A timed Phalen’s test predicts abnormal electrophysiology in carpal tunnel syndrome

Abstract Objective Previous studies reported variable sensitivity and specificity of the Phalen test. We investigated whether a timed Phalen’s test (TPT) could predict abnormal nerve conduction studies (NCS) results in carpal tunnel syndrome (CTS). Methods Patients with CTS were consecutively recruited. A neurologist confirmed the clinical diagnosis of CTS and recorded the TPT before NCS were performed. Another neurologist, blinded to the TPT, graded the severity of NCS. The TPT sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated. Results In total, 403 patients with 706 hands were recruited and diagnosed with CTS; 465 hands had positive TPT, and 611 hands showed abnormal NCS results. A positive TPT at ≤ 10 s had a specificity of 96.8% and a PPV of 96.6% in predicting abnormal NCS. The sensitivity and NPV of TPT were insignificant. Discussion A positive TPT at ≤ 10 s can be useful in predicting NCS abnormalities in patients with CTS.


| INTRODUC TI ON
The diagnosis of carpal tunnel syndrome (CTS) is based on clinical features and confirmed by nerve conduction studies (NCS). The sensitivity and specificity of NCS exceeds 85% and 95%, respectively (Jablecki et al., 2002). It has been suggested that physicians incorporate NCS along with clinical measures when assessing patients with CTS (Schrijver et al., 2005). One of these clinical tools is the wristflexion test described by Phalen in 1951(Phalen, 1951, 1970Phalen & Kendrick, 1957). A false-positive Phalen test has been reported in 20% of normal subjects as well as in other conditions including cervical radiculopathy, tendonitis, and ulnar neuropathy (Jablecki et al., 2002).
Subsequent studies revealed a wide range of sensitivity, ranging from 42% to 85%, and specificity, ranging from 54% to 98% (Brüske et al., 2002;D'Arcy & McGee, 2000). This variability limits the diagnostic utility of the Phalen test (D'Arcy & McGee, 2000). However, previous studies have not thoroughly explored the rapidity of symptoms provocation by the Phalen test in relation to NCS findings.

| Clinical assessment
A neurologist assessed patients for inclusion and exclusion criteria, recorded the TPT, and completed a data collection form before NCS were performed. Patients were asked to flex both wrists by placing the dorsal surface of the hands against each other while the elbows were flexed and arms were abducted. Wrist flexion was maintained at the maximum possible degree up to 90° for 60 seconds (s), or until positive results were reported. A positive TPT was defined as numbness or paresthesia in at least one of the median-innervated digits. Patients were instructed to immediately inform the examiner once the symptoms began. The examiner recorded the time at which patients reported the onset of symptoms for each hand.

| Electrodiagnosis of abnormal nerve conduction
Two trained technicians blinded to TPT results performed NCS following the protocol described previously (Alanazy, 2017). In brief, antidromic sensory NCS were conducted for median digits 2 and 4, and ulnar digits 4 and 5. Both nerves were stimulated 14 cm proximal to the recording electrode. Palmar orthodromic mixed studies were conducted by stimulating median and ulnar nerves in the palm and recording the respective nerve 8 cm proximal to the stimulation site. Motor NCS were conducted by using a belly-tendon montage and stimulating the respective nerve 7 cm proximal to the record-

| Analysis
Spearman correlation coefficients were used to assess the relationship between TPT, NCS severity, age, and diabetes. NCS results were dichotomized into normal (grade 0) and abnormal (grades 1-5). Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of a positive TPT at ≤10 , 15, 20, 25, 30, 35, 40, 45, 50, 55, and 60 s in predicting abnormal NCS were calculated. PPV indicates the likelihood that a positive TPT at each cutoff value would predict abnormal NCS. Negative predictive value indicates the likelihood that a negative TPT at each cutoff value would predict a normal NCS. Two-tailed p < .05 were considered significant. Data were analyzed using the software SPSS (version 23, Chicago, IL).

| D ISCUSS I ON
This study found that shorter Phalen times correlated with higher specificity and PPV in predicting the presence of abnormal NCS results, but not NCS severity. However, the sensitivity of the low TPT cutoff values was poor, and even at 60 s, it was within the range of the previously reported sensitivities (Brüske et al., 2002). Nevertheless, a positive TPT at ≤10 s had a high specificity and low false-positive rate (FPR = 1 − specificity, 3.2%), which indicate that it has value as a rule-in test in clinical diagnoses. Higher cutoff values can be chosen based on false-positive rates that would be tolerable for individual cases. The PPV remained above 90% for all cutoff values. PPV and NPV are influenced by disease prevalence. The prevalence of abnormal NCS in our study is consistent with that reported in the literature (Werner & Andary, 2011). Furthermore, the electrodiagnostic techniques and reference values used here are consistent with the recommendations by the American Academy of Neuromuscular and Electrodiagnostic Medicine (Werner & Andary, 2011).
Consistent with previous studies, (Nora et al., 2004) the TPT at all cutoff values was positive in a higher proportion of patients with moderate and severe NCS than in those with minimal and mild NCS; however, the correlation coefficient was negligible. Ansari