Perfusion index as a reliable parameter of vasomotor disturbance in complex regional pain syndrome.

ABSTRACT

BACKGROUND: Discrepancy between a patient's subjective thermal symptoms and objective measured temperature value is common in complex regional pain syndrome (CRPS) in clinical settings. For that reason, the diagnostic validity of a single temperature measurement method has been criticised. Several studies showed that the perfusion index (PI), which is derived from pulse oximetry, is a more sensitive and earlier indicator of sympathetic blockade than temperature measurement. METHODS: Forty-six participants (CRPS group, n=23; healthy control group, n=23) were enrolled. CRPS patients with subjective sensation of coldness on the affected side were chosen for this study. Temperature and PI were measured at the affected and contralateral hands in the CRPS group, and right and left hands in the healthy control group. To evaluate each method's efficacy for detecting thermal symptoms, differences in the parameters (temperature ΔT, PI ΔPI) were compared with receiver operating characteristic curves. RESULTS: ΔPI was significantly larger in the CRPS compared with the healthy control group (P=0.002), but ΔT was not (P=0.171). The accuracy of PI measurement for reflecting thermal symptoms was good (area under the curve for ΔPI 0.873, P<0.0001). The optimal cut off value of ΔPI was calculated as 0.50% with a sensitivity of 78.26%, whereas the sensitivity of ΔT using the conventional criterion of ≥1°C was only 34.78%. CONCLUSIONS: The accuracy of detection of subjective abnormal thermal sensations was superior for PI compared with the temperature measurement method. PI was a more sensitive measure for detecting vasomotor symptoms in CRPS compared with temperature.