Nancy Schoenborn, a geriatrician at Johns Hopkins University's School of Medicine, noticed that doctors increasingly are being told by their professional organizations to treat patients in the last decade or so of life differently. Less aggressive control of blood sugar and blood pressure makes sense for people with fewer years to go, the guidelines suggest. Screening tests for certain cancers probably won't be beneficial if a patient is unlikely to live at least an additional 10 years.

The emphasis on life span rather than age stems from the recognition that health varies widely in the last chapters of life, and age alone is a poor predictor of how a patient is doing. A sick 65-year-old and a healthy 80-year-old might each have nine years left.

These new rules, though, present doctors like Schoenborn with a problem. How exactly is she supposed to explain her treatment decisions to patients?

This question led her to start asking older Americans how they want to talk about mortality with their doctors. Her recent survey, published in Annals of Family Medicine, revealed some surprising results.

The online survey asked a nationally representative sample of 878 people aged 65 and up to consider whether they'd want to discuss their life span under a variety of scenarios. The first involved a hypothetical patient who was not in imminent danger of dying but was likely to die earlier than peers because of serious health problems and difficulty with typical daily activities.

Schoenborn had in mind the kind of patient she often sees, someone who transfers to her from a primary-care physician because of multiple chronic conditions and advanced age. This patient could have serious lung, kidney, heart, and mobility problems, but each condition is reasonably well-managed and seems stable. Unlike a person with, say, advanced cancer, this patient might not be thinking about life span, so the idea of discussing it might seem to "come out of left field."

Fifty-nine percent of those surveyed, whose age averaged 73, said they would not like to talk with their doctor about their expected life span under those circumstances. Within that group, 60 percent didn't even want the doctor to ask whether it was OK to discuss the topic and 88 percent were opposed to the doctor discussing their life expectancy with their families.

Knowing that more than a third of patients don't even want the subject of life span broached can make it even more challenging for doctors, who might not want to discuss it either. "From a primary-care provider standpoint, they don't have a lot of time," Schoenborn said. "They're already stressed out."

Willingness to talk about life expectancy increased as the likelihood of death rose. Forty-four percent wanted to talk about death when their doctor estimated they had two years left. That percentage rose to 62 when death was expected within a year. By the last six months, about three-fourths of those surveyed were willing to talk about life expectancy. However, 16.5 percent did not want to talk about their odds even when they had a month left. On the other end of the spectrum, 11 percent were up for the discussion when they were likely to live an additional 20 years.

College graduates were twice as likely as high school grads to say they'd want to talk with their doctor about life expectancy. Having experienced a life-threatening illness or having discussed a loved one's life expectancy also increased openness. And, people who thought doctors were good at predicting life span were more willing to talk. People who thought religion is important were considerably less likely to talk about life span.

Those who wanted to talk about mortality thought it would be helpful for planning. Those who didn't questioned doctors' estimating skill and worried the information might be depressing.

Schoenborn said that, in her practice, these conversations typically happen because patients or their family members bring up the topic or there's been a health crisis. When she discusses cancer screening or aggressive treatment near the end of life, she can make the point that time is a factor without explicitly discussing life span with reluctant patients. She talks about how long it takes to see a benefit from treatment or screening as well as the side effects or other consequences. A mammogram, for example, can lead to other testing. It may sap energy and time that could be used to see the kidney doctor or the physical therapist.

For older patients, the message from her survey is clear. If you want to know how long you're likely to live, you should probably bring it up yourself.

Visit Philly.com at www.philly.com