Callaghan BC, Kerber KA, Lisabeth LL, Morgenstern LB, Longoria R, Rodgers A, Longwell P, Feldman EL. Role of Neurologists and Diagnostic Tests on the Management of Distal Symmetric Polyneuropathy. JAMA Neurol. 2014;71(9):1143-1149. doi:10.1001/jamaneurol.2014.1279
Distal symmetric polyneuropathy (DSP) is a prevalent condition that results in high costs from diagnostic testing. However, the role of neurologists and diagnostic tests in patient care is unknown.
To determine how often neurologists and diagnostic tests influence the diagnosis and management of DSP in a community setting.
Design, Setting, and Participants
In this retrospective cohort study, we used a validated case-capture method (International Classification of Diseases, Ninth Revision screening technique with subsequent medical record abstraction) to identify all patients with a new DSP diagnosis treated by community neurologists in Nueces County, Texas, who met the Toronto Diabetic Neuropathy Expert Group consensus criteria for probable DSP. Using a structured data abstraction process, we recorded diagnostic test results, diagnoses rendered (before and after testing), and subsequent management from April 1, 2010, through March 31, 2011.
Main Outcomes and Measures
Changes in DSP cause and management after diagnostic testing by neurologists.
We identified 458 patients with DSP followed up for a mean (SD) of 435.3 (44.1) days. Neurologists identified a cause of DSP in 291 patients (63.5%) before their diagnostic testing. Seventy-one patients (15.5%) had a new DSP cause discovered after testing by neurologists. The most common new diagnoses were prediabetes (28 [6.1%]), vitamin B12 deficiency (20 [4.4%]), diabetes mellitus (8 [1.7%]), and thyroid disease (8 [1.7%]). Management changes were common (289 [63.1%]) and usually related to neuropathic pain management (224 [48.9%]). A potential disease-modifying management change was made in 113 patients (24.7%), with the most common changes being diabetes management in 45 (9.8%), treatment with vitamins in 39 (8.5%), diet and exercise in 33 (7.2%), and adjustment of thyroid medications in 10 (2.2%). Electrodiagnostic testing and magnetic resonance imaging of the neuroaxis rarely led to management changes.
Conclusions and Relevance
Neurologists diagnosed the cause of DSP in nearly two-thirds of patients before their diagnostic testing. Inexpensive blood tests for diabetes, thyroid dysfunction, and vitamin B12 deficiency allowed neurologists to identify a new cause of DSP in 71 patients (15.5%). In contrast, expensive electrodiagnostic tests and magnetic resonance imaging rarely changed patient care.
Disorders of the peripheral nervous system account for 1.5 million visits to neurologists annually, which is more than 10% of all visits.1 Diagnostic testing of these conditions by outpatient neurologists costs $357 million each year, with electrodiagnostic tests ($205 million, 57% of the total cost) and magnetic resonance imaging (MRI) ($135 million, 38% of the total cost) accounting for most of the costs.1 Peripheral neuropathy is the most common disorder of the peripheral nervous system, with a prevalence of 2% to 7% in the entire population, which increases to greater than 10% in elderly populations.2- 4 Not surprisingly, the evaluation of peripheral neuropathy can be costly, with most of the cost driven by electrodiagnostic and MRI testing.5- 7 Given the high aggregate costs associated with this evaluation, determining the value of these diagnostic tests becomes of paramount importance.
Understanding the role of neurologists in the care of patients with peripheral neuropathy is essential. Not all primary care physicians see the benefit of having a neurologist involved in the diagnosis and treatment of common neurologic diseases, such as transient ischemic attack, dementia, and Parkinson disease.8 As reimbursement incentives are realigned, such as the cognitive care bonus for primary outpatient care specialties (which excluded neurologists), the need for evidence to support the value of neurologists is evident and critical.9,10 Although data exist to support the role of neurologists in improving patient outcomes in stroke populations, few data exist to define the role of neurologists in the care of patients with peripheral neuropathy.11- 13
Current evidence supports routine fasting glucose, vitamin B12, serum protein electrophoresis, and glucose tolerance tests in the initial evaluation of distal symmetric polyneuropathy (DSP), the most common subtype of peripheral neuropathy.14 However, little is known about the value of electrodiagnostic tests and MRI in this evaluation. Similarly, to our knowledge, the value of neurologists in the diagnosis and management of patients with DSP has not been previously studied. The aim of this study was to determine the role of neurologists and diagnostic tests in the diagnosis and management of patients with DSP in a community setting.
This study was approved by the University of Michigan Institutional Review Board, and the informed consent requirement was waived. We attempted to identify all patients with a new diagnosis of DSP seen by community neurologists in Nueces County, Texas (retrospective cohort), from April 1, 2010, through March 31, 2011. Most of the county population lives in Corpus Christi, which is more than 140 miles from tertiary care centers in Texas. Because of the geographic location of this community, patients with common medical conditions are likely to receive their medical care within the county. Of the 11 practicing neurologists in Nueces County, 9 agreed to participate and 2 declined (one was retiring and the other did not regularly treat patients with neuropathy).
Patients were required to meet the Toronto Diabetic Neuropathy Expert Group consensus panel definition of probable neuropathy, which is 2 or more of the following criteria: neuropathic symptoms (self-report of pain, numbness, and/or tingling in the feet and/or legs), decreased distal sensation on neurologic examination, or decreased or absent ankle jerks.15 We also required documentation of a diagnosis of neuropathy in the medical record. We excluded patients who were seen in the hospital only, underwent electrodiagnostic testing only, or were previously diagnosed as having neuropathy by a neurologist in Corpus Christi.
We used a previously validated DSP case-capture method that involves screening all new patient visits for International Classification of Diseases, Ninth Revision (ICD-9) neuropathy symptom and diagnostic codes (250.60, 356.0-356.2, 356.4, 356.8, 356.9, 357.1-357.7, 357.82, 357.89, 357.9, 729.5 [pain in limb], and 782.0 [disturbance of skin sensation]) using the billing database of the outpatient office followed by medical record abstraction to confirm that the patients met our DSP definition.16 A previous study16 reported that this case-capture method has a sensitivity of 100% and a specificity of 88% for the classification of patients with DSP when compared with classification by a neuromuscular specialist. The case-capture method was used from April 1, 2010, through March 31, 2011.
Medical records were abstracted by a trained research coordinator (R.L.) using the entire outpatient medical record, from the initial visit and any subsequent follow-up visits within the next 1 to 2 years. Information abstracted included demographic characteristics, clinical characteristics (the 3 criteria for the Toronto consensus definition of probable neuropathy, time since symptom onset, family history of neuropathy, pain, weakness on examination, and the 5 warning signs of an atypical neuropathy, including acute, subacute, or relapsing presentation; asymmetry; non–length dependent; motor predominant; and prominent autonomic features), and all diagnostic tests ordered by the neurologist. We also documented the suspected causes identified by the neurologist at the initial evaluation and at last follow-up as recorded in the medical record. All management changes were documented from the first evaluation to the last neurologist visit. Potential disease-modifying management changes that were recorded included the following: recommending improved diabetes mellitus management, starting treatment with vitamins such as B12, encouraging diet and exercise, changing thyroid medications, recommending alcohol cessation, stopping use of medications thought to be neurotoxic, and starting prednisone therapy. Ongoing quality control comparing abstractions performed by the trained research coordinator (R.L.) and a neuromuscular specialist (B.C.C.) was performed throughout the study period.
We screened the medical records of 4890 patients and identified 831 patients who had an ICD-9 code for neuropathy. Of the 831 patients identified by the screen, 86 were excluded because the patients were seen in the hospital only, underwent electrodiagnostic testing only, or were previously diagnosed as having neuropathy by a neurologist from Corpus Christi. We also excluded 287 patients who did not meet the Toronto consensus for probable DSP. Thus, the final population was 458 patients with DSP. Population demographic characteristics are presented in Table 1. The mean (SD) age of the patients was 65.8 (12.9) years, and 258 (56.3%) were women. Medical records were available for a mean (SD) of 2.4 (1.6) visits per patient and 435.3 (44.1) days after the initial neurology evaluation. In the entire population, neurologists ordered electrodiagnostic testing in 353 patients (77.1%) and MRI of the neuroaxis in 65 (14.2%). Of the 4 American Academy of Neurology–recommended tests, neurologists ordered measurement of vitamin B12 levels in 177 patients (38.6%), measurement of fasting glucose levels in 56 (12.2%), serum protein electrophoresis evaluations in 127 (27.7%), and glucose tolerance tests in 20 (4.4%). Thyrotropin levels were measured in 144 patients (31.4%).
Before diagnostic testing, neurologists were able to determine the cause of DSP in 291 patients (63.5%). The most common cause identified before diagnostic testing was diabetes (233 [50.9%]), followed by thyroid condition (31 [6.8%]), alcohol (14 [3.1%]), chemotherapy (9 [2.0%]), and vitamin B12 deficiency (8 [1.7%]) (Table 2). A total of 167 patients (36.5%) had no clear cause of DSP before testing by the neurologists. Of these 167 patients, diagnostic testing by neurologists revealed a new cause in 45 patients (9.8% of total population). In total, the neurologist discovered a new cause in 71 patients (15.5%). The most common new diagnosis was prediabetes (28 [6.1%]), followed by vitamin B12 deficiency (20 [4.4%]), diabetes (8 [1.7%]), and thyroid disorder (8 [1.7%]) (Table 3). Other new conditions diagnosed by laboratory abnormalities included paraproteinemia and folic acid deficiency. The neurologists identified a new cause in 8 patients based on history alone, including DSP attributed to toxic medications, alcohol, inherited neuropathy, peripheral vascular disease, poliomyelitis, and steel-toed shoes. An additional 4 new causes were based on history and/or laboratory abnormalities, including renal disease, hypoglycemia, and the metabolic syndrome. Of note, 2 patients were no longer considered to have neuropathy after diagnostic testing.
A total of 289 patients (63.1%) had a least 1 management change at or after the initial neurologist evaluation (Table 4). The most common management change was altering medication regimens (262 [57.2%]). A total of 224 patients (48.9%) had a change to their neuropathic pain medications, with most of the changes involving calcium channel agonists (145 [31.7%]), such as gabapentin and pregabalin, tricyclic antidepressants (53 [11.6%]), and serotonin norepinephrine reuptake inhibitors (49 [10.7%]). A potential disease-modifying management change was made in 113 patients (24.7%). Improved diabetes management was recommended for 45 patients (9.8%), starting treatment with vitamins for 39 (8.5%), with most involving vitamin B12 (33 [7.2%]), encouraging diet and exercise for 33 (7.2%), and changing thyroid medications for 10 (2.2%). The next most common disease-modifying management changes were recommending alcohol cessation (8 [1.7%]) and stopping use of medications thought to be neurotoxic (4 [0.9%]). The only other potentially disease-modifying therapy was prednisone, which was given to 2 patients, one of whom also received methotrexate. Of these 2 patients, one had a known mixed connective tissue disease for which the patient was taking prednisone and had chronic DSP of unclear duration. A sural nerve biopsy specimen revealed a mild axonal neuropathy with no features suggestive of vasculitis. The prednisone was increased with little change in the neuropathy, although the patient only had 18 days of follow-up. The other patient had a subacute onset of DSP and was diagnosed by the referring physician as having Sjögren syndrome based on an elevated erythrocyte sedimentation rate and SS-A and SS-B antibodies. The electrodiagnostic study revealed an axonal sensory motor neuropathy. The neurologist prescribed prednisone for the patient, whose neuropathic symptoms substantially improved. Over time the patient was prescribed methotrexate because he was unable to tolerate weaning off the prednisone.
Electrodiagnostic studies (n = 368) led to a change in cause and/or management in 2 patients. In both these patients, the change in cause was from a neuropathy diagnosis to a nonneuropathy diagnosis. Neither of the 2 patients who received disease-modifying therapy with prednisone had a change in management based on electrodiagnostic studies. One did not have a study performed, and the other had the nonspecific finding of an axonal symmetric sensory motor neuropathy. Neuroaxis MRI did not lead to a change in management in any case, with the exception of alteration in pain medications. Diagnostic testing that most frequently led to a change in management included measurement of vitamin B12 levels, thyroid studies, and testing for diabetes (Table 3).
Before diagnostic testing, neurologists were able to determine the cause of DSP in nearly two-thirds of patients with new DSP presenting to community neurologists. After diagnostic testing, neurologists identified a new cause of DSP in an additional 71 patients (15.5%), with the most common new diagnoses being prediabetes, vitamin B12 deficiency, diabetes, and thyroid disease. Therefore, neurologists, with the aid of clinical evaluation and a few simple blood tests, were able to diagnose the underlying cause of DSP in nearly three-fourths of this population. Moreover, neurologists frequently made management changes, with most of these being alterations in neuropathic pain medication regimens. Other common management changes were recommending improved diabetes management, starting treatment with vitamins (usually vitamin B12), encouraging diet and exercise, and altering thyroid medications. On the other hand, other disease-modifying management changes were rare and did not result from the findings of electrodiagnostic tests or MRI of the neuroaxis. Electrodiagnostic tests and MRI rarely changed the management of patients with new DSP presenting to community neurologists.
Similar to previous studies,17,18 we found that diabetes is the most common cause of DSP, accounting for more than half the cases. However, compared with studies17,18 performed in tertiary care clinics and hospital settings, an even higher proportion of patients in this community sample have DSP attributable to diabetes (53% compared with 20%-30%). The next most common diagnostic category was idiopathic DSP, which accounted for more than one-third of cases before diagnostic testing and more than one-fourth after all tests were performed. Several investigations17- 19 have also described a similar proportion of idiopathic neuropathy, although none have been performed in a community neurology setting. On the other hand, alcohol was a much less frequent cause of neuropathy in our sample compared with other populations.17,18 Other common causes found in this population included thyroid dysfunction, prediabetes, and vitamin B12 deficiency.
Almost 10% of patients with an unclear diagnosis before evaluation had a new cause discovered during their diagnostic evaluation. Therefore, neurologists are able to diagnose not only the cause of DSP in nearly two-thirds of patients with history and examination alone but also a new cause in a large proportion of patients after diagnostic testing. Neurologists identified the cause of neuropathy in such a large proportion despite the fact that few patients received all of the tests recommended by the American Academy of Neurology. These results highlight the need to increase guideline adherence testing as has been demonstrated in previous nationally representative studies.6,7 The most common new cause of DSP was prediabetes, followed by vitamin B12 deficiency, diabetes, and thyroid dysfunction. All these new causes have the potential to lead to changes in management and are diagnosed with inexpensive laboratory tests. Other new causes that were discovered were rare and usually the result of new historical information and/or simple laboratory tests. The exception is that 2 patients were considered to have DSP before electrodiagnostic testing, but after these tests the patients were not considered to have neuropathy. Community neurologists identified the cause in three-fourths of patients with DSP through their clinical evaluation or inexpensive laboratory tests.
The most common management change that a community neurologist makes for patients with DSP is adjustment of neuropathic pain medications. Neurologists recommended changes in the pain medication for almost half the patients with DSP, which emphasizes the importance of the neurologist in the care of these patients and the frequency of pain in this prevalent condition. Furthermore, 85% of the changes in pain medication made by neurologists involved the 3 classes of neuropathic pain medications with the best levels of evidence to support their use (serotonin norepinephrine reuptake inhibitors, tricyclic antidepressants, and calcium channel agonists).20,21 Neurologists rarely prescribed nonsteroidal anti-inflammatory drugs or narcotics for neuropathic pain.
Neuropathic pain is often underrecognized and undertreated.22 Moreover, the management of neuropathic pain requires expertise and comfort with multiple medication classes. Neurologists can aid general practitioners not only in the discovery of the underlying cause of neuropathy but also in the management of the pain that is often the patient’s most debilitating symptom.
In contrast to pain management changes, disease-modifying interventions were less commonly initiated by community neurologists. The most frequent interventions included recommending enhanced diabetes management, starting treatment with vitamins such as B12, encouraging diet and exercise, and adjusting thyroid medications. Less common interventions included counseling to limit or abstain from alcohol or to stop use of toxic medications. The only other disease-modifying therapy that was prescribed in this population was prednisone in 2 patients. Although no firm conclusions can be made based on these 2 patients, neither of these patients had a change in management based on an electrodiagnostic test. The only changes in this study that could be attributable to electrodiagnostic testing were the reclassification of 2 patients to nonneuropathy diagnoses after testing. Similarly, MRI of the neuroaxis did not lead to significant changes in cause or management in this population. Although neurologists make important management changes, such as addressing diabetes, prediabetes, vitamin B12 deficiency, and thyroid dysfunction, other management changes are rare and not influenced by electrodiagnostic tests or MRI.
Limitations of the study include the retrospective cohort design and the use of medical record abstraction for data collection. Our ICD-9 case-capture technique may have missed DSP cases; however, a previous study16 found that this technique captures almost 95% of cases in this community. We were unable to study potential changes in cause and management that occur years after seeing a neurologist because the mean follow-up was only 435 days. Similarly, we do not know what the referring physician considered the cause of the DSP or what they would have done without the neurologists being involved; therefore, neurologists likely did not make the initial diagnoses in all cases. We were also unable to validate the neurologist’s diagnosis or assess the quality of the electrodiagnostic testing. This study was not designed to demonstrate whether diagnostic tests that changed the patient’s cause and/or management led to improved patient outcomes or whether patients with DSP who are treated by neurologists have better outcomes than those who are not. Of note, our study population had a low proportion of patients with warning signs of atypical neuropathy; these patients likely require more extensive testing than patients without these features, which should be left to the discretion of the neurologist. Specifically, patients with concern for inherited, vasculitic, or demyelinating neuropathy likely need electrodiagnostic testing. Similarly, some patients with DSP likely need MRI for other indications, such as concerns for spinal stenosis. Future studies are needed to define the clinical scenarios and circumstances in which electrodiagnostic tests and MRI are likely to change the cause of DSP and/or management of patients with DSP. Although many studies23- 28 support prediabetes as a cause of neuropathy, controversy remains. Furthermore, how the results generalize to other communities and clinical settings, such as tertiary referral centers, is unclear.
Neurologists influence the care of patients with DSP in important ways, including discovering new causes that result in potentially disease-modifying therapies, such as treatment of prediabetes, diabetes, vitamin B12 deficiency, and thyroid medications. Neurologists are also instrumental in the treatment of neuropathic pain and frequently make pain medication changes with the best evidence to support their use. More important, the neurologist’s clinical history and examination and a few simple blood tests provide the clues for identifying the underlying cause of most neuropathies. However, electrodiagnostic tests and MRI, which account for most of the expenditures in the evaluation of DSP,5- 7 rarely lead to changes in cause and management. Therefore, the role of these tests requires further study.
Accepted for Publication: April 21, 2014.
Corresponding Author: Brian C. Callaghan, MD, Department of Neurology, University of Michigan, 109 Zina Pitcher Pl, 4021 Biomedical Science Research Bldg, Ann Arbor, MI 48104 (firstname.lastname@example.org).
Published Online: July 21, 2014. doi:10.1001/jamaneurol.2014.1279.
Author Contributions: Dr Callaghan had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Callaghan, Kerber, Lisabeth, Morgenstern, Longoria, Longwell, Feldman.
Acquisition, analysis, or interpretation of data: Callaghan, Kerber, Longoria, Rodgers, Longwell.
Drafting of the manuscript: Callaghan.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Callaghan, Kerber, Lisabeth, Rodgers.
Obtained funding: Morgenstern.
Administrative, technical, or material support: Longoria, Longwell.
Study supervision: Callaghan, Morgenstern, Feldman.
Conflict of Interest Disclosures: Dr Callaghan reported receiving research support from Impeto Medical and performing center certifications for the ALS Association. Dr Kerber reported receiving speaker honoraria from the American Academy of Neurology and Munson Medical Center and serving as a consultant for the American Academy of Neurology and The Weinberg Group. No other disclosures were reported.
Funding/Support: This study was funded by National Institutes of Health grant R01 NS38916.
Role of the Sponsor: Funding source had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.