In 2010, Dr. David Ring, a hand surgeon at Massachusetts General Hospital (MGH), courageously published an article about performing the wrong operation at the wrong site on a patient. The patient was scheduled for a trigger finger release procedure, and instead received a carpal tunnel release. He and his colleagues analyzed the causes of the error and recommended solutions. MGH, and hospitals everywhere, have worked effectively to reduce the risk of “wrong site” surgery, through careful protocols that include marking surgical sites and taking time-outs while the plan for the procedure is carefully reviewed. Although good national data are understandably hard to come by, what data is available suggests that rising rates of wrong-site surgery through 2010 have fallen since the publication of this article.
But imagine a different circumstance. Suppose a patient who does not fully understand the treatment options, or the likely outcomes of those options, is scheduled for an operation. A national survey conducted by the Foundation before our merger found that people who had recently undergone elective surgery often were not fully informed about their procedures, the alternatives, and the harms and benefits; nor were they often asked for their opinions. If the patients scheduled for the surgery would have declined it had they been fully informed, then that too can be considered a medical error … “wrong patient” surgery. Al Mulley, one of our cofounders, has framed wrong-patient surgery as a diagnostic error. He and his colleagues coined the term “silent misdiagnosis” in a report for The King’s Fund, essentially stating that wrong-patient surgery is more subtle than wrong-site surgery. Let’s look at what this means. A patient’s condition may be correctly diagnosed and the procedure might be right for some patients with that condition, but the patient’s preferences for the operation have been misdiagnosed. The patient did not fully understand the procedure and was never asked his or her opinion about proceeding. Preference misdiagnosis can happen in reverse as well. A procedure can be withheld that a patient would actually want if the options were fully understood. Informed consent, as it is currently practiced, is inadequate to prevent errors due to preference misdiagnosis. In short, if the risks, benefits, and alternatives of a surgical procedure are not disclosed to the patient and the decision does not reflect the patient’s values and preferences, then the process of consent is neither consensual nor informed.
Preference misdiagnosis can occur for medical as well as surgical treatments, or even for diagnostic or screening tests. Whenever a medically reasonable test or treatment is given to a patient who, when fully informed, would decline it, or a test or treatment is withheld when a fully informed patient would have wanted it, preference misdiagnosis occurs. For every case of wrong-site surgery, there are literally hundreds of cases of wrong-patient surgery. In fact, in the Cochrane review of decision aids for tests and treatments, participants exposed to decision aids in randomized trials are not only better informed, but they also have about a 20% lower rate of surgery than in the control groups. From those results, a reasonable extrapolation is that perhaps 20% of patients who get preference-sensitive surgical procedures are the victims of preference misdiagnosis and wrong-patient surgery.
Just about everyone would agree that wrong-site or wrong-patient surgery are errors to be avoided. But what about medical treatments, where the stakes appear lower? We been working with a team of investigators at UCLA, led by Dr. Carolyn Crandall, to study preference misdiagnosis in the setting of osteopenia and osteoporosis. These conditions are marked by low bone mineral density, as reflected on bone densitometry tests. Bone density measured at the hip or spine from 1 to 2.5 standard deviations below the mean for younger women is called osteopenia, while bone density lower than 2.5 standard deviations below the mean are called osteoporosis. Affecting primarily, but not exclusively, postmenopausal women, osteopenia and osteoporosis affect 43 million and 10 million people in the United States, leading to about 2 million fractures and almost half a million hospitalizations. Hip fractures are the most devastating and independence-threatening fractures, and they are related to almost 200,000 nursing home admissions annually.
The risk of a major fracture due to weak bones is related not only to bone density but other factors such as age, parental history of hip fractures, and body mass. A well-validated calculator, called the FRAX, is available to predict women’s 10-year risk of any major fracture, and hip fracture specifically. Medications, usually in a class called bisphosphonates, reduce the risk of fractures by 30%–40% for women with osteoporosis, though less for hip fractures than for other fractures. The effectiveness of medications for women with osteopenia rather than osteoporosis, though, has not been established. Clinical practice guidelines from the National Osteoporosis Foundation, which has ties to the pharmaceutical industry, recommend treatment at a low threshold, meaning about 40% of women over 50 in the U.S. would be candidates for treatment. At the same time, there is evidence that many particularly high-risk women in the U.S. do not receive treatment.
But what do well-informed women with osteopenia and osteoporosis really want in terms of treatment? In our collaborative project with UCLA, women with bone density measurements that would qualify them for treatment according to the guideline would be recruited to find out. Participants’ physicians would make a preliminary recommendation to treat or not treat, based on the bone mineral density results. Women would receive a personalized estimate of their 10-year risk of a fracture using the FRAX calculator, as well as a Healthwise® Decision Point about taking a medication to lower this risk. Women would then tell us their treatment preference. The percentage of women who make a different treatment choice than their physicians would be our estimate of preference misdiagnosis.
We theorize that women would be sensitive to their risk of a fracture when making a decision about treatment. By educating and involving them, we hope to increase the proportion of high-risk women who accept and adhere to treatment, leading to fewer fractures. At the same time, we expect to reduce the number of low-risk women who are treated, reducing costs and side effects.
Our grant application in collaboration with our UCLA colleagues was favorably reviewed by the Agency for Healthcare Research and Quality, with numerous suggestions for making the project even better. We are resubmitting the grant this week, and we are very hopeful that we’ll be funded this time around and be able to proceed!