Background: Information on the benefits of therapeutic interventions can be ex-pressed in various ways, including relative risk reduction, absolute risk reduction,and number needed to treat (NNT). An alternative to such risk-based measures is postponement of an adverse outcome (eg, hip fracture in the case of osteoporosis).
Objective: The goal of this study was to examine whether laypersons' perception of the benefit of an osteoporosis therapy differs when it is presented in terms of the NNT to avoid 1 hip fracture compared with the duration of postponement of hip fracture.
Methods: This was a cross-sectional, randomized, controlled trial. Face-to-face interviews of a representative sample of the Danish population were conducted in respondents' homes. Respondents were randomized to receive information about the benefits of a hypothetical osteoporosis intervention either in terms of different magnitudes of NNT (10, 50, 100, or 400) or different durations of postponement of hip fracture (1 month, 6 months, 1 year, or 4 years). Participants were subsequently asked if they would consent to the intervention.
Results: A total of 1728 individuals were contacted at home and asked if they would take part in a face-to-face interview; 967 (56%) were successfully interviewed. The age (mean age, 44.5 years; range, 20-74 years) and sex distribution (51% male, 50% female) of the sample was similar to that of the general Danish population. Based on NNTs of 10, 50, 100, and 400, the proportions of respondents who said they would consent to the intervention were a respective 65%,61%, 63%, and 57%. Increasing NNT was not significantly associated with a lower proportion of consent (test for trend chi-square(1)= 0.75; P = NS). Forty-three percent of respondents indicated that the concept of NNT was difficult to understand, and 38% interpreted NNT in a way that was probably incorrect. In terms of postponement of hip fracture by 1 month, 6 months, 1 year, and 4 years, the proportions who said they would consent to the intervention were a respective 25%, 40%, 39%, and 53%. Increasing postponement of hip fracture was significantly associated with higher proportions of consent (test for trend chi-square(1)= 20.09;P < 0.001). Logistic regression analysis found that consenting to therapy was inversely associated with age (NNT: OR, 0.83; 95% CI, 0.72-0.96; postponement of fracture: OR, 0.84; 95% CI, 0.73-0.98) and with the magnitude of benefit presented in terms of postponement of fracture. No other variables were significantly associated with consent to therapy.
Conclusions: When laypersons were presented with brief information about the benefit of a hypothetical osteoporosis intervention and were then offered this therapy, their choices were sensitive to the magnitude of treatment benefit when it was presented in terms of postponement of hip fracture but not in terms of NNT. These findings suggest that it may be easier for laypersons to understand a potential treatment benefit in terms of postponement of fracture rather than NNT.