UW Radiology

Carpal Tunnel Syndrome

Carpal Tunnel Syndrome: Diagnosis and Treatment Trial

The primary goal is to study the effectiveness of surgery for patients with mild to moderate carpal tunnel syndrome (CTS). An important secondary goal is to study the ability of MRI to predict patient outcomes. We have designed a randomized controlled trial nested within a cohort study.

Principal Investigator

Jerry Jarvik, MD, MPH
Professor, Radiology and Neurosurgery
Section Chief – Neuroradiology
Adjunct Professor, Health Services

Research Questions

1. Determine if select patients with early, mild or moderate CTS benefit from early surgery compared with conservative therapy.

We will assess patient outcomes using clinical data (e.g. hand diagrams) as well as validated questionnaires for symptoms, functional status, satisfaction, and disability. We will obtain measures at baseline, before randomization, and then 1.5, 3, 6, 9 and 12 months. Our primary outcome measure will be the 12-month Carpal Tunnel Syndrome Assessment Questionnaire (CTSAQ) functional status index.

2. Compare the predictive value of MRNI with electrodiagnostic studies for determining clinical outcome following either conservative treatment or surgery.

Our hypothesis is that MRNI will predict better than EDS which patients will benefit from carpal tunnel release. Our primary imaging variable of interest will be nerve signal abnormality on T2-weighted images. Three sensory nerve latency differences (median, ulnar and radial) will be summed to form the combined sensory index; a sensitive, reliable, and specific parameter for diagnosing CTS. Subjects will have baseline MRNI and EDS. We will blind patients and clinicians to the MRNI, with clinical decisions made on the basis of the clinical findings and EDS. This will remove any potential influence of the MRNI on treatment decisions or patient outcomes. We will then compare how well the diagnostic tests predict primary and secondary outcomes.

3. Examine the predictive value of symptom, functional status, disability, and psychological factors alone and in conjunction with MRNI and EDS in patients with mild to moderate CTS.

There is evidence that depression, somatization, multiple pain sites and disability predict poor long term outcomes for low back pain as well as other pain conditions. However, little is known about the association of these factors with CTS outcomes. Thus, we will examine the strength of association of these factors, alone and in combination with the MRNI and EDS findings, with measures of symptoms and functional status at follow-up, using multivariate analytic techniques. This will be an exploratory analysis.

Research Highlights


Darshan Acharya MD
Jason Shewchuk MD
Patrick J. Heagerty PhD
Leighton Chan MD, MPH
Jeffrey G. Jarvik MD, MPH


To perform high resolution MR of the carpal tunnel in 92 subjects with mild to moderate carpal tunnel syndrome who were participants in one of two on-going prospective studies. All subjects had clinical evaluations and completed questionnaires that included the Carpal Tunnel Syndrome Assessment Questionnaire (CTSAQ), as well as measures of somatization and depression. 89 subjects underwent nerve conduction studies. High resolution images were obtained using phased-array wrist coils. Two readers, blinded to all clinical information, independently measured the median nerve and carpal tunnel at preselected anatomic points. Using these measurements, we derived nerve flattening ratios, (height/width) and the proportional change in nerve size within the carpal tunnel ({distal radio-ulnar joint} – {proximal metacarpals})


  1. MRI can reliably be used to quantitatively measure the median nerve in the carpal tunnel. Limitations include resolution at the edge of the coil, which may be responsible for poorer inter-reader reliability at the proximal metacarpal joint. Reliability was poorer for determining abnormally bright median nerve signal and the locations where the median nerve was ‘flattest’ and ‘largest’
  2. The nerve area at its largest point was positively associated with both symptom and function summary scores. EDS was not associated with either summary measure, but was associated with numbness. The measure having the greatest correlation with symptoms and function was the somatization score.

These images reflect an enlarged nerve containing prominent, irregular fascicles with increased T2 signal intensity. T1 and STIR T2 images of an abnormally flattened median nerve at its flattest point. MRI can reliably be used to quantitatively measure the median nerve in the carpal tunnel. Limitations include resolution at the edge of the coil, which may be responsible for poorer inter-reader reliability at the proximal metacarpal joint. Reliability was poorer for determining abnormally bright median nerve signal and the locations where the median nerve was ‘flattest’ and ‘largest’.


Jarvik, J. G. MD, MPH
Yuen, E. MD
Haynor, D. R. MD, PhD
Bradley, C. M. MS, MPH
Fulton-Kehoe, D. MPH
Smith-Weller, T. RN, MN, COHN-S
Wu, R. MD, MPH
Kliot, M. MD
Kraft, G. MD
Wang, L. MD
Erlich, V. MD, PhD
Heagerty, P. J. PhD
Franklin, G. M. MD, MPH

Neurology. 2002 June 11; 58(11) 1583-4.


To evaluate the reliability and diagnostic accuracy of high-resolution MRI of the median nerve in a prospectively assembled cohort of subjects with clinically suspected carpal tunnel syndrome (CTS).


Intra-reader reliability was substantial to near perfect (kappa = 0.76 to 0.88). Inter-reader agreement was lower but still substantial (kappa = 0.60 to 0.67). Sensitivity of MRI was greatest for the overall impression of the images (96%) followed by increased median nerve signal (91%); however, specificities were low (33 to 38%). The length of abnormal signal on T2-weighted images was significantly correlated with nerve conduction latency, and median nerve area was larger at the distal radioulnar joint (15.8 vs 11.8 mm(2)) in patients with CTS. A logistic regression model combining these two MR variables had a receiver operating characteristic area under the curve of 0.85.

Correlation of wrist area measured clinically with carpal tunnel area measured by MRI (Rsq = 0.2292).
From: Jarvik: Neurology, Volume 58(11). June 11, 2002. 1597-1602


Martin, B.I.
Levenson, L.M.
Hollingworth, W.
Kliot, M.
Heagerty, P.J.
Turner, J.A.
Jarvik, J.G.


To design a randomized clinical trial to compare surgical release to non-surgical treatment for patients with mild to moderate CTS. We will examine the association between outcome, as measured by symptoms and functional status, and baseline variables such as symptoms, function, occupational risk factors, EDS measures, demographics, signs and symptoms, and prior treatments. Our primary endpoint is at 12 months. An important secondary goal is to study the ability of MRI to predict patient outcomes.


A Randomized Control Trial offers the best chance of answering, in an unbiased fashion, the relative efficacies of surgery compared with conservative therapy for patients with mild to moderate CTS.


Grant G.A.
Britz, G.W.
Goodkin, R.
Jarvik, J.G.
Maravilla, K.
Kliot, M.

Muscle Nerve. 2002 Mar, 25(3):314-31.


To illustrate how standard and evolving magnetic resonance imaging (MRI) techniques provide additional information in dealing with some of the challenging peripheral nerve problems.


Several conclusions from this study can be drawn. First, MRI can be used to identify reliably the median nerve and other structures within the carpal tunnel. Second, the diagnostic accuracy of MRI compared to electrodiagnostic studies is only moderate in evaluating patients with CTS. This finding may not be all that surprising, as the two methods are evaluating different aspects of this disease. It must be kept in mind that subjects were diagnosed as having CTS (i.e., the gold standard) on the basis of clinical and electrodiagnostic criteria. MRI could prove useful in helping to predict which patients will respond best to medical or surgical treatment, a study that is currently being planned by our group. MRI may also be useful in diagnosing and guiding treatment of patients with clinical CTS who have normal electrodiagnostic studies, a group representing 10% of the subjects in this study. It will be very important to follow their clinical course in response to medical or surgical therapy. Finally, MRI may be helpful in diagnosing and treating patients with clinical CTS who have electrodiagnostic abnormalities that preclude localization of median nerve pathology to the carpal tunnel (e.g., patients with advanced CTS or a severe superimposed peripheral neuropathy).

Selected Bibliography

Martin BI, Levenson LM, Hollingworth W, Kliot M, Heagerty PJ, Turner JA, Jarvik JG. Randomized clinical trial of surgery versus conservative therapy for carpal tunnel syndrome. BMC Musculoskeletal Disord. 2005 Jan 18;6(1):2

Storm S, Beaver SK, Giardino N, Kliot M, Franklin GM, Jarvik JG, Chan L. Compliance with electrodiagnostic guidelines for patients undergoing carpal tunnel release.Arch Phys Med Rehabil. 2005 Jan;86(1):8-11; quiz 180.

Turner JA, Franklin G, Heagerty PJ, Wu R, Egan K, Fulton-Kehoe D, Gluck JV, Wickizer TM. The association between pain and disability. Pain. 2004 Dec;112(3):307-14.

Jarvik JG, Yuen E, Haynor DR, Bradley CM, Fulton-Kehoe D, Smith-Weller T, Wu R, Kliot M, Kraft G, Wang L, Erlich V, Heagerty PJ, Franklin GM. MR nerve imaging in a prospective cohort of patients with suspected carpal tunnel syndrome. Neurology. 2002 Jun 11;58(11):1597-602.

Jarvik JG, Yuen E. Diagnosis of carpal tunnel syndrome: electrodiagnostic and magnetic resonance imaging evaluation. Neurosurg Clin N Am. 2001 Apr;12(2):241-53.