Predictive factors for the longterm outcome of spondyloarthropathies

To determine the predictive factors of outcome in patients with spondyloarthropathy (European Spondyloarthropathy Study Group or Amor criteria) monitored by a single investigator.

Classification of longterm outcome on a 3-grade scale. Candidate predictive factors: presence or not of 12 clinical or biological variables during the first 2 years of the disease, collected by history at the time of the first visit. Univariate analysis to pick up the factors statistically correlated with severity and then odds ratio and 95% confidence interval (CI) for each variable were calculated.

Of the 328 patients with spondyloarthropathy, 151 had a followup of > or = 10 years and minor disease (81), severe (28), or moderate disease (42). Seven variables at entry were correlated with disease severity (odds ratio; CI 95%); hip arthritis (22.85; 4.43-118); erythrocyte sedimentation rate > 30 mm/h (7; 4.84-9.50); poor efficacy of nonsteroidal antiinflammatory drugs (8.33; 2.56-27.10); limitation of lumbar spine (7; 2-25); sausage-like finger or toe (8.45; 1.48-9); oligoarthritis (4.25; 1.38-13.10); onset < or = 16 years (3.47; 1.06-12.75). If none of these factors is present at entry a mild outcome can be predicted (sensitivity: 92.5%; specificity: 78%). If a hip is involved or if 3 factors are present, a severe outcome is predictable (sensitivity: 50%) and a mild disease practically excluded (specificity: 97.5%).

Predictive factors of poor or benign longterm outcome could be defined very early after onset of spondyloarthropathy in a set of patients monitored by one observer.