In Reply Dr Desai and colleagues raise concerns about our study in which we investigated the association between use of alendronate and hip fracture risk in older patients taking prednisolone. They question whether early adverse effects with alendronate would not have been captured. It is correct that only patients with at least 3 months of treatment with alendronate were included. This cutoff was chosen because a shorter treatment time would mean that patients had retrieved only 1 prescription, thus making it unclear whether they had actually started treatment, and because alendronate would be very unlikely to affect fracture risk with less than 3 months of treatment.1 As a result, adverse effects leading to immediate treatment discontinuation would not be captured using this study design. To study such adverse effects, patients starting and discontinuing alendronate after 1 prescription would have to be selected and matched to nontreated patients with similar comorbidity and other important risk factors and characteristics, which we believe could serve as the basis of another important but completely different study.
Kristian F. Axelsson, Anna G. Nilsson, Mattias Lorentzon. Alendronate and Hip Fracture in Patients Using Glucocorticoids—Reply. JAMA. 2017;318(17):1712. doi:10.1001/jama.2017.14295