Abstract

A conventional method of bacteriological examination of removed cerebrospinal fluid shunts was compared with another method which relies on microscopic and cultural examination of intraluminal fluid. Fifty-five shunts were tested. All eight cases of clinical shunt infection gave positive results with the latter method, whereas a further 23 shunts yielded positive cultures by the conventional method in the absence of clinical infection. The consequences of missed infections due to omission of microscopic examination and overdiagnosis using the conventional culture method are discussed.