Using Skin Surface Temperature to Differentiate Between Complex Regional Pain Syndrome Type 1 Patients After a Fracture and Control Patients with Various Complaints After a Fracture
From the Departments of *Pain Treatment, Medical Psychology and Psychotherapy, and Anesthesiology, Erasmus MC, University Medical Center, CA Rotterdam, The Netherlands.
Address correspondence and reprint requests to Sjoerd P. Niehof, BSc, Department of Pain Treatment, Erasmus MC, University Medical Center, Postbox 2040, 3000 CA Rotterdam, The Netherlands. Address e-mail to firstname.lastname@example.org.
OBJECTIVE: In this study, we assessed the validity of skin surface temperature recordings, based on various calculation methods applied to the thermographic data, to diagnose acute complex regional pain syndrome type 1 (CRPS1) fracture patients.
METHODS: Thermographic recordings of the palmar/plantar side and dorsal side of both hands or feet were made on CRPS1 patients and in control fracture patients with/without and without complaints similar to CRPS1 (total in the three subgroups = 120) just after removal of plaster. Various calculation methods applied to the thermographic data were compared using receiver operating characteristics analysis to obtain indicators of diagnostic value.
RESULTS: There were no significant differences in demographic data and characteristics among the three subgroups. The most pronounced differences among the subgroups were vasomotor signs in the CRPS1 patients. The involved side in CRPS1 patients was often warmer compared with the noninvolved extremity. The difference in temperature between the involved site and the noninvolved extremity in CRPS1 patients significantly differed from the difference in temperature between the contralateral extremities of the two control groups. The largest temperature difference between extremities was found in CRPS1 patients. The difference in temperature recordings comparing the palmar/plantar and dorsal recording was not significant in any group. The sensitivity and specificity varied considerably between the various calculation methods used to calculate temperature difference between extremities. The highest level of sensitivity was 71% and the highest specificity was 64%; the highest positive predictive value reached a value of 35% and the highest negative predictive 84%, with a moderate 0.60 area under the curve 0.65.
CONCLUSION: The validity of skin surface temperature recordings under resting conditions to discriminate between acute CRPS1 fracture patients and control fracture patients with/without complaints is limited, and only useful as a supplementary diagnostic tool.