these three pelvic sphincter muscles with reproducible latencies (Table ). This new technique utilizes a stimulator capable of delivering shocks of 200-1500V, with a stimulus duration of 0-5 ms (decay constant 0.05 ps). Increases in the latencies t o the EAS, EUS and PR muscles must be interpreted with caution since they subsume the increased terminal motor latencies measured by the direct technique described above. Conduction velocity can be calculated directly in the cauda equina nerve roots between the L1 and L4 levels after measurement of the inter-stimulus distances', but it is more relevant in most patients to use the spinal latency ratio, the ratio between the L1 and L4 latencies recorded t o the evoked responses in any one of the three muscles investigated, since this will be unaffected by any increase in the terminal motor l a t e n ~y ' ~. ' ~.The latter will be identical in the two measurements. The SLR will thus be increased in the presence of a lesion in the proximal part of the innervation, but will remain unchanged with a lesion located distally3312. Using this methodology we have been able t o recognize the existence of groups of patients in whom incontinence is due t o cauda equina disease, or to the combination of distal stretch lesions and proximal lesions, usually spondylitic in nature13.
Which test?
With practice all the tests described briefly above can be Carrie$ out in about 40min, but the experience of so many electrim1 shocks in a short time is too much to expect of most patients: The most useful tests, and the most easily tolerated, are the SFEMG and transrectal PNTML and PerNTML stimulation tests. Spinal stimulation can be deferred in most patients, and used only when there are clinical reasons for suspecting a more proximal lesion, as in patients suspected of having a neurological cause for their incontinence.