Effect of systematic implementation of vertebral fracture assessment (VFA) in patients with a recent non-vertebral fracture

Abstract

Background: Patients presenting at a fracture liaison service (FLS) with a prevalent non-vertebral fracture have a high prevalence of vertebral fractures. However, the diagnosis of a vertebral fracture can only be based on imaging of the spine. VFA is a low-radiation method for diagnosing a vertebral fracture at the same time as a DXA is performed.

Objectives: To analyse the effect of systematic implementation of VFA in FLS patients on the diagnosis of prevalent vertebral fractures.

Methods: Systematic use of VFA to diagnose prevalent vertebral fractures in 352 consecutive patients presenting at the outpatient FLS of the Maastricht University Medical Centre for fracture risk evaluation because of a recent non-vertebral fracture, compared to 1000 patients who received standard care before the introduction of the FLS (pre-FLS). The presence of vertebral fractures with VFA and X-ray was evaluated semi-quantitatively according to the Genant score.

Results: In pre-FLS patients, 5% had a DXA, and no patient had a VFA. VFA was performed in 352 consecutive patients (mean age 65 yrs, 71% women) who were able and willing to participate in the FLS and had additional X-ray of the spine when indicated. Sixty patients (17%) had one or more vertebral deformities ≥ grade 1 on VFA (p<0.001 versus pre-FLS) and VFA was not possible to interpret because of bad image quality in 7 patients.

Of these 60 patients, 30 (50%) had one or more moderate (grade 2) or severe (grade 3) vertebral deformities on VFA (9% of all patients). In 37 patients (11% of all patients) an additional X-ray was performed because of doubt about the degree of deformity (n=30) or an insufficient image quality (n=7) on VFA. Of these patients, 32 had a vertebral fracture ≥grade 2. As a result, 62 patients were diagnosed with a prevalent vertebral fracture (18% of all patients) and 22 of them (7% of all patients) had 2 or more vertebral fractures. Vertebral fractures were diagnosed in 23% of patients with osteoporosis (any T-score in spine or hip <-2.5), in 17% of patients with osteopenia and in 2% of patients with normal BMD.

Conclusions: Systematic implementation of VFA in FLS assessment resulted in the diagnosis of one or more previously unknown vertebral fractures in one out of five patients who attended the FLS with a recent non-vertebral fracture. In 11% of all patients an additional X-ray was performed to confirm or exclude a vertebral fracture.