The Utility of Inpatient Electromyography/Nerve Conduction Studies: Quality Over Quantity Saves Quite a Quota

Objective: To predict the utility of changing inpatient management with EMG/NCS by admission criteria, EMG/NCS indication, duration of symptoms prior to testing, presence of artifacts, and location of testing. Background: Inpatient EMG/NCS cost more than outpatient studies at our center due to facility fees. Professional opinion suggests inpatient studies have more artifacts, are performed too early from symptom onset, and often are irrelevant to admission criteria, prompting this retrospective cross-sectional study. Design/Methods: Charts (n=83) were examined for admission criteria, EMG/NCS indication, duration of symptoms prior to EMG/NCS, location of testing, and presence of artifact (edema, 60 hertz interference, or cold limbs). “Change in management” occurred if medications were started/stopped, workup was broadened/narrowed, or early discharge occurred. Cost was compared with t-testing, and categorical variables were studied with Chi-Squared statistics. Results: Cost was less for outpatient EMG/NCS (μ=$124.69±$40.12) compared to inpatient tests (μ=$519.17±$174.37) statistically (T=1.98, p=1.6E-37). Matching indications for admission and EMG/NCS positively predicted change in management (OR=9.4,χ2=6.19, p=0.013). Presence of symptoms ≥3 weeks prior to testing positively predicted change in management (OR=3.7,χ2=4.99, p=0.025). Artifact was less likely with ward EMG/NCS than PCU/ICU testing (OR=0.18,χ2=5.15, p=0.023). The presence of artifacts did not negatively affect inpatient management (χ2=0.085, p =0.77), though studies were often repeated due to artifacts. Results of EMG/NCS being normal or abnormal favored abnormal tests changing inpatient management only slightly (OR=1.04,χ2=4.82, p=0.028). Conclusions: Inpatient EMG/NCS are more expensive than outpatient tests at our center. To optimize quality of data and chances of changing management, the admission criteria should match the EMG/NCS indication, the patient’s symptoms should be present for ≥3 weeks prior to testing, and testing should be performed in a ward to reduce artifacts. While artifacts did not statistically negatively impact changing inpatient management in this study, artifacts resulted in repeat testing. Neither normal nor abnormal testing predicted change in management appreciably over the other without appropriate clinical context. Disclosure: Dr. Miller has nothing to disclose. Dr. Bodkin has received research support from Mallinkrodt Pharmaceuticals, Alexion Parmaceuticals and Amicus therapeutics..