Background Candida infections affect outcome after injury. Candida ant
igen titres were used to detect these infections early. This study was
undertaken to correlate Candida antigen titre dilution with conventio
nal injury scoring and outcome after severe injury. Methods Candida an
tigen titres were determined by agglutination when clinically apparent
source(s) of Candida were noted, when Candida was grown in culture of
body fluids, or when unexplained clinical deterioration occurred. The
findings were compared with the Injury Severity Score (ISS). Results
Seventy-five seriously injured adults (median ISS 25 (range 18-50)) de
veloped raised Candida antigen titres. Multivariate analysis showed th
at age and Candida antigen titre correlated significantly with mortali
ty, but not with each other. Culture evidence of Candida, or lack ther
eof, did not correlate with Candida antigen titre or mortality. Sixtee
n of 75 patients died, 14 from bacterial sepsis and none from Candida
infection. Conclusion In seriously injured adults, the mortality rate
is related to raised Candida antigen titres. The association between C
andida antigen titre and mortality, although real, remains unexplained
.