L. Leibovici et al., PROGNOSTIC-SIGNIFICANCE OF THE NEUTROPHIL COUNT IN IMMUNOCOMPETENT PATIENTS WITH BACTEREMIA, Quarterly Journal of Medicine, 88(3), 1995, pp. 181-189
To examine the prevalence of neutropaenia in immunocompetent, bacterae
mic patients, and whether it carries an independent risk for mortality
, we surveyed 2096 bacteraemic patients without malignant diseases, an
d who were not receiving cytotoxic drugs. The granulocyte count on the
day of the first positive blood culture was <1 x 10(9) cells/l in 33
patients (1.7%, group 1); 1.0-4.0 x 10(9) cells/l in 154 patients (7.9
%, group 2); 4.0-8.0 x 10(9) cells/l in 564 patients (29%, group 3); 8
.0-20.0 x 10(9) cells/l in 1034 patients (53%, group 4); and >20.0 x 1
0(9) cells/l in 163 patients (8.4%, group 5). The mortality rates in t
he five groups were 39.4%, 18.8%, 18.1%, 25.7% and 25.8%, respectively
(p=0.0001). The main pathogens in group 1 were Staphylococcus aureus
in 25% of patients and Pseudomonas sp, in 23%. Mortality in group 1 pa
tients was higher than in the other patients (odds ratio 1.4, 95% CI 1
.1-1.9]. Mortality was also significantly higher in group 2 patients w
ith high blood urea nitrogen. The percentage of neutropaenic, septic p
atients without known risk factors for neutropaenia is small, but thei
r mortality is high. Overall mortality in patients with relative neutr
opaenia (1.0-4.0 x 10(9) cells/l) is low, but a subgroup of patients w
ith high blood urea nitrogen is at considerable risk for a fatal outco
me. High leucocyte counts are also a marker of increased risk for mort
ality, but this association is not an independent prognostic factor.