SUMIT R. MAJUMDAR, MD,
Associate Professor,
Department of Medicine,

Download PDF View Table of Contents → RELATED ARTICLE

University of Alberta,
2E3.07 Walter Mackenzie Health Sciences Centre,
University of Alberta Hospital,
8440-112th Street,
Edmonton, Alberta T6G 2B7, Canada
E-mail: me2.majumdar@ualberta.ca

Supported by salary awards from the Alberta Heritage Foundation for Medical Research (Health Scholar) and the Canadian Institutes of Health Research. Address reprint requests to Dr. Majumdar.

Osteoporosis is a common, chronic, and costly condition. Both the incidence and costs of osteoporosis-related complications are expected to increase by 50% over the next 2 decades¹. The most devastating complication of osteoporosis is hip fracture: it is associated with substantial mortality, and among those who survive, with marked loss of function and independence. What is not well recognized is that patients with a hip fracture have a marked increase in risk of future fractures. The rate of any clinical fracture in those who survive hip fracture is about 10% per year², and the rate of recurrent hip fracture approaches 15% over 4 years³.

Unfortunately, practice audits also suggest that rates of osteoporosis treatment in the 6 to 12 months after hip fracture are less than 10%–20%, even though 80% of these patients will have low bone mass⁴. In this issue of The Journal, Cadarette and colleagues describe a 10-year (1995–2004) time-series in a population-based cohort of elderly hip fracture patients in Pennsylvania and show us that quality of osteoporosis care is less than optimal, although (at the least) there has been some improvement over time⁵. They report improvements in 6-month postfracture osteoporosis treatment rates from 7% in 1995 to 31% in 2004. While this represents a 4-fold increase over a decade, it also illustrates a persistent care-gap — better termed in this case as a yawning chasm — between what we know we ought to do for our patients and what we actually help them achieve. Certainly by 2004, we might have hoped to achieve osteoporosis treatment rates of at least 50% among survivors of hip fracture. However, judging by the results of Caderette, et al, we are not even close to this benchmark. Still worse, their study most certainly overestimates what we are achieving and exposes several disturbing trends.

A better way to estimate how well we recognize the threat of osteoporosis and recurrent fracture in this vulnerable population is to examine treatment-naive patients. This approach is necessary because refilling a prescription for an osteoporosis medication that was taken prefracture tells us little about the treating physician's recognition or intent related to subsequent fracture prevention. Accepting the evidence-free zone for patients who fracture while on treatment and the lack of recommendations for this population in consensus-based treatment guidelines, we should still expect some therapeutic intensification, either switching or adding osteoporosis treatments. Instead, Caderette, et al found that 82% continued the same osteoporosis treatment 6 months after fracture, and use of combination treatments was virtually nonexistent⁵. This suggests considerable clinical inertia, defined as "the failure of healthcare providers to initiate or change treatment when the health status of the patient indicates such action is necessary⁶." This inertia is best exemplified in treatment-naive patients, where we see rates of osteoporosis treatment post-hip fracture have increased from 4% in 1995 to only 17% in 2004⁵. To provide the most informative and interpretable comparative data possible, I believe the focus of future studies of prescribing trends in fracture patients (and for secondary prevention in most chronic conditions presenting with acute exacerbations) should be on treatment-naive patients.

The use of longitudinal data to examine the quality of care delivered to hip fracture patients also reveals 3 disturbing trends within the overall picture of very modest improvements described by Cadarette, et al⁵. First, their data demonstrate a leveling-off of improvements in treatment over the last 3 years of their study, from 2002 to 2004. Given the undertreatment already documented, it is distressing that improvement has now ceased. I cannot speculate why this might be the case and am unaware of any other longitudinal studies that could either support or refute these data. If this finding is replicated elsewhere, it suggests that we cannot count on even the slow and haphazard but inexorable improvements in quality of care that our healthcare system usually achieves to help these patients.

The second trend is the virtual neglect of male patients. What is most troubling is how large and refractory this disparity appears to be. Cadarette found that men were always about one-third as likely to be treated as women; and over 10 years, rates of treatment postfracture have improved little: from less than 1% to about 5%⁵. Why? Most likely, it is because of a profound attitudinal barrier on the part of both physicians and male patients that osteoporosis is a disease only of elderly women^5,7.

As a specialist, perhaps the most disturbing trend uncovered by Cadarette relates to the performance of specialist physicians who should be interested in bone health (i.e., rheumatologists, endocrinologists, geriatricians, obstetrician/gynecologists, and orthopedic surgeons). Collectively, this group should be in the vanguard of implementing guidelines and quality improvement efforts in this area. How are these specialists doing? Except for geriatricians, who have not improved over time, the other specialists have steadily and significantly decreased their rates of prescribing osteoporosis treatment for hip fracture patients. Indeed, 2 of these groups seem less likely than the others to prescribe effective treatment with bisphosphonates; instead they tend to use treatments without any proven nonvertebral fracture benefit — namely inhaled calcitonin (by geriatricians) and raloxifene (by obstetrician/gynecologists)⁵. I suspect that this well-intended but non-evidence-based pattern of practice is at least partly related to marketing by the pharmaceutical industry⁸. For example, the promotional message for calcitonin to geriatricians might highlight ease of administration, minimization of drug interactions and reduced polypharmacy, and absence of adverse effects for elderly patients; for raloxifene to obstetricians/gynecologists, the message might emphasize the putative additional benefits in terms of women's cardiovascular health and breast cancer reduction; and both promotional messages would necessarily need to minimize the fact that neither drug reduces nonvertebral fractures.

However we have arrived here, clearly there must be many barriers to improving the quality of osteoporosis care for hip fracture patients: at the level of the system, the physician, and the patient^4,9,10 (Table 1). Similarly, any successful intervention will need to be "multifaceted" to overcome these multiple barriers. This is borne out in the results of the only 2 valid randomized quality-improvement trials directed at osteoporosis in hip fracture patients I am aware of. In the first, Gardner, et al randomized 40 patients to education, a set of prompts to give their primary care physicians, and reminder calls; and 40 patients to usual care¹⁰. This might be considered a patient-directed and single-faceted intervention. On an intention-to-treat basis, their intervention was unable to increase treatment with bisphosphonates (25% for intervention vs 15% for controls; p = 0.26)¹⁰.

Table 1. Barriers to osteoporosis treatment in patients with hip fracture.

In the second study, my colleagues and I reported a randomized controlled trial: the intervention was a hospital-based case-manager⁴. The case-manager educated hip fracture patients, facilitated bone mineral density tests, arranged prescriptions of weekly oral bisphosphonates for patients with low bone mass, and communicated all results and plans to the primary care physician of record. This intervention was compared with usual care in 220 patients. The intervention increased bisphosphonate treatment rates to 51% versus 22% for usual care (p < 0.001), and it increased rates of "appropriate care" (i.e., receipt of a bone mineral density test and bisphosphonate treatment if bone mass was low) to 67% versus 26% for usual care (p < 0.001) within 6 months of fracture. Although this may seem like a labor-intensive intervention, in a formal micro-costing study we reported the intervention took 70 minutes and cost US $50 per patient — about one-fifth the cost of one year of generic alendronate therapy⁴.

In summary, the article in this issue of The Journal defines usual-care rates of osteoporosis treatment after hip fracture, and clearly demonstrates that we cannot rely on the healthcare delivery system and the collective good intentions of physicians and allied health professionals to improve things without some help. As with drugs or devices, knowledge translation interventions will need to be designed, implemented, and evaluated in randomized controlled trials (or at least in careful quasiexperimental studies of reasonable internal validity, such as the before-after study, with concurrent controls or interrupted time-series analyses) and then proven effective and efficient before widespread adoption. Perhaps future efforts will be easier since we now have randomized controlled trial evidence that treating patients with hip fracture with bisphosphonates (albeit intravenous annual zolendroic acid) reduces both recurrent clinical fractures and mortality¹¹. Most importantly, hospital-based specialists interested in bone health (rheumatologists, endocrinologists, geriatricians, hospitalists, and orthopedic surgeons) must work together, gather needed momentum to overcome clinical inertia, and help the rest of us do the right thing right for our elderly patients with hip fracture.

REFERENCES

Search PubMed for: