Pursuing Palliation

We’ve known for a long time that breaking a hip can be devastating for an older person. It often leads to a loss of independence and it's associated with a substantial increase in the risk of both nursing home placement and death. While the rate of hip fractures declined in the US between 1990 and 2010, it remains stubbornly high: every year, 300,000 people fall and fracture a hip. A new study suggests the consequences of a hip fracture are especially severe in people who live in nursing homes.  In light of the high rate of disability and death in the 6 months after a hip fracture, an accompanying editorial recommends that breaking a hip in the nursing home signal the need for palliative care. Absolutely—but many of the patients reported in the new study should have been getting palliative care long before their fateful fall.

The authors of the study in JAMA Internal Medicine identified 725,000 Medicare fee-for-service patients who broke a hip during a 4-year period between 2005 and 2009; just over 60,000 of them, or 8%, lived in a nursing home prior to sustaining a hip fracture. It was these 60,000 who were the basis of the study:  the vast majority were women (75%), white (92%) and demented (91%). They also tended to have multiple chronic diseases and to have difficulty with many basic daily activities, which is hardly surprising, as their illnesses and their impairments were precisely why they lived in the nursing home.

After their hip fractures, many patients went downhill quickly. Fully 36% were dead within 6 months. Among those who survived to the 6 month mark, over a quarter had become totally dependent in walking. By the time a year had elapsed, just under half of the residents who had fractured their hip had died. Only 1 in 5 of those who started out being fairly independent in walking had regained their previous level of function. Risk factors for death or disability included male sex, advanced age, white race, multimorbidity, cognitive impairment, and dependence in basic activities. Of note, patients treated non-operatively (11.8% of the sample) did particularly poorly in terms of physical functioning and survival.

These findings fit with the results of earlier, smaller studies. But I couldn’t help asking, as I read this dismal portrait of life-after-hip-fracture-in-the-nursing-home, what happens to people who live in a nursing home who don’t break a hip? How many of them decline? The study in JAMA Internal Medicine had no control group, no nursing home residents with a comparable degree of physical and mental impairment, cared for during the same time period, who happened not to sustain a hip fracture.

I turned up another study also published this year entitled “Natural course of dependency in residents of long-term care facilities: prospective follow-up study.” The study was carried out in Dutch nursing homes and it included both people who had hip fractures along with those who did not. The population was reasonably similar to the US nursing home Medicare population: 75% were women and their mean age was 84. The Dutch have a system for reporting dependency that is different from (and more sophisticated than) the American approach: they report on the “Care Dependency Scale (CDS),” a 15-item scale that grades patients from 15 (totally dependent) to 75 (almost independent in eating, dressing, walking, and dressing). What they found was that among 890 Dutch nursing home residents, just under 15% had died at 6 months and another 18% had died by one year. But the degree of dependency at baseline was strongly predictive of the outcome, with higher dependency leading to greater likelihood of death, even after correcting for gender, age, and the presence of diagnoses such as cancer or dementia. Among residents in the low CDS group (that is, the most independent people), 20% actually improved over a 12-month period and the vast majority (80%) remained unchanged. Among residents in the highest CDS group (the most dependent people), no one improved, 64% remained stable over 6 months, and 36% got worse. The middle group, not surprisingly, fell between these extremes.

How did the nursing home residents in the Dutch study (some unknown fraction of whom fell and broke a hip) compare to the nursing home residents in the American study (who were selected based on their having fallen, broken a hip, and been hospitalized)? Those with hip fracture did worse: over the course of a year, 80% deteriorated markedly in their independence, compared to only 36% of the worst-off Dutch residents; and by 6 months, 36% had died, compared to about 20% of the worst-off Dutch.

These comparisons are approximate at best, but they give some hint of what the Medicare nursing home residents might have expected if they hadn’t broken a hip. So while a hip fracture does represent a turning point, admission to a nursing home in the first place also represents a turning point, a transition to life’s final chapter. Thus while it is eminently reasonable to offer palliative care alongside of surgery to all nursing home residents who break a hip, it would be even better to offer palliative care alongside conventional medical care to everyone in the nursing home.