For an elderly person, a bad fall can mark the difference between walking and wheelchair, between living in one’s own home and moving to a nursing home. It can be a life-changing event.

Or it can be more like what happened to Marjorie Keyishian.

In the blackout after Hurricane Sandy, Mrs. Keyishian toppled down the stairs in her Morristown, N.J., home and broke three vertebrae in her neck. She spent a week in the hospital, a week in a rehab facility and six months in a stiff neck collar; she complains that her balance still isn’t what it was. But a year later, at 78, she’s back to twice-weekly yoga classes and daily 45-minute walks through a nearby park. “It’s important for the mind as well as the body,” she told me in an interview.

Understanding which outcome is more likely after a fall injury would be a great benefit to older people and those who take care of them. Falls remain dishearteningly common, despite prevention efforts: About one person in three over age 65 will fall each year, according to the Centers for Disease Control and Prevention, and 20 percent to 30 percent sustain moderate to severe injuries as a result.

But afterwards, who gets better and how quickly? Do we take those injured by falls to physical therapy and hope for the best — or start pricing motorized scooters? Put grab bars in the bathroom or put the house on the market?

A group of researchers at Yale, closely following a group of 754 older adults for nearly 14 years, has monitored disability before and after a fall injury and found that recovery is more predictable than we might think. The big clue, obvious only in retrospect: People with only minor disabilities, or none, before the fall are far more likely to recover, either quickly or gradually. Among those already severely disabled, the prognosis is much more grim.

“When you read reports of how devastating a hip fracture can be, it’s because those patients are all lumped together,” said Dr. Thomas M. Gill, a geriatrician and a lead author of the study, just published in JAMA Internal Medicine. “For doctors and patients and families, it’s important to get an idea of what trajectory someone was on before.”

The people in their study, all over age 70 and living independently without disability when researchers began following them in 1998 and 1999, were interviewed by phone every month and visited at home every year and a half through June 2012. During that period, 130 fell and were hurt seriously enough to require hospitalization, at an average age of 86. Almost half (62) broke a hip; the rest suffered head trauma, lacerations, or other injuries and complications.

The interviewers had recorded information on their ability to perform 13 activities: basic self-care, like bathing and dressing; so-called instrumental activities, like shopping and managing finances; and mobility (walking, climbing stairs, driving).

Based on those reports, the researchers divided the group into five “trajectories” pre-fall. About 12 percent of these older people had no disability and another 26 percent just “mild” disability. Most were classified as having “moderate” (26 percent) or “progressive” disability (18 percent). Nearly 18 percent were already severely disabled before they fell.

And afterward? “The role of function before the fall is of critical importance,” Dr. Gill said. “Those who have relatively high levels of functioning have a good opportunity at a meaningful recovery.” Only participants with no or mild disability – like Mrs. Keyishian — recovered rapidly, meaning they regained most of their functional ability within six months.

But the probability of recovering diminished with increasing disability before the fall. Among those with moderate disabilities, only a third recovered within a year.

And “if you start off severely disabled, you’re not going to get that rapid or even gradual recovery,” Dr. Gill said. “The die is cast.” In this study, severely disabled seniors were least likely to regain any independent function.

(The type of injury mattered, too. “The folks with hip fractures did worse, on average,” Dr. Gill said. “It’s probably the most severe fall-related injury.”)

Having a realistic sense of how likely someone is to recover after a bad fall can help doctors guide their treatment and help families know what to expect. For those with scant disability before the fall, “you really want to be aggressive,” Dr. Gill said. “They have the ability to regain much of the independence they lost.”

With so little prospect of recovery for those with greater disability, however, “perhaps the goals of care should shift,” Dr. Gill suggested. Palliative care — helpful in reducing the pain that often accompanies fall injuries — might make more sense, while extended physical therapy might be of little help.

I could imagine this being a difficult reality for a physician to discuss. But as Dr. Gill pointed out, “We are obligated to provide what we know about prognosis. That’s what the field of geriatrics is designed to do, to engage patients and caregivers in honest discussion so we can make decisions that are informed and in the best interests of patients and families.”

On the other hand, there’s Mrs. Keyishian. If she’d been disabled before the fall and an honest doctor explained that her chances of regaining independence were slim, “I wouldn’t have listened,” she said. “I would have been sure that if I pushed hard enough, I’d get it all back.”