Personalizing care when life expectancy is short

Dr. Eric B. Larson explores how prevention, testing, and treatment must be tailored to meet the needs of patients in late life

by Eric B. Larson, MD, MPH, Kaiser Permanente Washington Health Research Institute executive director, and Kaiser Foundation Health Plan of Washington vice president for research and health care innovation

A few years ago, one of my relatives took her aged mother to a clinic for a blood transfusion. Along with other standard procedures, the clinic staff ordered an electrocardiogram to test for heart problems.

Fortunately, my relative is not the quiet type. “Excuse me,” she asked, “but why are you ordering this test?”

“It’s routine,” the nurse responded.

“That may be true, but how is this going to help her?” my relative asked. “My mother is 95 years old. If you find something wrong, are you going to do heart surgery or not do the transfusion?”

With this, the nurse consulted her mother’s doctor and the order was withdrawn.

My story illustrates a problem that happens too often in U.S. health care: Providers doing otherwise routine tests and treatments on older patients without considering the potential benefit in light of the patients’ clinical status and, especially, their life expectancy. The providers may falsely believe that any form of prevention and health screening is good. They fail to recognize that benefit depends on an individual’s health status and how long they can expect to live.

True, some forms of preventive care make sense no matter the patient’s age — especially when they result in immediate good. Screening for alcohol use, which can increase a person’s risk for falls, for example, is prudent at any age — and especially for older people with weak bones or trouble balancing.

Annual flu shots also make sense for all, even though flu vaccines are generally less effective at protecting older people. That’s because people of all ages are less likely to get the flu when they have the vaccine than when they don’t. And widespread vaccination helps establish “herd immunity,” reducing the likelihood of an epidemic.

More harm than good?

But we need to recognize some forms of preventive care are intended to prevent problems years into the future — precluding any benefit for people with a short life expectancy. Some prevention and screening can cause more harm than good for these populations, particularly if such care leads to harmful overtreatment.

One example is mammography for breast cancer screening, which can sometimes produce false positive results leading to unnecessary biopsies. Another is colonoscopy to screen for colorectal cancer, a disease that progresses quite slowly. Because of this — and because of the seriousness of colonoscopy risks such as perforation of the colon, bleeding, or incontinence in older people — the test is not recommended for most people over age 75.

Many other seemingly harmless procedures carry high risks in older populations as well. One example: I had a 90-year-old friend who went to an emergency room with chest pain. Unfortunately, during placement of a Foley catheter in his bladder, a routine procedure during a test to diagnose heart disease, he suffered irreparable damage to his urethra. While the test turned up nothing in this otherwise healthy patient, his injury required permanent catheterization. After that, he no longer felt comfortable doing things he used to enjoy, like going out to concerts and restaurants. Sadly, all this could have been avoided if the emergency room doctors had acknowledged that my friend was too old to benefit from any heart procedure the test might have indicated.

What’s at the root of such failure to tailor care based on a patient’s individual needs and life expectancy? The causes are many, including:

  • Provider incentives: Most U.S. health systems use HEDIS (a tool to measure individual providers’ performance on various dimensions of care, including prevention). While such indicators are important, many fail to consider patients’ diverse needs, especially in advanced age.
  • Inertia: Many providers are trained to follow systems by rote, not paying attention to the needs of the individual patient.
  • Lack of clarity: Providers might not know what’s good for individuals based on their life expectancy and the time needed for a particular intervention to show benefit.
  • Communication barriers: Doctors and patients are often uncomfortable discussing changes based on life expectancy.

An excellent framework

But we have guidance to address these issues. A recent article in the Journal of the American Geriatric Society by Sei J. Lee, MD, and Christine Kim, RN, GNP, of the University of California, San Francisco, presents an excellent framework for determining life expectancy and time to benefit for various interventions.

“When life expectancy (LE) is substantially longer than time to benefit (TTB), the intervention should be recommended because it is more likely that the intervention will benefit than harm the individual,” the pair write. “Conversely when TTB is shorter than LE, the intervention should not be recommended because it is more likely to harm than benefit the individual.” When LE is close to TTB, the benefits and harms are similar and personal preferences should guide decision-making, they add.

Lee and Kim further explain that their framework also fits the treatment of chronic illness care. Take hypertension, for example. New and somewhat controversial guidelines emphasize tighter control. But common treatments, especially when intensified to achieve lower blood pressure, can quickly cause older people to have dizziness and falls. Meanwhile, it may take several years before many older people might benefit from lower risk for cardiovascular disease. So we may be putting people at risk for devastating injury by giving them medication that — given a short life expectancy — may not be of benefit.

What’s the alternative to cookie-cutter approaches? Designing and delivering personalized care that fits each patient’s health status, life expectancy, and preferences.

Much research is needed to develop and implement appropriate panels of preventive tests for individuals as they reach old-old age.

We also need to help providers and their patients to have clear, honest conversations about such matters. It’s not easy to tell a patient, “I don’t recommend this test because you may not live long enough to benefit from it.” But we have sensitive, compassionate ways to raise such topics. Suggestions from authors Lee and Kim include: “This test isn’t going to help you live longer,” or “This test is more likely to hurt you than help you.”

And what about older patients and their caregivers? I believe all would do well to always ask the question my family member did: “How is this going to help?”


Learn more about Kaiser Permanente Washington Health Research Institute. Sign up for our free monthly newsletter. An earlier version of this article first appeared in LinkedIn Pulse.