Risk of infection in patients who have co-placement of a percutaneous endoscopic gastrostomy and a ventriculoperitoneal shunt: a retrospective study of 23 patients

• Published in: International Congress series 2004-02, Vol.1259, p.373-373
• Main Authors: Nabika, Shinya, Oki, Shuichi, Sumida, Masayuki, Isobe, Naoyuki, Kanou, Yukari, Watanabe, Yosuke
• Format: Article
• Language: English
• Subjects: Infection; Percutaneous endoscopic gastrostomy; Ventriculoperitoneal shunt
• ISSN: 0531-5131; 1873-6157
• DOI: 10.1016/S0531-5131(03)01820-X

Objectives: Patients with severe neurologic impairment requiring tube feeding may display concomitant hydrocephalus. Co-placement of a percutaneous endoscopic gastrostomy (PEG) and ventriculoperitoneal (VP) shunt is currently standard in such cases. The present study investigated factors involved with shunt infection in patients who had undergone PEG and VP shunt procedures. Methods: Medical records for 23 patients with a PEG and cerebrospinal fluid (CSF) shunt were retrospectively reviewed. Correlations between shunt system infection and factors potentially contributing to infection were analyzed. Potential factors included the order of installation of the PEG and shunt, position of the abdominal shunt catheter, diabetes mellitus, tracheostomy, and activities of daily living. Results: Twelve patients received a CSF shunt after PEG insertion, while 11 underwent PEG after shunt placement. Four patients experienced shunt infection, and three of these required shunt revision. Three of these four underwent shunt placement after PEG insertion. The period between PEG insertion and CSF shunt placement for these three patients was 18, 19 and 25 days, shorter than the mean period of 30.0 days. Conclusion: The incidence of shunt infection in patients receiving PEG before CSF shunt placement was slightly higher than incidences reported elsewhere. A relatively short period between the placement of the PEG and subsequent CSF shunt seems to represent a contributing factor for shunt infection. Other factors, including existing tracheostomy, displayed no correlation with shunt infection.