October 25, 2007

Don't Fall for It

In an earlier posting, “Break a Leg,” I recommended that frail older people use hip protectors to prevent hip fractures. The data, at that time, were suggestive but not conclusive. A recent study, unfortunately, definitively argues against the benefits of hip pads. In an article published in the Journal of the American Medical Association, Dr. Douglas Kiel and his colleagues reported on a large randomized trial that used a very elegant approach to answer a difficult and important question (D Kiel, J Magaziner, S Zimmerman et al, “Efficacy of a Hip Protector to Prevent Hip Fracture in Nursing Home Residents,” Journal of the American Medical Association 2007; 298: 413-422).

Earlier studies were plagued by all sorts of problems: people didn’t like wearing the hip pads, leading to what was called “poor compliance,” making it hard to analyze the results. The hip pads studied were made of differing materials and some of the pads may actually have increased the chance that a fall would cause an injury to the hip bone itself rather than merely to the surrounding soft tissue. This new study used a hip protector that had been shown to have the desired biomechanical properties. It was conducted in nursing homes, where the frailest elders live, the people with the highest risk of falling and fracturing a hip. And the investigators did something very ingenious: they used each nursing home resident as his own control. Every individual in the study wore a one-sided hip pad on either the right or left hip. The researchers would then observe all the residents over time to see whether the protected hip was any less likely to be fractured than the unprotected hip.

The study was carried out in 1042 individuals with a mean age of 85 who lived in one of 37 nursing homes located in Massachusetts, Missouri, or Maryland. What the authors found was that the nursing home residents were just as likely to fracture the protected hip as the unprotected one. In fact, the study was stopped early because preliminary evidence was so overwhelming that the hip pads, contrary to everyone’s hopes and expectations, simply did not work. Even when the analysis was restricted to the 334 nursing home residents who wore the pad more than 80% of the time, there was still no difference in fracture rates between the protected and the unprotected hip.

It would have been great to be able to say, at last, that hip pads really work. But at least we now have the necessary information with which to conclude that they don’t. Now we know that we shouldn’t spend money on hip pads. I have to retract my earlier recommendation. It’s time to look for other strategies.

October 15, 2007

A New Test for Alzheimer's Disease: Hype or Hope?

The New York Times reported today on “a blood test” for Alzheimer’s disease (Andrew Pollack, “Progress Cited in Alzheimer’s Diagnosis,” NYT October 15, 2007). Is there actually such a test? If not, is one imminent?

Don’t hold your breath. I’m old enough to remember the eye drop test for Alzheimer’s and the skin biopsy for Alzheimer’s, neither of which panned out, not to mention the serum beta amyloid test and a few spinal fluid tests. What the news media are reporting appeared in the on-line edition of Nature Medicine yesterday in a letter-to-the-editor, not in a peer-reviewed article (Sandip Ray, Markus Britschgi, Charles Herbert et al, “Classification and Prediction of Clinical Alzheimer’s Diagnosis Based on Plasma Signaling Proteins,” Nature Medicine online, October 14, 2007). This means the findings have not been written up in an article and subjected to careful scrutiny by other scientists who are experts in the field. The letter, penned by 25 scientists, notes that the group has been studying over 100 different proteins found in the blood of patients with a diagnosis of Alzheimer’s disease to see if some combination of them might serve as a diagnostic test. They have found that a particular pattern involving 18 different proteins successfully classified 8 out of 9 patients with Alzheimer’s disease (confirmed at autopsy) as having the condition. The results, while interesting, are far too preliminary to be of interest to the general public.

If we did have a test that could accurately diagnose Alzheimer’s disease, would it matter to the millions of people with cognitive impairment and their families? Probably not. Right now, physicians can already diagnose Alzheimer’s disease with 90% accuracy, based on a careful history and physical examination together with existing laboratory tests. Even more crucial, while it is useful to diagnose dementia (whatever the cause) so as to begin planning for the future, it’s only important to distinguish Alzheimer’s disease from other forms of dementia to the extent that we have good treatments specific to Alzheimer’s. Today, the only “treatment” we have for early Alzheimer’s disease is a group of drugs known as cholinesterase inhibitors—drugs such as Donepezil (Aricept). This medicine can temporarily improve cognitive function a very modest amount. Its effectiveness is so questionable that the British NICE (National Institute for Health and Clinical Excellence) recommends against its use in patients with mild dementia (see my earlier blog posting, “Americans, Alzheimer’s, and Aricept,” February, 2006). It doesn’t cause any harm if it’s administered to patients who prove to have a non-Alzheimer’s dementia, such as vascular dementia.

One day, when we have good treatments for early Alzheimer’s disease, it will be important to make the diagnosis early and accurately. Today, what a good doctor can do is good enough.