Objective: To attempt to characterize reflex sympathetic dystrophy (RS
D) and to determine factors that would predict a response to sympathet
ic block. Design: We undertook a retrospective analysis on 396 patient
s with chronic limb pain referred for autonomic testing during a 5-yea
r period. Material and Methods: Clinical endpoints were relief of pain
after sympathetic block and a composite RSD diagnostic probability sc
ore, based on the clinical attributes of allodynia, protopathia, swell
ing, and vasomotor alterations. We compared the results of three auton
omic tests-resting sweat output (RSO), resting skin temperature (RST),
and quantitative sudomotor axon reflex test (QSART). Results: Increas
ed RSO predicted the diagnosis of RSD with 94% specificity, and the sp
ecificity was 98% when RSO was considered in conjunction with an abnor
mal QSART result, the best laboratory correlate (P = 0.003) of the cli
nical diagnosis. Shorter duration of pain correlated with a warmer lim
b (P<0.001), even in the absence of RSD. Response to single sympatheti
c block did correlate with the diagnosis (P = 0.031) but correlated mo
st significantly with short duration of pain in the arm (P = 0.001) an
d laboratory findings in the leg, where increased RST (P<0.001) and QS
ART (P<0.001) were near-perfect predictors of response. Conclusion: Sw
eating abnormalities correlate strongly with the clinical syndrome of
RSD, and alterations in RST may be superior to clinical findings in pr
edicting the response to sympathetic block. The findings provide physi
ologic support for the unproven view of a natural disease progression
(''stages''), with better treatment response and a warmer extremity in
itially. Because certain physiologic trends occur in all patients, gen
eral alterations of autonomic function with pain are suggested.