Presentation (CTI)

OBJECTIVES: Opposing conclusions of reviews of osteoporosis screening have resulted in different screening recommendations across high-income countries. We aimed to explain reasons for differences in conclusions across reviews based on a largely shared evidence base, to help the UK National Screening Committee interpret the evidence for policy.
METHODS: A systematic review of reviews of osteoporosis screening was conducted searching nine electronic databases in October-November 2025. Each review and the studies included were extracted and quality appraised by two independent reviewers. The reviews were compared across four key domains: scope of review, included evidence base, numerical data and author conclusions.
RESULTS: Five systematic reviews were compared. The reviews had similar scope, comparing screening strategies versus usual care in asymptomatic populations. Three randomised controlled trials were the primary evidence source, and most reviews conducted meta-analyses. Numerical estimates from meta-analyses were consistent across reviews in terms of direction of effect and statistical significance. Author conclusions on screening effectiveness pointed in opposite directions, ranging from no benefit to effective reductions in fractures. Divergent conclusions reflected differences in interpretation and presentation of the same evidence. Favourable conclusions of screening were associated with pooling of small effects across heterogeneous trials, emphasising statistically significant relative effects over clinically meaningful absolute effects, and reliance on the per-protocol analyses among women who engaged in screening.
CONCLUSIONS: This review highlights that repeated reviews of the same evidence base offered limited additional value, with differences in conclusions reflecting variation in interpretation of similar underlying findings. Decisions on whether a new review is needed, particularly for a different policy setting, should focus on the underlying results within existing reviews rather than author interpretations alone. Future work could explore frameworks to help policy-makers assess both the relevance of existing reviews and the circumstances in which new evidence syntheses are likely to add value.