Adenoviruses are associated with endemic and epidemic acute conjunctiv
itis, large nosocomial outbreaks reflecting virus transmission on unwa
shed hands or inadequately sterilised ophthalmic instruments. The poly
merase chain reaction (PCR) proved more sensitive than antigen detecti
on by immune dot-blot test for the rapid diagnosis of ocular adenoviru
s infection (sensitivities in a retrospective study 112/123 (91%) vers
us 72/123 (59%), P < 0.001). Indeed, in a prospective comparison, DNA
amplification and virus isolation generated similar numbers of positiv
e results (34 versus 32), though five PCR positive results were possib
ly false positives. The sensitivity of the PCR was largely independent
of adenovirus subgenus or serotype, though reduced sensitivity with s
ubgenus B strains could not be excluded. Specimen preparation for DNA
amplification using a simple lysis buffer proved more effective than p
henol-chloroform extraction. The immune dot-blot test gave unavoidable
false positive results, but with the PCR this problem could be minimi
zed by technical modifications. The PCR could replace antigen detectio
n and virus isolation as the initial test for adenoviruses in conjunct
ival swabs, with cell culture only being retained for adenovirus serot
yping in PCR positive specimens and for other viruses such as herpes s
implex. (C) 1995 Wiley-Liss, Inc.