The selection of the comparators was appropriate. The choice of no screening as the background comparator allowed the assessment of the additional value of the screening/treatment branch of the model. The authors stated that other oral bisphosphonates (like risedronate) might have a similar cost-effectiveness profile.
Although clinical data were not retrieved by means of a systematic review of the literature, valid sources appeared to have been selected. Treatment effect and safety data for alendronate were taken from clinical trials and meta-analyses that should ensure high internal validity. The baseline fracture risk was obtained from an observational study presumably conducted in the authors’ setting which was representative of the patient population under analysis. Extensive sensitivity analyses on all clinical parameters were conducted.
QALYs were an appropriate benefit measure, as they captured the impact of the treatment on disease for both survival and quality of life. Limited information was provided on the derivation of utility valuation, which would have been useful to judge the validity of sources and instruments used.
The costs reflected the perspective of the third-party payer, but this was not explicitly stated by the authors. The economic analysis was not extensively reported, as no clear information on data sources was given. The cost of dual-energy x-ray absorptiometry was reported and the cost of alendronate was varied (from $20 to $800) to assess its impact on cost-effectiveness results. Other costs were not reported. Reflation exercises in other time periods were feasible as the price year was explicitly reported.
Analysis and results:
An incremental approach was appropriately used to synthesise costs and benefits; the estimates were clearly reported. A simplified schematic of the model was provided to represent the key transitions of women over time. The authors stated that model outcomes (fracture and life expectancy predictions) were validated using published US data. Both deterministic and probabilistic sensitivity analyses were performed to investigate uncertainty underlying selected inputs. The results of the sensitivity analyses were clearly reported. The authors acknowledged some limitations of their analysis, such as the exclusion of some potential adverse events for alendronate or the possibility of lower compliance to treatment than they assumed. The study results might be relevant for other settings with similar relative prices and baseline fracture risk.
The analysis considered a wide range of scenarios. The authors’ conclusions appear robust.