In a 2012 report for The King’s Fund, Foundation cofounder Al Mulley and colleagues described the problem of misdiagnosis of patients’ preferences. “Preference misdiagnosis” may be the most common form of medical error in health care. It can occur when a person has more than one medically reasonable course of action open to him or her. A clinician, sometimes supported by recommendations from clinical practice guidelines, may prescribe or withhold a treatment based on what the evidence best supports and what he or she thinks the patient would want. However, research shows that among fully informed patients who understand a certain treatment’s possible benefits and risks, some would want that prescribed treatment, and some would not want that treatment prescribed. While there is evidence that preference misdiagnosis occurs, its epidemiology and solutions have not been well described.
We recently teamed up with colleagues including Dr. Carolyn Crandall at UCLA to apply for a large grant to study this problem. The grant submission was in response to an Agency for Healthcare Research and Quality (AHRQ) call for proposals addressing medical error. We proposed studying preference misdiagnosis in the setting of treatment for low bone mass among older women. Low bone mass, including both osteoporosis and osteopenia, is a risk factor for fractures, particularly among postmenopausal women. These fractures, especially hip fractures, can have devastating effects. Effective preventive treatments are available, at least for women with established osteoporosis. In contrast, the antifracture efficacy of treatment is not proved for women with osteopenia. The most commonly used clinical practice guideline, from the National Osteoporosis Foundation (NOF), recommends initiating treatment for women who have 1) osteoporosis regardless of predicted fracture risk or 2) osteopenia with an estimated fracture risk above certain thresholds according to a fracture-risk calculator. The most common treatment for low bone mass is a bisphosphonate medication in addition to calcium, vitamin D, and exercise. Roughly 40% of women in the United States 50 years of age and older would be recommended for treatment for low bone mass according to the NOF guideline.
In our grant, the preference misdiagnosis rate is defined as the proportion of women who, among women who are clinically eligible to take a bisphosphonate to reduce their future risk of fractures and who are fully informed about their absolute risk of a future fracture and the potential benefits and harms of treatment using a Healthwise® Decision Point, would make a different treatment decision than their clinicians.
What are the implications of preference misdiagnosis in treatment decisions for low bone mass? First, prescribing treatments without considering the preferences of informed patients violates the ethical principle of autonomy. Second, “one-size-fits-all” treatment may risk undertreatment of women at higher risk for fracture with probabilities of future fractures that are unacceptable from the women’s perspectives. Third, one-size-fits-all treatment for the greater number of women at lower risk for fracture may risk overtreatment with probabilities of side effects and costs that are unacceptable from the women’s perspectives.
The overall goal of our project is to measure the preference misdiagnosis rate among women who are 50 years and older with newly found low bone mass and who would be candidates for treatment according to the NOF guideline. This study would take place in a set of outpatient practices at the David Geffen School of Medicine at UCLA.
Preparing this grant for submission in June reflected a great collaboration across many teams at Healthwise, as well as with the UCLA research team. The proposal will be reviewed by an AHRQ study section in October. If favorably reviewed, a final funding decision would be made next February. So stay tuned!