Statistics from Altmetric.com
Previous studies show that electrodiagnostic tests and MRIs are frequently ordered in the initial evaluation of neuropathy.1 ,2 However, American Academy of Neurology (AAN) guideline-supported tests, particularly the glucose tolerant test (GTT), are often omitted.1 Recent evidence suggests that electrodiagnostic studies and MRIs are the primary drivers of expenditures associated with neuropathy testing despite limited data to support their use.3 However, these results are based on Medicare claims; therefore, it remains unclear if this observation applies to other populations and when using a more rigorous case definition of neuropathy. Furthermore, Medicare claims do not provide detailed clinical information that would allow for investigation of patient-level factors associated with utilisation and expenditures.
Our aim was to determine utilisation and expenditures in the evaluation of a new diagnosis of distal symmetric polyneuropathy (DSP) by community neurologists using a population-based design and a strict case definition. We also sought to determine which patient and physician factors were associated with testing expenditures, electrodiagnostic and MRI utilisation.
We attempted to capture all new patients with DSP seen by community neurologists in Nueces County, Texas as previously described.4 Patients were required to meet the Toronto consensus panel definition of probable neuropathy and have a documented neuropathy diagnosis. From 1 April 2010 to 31 March 2011, we used a validated International Classification of Diseases 9 case capture technique to screen all new patient visits for cases followed by medical record abstraction to confirm that they met our DSP definition.5 The aetiology at the time of the initial visit to the neurologist was determined by the neurologist's documented assessment. Utilisation was documented separately for the referring physician and for the neurologist. For neuropathy-related expenditures we used the Medicare physician fee and clinical laboratory schedules. Descriptive statistics were used to describe population demographics and clinical variables, utilisation and expenditures. Multilevel (patients nested within physicians) linear and logistic regression were used to investigate the association between patient-level variables and testing expenditures, electrodiagnostic and MRI utilisation by neurologists. As a measure of the importance of the physician effect on our outcomes, we estimated the percentage of the variance in utilisation attributable to physicians using the intraclass correlation coefficient from a model with only a physician-specific random effect. Our models were also used to calculate the percentage of total variance attributable to patient-level characteristics.
We identified 458 patients meeting our DSP criteria as previously described.4 Demographics of the population are presented in online supplementary table S1.
Utilisation of the AAN recommended tests, electrodiagnostic testing and MRIs of the neuroaxis are presented in figure 1A (entire population) and B (unknown aetiology prior to diagnostic testing).
The total mean (SD) testing expenditures were US$892 (507) per patient with electrodiagnostic testing (US$553, SD 306) and MRIs (US$203, SD 363) accounting for 62% and 23% of the total testing expenditures, respectively. All other tests resulted in a mean of US$136 (141) in testing expenditures accounting for 15% of the total.
In fully adjusted models, age was the only patient-level variable associated with testing expenditures and MRI utilisation by neurologists (see online supplementary table S2). Physicians accounted for 41.9% of the variation in diagnostic expenditures, 40% in electrodiagnostic utilisation and 13.6% in MRI utilisation. Patient-level factors accounted for 14% of the variance in diagnostic expenditures, 25.7% in electrodiagnostic utilisation and 17.9% in MRI utilisation.
Electrodiagnostic testing was performed in over 80% of patients, almost all of which were ordered by neurologists. This finding indicates that neurologists consider electrodiagnostic testing an essential component of the initial evaluation of DSP regardless of the clinical scenario. This practice is in concert with recent AAN quality measures, which state that all patients with DSP should undergo an electrodiagnostic test unless they have an obvious cause for neuropathy and highly suggestive symptoms and signs of neuropathy that cannot be attributable to another condition. However, electrodiagnostic tests accounted for 62% of overall testing expenditures and very little evidence supports or refutes their routine use in patients with DSP. Therefore, future work needs to clarify if all patients with DSP should undergo these tests or if only a subset benefit.
MRIs of the neuroaxis were another large driver of expenditures (23%), being performed in 28% of patients. Unlike electrodiagnostic tests, MRIs have a less intuitive role in the evaluation of DSP. Thus, it remains to be seen what is driving MRI utilisation in DSP populations, and whether the utilisation is appropriate. Possibilities include a high frequency of comorbid conditions that require MRI, concern for a central nervous system localisation causing neuropathic symptoms and signs, and increased visits to physicians leading to more opportunities for testing.
Considering AAN guideline-supported tests, only 7.2% of those with an unclear aetiology prior to diagnostic testing had undergone all recommended tests. The GTT was only ordered in 11.4% of this population. While B12 and serum protein electrophoresis (SPEP) were ordered more frequently, 31% with an unclear aetiology did not receive B12 testing and 49% did not receive SPEP testing. Overall, this data suggests an opportunity to improve guideline concordant DSP testing.
Despite detailed demographic and clinical information, age was the only patient-level factor significantly associated with testing expenditures, electrodiagnostic or MRI utilisation by neurologists. This result implies that the individual patient-level factors assessed in this study are not major drivers of the extent of the diagnostic evaluation of DSP. Conversely, physicians account for a dominant proportion of the variability in testing expenditures and electrodiagnostic testing. Therefore, the physician that a patient chooses is more important than the patient's specific clinical scenario in determining the extent of the DSP evaluation. These results indicate a need for clinical decision support for DSP testing, particularly for electrodiagnostic tests.
In conclusion, multiple lines of evidence (nationally representative survey, Medicare claims, and now this study) demonstrate that electrodiagnostic tests and MRIs are frequently ordered in the diagnostic evaluation of DSP and account for more than 80% of the expenditures.1–3 Yet, no prior studies have defined the value of these studies in the evaluation of DSP. Therefore, future studies are needed to determine in which scenarios these tests impact patient management.
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Data supplement 1 - Online supplement
Contributors BCC was involved in study design and statistical analysis, and wrote the manuscript. KAK helped with study design, statistical interpretation, and contributed to the manuscript. MB helped with the statistical analysis plan and interpretation, and contributed to the manuscript. RL performed the screening and medical chart abstraction and participated in the study design. AR assisted in the statistical analysis. PL contributed to the manuscript. LBM, ELF, and LDL were involved in the planning of the project, interpretation of the statistical analysis and contributed to the manuscript. PL, Frank Bonikowski and J Felipe Santos helped coordinate this project at their respective sites.
Funding BCC was supported by a NIH T32 grant and an American Diabetes Association (ADA) Junior Faculty Award and is currently supported by NIH K23 NS079417. BCC and ELF are supported by the Katherine Rayner Program and the Taubman Medical Institute. KAK is supported by NIH/NCRR K23 RR024009 and AHRQ R18 HS017690. LBM is supported by the following NIH funding (significant): R01NS38916, R01NS062675, U01NS056975, U01NS062835, R18HS017690, R01NS073595, and R01HL098065. He also receives significant research support from St. Jude Medical. ELF is supported by NIH R24 DK082841-01 and NIH UO1 DK076160. LDL is supported by NIH/NINDS R01 NS38916, NIH/NINDS R01 NS062675, NIH/NHLBI R01 HL098065, and NIH/NINDS R01 NS070941.
Competing interests KAK received speaker honoraria from the American Academy of Neurology and Munson Medical Center, and served as a consultant for the American Academy of Neurology, and The Weinberg Group. KAK and LBM also provide expert medical legal work (modest) that does not involve industry.
Ethics approval University of Michigan IRB.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.