Background: ShuntCheck (Neuro Diagnostic Devices, Inc., Trevose, Pennsylvania) is a new device designed to detect cerebrospinal fluid (CSF) flow in a shunt by sensing skin temperature downstream from a region of CSF cooled by an ice cube.
Objective: To understand its accuracy and utility, we evaluated the use of this device during routine office visits as well as during workup for suspected shunt malfunction.
Methods: One hundred shunted patients were tested, including 48 evaluated during possible shunt malfunction, of whom 24 went on to surgical exploration. Digitally recorded data were blindly analyzed and compared with surgical findings and clinical follow-up.
Results: Findings in the 20 malfunctioning shunts with unambiguous flow or absence of flow at surgery were strongly correlated with ShuntCheck results (sensitivity and specificity to flow of 80% and 100%, respectively, P = .0007, Fisher's exact test, measure of agreement κ = 0.8). However, the thermal determination did not distinguish patients in the suspected malfunction group who received surgery from those who were discharged without surgery (P = .248 by Fisher's exact test, κ = 0.20). Half of the patients seen in routine office visits did not have detectable flow, although none required shunt revision on clinical grounds. Intermittent flow was specifically demonstrated in one subject who had multiple flow determinations.
Conclusion: Operative findings show that the technique is sensitive and specific for detecting flow, but failure to detect flow does not statistically predict the need for surgery. A better understanding of the normal dynamics of flow in individual patients, which this device may yield, will be necessary before the true clinical utility of non-invasive flow measurement can be assessed.