Welcome to Facts on the Ground a new series brought to you by Natan Cramer, Pediatric Emergency Medicine fellow from the University of Pittsburgh. Facts on the Ground is designed to be a concise literature review that helps answer common clinical questions.

The Article

The Use of Ultrasound-Measured Optic Nerve Sheath Diameter to Predict Ventriculoperitoneal Shunt Failure in Children
Lin, et al.
Pediatric Emergency Care, 2019

Objective

To evaluate the accuracy of Optic Nerve Sheath Diameter (ONSD) ultrasound in predicting ventriculoperitoneal (VP) shunt failure defined as either advanced image confirmed (CT/MRI) failure or a neurosurgical impression consistent with failure (with subsequent shunt revision).

For more background on Optic Ultrasound check out this link from SonoGuide.

Learn all about CSF shunt complications in this episode of PEM Currents, the Pediatric Emergency Medicine Podcast.

Study Design

This was a prospective convenience sample of patients with VP shunts between the ages of 0-21 years presenting with symptoms concerning for increased ICP who underwent both advanced diagnostic brain imaging and optic nerve ultrasound. It excluded patients with ocular/facial trauma, patients that were hemodynamically unstable, and patients needing emergent neurosurgical evaluation. The final outcome (shunt revision) was compared between positive and negative ultrasounds with advanced imaging results and neurosurgical impression.

Ultrasound was performed by 9 providers (pediatric emergency medicine attendings and residents) trained for this indication. Each subject had two ONSD measurements performed and the mean was evaluated. Shunt failure on ultrasound was defined as a mean ONSD greater than 4.0mm and 4.5mm in children younger and older than 12 months respectively.

Outcome

There were 32 patients enrolled. Sensitivity and specificity for ultrasound was overall low when compared to both advanced imaged confirmed and neurosurgical impression confirmed shunt failure. However, the negative predictive value (NPV) was 95% (Cl: 76.4-99.1%) when compared to neurosurgical impression of failure. NPV compared to advanced image confirmed shunt failure was 90% (Cl: 69.9-97.2%)

Weaknesses

The small sample size limits the generalizability of the study, specifically as there was only one child under the age of one year it is difficult to apply the study to the infant population. Prospective study without power analysis limits the utility as a definite screening tool. Baseline ONSD of asymptomatic patients with VP shunts was not assessed and thus normal measurements may be differ from the regular population. Therefore, the determination of what is considered an “abnormal” ONSD measurement may be skewed.

The Bottom Line

Ultrasound at this time cannot be used to reliably predict who has a true shunt failure requiring revision. It has stronger use in helping to rule out a shunt failure; in conjunction with other pieces of clinical data.