Timothy R. Dillingham

Identifying lumbosacral radiculopathies: An optimal electromyographic screen

ObjectiveThe objective of this study was to determine prospectively the optimal electromyographic screening examination of the lower limb that ensures identification of those lumbosacral radiculopathies that can be electrodiagnostically confirmed, yet minimizes the number of muscles studied. DesignA prospective multicenter study was conducted from May 1996 to September 1997. Patients with suspected lumbosacral radiculopathy referred to participating electrodiagnostic laboratories were recruited and examined by needle electromyography using a standard set of muscles. Patients with electrodiagnostically confirmed lumbosacral radiculopathies were selected for analysis. Various muscle screens were tested against this group of patients with radiculopathies to determine the frequency with which each screen identified the patient with radiculopathy. ResultsThere were 102 patients identified. When paraspinal muscles were one of the screening muscles, four-muscle screens identified 88–97% of the radiculopathies, five-muscle screens identified 94–98%, and six-muscle screens 98–100%. When paraspinal muscles were not part of the screen, identification rates were lower for all screens, and eight distal muscles were necessary to identify about 90% of the radiculopathies. ConclusionsSix-muscle screens with paraspinal muscles yielded consistently high identification rates. Studying additional muscles produced no improvements in identification.

Identification of cervical radiculopathies: Optimizing the electromyographic screen

Dillingham TR, Lauder TD, Andary M, Kumar S, Pezzin LE, Stephens RT, Shannon S: Identification of cervical radiculopathies: optimizing the electromyographic screen. Am J Phys Med Rehabil 2001; 80:84–91. ObjectiveTo determine the optimal electromyography screening examination of the upper limb that ensures detection of those cervical radiculopathies, which can be electrodiagnostically confirmed, yet minimizes the number of muscles studied. DesignA prospective multicenter study was conducted from May 1996 to September 1997 at five institutions. Patients who were referred to participating electrodiagnostic laboratories with suspected cervical radiculopathy were recruited. A standard set of muscles were examined by needle electromyography. Patients with electrodiagnostically confirmed cervical radiculopathies, based on electromyography findings, were selected for analysis. Muscle screens were tested against this group to determine whether the screen identified the patients with radiculopathy. ResultsThere were 101 patients with cervical radiculopathies representing all cervical root levels. When paraspinal muscles were one of the screening muscles, five muscle screens identified 90% to 98% of radiculopathies, six muscle screens identified 94% to 99%, and seven muscle screens identified 96% to 100%. When paraspinal muscles were not part of the screen, eight distal limb muscles recognized 92% to 95% of radiculopathies. ConclusionThis study demonstrated that six muscle screens including paraspinal muscles yielded consistently high identification rates. Studying additional muscles led to marginal increases in identification.

Effect of history and exam in predicting electrodiagnostic outcome among patients with suspected lumbosacral radiculopathy.

OBJECTIVE To determine the extent to which the history and physical examination predict the outcome of the electrodiagnostic (EDX) evaluation in patients with suspected lumbosacral radiculopathy. DESIGN Data for 170 subjects referred for low-back and lower limb symptoms were prospectively collected at five EDX laboratories. The sensitivity, specificity, positive and negative predictive values, and odds ratios were determined for symptoms and neurologic signs. RESULTS Symptoms were not significantly associated with an EDX study or a lumbosacral radiculopathy. The physical examination was better at predicting that an EDX study would be abnormal in general than it was at predicting a lumbosacral radiculopathy in particular. Of those subjects with normal physical examinations, 15%-18% still had abnormal EDX findings. CONCLUSIONS In a population of patients referred for an EDX study, the history and physical examination alone cannot reliably predict electrodiagnostic outcome.

Musculoskeletal Disorders in Referrals for Suspected Lumbosacral Radiculopathy

Cannon DE, Dillingham TR, Miao H, Andary MT, Pezzin LE: Musculoskeletal disorders in referrals for suspected lumbosacral radiculopathy. Am J Phys Med Rehabil 2007;86:957–961. Objective:An electrodiagnostic evaluation is often requested for patients with suspected lumbosacral radiculopathy. Although musculoskeletal disorders can produce lower-limb symptoms, their prevalence in this referral population is unknown. The purpose of this study was to determine the prevalence of common lower-limb musculoskeletal disorders in patients referred for lower-limb electrodiagnosis and determine whether these findings predict study outcome. Design:Subjects undergoing an electrodiagnostic study for lower-limb symptoms and suspected lumbosacral radiculopathy constituted the sample. A standardized clinical and electrodiagnostic evaluation was performed for all patients. Results:There were 170 subjects included. The mean age was 52 (SD = 17), and 45% were males. The total prevalence of musculoskeletal disorders in the sample was 32%. The prevalence in those with a normal study was 55% as compared with 21% in those with lumbosacral radiculopathy (P < 0.0001). Conclusions:Musculoskeletal disorders are common in patients suspected of having lumbosacral radiculopathy. The high prevalence among both patients with normal studies and those with radiculopathy and other disorders limits the usefulness of this information in predicting study outcome. In particular, it is common for patients to have two or more problems and the presence of a musculoskeletal disorder should not preclude such testing.

Electrodiagnostic services in the United States

To identify patterns of service provision in the United States, electrodiagnostic encounters were characterized by analyzing data from the 1998 MarketScan Commercial Claims & Encounters Database (The MEDSTAT Group). Electrodiagnostic‐related claims were identified using CPT codes for nerve conduction studies, electromyography, and somatosensory evoked potentials. Over 130,000 electrodiagnostic claims representing 48,207 separate electrodiagnostic encounters were identified. Neurologists accounted for the highest percentage of physician providers (23.6%), followed closely by physiatrists (22%). Physiatrists performed the services at a higher rate than any other provider (886.3 studies per 100 practicing physiatrists). Electrodiagnostic encounters by nonphysician providers accounted for 16.9% of all studies. Physical therapists were the dominant nonphysician providers of electrodiagnostic services (9.3%), followed by podiatrists (5.5%). Physician providers had a median of three claims per encounter compared with one claim per encounter for nonphysician providers. The relatively large proportion of studies conducted by nonphysicians, as well as their limited extent of testing, raises concerns about the quality of care for these patients. These quality issues should be addressed in future studies. Muscle Nerve 29: 198–204, 2004

Cervical radiculopathies: Relationship between symptom duration and spontaneous EMG activity

The purpose of this multicenter study was to prospectively examine whether denervation in paraspinal muscles (PSM) and in other major proximal and distal muscles is related to symptom duration in cervical radiculopathies (CRs). Information was collected on 93 electrodiagnostically confirmed CRs using standardized history, physical examination, and electromyographic (EMG) screens. Multivariate, maximum‐likelihood estimates showed no evidence of correlation between PSM spontaneous activity and symptom duration. Symptom duration was also nonsignificant in eight of the remaining nine upper limb muscles analyzed. We conclude that the probability of spontaneous activity is not related to symptom duration. Clinicians, therefore, should refrain from interpreting electrodiagnostic findings based upon duration of symptoms.

Under-recognition of polyneuropathy in persons with diabetes by nonphysician electrodiagnostic services providers.

Dillingham TR, Pezzin LE: Under-recognition of polyneuropathy in persons with diabetes by nonphysician electrodiagnostic services providers. Am J Phys Med Rehabil 2005;84:399–406. Objective:Healthcare providers commonly refer patients to physiatrists and neurologists for electrodiagnostic testing when they have symptoms suggestive of a peripheral nerve disorder. Published practice guidelines specify that electrodiagnostic medicine consultants should possess special neurologic and procedural training in this area. We recently found that despite these practice guidelines, physical therapists, chiropractors, and podiatrists perform 17% of electrodiagnostic studies in the United States. These findings prompted the current investigation examining electrodiagnostic care across different providers for an important target population—persons with diabetes. Design:A retrospective cohort of patients with diabetes who underwent electrodiagnostic testing in 1998 was identified in the MarketScan Commercial Claims & Encounters Database (The MEDSTAT Group) using CPT and ICD9CM codes. This database represents the healthcare claims for 16 million Americans in private and employer-based health plans. The outcome of interest was the rate of polyneuropathy identification across different providers, controlling for patient characteristics. Results:There were 6381 electrodiagnostic encounters for persons with diabetes in 1998. Polyneuropathy identification rates were highest for physiatrists, osteopathic physicians, and neurologists (12.5%, 12.2%, and 11.9%, respectively). Podiatrists and physical therapists identified 2.4% and 2.1%, respectively, as having polyneuropathy—rates about one sixth that of physiatrists and neurologists despite controlling for casemix differences. Nonphysician providers who did not recognize polyneuropathy performed almost exclusively EMG testing (>90%) at the expense of nerve conduction studies. Conclusions:This study raises concerns about the quality of electrodiagnostic testing by nonphysician providers for persons with diabetes. These results should prove useful for physicians, third-party payers, and health policy makers when confronting issues related to provision of electrodiagnostic services.

Symptom duration and spontaneous activity in lumbosacral radiculopathy

OBJECTIVES A long-held notion in the electrodiagnostic literature is that paraspinal muscles tend to show spontaneous activity (fibrillations and positive sharp waves) on needle electromyography, early on in a lumbosacral radiculopathy, and that more distal muscles become abnormal later in the disease process. The purpose of this study was to determine whether paraspinal muscle and other major proximal and distal muscle spontaneous activity is related to a lumbosacral radiculopathy symptom duration. METHODS A multicenter, prospective study that collected standard information on history, physical examination, and electrodiagnostic findings in patients with electrodiagnostically confirmed lumbosacral radiculopathies was undertaken. RESULTS Multivariate probit analyses of 96 patients identified with a lumbosacral radiculopathy showed no evidence of correlation between spontaneous activity in the paraspinal muscles and symptom duration. Symptom duration was also nonsignificant in nine of the remaining ten lower limb muscles analyzed. CONCLUSION These findings emphasize the limitations of using symptom duration when interpreting electrodiagnostic findings in lumbosacral radiculopathy.

Relationship between muscle abnormalities and symptom duration in lumbosacral radiculopathies.

Dillingham TR, Pezzin LE, Lauder TD: Relationship between muscle abnormalities and symptom duration in lumbosacral radiculopathies. Am J Phys Med Rehabil;1998;77:103–107 A long held notion in the electrodiagnostic literature is that paraspinal muscles tend to show electromyographic abnormalities early on in a lumbosacral radiculopathy and that more distal muscles become abnormal later in the disease process. The purpose of this study was to determine whether paraspinal muscles and other major proximal and distal muscle abnormalities are related to lumbosacral radiculopathy symptom duration. A multivariate logit analysis of 139 patients (retrospectively identified) with electrodiagnostically confirmed lumbosacral radiculopathies was used to test these hypotheses. Maximum likelihood estimates showed no evidence of correlation between abnormal paraspinal muscles and symptom duration. Symptom duration was also insignificant for the remaining five lower limb muscles analyzed. We conclude that the probability of having electromyographic abnormalities is not related to symptom duration. A prospective study is needed to confirm these findings. Nonetheless, clinicians should use caution when interpreting electrodiagnostic findings based on symptom duration.

Electrodiagnostic consultation and identification of neuromuscular conditions in persons with diabetes.

Introduction: Although the American Association of Neuromuscular and Electrodiagnostic Medicine recommends that electrodiagnostic procedures should be performed by physicians with specialty training, these procedures are increasingly being performed by non‐specialists. Methods: We used a nationally representative sample of Medicare beneficiaries with diabetes who used electrodiagnostic services in 2006 to examine whether specialists and non‐specialists were different in the rates of identifying common neuromuscular conditions. Results: Specialists (neurologists and physiatrists) performed 62% of electrodiagnostic consultations; non‐specialist physicians and non‐physicians performed 31% and 5%, respectively. After adjusting for age, race/ethnicity, diabetes severity, and comorbidities, specialists were 1.26–9 times more likely than non‐physicians to diagnose polyneuropathy, lumbosacral radiculopathy, cervical radiculopathy, carpal tunnel syndrome, and ulnar neuropathy. Almost 80% of electrodiagnostic studies performed by specialists included electromyography testing; fewer than 13% by non‐specialists did. Conclusions: Inadequate use of electromyography and fewer specific diagnoses suggest that many non‐specialists perform insufficiently comprehensive electrodiagnostic studies. Muscle Nerve, 2011