When she filled out the advance directive form, Lisa Curcio, who has metastatic breast cancer, made it clear that she does not want aggressive measures as the end nears.
"If my body can't function on its own, then my body is telling me it's time to go," said Curcio, a 44-year-old Levittown accounting clerk. She's currently responding well to treatment but knows that her cancer is likely to be deadly.
Faced with the same questions, Matthew Bellina, a 32-year-old with the fatal neurological disease ALS, said he wanted to go all out. The Holland, Pa., man has two young boys and he wants every last second with them, even if he can only move one eyelid.
"I just feel like a fraction of a dad is better than no dad," he said.
The two Penn Medicine patients are part of a clinical trial that is analyzing how the wording of advance directives changes patient choices and outcomes. It is among the portfolio of studies overseen by the 4-year-old FIELDS (Fostering Improvement in End-of-life Decision Science) program, an effort to look at how we die in a more scientific way.
"Everyone recognizes the problems that pervade end-of-life care and, because of that broad recognition, everyone is interested in a solution," said Scott Halpern, a critical care medicine doctor and director of FIELDS.
The problem, he said, is that lots of people have jumped in with well-intended, intuitively appealing programs that may not work. Those could crowd out opportunities for more effective approaches.
Among his targets are Five Wishes, Respecting Choices and the Conversation Project, popular programs meant to help people make end-of-life choices.
"None of these things are backed by much evidence," said Halpern, who, in addition to the M.D., has a Ph.D in epidemiology and a master's in bioethics. While there is scientific support for outpatient palliative care programs, the evidence is weaker for the burgeoning number of inpatient programs, he said. He is most critical of POLST (Physician Orders for Life-Sustaining Treatment) programs because they are inflexible and illness is unpredictable.
Halpern made his case recently in the New England Journal of Medicine for funding more of the kind of research he does. In his essay, he argued that growing political acceptance of end-of-life planning, including the government's new willingness to pay physicians for talking with patients about it, calls for evidence on what actually helps patients get the kind of care they want while minimizing trauma to families.
He said FIELDS is the country's only program devoted to applying the principles of behavioral economics, in essence the study of how people make choices, to end-of-life care.
One FIELDS study found that 77 percent of patients chose "comfort care" in an advance directive where that was the "default" or preselected option. The percentage fell to 43 when the questions were written to make life extension the default choice.
Another study made completing an advance directive part of the orientation process for new Penn employees. The completion rate was twice as high when workers had to actively decline to fill out the form as it was when they were told it was optional.
One that is not yet published measured how good doctors and nurses were at predicting how patients would be doing six months after an ICU stay. "They couldn't predict patients' quality of life well at all," Halpern said.
The FIELDS team is now involved in a big project with Ascension Health, a large health system based in St. Louis. This summer, 11 hospitals will switch from the current system, in which patients or doctors must ask for a consultation with a palliative-care expert, to one in which seriously ill patients will routinely be offered such treatment.
Halpern is in talks about more studies with Lynn Schuchter, chief of hematology/oncology at Penn's Abramson Cancer Center. "What he's doing at the center is a new way of affecting health care, and I think it's really powerful," she said.
One study will use patients' electronic medical records to prompt oncologists with a question, such as, "Would you be surprised if this patient is alive six months from now?" The researchers will then look at the use of hospice and palliative care.
Another will give doctors report cards on how many of their patients got chemotherapy in the last two weeks of life compared with patients of their peers.
Ultimately, Halpern would like to measure quality of end-of-life care. But using typical outcomes such as survival is obviously problematic.
He thinks most people want to live as long as possible, as long as there are not adverse consequences, such as excessive suffering or serious emotional, financial or psychological pain for their families. Yet the likelihood of suffering, he said, is higher with more aggressive care.
Most of us, though, "don't have the foggiest idea about how to make specific health-care choices that promote good lives and deaths," Halpern said.
We know whether we prefer vanilla and chocolate ice cream because we've had plenty of opportunity to taste them. But most people have very little experience with end-of-life decisions, and even less understanding of how the alternatives would feel.
Halpern thinks patients are most likely to make stable decisions about their preferences once they have experienced serious illness.
It runs counter to what many experts on dying suggest, but Halpern, who is 42 and has no chronic illnesses, does not have an advance directive. "I don't think anyone like me should," he said, adding that he doesn't think he's in a position to guess where he'd want to set limits if he were sick.
"Sure, I could get hit by a bus tomorrow, but, less than 1 percent of people who die, die that way."
Curcio and Bellina know what it's like to be sick and to live with a dire prognosis.
They are part of a trial to look at how default wording in advance directives affects choices, hospital utilization and bereavement. Neither had any idea which set of questions they'd received, even after a Penn staffer called to discuss their decisions and make sure they'd said what they wanted to say. (Both had gotten the comfort-care default.)
Curcio trusts that her best friend, a nurse, will make good decisions for her if she's unable to make them herself.
"She'll know in her heart what to do," Curcio said.
Bellina, a former Navy jet pilot who has long lived in the shadow of death, was equally sure that the wording of questions wouldn't change his answers.
"My life situation is dictating the decisions I make very strongly, so I don't know how much wording would make a difference," he said. "It's very cut and dried with me."