LAUSR

Background Patient decisions to take preventative treatments for osteoporosis depend on their perceptions of fracture risk, medication effect sizes (ES) of benefits and harms. However, physicians and lay persons may have differing perceptions of risks and medication efficacy. Both tend to overestimate medication benefits. This study surveyed at what risk physicians would prescribe and lay persons would be willing to take bisphosphonates, the minimum ES both groups do demand and the physicians estimates of the actual benefit of bisphosphonates. Design Cross-sectional online questionnaire survey. Methods Respondents were confronted with a case vignette with an osteoporotic patient (10-year femoral fracture risk: 32%). They were asked at what threshold of 10-year-risk of femoral neck fracture they would prescribe or take a drug. They were asked for the minimum ES (absolute risk reduction, ARR) they demand from bisphosphonates to prescribe or take them. Physicians were asked to provide their estimate of the actual ARR of bisphosphonates. Results 114 physicians and 140 lay persons answered (convenience sample/snowball distribution). The 10-year-risk threshold of lay persons (Mdnlay = 60%) willing to take medication was twice as high as the physicians’ threshold (Mdnphy = 30%) to prescribe it (p < .001). The median minimum ARR physicians demanded for bisphosphonates prescription was 17%, whereas lay persons demanded 22% (p < .001). Physicians estimated the actual ARR of bisphosphonates to be 12%. This estimated effect size was below their own minimum threshold for prescription. Conclusions Lay persons tolerate a higher fracture risk and demand a larger benefit of antiosteoporotic medication for fracture prevention than physicians. Physicians demand higher minimum benefits than their own estimates which in turn are above the benefit evidence suggests. Physicians should be more familiar with ES of antiosteoporotic drugs concerning patient outcomes and actively advise lay persons before preventive treatment decisions are taken.