Inexact Science: Is Patient Eligible for Medical Aid in Dying?

Roxanne Nelson, RN, BSN

February 25, 2020

BERKELEY, California — For patients to be eligible for Medical Aid in Dying (MAID), they must have a prognosis of less than 6 months to live, and the mental capacity to make an informed decision. These are legal requirements in the states that allow the procedure. But how easy is it for physicians to be sure that patients meet these requirements?  

In a session held during the recent inaugural  National Clinicians Conference on Medical Aid in Dying, experts explored these two essential issues and the challenges that clinicians may face when evaluating patients.

How Long Do I Have, Doc?

"One of the most challenging issues in working with patients is that of prognosis," said Lynette Cederquist, MD, professor of medicine and director of clinical ethics at the University of California, San Diego.

She oversees the aid in dying program at UCSD.

"The first patient who ever requested this at our institution went to his oncologist the day the legislation passed (in June 2016) and told him that he wanted it the day it became available. His oncologist eventually gave him a prescription, and the patient commented that he didn't want to take it until November 2016 because he wanted to see who was elected president," she told the audience.  

"This particular gentleman ended up living a year and a half after he received the prescription. So this goes to show our uncertainty," she said.

Cederquist told conference attendees that the two most distressing patient visits she had ever experienced was when she had to tell these patients that they didn't meet the prognostic criteria.

"I walked away feeling terrible, they walked away feeling terrible, and it's something I've grappled with," Cederquist said.

"It's written into the legal code and looks straightforward, and we chose that [6 months left to live threshold] because it works for hospice, but I don't think people realize how inexact a science this is," she added.

Prognostication is a process rather than an event, and prognostic factors evolve over time with disease. The accuracy of making a prognosis varies according to the actual definition of accuracy, the patient population, and the time frame of the prediction, she explained.

The more precise we try to be, the more wrong we will be. Dr Lynette Cederquist

"Cancer prognosticating is more accurate than with many other illnesses, but even that is wildly inaccurate," she noted. "And the tighter the time frame, and the more precise we try to be, the more wrong we will be."

How to Approach Prognostic Question

There are three basic ways of approaching the prognostic question.

The first is to ask how long the patient is estimated to live, Cederquist commented. That is usually framed by estimating if they have hours to days, days to weeks, weeks to months, or months to years. "And that is about how accurate we can be," she said.

Another way to go about this, Cederquist continued, is "to ask yourself if you would be surprised if this patient died within 6 months. And that, again, is just a broad way to think about it."

The third approach is to estimate the probability that the patient will live 6 months. "Is it 50%, 20%, 80% — what are the odds that we're looking at?"

As for the accuracy of prognosticating, Cederquist noted that physicians tend to overestimate more than underestimate.

A review of physicians' survival predictions in terminally ill cancer patients, for example, found that clinicians were correct to within a week in only 25% of cases, and overestimated survival by at least 4 weeks in 27%. (BMJ. 2003; 327:195-198)

"I just think that's a bias of being human beings, that we are going to be overly optimistic," she said. "And another component is that the longer a physician has been caring for a patient, the more biased they are going to be."

There are available tools that are used to estimate patient prognosis, but most contain subjective parameters. Validation of the tools has been variable and comparisons between tools has been unsystematic. "That leaves a lot of room for uncertainly," Cederquist noted. "Some utilize complex statistics, and that makes them less useful in routine practice and more applicable in research."

In addition, prognostication in many studies looks at patients with survival under 3 months. "There aren't a lot of good studies that look past that, so 6 months is a nebulous area," she said. "Given the problems of prognostication, should we have a different strategy than [we have] for hospice?"

There is one very significant difference between MAID and hospice, she emphasized. If a patient enters hospice and survives longer than 6 months, "it's no big deal. Nothing lost, nothing gained."

"But with aid in dying, the decision to prognosticate for less than 6 months results in an irreversible decision," she said. "And legislation does not stipulate how close to 6 months it needs to be."

More Questions Than Answers

Overall, there are more questions than there are answers, and Cederquist proposed some "food for thought" to attendees:

  • Should prognostication for purposes of aid-in-dying be more stringent than for hospice referrals?

  • Must we abide by the specialist's prognostication as the final say? The oncologists or neurologists, who may have been caring for these patients for a long time, may be overestimating survival.

  • Would it be more reasonable if the law allowed a prognosis of less than a year? Would that ever happen?

  • Should we allow aid in dying for chronically ill but not terminally ill individuals, including those with severe mental illness?

"These are the issues that many of us trying to prognosticate grapple with," she said.

"Can we envision a world where, rather than pick a time frame, we can make patients eligible with these criteria — they have progressive active illness for which no treatment is planned. Is that something that would ever be feasible? I think many of us would advocate for that."

"These are things to think about," she concluded.

Assessing Capacity

The other requirement for a patient to be eligible for MAID is mental capacity to make medical decisions.

Capacity is a functional assessment and a clinical determination concerning a specific decision and forms the basis of informed consent.

This is not to be confused with competence, even though the two are often used interchangeably, explained psychiatrist Lawrence Kaplan, DO, director, Consultation-Liaison Service, University of California, San Francisco.

"Capacity can be determined by the physician while competence is determined by the courts," explained Kaplan, who also works in the Psycho-Oncology Department at the UCSF Helen Diller Family Comprehensive Cancer Center.

However, capacity doesn't remain static. "Capacity can change within minutes, hours, days, or weeks," he emphasized.

With the exception of Montana, all states that permit MAID require two physicians to agree on the diagnosis and the patient's mental status. If either physician suspects a psychiatric/psychological disorder, then a referral for counseling must be made before a prescription can be written.

"This is to make sure that the patient is capable of making a medical decision," he explained, "And to make sure that judgment isn't impaired due to a mental disorder."

The question does arise whether evaluations for capacity among patients seeking aid in dying should be held to a different standard, as compared with capacity for making other types of medical decisions.

The bar is much higher for MAID, as the capacity/consent process is outlined by state law and a consultation by a second physician in mandatory, said Kaplan.

The most pronounced difference between MAID and other medical decisions, however, is the ultimate goal of that decision. "In MAID the patient's goal is death, while for other medical decisions, the patient is usually looking at different health outcomes," Kaplan said.

Assessing risk and benefit is also quite different. Generally, he explained, a "balancing approach" is used when assessing capacity, in which autonomy is being balanced against the best interests of the patient.

"The level of capacity that is required is generally commensurate and proportional to the seriousness of what is at stake and concordant risks/benefits," he said. "This approach balances patient autonomy and the risks and benefits, or risk–benefit ratio, of the medical decision in question."

Death is both the benefit and risk, with or without life-ending medications. Dr Lawrence Kaplan

What is unique to MAID is that the benefit–risk ratio is relatively equivalent, as death is both the benefit and risk, with or without life-ending medications," he said. "This reduces the level of capacity required and also increases the margin of error in performing such assessments."

Eligible or Not?

Assessing capacity can be challenging when a patient has a history of mental illness. Kaplan gave examples of two patients that he had consulted on, and the final conclusion that was reached.

The first case involved a 61-year-old man with metastatic prostate cancer. He also had a history of bipolar disease, catatonia, and had recently been hospitalized for psychiatric reasons.  

The prostate cancer had spread, with invasion into the bladder. After undergoing many rounds of treatment with various drugs, he experienced significant disease progression and declined further therapy.

He decided to pursue MAID, and said that his decision was related to his mother's death from colon cancer and her last days of life. She was "a very private person, did not want someone changing her diaper and people being around her all the time," he said.

He had been diagnosed with his prostate cancer just 3 weeks after his mother's death. 

Although he had a history of mental disorders, a psychiatric assessment  revealed that his impairments were primarily in attention and delayed recall, and he met the criteria for having capacity for medical decision making in this context.

"It is important to monitor cognitive and mood states as well as the physical when we get into the 6-month prognosis," said Kaplan. "Just having a history of mental illness doesn't preclude the patient, but if it impairs their judgment then it may, especially if it's a form of suicidal behaviors."

Case 2 was a 52-year-old man with left-sided, parietal glioblastoma multiforme, with worsening right sided weakness and hemianopia, and nonfluent expressive aphasia. He had a history of depressive disorder, generalized anxiety disorder, neurocognitive disorder, and alcohol use disorder while in remission, and was on multiple psychotropic medications.

During his evaluation, he had limited ability to participate in the interview because of severe expressive aphasia and little ability to respond to simple "yes/no" type questions or with head gestures. Written communication was also very limited.

While indicating that his family was very supportive and aware of his request for medication to end his life, the patient had also gestured that he felt like a burden on his family, although it could not be determined if this was his primary reason for requesting MAID.

"It wasn't clear if he was feeling coerced or not and I tried every trick in the book," said Kaplan. "I couldn't use any of the tools we have available to assess him."

There was some evidence of delirium and he was unable to articulate a general rudimentary understanding of MAID. In the end, he did not meet the criteria for having capacity for medical decision making, and other methods were explored for hastening death with palliative care, including cessation of oral intake. "The patient passed away comfortably with family at the bedside 1 week later," he said. "He probably had capacity at some point and maybe we should be intervening sooner."

He added that there is always significant room for interpretation and physician discretion within the framework. "But one key element we need to consider when prescribing the medication: We might give it when they have capacity but it may be ingested when they don't. These are things to think about."

For more from Medscape Oncology, join us on Twitter and Facebook


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.