Can ultrasound imaging be used for the diagnosis of carpal tunnel syndrome in diabetic patients? A systemic review and network meta-analysis.

Abstract

High-resolution ultrasound (US) becomes a reliable tool for diagnosing carpal tunnel syndrome (CTS), but whether it can be applied to patients with preexisting diabetes mellitus (DM) remains unclear. We searched PubMed and Embase and systemically reviewed studies exploring the median nerve CSAs at the wrist level by US imaging. Nine studies enrolling at least one subgroup comprising patients with both DM and CTS were included for network meta-analysis. The primary outcome was the inter-group difference of the wrist-level median nerve CSA. The median nerve size at the wrist level was larger in patients with only CTS than in patients with only DM [CSA difference = 3.14 mm2, 95% confidence interval (CI) 1.92-4.35]. Patients with DM and CTS had a slightly enlarged median nerve CSA than did patients with only CTS, but the difference was not statistically significant (0.52 mm2, 95% CI - 0.54 to 1.59). According to rank probabilities, median nerve CSAs in patients with DM and CTS were likely to be ranked as the largest, followed by patients with only CTS, patients with only DM, and healthy controls. Furthermore, median nerve CSAs seemed smaller in patients with than without diabetic polyneuropathy. Although DM causes swelling of the median nerve at the wrist level, patients with CTS have a larger CSA regardless of preexisting DM. The add-on effect of DM on median nerve CSAs in patients with CTS is limited. Diabetic polyneuropathy tends to result in less swollen median nerves in the CTS population.