Clinical Indicators of Pediatric Shunt Malfunction: A Population-Based Study From the Nationwide Emergency Department Sample.

Abstract

The majority of the previous literature on clinical predictors of shunt malfunction is from the neurosurgical data, looking at the symptoms of patients who had surgery. Because common childhood illnesses are filtered from these samples, the prevalence of shunt malfunction is markedly higher than it would be for the pediatrician's office or emergency department (ED). Clinical predictive values obtained from a representative population can better inform clinical judgment in these environments. A retrospective analysis of the Nationwide Emergency Department Sample (2006-2015) was performed. International Classification of Diseases, Ninth Revision, Clinical Modification Diagnosis/Procedure Codes were used to identify pediatric (≤20 years of age) ED visits with the presence of a cerebrospinal fluid (CSF) shunt (V45.2). Shunt malfunction was defined as any condition resulting in surgical revision (02.41, 02.42, 02.43). Multivariable logistic regression was used to examine the associations between shunt malfunction, demographic factors, and clinical presentation. There were 74,552 observations for ED visits by pediatric patients with a CSF shunt between 2006 and 2015, of which 12.8% (9,560) required shunt revision. Positive predictive values for clinical indicators were reported along with the results of multivariable logistic regression. We identified peritonitis, papilledema, and oculomotor palsies as the strongest clinical indicators for shunt malfunction in pediatric ED visits with a CSF shunt. We found that patients presenting with headache, nausea/vomiting, convulsions, or fever were more likely to have an etiology other than shunt malfunction. Thus, after an appropriate shunt evaluation, other sources of symptoms should be investigated.