Nicole Francoise BOUPDA
Dr.med.
Significance of coagulase-negative staphylococci isolated from blood cultures of pediatric
patients
Geboren am 09. 03. 1972 in Douala, Kamerun
Reifeprüfung am 10. 08. 1988 in Douala, Kamerun
Studiengang der Fachrichtung Medizin vom SS 1989 bis SS 1998
Physikum am 21. 03. 1991
Klinisches Studium in Heidelberg
Praktisches Jahr in Heidelberg/Mannheim
Staatsexamen am 23. 05. 1995 an der Universität Heidelberg
Promotionsfach: Hygiene
Doktorvater: Prof. Dr. med. Heinrich Konrad Geiss
The problem which was posed at the beginning of the actual study was the following: Is it possible to
distinguish coagulase-negative staphylococci (CNS)-sepsis isolates from contaminants in order to
diagnose a CNS-sepsis?
The study was realized with a patient population of 83 children, whose charts were reviewed
regarding the sepsis diagnosis. The decision to make a diagnosis was strongly dependent of 3 major
and 5 minor clinical criteria. In this way, 96 disease episodes were considered, of which 33 turned to
be sepsis and 63 non-sepsis episodes. The belonging 100 CNS-isolates were then examined regarding
their microbiological properties.
At the end of this study, we agree that CNS are potential pathogens for sepsis. The mechanisms of
pathogenicity are not all cleared yet. We could only examine a few aspects of the virulence of these
bacteria, finding that S.epidermidis was the most frequent species isolated as well during sepsis as
non-sepsis episodes. The predictive value positive for this species and concerning the sepsis diagnosis
was poor (38,3%), however we could prove that a contamination was more probable (84,2%) when
the species isolated was another CNS species.
Neither the biotype nor slime nor DNAse production could give a clinical significance to the CNS.
The two last properties were found as well among sepsis isolates as among contaminants, but they
were uncommon in general. It is interesting to note that 2/3 of the CNS isolates which didn’t produce
any of the 2 substances were contaminants, which can be explained by the fact that our CNS-
contaminant population was much greater than the CNS-sepsis isolate one.
Generally, the sepsis isolates were more resistant to antibiotics than contaminants. A sepsis could be
diagnosed in nearly half of the cases if the isolate was resistant to at least 5 antibiotics.
Comparatively, a CNS species which was susceptible to more than 8 antibiotics led to a
contamination in 3/4 of the cases. However, this criterion wasn’t statistically significant either.
It was also observed that resistant strains belonged to disease episodes (sepsis or non-sepsis) with a
hospital stay longer than 2 days. The resistance increases with the duration of hospitalization.
Our study confirms that none of the parameters cited above allows to diagnose a CNS sepsis with
100% confidence, and sepsis is still first of all a clinical diagnosis. In order to define true CNS-sepsis,
the clinician should primarily rely on the clinical picture the patient presents. Criteria like „number of
blood cultures positive for CNS“ and „increased CRP-value“ may be of particular help to distinguish
true pathogens from contaminants.
Although it should be continued to investigate about the CNS virulence, it shouldn’t be forgotten that
they are frequent contaminants, and each person drawing bloodcultures should aim at avoiding
contamination.
