^1,2 Similarly, a recent legislative effort to reimburse practitioners for advance care planning visits was portrayed by politicians as government-sponsored euthanasia.³ Finally, empirical studies of patients, their families, the general public, and even clinicians suggest that many people are not clear about the boundaries between legal and illegal health care practices at the end of life.^4–9

Given this confusion, physicians who care for patients near the end of life may be at risk for allegations of euthanasia or formal charges of murder. (We use the terms “euthanasia” and “murder” interchangeably in this article; “murder” is the term frequently used by accusers, and it is the formal charge brought by districts attorneys in the legal system.) This may be especially true for palliative care practitioners as they care for patients near the end of life, stop life-sustaining treatments, and prescribe medications such as opiates to treat dying patients' symptoms (a practice that has been shown to be safe and beneficial, and one that does not hasten death in patients with advanced disease).^10–12 Although physician opinions and practices around assisted suicide and euthanasia^6,13 have been reported, no studies have examined the prevalence of euthanasia accusations leveled at physicians who believed they were providing end-of-life care in a legal and ethically acceptable manner. We undertook a national survey of physicians to clarify their experiences with palliative treatments that may be misconstrued as hastening death, and to illuminate their experiences with allegations and formal investigations of euthanasia or murder.

Methods

Design overview

We performed a national online survey of the physician members of a national hospice and palliative care professional organization. Contact information from the 2009 membership roster of this organization was abstracted and then supplemented with publically available information from the Internet to confirm e-mail and mailing addresses of physicians as necessary. We generated a 50% random sample (using statistical software SAS version 9.2; SAS Institute Inc., Cary, NC), and contacted all of these individuals via personalized e-mail. The e-mail provided a link to an online website (created specifically for this project) through which the participants accessed the password-protected survey. To access the survey, the respondents first had to read a short statement that served to inform them of the nature of the study, its risks, and its benefits. This statement included describing the part of the survey that asks the physicians questions about “whether you have ever been formally investigated for hastening death when that was not your intention.” Subjects who completed the survey were eligible to receive one of fifteen $250 cash rewards. We sent nonresponders two follow up e-mails and three letters by U.S. mail. Because the surveys were rendered anonymous at the time of completion, the study was approved as exempt from the need for written informed consent by the institutional review boards of Mount Sinai School of Medicine and Baystate Medical Center. Due to the sensitive nature of the survey, however, the investigators obtained a Certificate of Confidentiality from the National Institute of Health. We informed the subjects that this certificate protected the investigators from being legally compelled to reveal any information that might identify them or threaten their confidentiality.

Instrument design

We created a survey instrument based on qualitative interviews performed with 20 clinicians who had been formally accused of euthanasia or physician-assisted suicide when this had not been their intention. The instrument was developed by three of the authors (LC, LG, EG); reviewed with experts in the field of palliative medicine, ethics, and health care law; and pilot tested with a convenience sample of six clinicians. The final survey included information about physician demographics and practice characteristics and their views on legal medical practices that might increase or decrease physician risk for euthanasia accusations. Physicians who had been formally accused or investigated for hastening a patient's death were asked additional questions about what they considered to be their single most serious case, including details about the investigation, descriptions of the patient and accuser, and the ultimate resolution. The survey consisted mostly of closed-ended questions where the respondent could choose as many options as applied, followed by a series of open-ended “free response” questions at the conclusion. (The survey can be requested in its entirety from the authors).

Statistical analysis

The first series of e-mails were sent in October 2009, and the survey was closed in May 2010. Response rates were based on calculations derived from recommendations by the American Association of Public Opinion Research.¹⁴ Individuals for whom we had incorrect contact information (either e-mail or letter returned) were removed from the denominator as ineligible.

Statistical analyses were performed using SAS version 9.2. Continuous measures were presented as means with standard deviation (SD) and discrete data as counts with proportions. To evaluate the relationship between physician demographic and practice characteristics with the likelihood of being formally accused of hastening death we used χ² test, Fisher's exact text, or t tests as appropriate. Alpha was set at 0.05 and all analyses were two-tailed.

Role of the funding source

The funders played no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; or preparation, review, or approval of the manuscript. The findings and conclusion in this manuscript are those of the authors, and they are not representative of the individual author's institutions, the Department of Veterans Affairs, the Greenwall Foundation, the National Institute on Aging/National Institutes of Health, or the United States government.

Results

Response rate and characteristics of the respondents

After removing the 123 individuals for whom we had incorrect contact information (returned e-mails or letters) our pool of potentially eligible subjects was 1274. Of these, 34 were later found to be ineligible because they were not physicians, 10 started but did not complete the survey, and six individuals refused to participate. A total of 663 (53%) physicians completed the survey (Table 1). Seventy-five percent of respondents were trained and/or board-certified in palliative medicine, and the respondents had practiced in this field for a mean of 8.3 years. In the previous year, 70% had practiced in a hospice setting, 78% in a hospital, and 41% in a long-term care setting. Eighty-four percent reported having cared for more than 100 hospice or palliative care patients in the previous 5 years.

Table 1.

Characteristics of Responding Physicians (n=663^a)

Mean age (SD)
	49.1	10.7
Sex
Male	364	55.5.%
Specialty
Internal Medicine	363	54.8%
Family Practice	186	28.1%
Pediatrics	39	5.9%
Other	154	23.2%
Subspecialty trained and/or certified
Palliative Medicine	495	74.7%
Oncology	68	10.3%
Geriatrics	127	19.2%
Other	133	21.0%
None	60	9.0%
Mean (SD) number of years practicing palliative medicine
	8.3	8.8
Clinical practice setting
Hospice	467	70.4%
Hospital	516	77.8%
Nursing/Long-Term Care Facility	269	40.6%
Other	186	28.1%
Number of palliative care patients cared for by physician in last 5 years
<50	45	6.8%
50–99	59	8.9%
100–499	267	40.5%
500 or more	289	43.8%

^aIn some categories, respondents could choose more than one answer, so some categories may total more than 663 and some percentages may add to more than 100.

More than half of the respondents had had at least one experience in the last 5 years in which a patient's family, another physician, or another health care professional had characterized palliative treatments as being euthanasia, murder, or killing (Table 2). One in four respondents indicated that at least one personal friend or family member, or a patient had similarly characterized their treatments. Twenty-eight percent of physicians believed their personal risk of being investigated or formally accused of euthanasia was 1/1000 patients, 35% believed the risk was 1/10,000 patients, and 31% believed the risk was 1/100,000 patients. Respondents rated terminal (palliative) sedation and stopping artificially delivered nutrition and hydration as presenting the greatest risk of being misperceived as euthanasia, and the use of opiates, benzodiazepines, and barbiturates for symptom management as representing the lowest risk (Table 3). Fifty-five percent of respondents estimated that fewer than 10 palliative care or hospice physicians had been formally investigated for hastening death in the previous 5 years.

Table 2.

Physicians Responses to the Question: In the last 5 years, how often have any of the following people characterized a palliative care treatment or intervention that you recommended or implemented as “euthanasia” or “murder” or “killing?”

Frequency with which characterization occurred	A patient characterized treatment as…	A patient's friend or family member characterized treatment as…	A friend of family member of the responding physician characterized treatment as…	Another physician characterized treatment as…	Any other health care professional characterized treatment as…
Never	491 (75%)	268 (41%)	488 (75%)	292 (45%)	282 (43%)
1–2 times	93 (14%)	182 (28%)	104 (16%)	195 (30%)	191 (29%)
3–5 times	36 (6%)	113 (17%)	46 (7%)	104 (16%)	117 (18%)
6 or more times	34 (5%)	89 (14%)	15 (2%)	63 (10%)	62 (10%)

Percentages are column percentages representing frequency with which individuals in each category have characterized a physician's behavior in these manners. Columns may not add to 100 due to rounding.

Table 3.

Risk That a Palliative Care Treatment Might be Misperceived as Euthanasia

Action that might be misperceived	Mean rating of risk	SD	Actual number of physicians who were accused based on this action	Percentage of physicians who were accused based on this action
Total sedation (the application of pharmacotherapy to induce a state of decreased or absent awareness [unconsciousness] in order to reduce the burden of otherwise intractable suffering)	4.1	1.1	2	8%
Stopping artificially delivered nutrition/hydration	3.6	1.1	0	0%
Stopping oral nutrition/hydration in a patient who can eat/drink when requested by the patient	3.3	1.2	0	0%
Use of palliative and sedative medications in the process of discontinuing mechanical ventilation	3.2	1.3	6	24%
Stopping dialysis	3.1	1.2	0	0%
Use of barbiturates for symptom treatment	2.9	1.1	2	8%
Use of opiates for symptom treatment	2.8	1.2	13	54%
Use of benzodiazepines for symptom treatment	2.3	1.0	1	4%
Other	N/A	N/A	6	24%

Left side of table is respondents' answers to the query: “For the following activities, please rate each action performed by a hospice or palliative care physician in terms of its risk for being misperceived by a patient's family member or other staff as deliberately hastening a patient's death, with the highest risk of misperception being a 5 and the lowest being a 1.” The rightmost two columns show how often the 25 individuals who were formally accused of hastening a patient's death stated that this action contributed to their being accused. Accused physicians could answer that more than one action contributed to their being accused.

Twenty-five physicians (4% of respondents) reported having been investigated for hastening a patient's death, even when that had not been their intention or motivation. We found no statistically significant relationships between demographics, setting of practice, number of palliative medicine patients treated in the 5 years before the survey, or palliative medicine training of the respondents, and whether they had or had not been investigated (data not shown). As shown in Table 3, the use of opiates for treatment of symptoms was the most common precipitant of accusations (13 accusations), followed by use of medications while discontinuing mechanical ventilation (six accusations). Two accusations related to palliative sedation, and none related to stopping artificial hydration and nutrition. Ten (40%) of the accusations related to actions that took place in the hospital setting, eight (32%) in a hospice organization, four (16%) in a nursing home or long-term care facility, and three (12%) took place in another setting. Twenty physicians reported that they knew who had brought the accusation against them; the accusers were most commonly members of the health care team (total of eight cases: physicians in four cases and single instances involving a nurse, facility administrator, social worker, and ancillary staff) and patients' adult children (six cases).

Whereas five physicians reported ongoing investigations, 14 had been completely cleared, and no physician reported having been found guilty in a court of law. Despite not having been found guilty, consequences of the investigations included: losing Drug Enforcement Agency (DEA) registration, having a medical license suspended, settling for a monetary amount, and having to move to another state. Respondents could choose that cases were tried in more than one venue; of those who reported where the case had been tried, nine respondents reported that the investigations had conducted by their institutions, nine by a state medical board and six by the local or state attorney general's office.

Accused respondents experienced substantial worry and anger from the investigation (Table 4). The most common concerns included: damage to their reputation (16 physicians), financial costs (10), loss of medical licensure (eight), and imprisonment (five). In free-text responses, seven physicians commented that excellent documentation of their rationale for treatments in the medical record was exculpatory. Table 5 provides descriptions of selected cases in which physicians were accused.

Table 4.

Emotional Reactions of Accused Physicians

Reaction to investigation	Mean rating	SD
Worry	2.7	1.0
Anger	2.6	1.0
Anxiety	2.4	1.0
Sense of Isolation	2.1	1.3
Depressed Mood	2.0	1.1
Shock	2.0	1.2
Sleeplessness	1.9	1.0
Self-Doubt	1.9	1.1
Fatigue	1.7	1.0
Difficulty Working	1.6	1.0
Embarrassment	1.6	0.9

Respondents answered using a Likert scale of 1 to 4 where 1=none, 2=mild, 3=moderate, and 4=severe. N=25 for all rows (except fatigue where N=24 as one respondent skipped the question).

Table 5.

Descriptions of Selected Cases of Hospice and Palliative Medicine Physicians Being Accused of Hastening a Patient's Death

I used haloperidol, which spouse alleged was contrary to his expressed wishes, and that this was an action that directly contributed to the patient's death. I maintained (then and now) a process of meticulous documentation as this experience reinforced to me the importance of a thorough medical record. Times on all notes and the thoroughness of communications were especially important in making this a record that could be defended. My legal counsel complimented me for the excellent record.

Patient with post-polio syndrome chose to stop ventilator support after 10 months. She required opiates to control dyspnea, and lived about 36 hours. The doses of opiates were questioned as possibly contributing to her death, but the Ethics Committee found that practices were appropriate and ethical.

I wrote a prescription for methadone 2.5 mg Q 8 hours×3 days. Pharmacy mis-filled as 25 mg and patient took 4 doses. Family noted changes, and declined acute care evaluation. Patient succumbed 7 days later on hospice inpatient unit. Plaintiffs retained counsel who urged criminal investigation.

I believe the investigation was started because the coroner ruled cause of death “morphine toxicity” in a patient with a fall who sustained a C2 fracture one week prior. The patient had severe pain from the initial injury until death one week after injury. I think this took place due to required state law to report any injury that may have resulted in possible death. The coroner ruled morphine toxicity as primary cause of death with secondary cause being C2 fracture of neck.

It took place because a state health department investigator, who is a pharmacist, took the position that any doses of fentanyl above the LD50 level is definitive proof of intent to kill. He was unwilling to listen to reason, and made public accusations that he was an "expert" on these medications and called me a murderer to many levels of hospital administrators, and to many representatives of the health department. The health department refused to stop this individual from his unfair public accusations. When a physician member of the health department finally reviewed the case (after 2 months) he determined that there was no wrongdoing, but by that time the damage had been done.

A naive and overzealous coroner was systematically investigating several local hospice deaths a few years back. He based his concern on toxicology results (morphine blood levels) without taking history, diagnosis, opioid tolerance, and goals of care into consideration. The investigation went nowhere, and was really only a minor nuisance and annoyance to me since it was clear from the outset that it was going nowhere and that neither I nor my staff had done anything wrong.

I stopped labs in a patient who was still on digoxin.

Discontinuation of ventilator for a child dying at home.

Patient did not die, but son accused me of trying to kill patient with opiates for dyspnea. I was thankful that I was meticulous in my dictation of every drug order and why I did it. I even dictated my exact conversion calculations for the opiates, which I think was helpful when it went to review by the anesthetist and the pharmacist.

Note: Responses were entered as “free text” by physicians to a series of questions relating to the case, including questions asking what they had learned from the experience, what actions they think had improved or worsened their situation, as well as a “general comments” free-text response. The responses to these questions have been aggregated in the table, and some elements have been changed in this reporting to assure anonymity of the respondent as well as to assure clarity in the case description.

Discussion

In this national survey of hospice and palliative medicine physicians, we found evidence that well-accepted, ethical, and legal end-of-life care practices continue to be commonly misperceived as being euthanasia or murder and the accusers are often other health care professionals. A small proportion of physicians were investigated for hastening death when that had not been their intention. Physicians practicing palliative medicine misidentify the medical practices that place them in jeopardy. Accused physicians experience significant emotional distress, professional havoc, and personal turmoil.

The ethical principles that underpin palliative care are respect for autonomy and beneficence. Autonomous patients may forgo or withdraw life-sustaining treatments. For patients who lack decision-making capacity, surrogates may direct medical treatments toward comfort and relief of suffering as long as it is in line with the patient's known goals and desires for medical care. It is both legal and ethical for physicians to administer drugs such as opiates and benzodiazepines to treat suffering, even if death may be hastened, provided that the physician's intent is to relieve symptoms, not shorten life. Indeed, misperceptions around the use of these medications for pain and other symptoms persist despite multiple studies showing they are safe, beneficial, and, when used correctly, they have benefit profiles that far outweigh adverse effects.^10–12 What is paramount is the physician's intention. As long as medical treatment is proportional, directed toward the amelioration of the symptoms, and the physician's intention is not to hasten death, both the judicial system and most major religious traditions acknowledge the correctness and compassion of these practices. This principle, called the rule of double effect (RDE), is based on the principle of beneficence and was supported in Vacco v. Quill,¹⁵ the 1997 U.S. Supreme Court decision that found no constitutional barrier to states criminalizing assisted suicide. Although the Supreme Court in the combined cases of both Glucksberg (the case of Washington v. Glucksberg argued that Washington State's ban on assisted suicide violated the Fourteenth Amendment) and Vacco rejected a right to assisted suicide, the cases have been read to suggest a right to palliative care, including palliative sedation and pain relief that may hasten death.¹⁶ Although legal precedents on this matter are clear, they cannot prevent conflicts and misunderstandings over terminal care practices or settle the discord in our society as to what constitutes humane end-of-life care.

Against the backdrop of controversies such as the Schiavo case and the furor around a legislative effort to reimburse practitioners for advance care planning visits, our findings suggest that educational efforts have been insufficient to stem conflicts in the arena. This is disheartening given the advances over the last decade in promoting care that eases suffering for patients with serious illness.^17,18 It is possible that even with education some health care providers, patients, or family members would continue to believe these palliative treatments are wrong, because they do not agree with the underlying moral principles.

Physician education efforts should continue around the rationale for the most common palliative treatments—stopping life-sustaining treatments and prescribing opiates—not just those perceived as higher risk, such as palliative sedation. Although hospice and palliative medicine physicians may be comfortable with these practices, it is important that they maintain open lines of communication to other health care practitioners as well as meticulous documentation as to the purpose of the medications they use and treatments they provide. Efforts to clarify and explain the intent of palliative medicine treatments seem particularly important given that the majority of the formal accusations reported were brought by other members of the health care team. As one physician noted, explicit documentation about the rationale for opiate use and even the recording of dosage calculations turned out to be essential in justifying his clinical care.

None of the physicians in this sample who had been formally accused were found guilty. They nevertheless experienced substantial emotional distress and concern over dreaded outcomes during the course of the investigation, and even when cleared, many had their professional careers disrupted or damaged. What may be useful, in addition to education about the importance of documentation, is the publication of additional consensus statements and position papers differentiating euthanasia or murder from commonly employed ethical and legal palliative medicine practices. The field may also consider a unified and coordinated effort to provide accused clinicians with legal expertise and support from leaders and professional organizations in order to expedite the process of exoneration and potentially mitigate the psychological and financial expense of these investigations. Creating methods to seek more mediation and less adversarial professional and legal proceedings for the resolution of these conflicts should be a goal for health care professionals, health care administrators, medical boards, and regulatory bodies overseeing health care law. Further education of health care professionals and the public is of paramount importance.⁷

Limitations in this study that leave uncertainty about the generalizability of the findings include our response rate of 53% and the inclusion of only physicians who practice hospice or palliative care. Our response rate does not appreciably differ from other physician surveys about care for patients near the end of life.^19,20 Any physician who works with patients near the end of life may be the potential target of murder allegations if their motivations are not understood by families and other members of the health care team; we chose to survey palliative care physicians because we believed they would have the greatest exposure to these cases and thus have the greatest breadth of experiences. However, any physician, nurse, or other practitioner who works with dying patients may be a potential target of murder allegations if their motives are doubted or misunderstood by members of the health care team and families.⁸ Finally, although the introduction to our survey stated that we were asking about accusations when the intent was not murder or euthanasia, we did not include a question formally asking what the intent of the action was for which the physician had been accused.

In this national study of physicians who practice hospice and palliative medicine, we found that there remains a public misperception of their practice as being euthanasia, and, at times, this leads to accusations, formal investigations, and even criminal trials for murder. Although this study surveyed only hospice and palliative medicine physicians, all clinicians who care for patients near the end of life face the possibility of allegations and investigation. Recognition of the prevalence of accusations and identification of the specific practices that put physicians at risk are the first steps in improving the care of patients with life-threatening illness.

Acknowledgments

The authors would like to thank the physicians who participated in our survey. This work was supported in part by a grant from the Greenwall Foundation. Dr. Goldstein is supported by a Mentored Patient-Oriented Research Career Development Award from the National Institute on Aging (1K23 AG025933-01A1). Drs. Ganzini and Goy received support through the use of resources and facilities at the Portland VAMC. Portions of these data were presented at the Annual Assembly of the American Association of Hospice and Palliative Medicine in Vancouver, BC, in February 2011. Drs. Goldstein, Cohen, Arnold, and Ganzini had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Author Disclosure Statement

No competing financial interests exist.