The societal burden associated with osteoporotic fractures is already great[1
], but is expected to increase further with aging of the population [2
]. Randomized trials have conclusively shown that use of bisphosphonate therapy in postmenopausal women can reduce their risk of osteoporotic fractures. Postmarketing studies have established a favorable balance of benefit over harms [3
]. Limited start and adherence to bisphosphonates in at-risk patients, however, reduces this beneficial impact [4
]. When bisphosphonates are prescribed, about 50% of patients discontinue therapy within one year of its prescription [8
]. Given current estimates of the prevelance of at-risk people -- the National Osteoporosis Foundation estimates that 44 million people in the United States age 50 years or over have either osteoporosis or osteopenia [10
] -- the potential morbidity, mortality and expense associated with bisphosphonate nonadherence is significant.
Poor adherence appears related, in part, to undisclosed and unexplored patient values and preferences [11
], as well as gaps in patient understanding of the efficacy and safety of pharmacologic treatment [13
]. One way to facilitate patient understanding and to elicit patient preferences is to use decision aids in consultations. Decision aids are tools that help patients participate in choosing among management options by providing them with information about the relevant features of the available choices. A systematic review of 55 randomized trials of decision aids vs. usual care, pamphlets, or education revealed that, overall, decision aids are effective at improving knowledge acquisition and reducing so-called decisional conflict (i.e., uncertainty about the choice, ignorance about the pros and cons of each option, pressure to make a particular choice, and effectiveness of the decision) [14
]; their effect on medication adherence remains largely unknown.
To our knowledge, our group has produced the only randomized trial evidence supporting improvements in adherence to preventive medications, using a decision aid about statins in patients with type 2 diabetes [15
]. In this case, clinicians and patients found the tool, Statin Choice
, acceptable; and, compared with no intervention, patients receiving the decision aid had more knowledge about and more accurate expectations of potential benefits and harms of statin use. They also reported significant reductions in decisional conflict and showed a 3-fold increase in the odds of self-reported adherence to statins at 3 months.
The timing for OSTEOPOROSIS CHOICE
has coincided with a paradigm shift in the management of osteoporosis [10
]. The prior standard approach relied most heavily on bone mineral density (BMD) results as measured with dual-energy X-ray absorptiometry (DXA) to inform practice guidelines and treatment decisions [16
]. While BMD T-scores can identify high risk patients, they are both insensitive and nonspecific [18
] since they do not address other pertinent factors related to risk of fracture [19
]. Moreover, the actual probability of experiencing an osteoporotic fracture with any given T-score is not immediately evident or available [20
]. With the arrival of FRAX, the World Health Organization's calculator of 10-year osteoporotic fracture risk [21
], treatment recommendations can now be commensurate with the probability of a hip fracture or any major osteoporotic fracture (hip, spine, distal forearm, humerus). FRAX risk estimates are based on population-based cohort studies. These studies generated models that incorporate a patient's BMD at the femoral neck, when available, and clinical risk factors. These factors include age, sex, ethnicity, body mass index, personal and parental history of prior osteoporotic fracture, long term use of oral glucocorticoids, rheumatoid arthritis or other secondary causes of osteoporosis (e.g. type 1 diabetes, premature menopause, chronic malabsorption, hyperthyroidism, hypogonadism), and current alcohol and tobacco use [19
]. Current osteoporosis practice guidelines incorporate the use of FRAX [10
], but optimal use of this tool is still evolving [20
At the onset of this project, the best available decision aid for osteoporosis treatment was 35 pages long [22
], did not incorporate absolute fracture risk estimates, and its efficacy remained unexplored. In contrast, OSTEOPOROSIS CHOICE
• is a one page, paper-based decision aid;
• incorporates a specific patient's 10-year risk of an osteoporotic fracture estimated using the FRAX online calculator http://www.shef.ac.uk/FRAX/
• is designed for use between patients and their clinicians.
The aim of this study is to evaluate the effect of an individualized fracture risk calculator and decision aid for postmenopausal women at risk of osteoporotic fractures, OSTEOPOROSIS CHOICE, on knowledge transfer, the quality of the decision making process, and on patient adherence to medications. Given our choice of study design, i.e., randomization at the patient level, we also will seek to evaluate the potential for contamination and poor fidelity in the use of the decision aid in this setting. Here we present the protocol we are using to conduct this trial.