System Breakdown

Abstract & Commentary

Synopsis: Most postmenopausal women who suffer a fracture do not receive treatment for osteoporosis.

Source: Andrade SE, et al. Arch Intern Med. 2003;163: 2052-2057.

Osteoporotic fractures are an important contributor to morbidity and mortality.1,2 Patients with hip fractures run a 10% risk of disability and 19% require institutionalization. They account for 140,000 nursing home admissions annually. Fractures can be a source of chronic pain. They interfere with the activities of daily living and have a negative effect on quality of life. The US price tag for osteoporotic fractures is estimated to be $17.5 million annually. Among postmenopausal women, 90% of fractures are associated with osteoporosis.3 Compared to a person who has neither osteoporosis nor a history of fracture, a person with both has an increased risk of sustaining another. There exist several medications (alendronate [Fosamax®], calcitonin [Miacalcin®], calcium, estrogens, raloxifine [Evista®], risendronate [Actonel®], and teriparatide [Forteo®]) approved by the Food and Drug Administration for the prevention of osteoporosis. Treatment with these drugs usually results in an increase in bone mineral density and reduced risk of fracture within a year or two. It stands to reason, then, that physicians should treat postmenopausal women who suffer fractures for osteoporosis. Is this reality?

Andrade and colleagues address that question by mining the databanks of 7 geographically diverse health maintenance organizations (HMOs). Looking at data from 1994 to 1996, they identified 3492 women, at least 60 years old, who suffered a wrist, vertebral, or hip fracture. They excluded those who had concurrent cancer diagnoses (to exclude pathologic fractures), multiple myeloma, and major trauma.

The primary outcome was a dispensing of a drug (estrogen, bisphosphonate, or calcitonin) that treats osteoporosis. The breakdown (pardon the pun) was 1572 (45%) hip fractures, 1620 (46%) wrist fractures, and 300 (9%) vertebral fractures. In the 90-day period before their fractures, 2995 of these women were enrolled in their HMOs. Of these, only 390 (13%) were receiving a drug indicated for osteoporosis. In the year after their fractures, 822 of 3492 (24%) were receiving such a drug.

When examining rates by fracture site, women with vertebral fractures were more likely (44%) to receive an osteoporosis drug than women with wrist (23%) or hip fractures (21%). The older a woman was, the less likely she was to receive medications, 33% for ages 65-70 years old vs 15% for those 80 or older. There was an increase in dispensing when comparing rates for 1994 (22%) and 1996 (26%). There were 2605 women who were not receiving an osteoporosis drug before their fracture; only 353 (14%) were receiving one in the year afterward. Of the 390 women who were receiving a drug before their fracture, 365 were in the following year.

Comment by Allan J. Wilke, MD

Not much to cheer about here. As primary care physicians, our goal is primary prevention; secondary prevention is second best. This study indicates that for postmenopausal osteoporosis we are not addressing either effort. We have not corralled the horse, let alone closed the barn door before it got out. What I find very troubling is that this population was drawn from HMOs who had prescription drug benefits and who, presumably, had greater access to health care and screening for osteoporosis. What would the results have been if the population had been our run-of-the-mill Medicare patients? Another consideration is that it is likely that the women receiving estrogen in the 90-day period before a fracture (353 of 2995) were not receiving it exclusively for bone health. And this was during the time before the Women’s Health Initiative (WHI), when many of us were enthusiastically endorsing the use of estrogen for better bones! (By the way, the WHI demonstrated that estrogen plus progesterone did increase bone mineral density and reduced the risk of fracture in their study group.4 The researchers decided that the absolute risk reduction, 2.5%, was insufficiently robust to outweigh the adverse effects of hormone replacement therapy, even in the group of women at high risk for fracture.)

Andrade et al speculate why undertreatment was so prevalent and came up with several plausible causes. They included "clinical inertia" and physician skepticism regarding the evaluation of osteoporosis and the efficacy of treatment. However, the reasoning that I find most compelling is health care system dysfunction and the breakdown of communication between the physician who treats the fracture and the physician who provides continuity of care. Simonelli queried orthopedic surgeons and primary care physicians about who was responsible for what in the care of patients hospitalized with a fragility fracture.5 The orthopedic surgeons were in agreement that the primary care physician owned the responsibility for postfracture treatment of osteoporosis. This kind of specialty territoriality can only contribute to patients falling through the cracks and sustaining another fracture (again, pardon me).

Are there any reasons to doubt the findings of this study? It was conducted retrospectively and examined databases from 1994 to 1996. Could we be doing a better job in 2003? Not likely. Feldstein looked at her HMO population over the years 1998 and 1999.6 Of the 2804 men and women, aged 50 to 89 years, who sustained a fracture, only 34.7% were on an osteoporosis medication and only 4.6% had new use after the incident fracture. When looking at women alone, the rate of treatment was better (42.4%) but still suboptimal. More than 90% of this population had a prepaid drug benefit.

There are several clinical guidelines that recommend routine screening and treatment of asymptomatic osteoporosis in postmenopausal females.7-10 The bottom line is that we must increase our efforts to identify patients at risk for osteoporosis and prescribe treatment. This is a situation where an electronic medical record could be enormously valuable: "Computer, please prepare me a list of all my patients who are 65 years or older and who have not had a DEXA scan." If only practice were this easy.

Dr. Wilke, Assistant Professor of Family Medicine, Medical College of Ohio, Toledo, OH, is Associate Editor of Internal Medicine Alert.


1. Melton LJ, et al. J Bone Miner Res. 2003;18:1139-1141.

2. Roberts SE, Goldacre MJ. BMJ. 2003;327:771-775.

3. Melton LJ, et al. J Bone Miner Res. 1997;12:16-23.

4. Cauley JA, et al. JAMA. 2003;290:1729-1738.

5. Simonelli C, et al. Mayo Clin Proc. 2002;77:334-338.

6. Feldstein A, et al. Arch Intern Med. 2003;163: 2165-2172.

7. U.S. Preventive Services Task Force. Ann Intern Med. 2002;137:834-839.

8. NIH Consensus Development Panel on Osteoporosis Prevention, Diagnosis, and Therapy. JAMA. 2001;285:785-795.

9. Meunier PJ, et al. Clin Ther. 1999;21:1025-1044.

10. Brown JP, et al. CMAJ. 2002;167(10 Suppl):S1-S34.