METHODS: 167 patients with CRPS who underwent CT of their affected limbs were retrospectively reviewed. After conducting univariable analyses where regional osteopenia determined by CT was dependent and other clinical factors were independent variables, the first multivariable analysis assessed risk predictors associated with regional osteopenia in CRPS. Next, after conducting univariable analyses where sympathetic block response was dependent and others, including regional osteopenia, were independent variables, the second multivariable model predicted factors associated with the response to sympathetic blockades, followed by receiver operating characteristic curve analysis.

RESULTS: Among 116 patients, regional osteopenia was identified in 72.2% early (<1 year) and 52.5% persistent CRPS. In the first multivariable analysis adjusted for age, sex, body mass index (BMI), and other variables with p values<0.1 from initial univariable analyses, older age (OR 1.06; 95% CI 1.02 to 1.10) and positive three-phase bone scintigraphy (TPBS) (OR 3.94; 95% CI 1.46 to 10.66) were significantly associated with regional osteopenia. In the second multivariable model adjusted for age, sex, BMI, and other variables with p values<0.1 from univariable analyses, early phase (OR 5.49; 95% CI 1.44 to 20.88), regional osteopenia (OR 5.11; 95% CI 1.49 to 17.53), and positive TPBS (OR 6.30; 95% CI 2.21 to 17.93) were significantly associated with positive responses to sympathetic blockade in CRPS, showing excellent performance characteristics with a predicted probability>0.358 (sensitivity 0.86; specificity 0.76).