SummaryWe evaluated the value of VFA in the identification of vertebral fractures using a retrospective study and a meta-analysis. Performance of VFA was adequate in the meta-analysis although this was not demonstrated in our centre. We recommend checking the performance of VFA tools before exclusively relying on this tool.IntroductionVertebral fractures are traditionally diagnosed using conventional radiographs of the spine. Vertebral fracture assessment (VFA) has been advocated as an alternative tool in the diagnosis of these fractures.MethodsWe conducted a retrospective study as well as a systematic review and a meta-analysis to evaluate the performance of VFA compared to conventional spinal radiography in patients who had sustained a fracture and thus at risk for osteoporosis. A risk of bias analysis was also performed.ResultsThe diagnostic study included 542 patients (25% male) with fractures. The sensitivity of low-radiation VFA to detect a patient with a vertebral fracture ≥ Genant grade 2 was 0.77 and its specificity 0.80. Two hundred ninety-seven (55%) patients had ≥1 and 135(25%) ≥3 unevaluable vertebrae. The systematic review identified 16 studies including a total of 3238 subjects (19% male) with a mean age range of 45 to 74 years. Seven studies had a low risk of bias and 9 had an intermediate risk, mainly due to not consecutively including patients. The pooled sensitivity of VFA to detect a patient with a vertebral fracture ≥Genant grade 2 was 0.84 (95% CI, 0.72–0.92) and specificity 0.90 (95% CI, 0.84–0.94).ConclusionsOur findings from the meta-analysis suggest an adequate performance of VFA for the detection of vertebral fractures. However, we could not demonstrate these findings in our center, especially the specificity. Our data advocate caution with exclusively relying on VFA in the assessment of vertebral fractures without identifying performance and potential limitations of the technique.Electronic supplementary materialThe online version of this article (doi:10.1007/s00198-017-4137-6) contains supplementary material, which is available to authorized users.

No abstract