Introduction - The role of clinical, MRI and neurophysiological parame
ters in predicting the outcome of acute transverse myelitis (ATM) is r
eported. Materials and methods - Thirty-one patients with ATM were sub
jected to clinical, MRI, somatosensory and motor evoked potential stud
ies in both upper and lower limbs and concentric needle electromyograp
hy. The outcome was defined at the end of 6 months into poor (Barthel
Index score < 12) and good (greater than or equal to 12). The relation
ship of various prognostic variables was evaluated by biserial correla
tion coefficient and stepwise discriminant analysis. Results - The mea
n age of the patients was 30.4 years and 7 were females Fifteen patien
ts had good and 16 had poor outcome. The variables significantly relat
ed to the outcome included severity of weakness, denervation on EMG an
d unrecordable central motor conduction time to tibialis anterior (CMC
T-TA) and tibial somatosensory evoked potentials (SEPs). Combination o
f severity of weakness and EMG had 90.3% predicting power. Addition of
central sensory conduction time (CSCT) or central motor conduction ti
me (CMCT) did not offer further advantage. Conclusion - Severity of we
akness and denervation on EMG are most useful for predicting the outco
me of ATM at 6 months although in early stage motor and somatosensory
evoked potentials may be used instead of EMG.