Sensitivity and specificity of different conduction block criteria
Introduction
Recognition of multifocal motor neuropathy with conduction block (Lewis et al., 1982) as a treatable disease (Pestronk et al., 1988) and early treatment of acute inflammatory demyelinating polyneuropathy (AIDP) have revived interest in conduction block and temporal dispersion, long known indicators of demyelination (Brown and Feasby, 1984, Cornblath et al., 1991). Proximal nerve stimulation evokes slightly lower but broader responses than distal stimulation because velocity differences become more notable if the excitation front has to travel over a longer distance. A conduction block augments this physiological amplitude decline. Estimates for its lower limit of normal in healthy controls varied between 0.7 and 0.8 (Felsenthal and Teng, 1983, Brown and Feasby, 1984, Olney and Miller, 1984, Taylor, 1993, Oh et al., 1994). Healthy nerves neither account for measurement difficulties in small responses through decreased signal to noise ratio, contaminating spontaneous activity, or movement artifacts, nor for increased interphase cancellation through altered motor units, which may cause amplitude decays of up to 50%, as suggested by computer simulation (Rhee et al., 1990). Proximal/distal amplitude declines were significantly higher in ulnar nerves of patients with axonal neuropathies than among healthy controls (Olney and Miller, 1984). Motor unit restructuring and slow conduction increase interphase cancellation: both factors are present in CMT1. Proximal/distal amplitude declines exceeded 50% in neural muscular atrophy Charcot–Marie–Tooth type I (CMT1) (Uncini et al., 1993). Therefore, conduction block criteria probably have to be very strict if they have to distinguish between CMT1 and acquired chronic demyelinating polyneuropathies (CIDP). Less strict criteria may be sufficient in early AIDP when motor unit remodeling and conduction slowing have not yet occurred (Brown and Feasby, 1984).
At present, uncertainty about the presence of a conduction block remains if the proximal/distal amplitude ratio is within the range of the published guidelines which vary from 0.4 to 0.8 (Cornblath et al., 1991). It can only sometimes be resolved by attention to qualitative aspects (Cornblath et al., 1991). Persistence over time (Lange et al., 1993) or abrupt local amplitude declines suggest a real conduction block. However, conduction blocks may be less focal in disseminated demyelination, and they are only transient in AIDP. Motor units that respond to distal stimulation but are not recruited voluntarily (Cornblath et al., 1991) confirm the presence of a conduction block, but this approach requires advanced methodology and optimal patient cooperation. Therefore, well founded quantitative criteria remain desirable. The effects of interphase cancellation are ignored if the control sample providing the normal limits for the proximal/distal amplitude ratio contains only healthy controls. Nerves affected by conduction delays and by motor unit re-structuring have to be included as well. We had the opportunity to harvest a large EMG database for such nerves. We compared limits estimated from healthy and diseased controls. Conduction block criteria based on these limits and on the arbitrary limit of 0.5 (Rhee et al., 1990) were compared in respect to sensitivity in AIDP, CIDP and MMNP. ALS and CMT served to study their specificity. We also considered proximal/distal area and duration ratios. This field study on data from routine EMG should show how different conduction block criteria work in everyday EMG practice.
Section snippets
Patients and methods
Our database included all routine EMG studies from 1990 to 1996. Data were entered on-line and used for the report. Referral, EMG, and final diagnosis after complete inpatient evaluation were reviewed for all patients. The number of nerves measured per patient varied. Therefore, we report sensitivity and specificity not only per patient but also per nerve. Forty-nine patients without neuromuscular disease after clinical and paraclinical evaluation (19 patients with psychogenic conditions, 20
Heterogeneity of the control nerves
Amplitude, area, and duration ratios were independent of age (Spearman rank correlation, P>0.05), but differed between median, ulnar, and peroneal nerves, requiring separate tolerance limits for these nerves (Table 1, criterion A). The standard deviation of the proximal/distal amplitude ratio in nerves with responses below 1 mV was about twice as high as in nerves with responses above 2 mV (Table 1), requiring separate tolerance limits for small and large responses (Table 1,*). There were no
Discussion
Our data suggest that stricter conduction block criteria should be used when neurogenic restructuring or slowed conduction have to be taken into account and if the responses are small. Inevitably, our retrospective approach involved more methodological variation than a prospective study. On the other hand, it closer resembles routine practice. Our EMG lab serves a neurological intensive care unit caring for many GBS patients and an outpatient clinic for neuromuscular diseases dedicated to ALS
References (23)
- et al.
Conduction block and segmental velocities in carpal tunnel syndrome
Electroenceph clin Neurophysiol
(1997) - et al.
Conduction block without demyelination following acute nerve infarction
J Neurol Sci
(1988) - et al.
Sensitivity and specificity of diagnostic criteria for conduction block in chronic inflammatory demyelinating polyneuropathy
Electroenceph clin Neurophysiol
(1993) Research criteria for diagnosis of chronic inflammatory demyelinating polyneuropathy (CIDP)
Neurology
(1991)Diagnostic considerations in Guillain–Barré syndrome
Ann Neurol
(1981)- et al.
Intravenous immunoglobulin treatment in patients with motor neuron syndromes associated with anti-GM1 antibodies: a double-blind placebo controlled study
Neurology
(1994) - et al.
Chronic inflammatory demyelinating polyneuropathy. Clinical characteristics, course, and recommendations for diagnostic criteria
Arch Neurol
(1989) - et al.
Conduction block and denervation in Guillain–Barré polyneuropathy
Brain
(1984) Electrophysiology in Guillain–Barré syndrome
Ann Neurol
(1990)- et al.
Conduction block in clinical practice
Muscle Nerve
(1991)
Changes in duration and amplitude of the evoked muscle action potential (EMAP) over distance in peroneal, median, and ulnar nerves
Am J Phys Med
Cited by (22)
Electrodiagnosis of peripheral neuropathy
2012, Neurologic ClinicsCitation Excerpt :The general NCS finding that establishes conduction block in a motor nerve is a reduced amplitude (and area) of the CMAP with proximal stimulation and relatively large amplitude when the nerve is stimulated distally (Fig. 1). The degree of amplitude difference between proximal and distal sites that is necessary to define conduction block has been debated, and many criteria exist.5 An inherent difficulty in defining conduction block is the inability to know the degree to which phase cancellation due to temporal dispersion accounts for reduction of the CMAP amplitude obtained with proximal stimulation.
Physiology and function
2006, Handbook of Clinical NeurophysiologyPathophysiology inferred from electrodiagnostic nerve tests and classification of polyneuropathies. Suggested guidelines
2005, Clinical NeurophysiologyMultifocal Motor Neuropathy and Conduction Block
2005, Peripheral Neuropathy: 2-Volume Set with Expert Consult BasicMultifocal Motor Neuropathy and Conduction Block
2005, Peripheral NeuropathyMultifocal motor neuropathy
2004, Handbook of Clinical NeurophysiologyCitation Excerpt :Rhee et al. (1990) set a criterion of the proximal-to-distal amplitude ratio being less than 0.5 using computer simulation. Pfeiffer et al. (2000) have compared various criteria for diagnosing conduction block and found that a ratio of <0.5 for nerves with distal responses above 1 mV and a limit of 0.36 for smaller responses were the best for excluding ALS. AAEM criteria for conduction block (Table 1) do not set a definite criterion of amplitude reduction for the proximal segment from Erb's point to the axilla.