ELECTRODIAGNOSTIC TESTING: BLIND SPOTS, PITFALLS, AND

OVERCALL

Timothy R. Dillingham, M.D.William J. Erdman Professor and Chairman

Department of Physical Medicine and RehabilitationChief Medical Officer

Penn Institute for Rehabilitation MedicineThe University of Pennsylvania

OBJECTIVES: LIMITATIONS, PITFALLS

• Understand the important limitations in electrodiagnostic medicine

• Nerve conduction reference values not well standardized– Interrater reliability low without standards

• Understand EMG has modest sensitivity, but very high specificity – complementary to MRI

• Overcall in its many forms– Polyphasics, excessive testing, poor technique

NERVE CONDUCTION STUDIES

Nerve Conduction Challenges with Performance

• Skin prep minimizes impedance missmatch between active and reference electrodes

• Overstimulation of adjacent nerves

• Stimulus artifact – optimize ground

• Captured motor unit can look like a SNAP– Repeatability over three tracings– Be careful with averaging, can “Lock In” artifact

NERVE CONDUCTION PITFALLS

• Multiple Entrapments found, eg; median, ulnar– MOST LIKELY A POLYNEUROPATHY– STUDY A LOWER LIMB

NERVE CONDUCTION TESTING

– Too many tests, increasing probability of one false positive

– If you do 7 tests; 16% chance of a false positive– Probability of TWO false positives with 7 tests is

1%• Two findings is MUCH more compelling for a

diagnosis

– Use caution with overcalling with just one abnormality

NORMATIVE DATA: REFERENCE VALUES

• No standardization in our field regarding reference values and techniques for Nerve conduction testing

• Older norms plagued by;– Low sample sizes– Statistics not matching non- Gaussian

distributions– Analog machines– Poor study designs

NORMATIVE DATA:BETTER REFERENCE VALUES

AANEM NORMATIVE DATA TASKFORCE

• ALL YOU NEED TO KNOW:

• RALPH BUSCHBACHER, MD• NORMATIVE STUDIES IN THE 1990’s

AND 2000’s

• TOWER ABOVE MOST OF THE LITERATURE REGARDING NORMS

ELECTRODIAGNOSIS:PITFALLS

EMG

• Motor units can look like PSWs depending on geometric position of needle

• PSWs and FIBs are REGULARLY FIRING POTENTIALS

• Don’t move the needle too quickly, thrust and stop

• Caution calling radiculopathy based upon polyphasicity

EMG

• Do not assess motor unit morphology in PSM, no norms

• Cervical PSM low false positives IF:

• Lumbar PSM low false positives IF:– IMPOSE CRITERIA OF REGULAR FIRING

• PSM - ONLY LOOK FOR: FIBS, PSW, CRDs

PARASPINAL MUSCLE EMG: PREVALENCE OF FIBRILLATIONS IN NORMALS

• Dumitru, Diaz, and King (2001)

• Prospective study 50 normals L4/L5 levels

• Monopolar needle, recorded potentials

• Examined firing rate and rhythm

• Fibrillation inclusion criteria; regular firing rate

• 4% false positive fibrillations in paraspinal muscles

NEUROLOGICAL CONDITIONS MIMICKING CERVICAL RADICULOPATHY

• Entrapment/Compression neuropathies– Median, Radial, and Ulnar

• Brachial Neuritis• Multifocal Motor Neuropathy• Need Extensive EDX study to R/O other conditions• ALS if upper motor neuron signs and

atrophy/weakness• Myopathy, Myasthenia Gravis

NEUROLOGICAL CONDITIONS MIMICKING

LSR• Diabetic Amyotrophy

• Mononeuropathies– Femoral– Tibial– Common Peroneal

• Need Extensive EDX study

EMG SENSITIVITIES FOR LUMBOSACRAL

RADICULOPATHIES

• Varies widely

• Ranges from about 50% to 80%

• Various diagnostic standards

Study Sample size

Gold standard EMG sensitivity %

Lumbosacral radiculopathy

Weber and Albert [55]

Nardin et al [28]Kuruoglu et al [8]Khatri et al [56]Tonzola et al [57]Schoendinger [58]Knutsson [45]Young et al [3]Linden and Berlit [3]

4247

1009557

10020610019

Clinical+imaging HNP

Clinical

Clinical

Clinical

Clinical

Surgically proven

Surgically proven

Clinical an imaging

Myelography and CT

605586644956798478

EMG SENSITIVITY FOR CERVICAL

RADICULOPATHIES

• Varies widely

• About 50% to 70%

• Usually clinical and/or myelographic

Study Sample size

Gold standard EMG sensitivity %

Lumbosacral spinal

stenosis Hall et al [46] Johnsson et al [59]

6864

Clinical+myelogram

Clinical+myelogram

9288

Cervical radiculopathy Berger et al [60] Partanen et al [61] Leblhuber et al [9] So et al [62] Yiannikas et al [18] Tackman and Radu [15] Hong et al [63]

187724142020

108

Clinical

Intraoperative

Clinical+myelogram

Clinical

Clinical/radiographic

Clinical

Clinical

61676771509551

SPECIFICITYTong, Haig, Yamakawa, Miner. Amer J of PM&R 2006

• Assymptomatic volunteers 55 and older• Standardized monopolar EMG with blinded researcher

– Five leg muscles and paraspinal muscles• When only PSW and Fibs were considered abnormal

– By criteria: i)Two limb muscles plus PSM or ii)two limb muscles or iii) one limb and PSM

– 100% specificity• If use polyphasicity

– 97%, 90%, 87% specificity • EMG has excellent specificity for LSR

CAVEATS AND LIMITATIONS

• Needle EMG is not an effective screening test alone (Radiculopathy)

• MRI better screen for structural causes

• Better specificity-Diagnosis confirmation

• Motor Axonal loss is necessary for fibs

• A purely sensory radiculopathy will not result in FIBS on EMG

EMG

• Hard signs, better interrater reliability– FIBS– PSWs– CRDs

• Softer signs, more tendency to overcall– Polyphasicity– Reduced recruitment– Amplitudes of MUAPs

EMG

• Normative data for needle EMG are rare and old– Mostly for CONCENTRIC NEEDLES– Monopolar needles – Chu textbook

• Wide range of normal variation. The extremes are clearer– Amplitudes over 8K with reduced recruitment– Firing ratio of 20Hz

FALSE NEGATIVE (no fibs or PSW) ON EMG IN RADICULOPATHY

• Sensory root involvement only

• Motor root involvement without axonal loss– Demyelination, conduction block

• Motor axonal loss balanced with reinnervation

MUSCLE INJURY CAUSING FIBRILLATIONS

Partanen et al 1982 Muscle & Nerve

• Study of 43 patients with EMG before and after Muscle biopsy

• 50% had fibrillations 6-7 days after biopsy

• At 16 days 100% had fibrillations

• Fibrillations persisted up to 11 months post biopsy

Symptom Duration is not Related to Fibrillation Potentials

• Long held notion in the electrodiagnostic literature regarding radiculopathies

• Paraspinal (PSM) muscles denervate first, then more distal

• Reinnervation thought to occur first in PSM then distal

• No evidence to support this model

SYMPTOM DURATION AND EMG FIBRILLATIONS

Dillingham et al, 1998, 1998, 2000; Pezzin et al 1999

• Four separate investigations– Two retrospective (Cervical and Lumbosacral)– Two prospective (Cervical and Lumbosacral)

• Probability of finding fibrillations in a muscle (proximal or distal) was not related to symptom duration.

• Simplistic model of symptom duration doesn’t explain the complex pathophysiology of radiculopathies and their EMG correlates

PROGRESSION OF CERVICAL SPONDYLOTIC CORD COMPRESSION

Bednarik et al Spine 2004

• Symptomatic cervical radiculopathy and EMG showing motor axonal loss in 2 myotomes predicted with 90% accuracy those who progressed to symptomatic myelopathy.

• Odds ratio 12.5 (p<0.001) for EMG

Needle electromyography predicts outcome after lumbar epidural steroid injection (LESI)

• Annaswamy TM, Bierner SM, Chouteau W, Elliott AC. Muscle Nerve. 2012

• 89 subjects prospectively studied

• Predictor variables regarding response to LESI

• Abnormal EMG is a strong and independent predictor of a positive response to LESI – long term pain relief

CONCLUSIONS

• EMG and NCS are excellent diagnostic tests

• Use solid reference foundation

• Realize limitations

• Don’t overcall