Prediction Models to Estimate the Future Risk of Osteoarthritis in the General Population: A Systematic Review
Authors' conclusion
Despite growing interest in multivariable prediction models for incident OA, focus remains predominantly on the knee, with reliance on data from a small pool of appropriate cohort data sets, and concerns over general population applicability.
Study context
- Population
- General population longitudinal studies deriving, comparing or validating multivariable models to predict individual risk of incident osteoarthritis (OA) (clinical or imaging criteria); EMBASE, PubMed, Web of Science, inception to December 202126 studies · 31 models
- Outcome
- Performance and applicability of multivariable prediction models for incident OA
Methodological detailsnarrative synthesis, paired independent reviewers, risk-of-bias assessment
- Narrative synthesis
- Paired independent reviewers
- Risk-of-bias assessment
Pooled result
As reported in the abstract
Multivariable prediction models for incident OA
- Final models by site: knee
- 23
- Final models by site: hip
- 4
- Final models by site: hand
- 3
- Final models by site: any-site OA
- 1
- Outcome events per model, median (range)
- 121.5 (27-12,803)
- Prediction horizon, median (range), years
- 8 (2-41)
- Predictors per model, median (range)
- 6 (3-24)
- Discrimination after validation (AUC), generally acceptable to excellent
- 0.70-0.85
Original publication
Prediction Models to Estimate the Future Risk of Osteoarthritis in the General Population: A Systematic Review. Arthritis Care Res (Hoboken). Published 25 January 2023. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →