Calcaneal quantitative ultrasound -- a diagnostic test that assesses the density and quality of bone in the heel using high-frequency sound waves -- is attracting increased interest as a screening tool for osteoporosis, and has been suggested as a potential improvement over the most commonly used test, known as DXA, because of ultrasound's greater accessibility and lower cost.

If ultrasound were found to be as accurate as DXA in detecting osteoporosis, it could significantly expand screening and identify the disease in millions of people who have it but don't know it.

For now, however, that's still a big if, according to a study led by CHP/PCOR trainee Smita Nayak and published in the June 6 Annals of Internal Medicine. The study -- the first to synthesize all published research on the accuracy of heel ultrasound vs. DXA -- found that current evidence indicates that heel ultrasound results are insufficient to rule in or rule out a diagnosis of osteoporosis as defined by World Health Organization (WHO) criteria.

Calcaneal ultrasound still holds promise, Nayak said, but further research is needed before it can be recommended for widespread use. "Until we have more information, we should stick with what we know," she said. "DXA may not be a perfect screening tool, but it's a good standard for determining who is most likely to benefit from osteoporosis treatment."

According to a 2004 study, an estimated 10 million Americans have osteoporosis, but only half of them have been diagnosed. DXA, short for dual-energy x-ray absorptiometry, detects osteoporosis by measuring bone mineral density (BMD). In 1994, when DXA was one of the few osteoporosis screening tests available, the WHO developed an operational definition for osteoporosis of a BMD 2.5 standard deviations (T-scores) or more below the mean for healthy young women. Since this definition was developed in the context of DXA, DXA has become the de facto gold standard for diagnosing osteoporosis.

While DXA is considered a reasonably good screening test for osteoporosis, there are reasons why heel ultrasound might be a better option. Ultrasound costs four times less -- about $36 per test compared with $139 for DXA. Ultrasound is also small and portable (the machine is about the size of a large briefcase) so it can be brought to hard-to-reach populations such as elderly homebound individuals or those living in remote areas. That portability confers a significant advantage over DXA, a large, expensive piece of equipment typically found only in hospitals or large clinics.

"People who don't live near a referral center may have to travel many miles for a DXA test. That can be a substantial barrier to screening," Nayak said. "With ultrasound, more patients could get screened right in their doctor's office, or the machine could be brought to them."

To determine whether heel ultrasound might be a viable alternative for osteoporosis screening, Nayak and colleagues (including CHP/PCOR investigators Ingram Ingram Olkin, Hau Liu, Michael Gould, Douglas K. Owens and Dena Bravata) reviewed all published studies on the subject, ultimately selecting 25 that met their inclusion criteria. Analyzing the studies was challenging due to a few factors. First, while there is a standard cutoff threshold for DXA used to determine who has osteoporosis (T-score of -2.5 or below), calcaneal ultrasound has no such standard. Thus, the studies evaluating ultrasound for osteoporosis detection used different cutoff thresholds, making it difficult to determine the accuracy of ultrasound machines as a group, versus DXA.

The researchers overcame this hurdle by developing regression models that allowed them to determine the relationship between the sensitivity and specificity of the ultrasound machines studied and the cutoff thresholds used. Using these models, "We were able to make sense of a body of literature that had findings that seemed to be all over the map. Much of the variation between study findings was related to the use of different ultrasound thresholds," Nayak said.

According to the researchers' analyses, "the sensitivity and specificity of calcaneal ultrasound at commonly used cutoff thresholds seem to be too low to conclusively rule out or rule in DXA-determined osteoporosis" for the kind of patients typically seen in clinical practice. Those who test positive with ultrasound, therefore, would still require a DXA test to confirm their diagnosis, and ultrasound would fail to identify some individuals who would meet the DXA criteria for osteoporosis.

Complicating matters further, while studies have shown that heel ultrasound predicts an individual's risk of fracture nearly as well as DXA, it is unknown whether patients identified as having high fracture risk by ultrasound would benefit from treatment as much as those identified by DXA. That's because all clinical trials of osteoporosis drugs in patients without previous osteoporotic fractures have studied patients selected by DXA criteria.

To determine whether patients identified as having high fracture risk by ultrasound would respond as well to drug therapies, new clinical trials would be needed, which would enroll ultrasound-identified patients. Such trials -- which have been contemplated but not conducted -- could determine whether heel ultrasound could be used as a stand-alone tool to identify individuals likely to benefit from osteoporosis therapy, without having to use DXA as a reference standard.

Even without such trials, and even if heel ultrasound is somewhat less accurate than DXA, the question remains whether ultrasound -- given its lower cost and broader accessibility -- might still be a cost-effective screening tool compared with DXA. That is the subject of another study Nayak is conducting.