Reply to Tubiana et al
0 Clinical Infectious Diseases
Reply to Tubiana et al
TO THE EDITOR—We thank Dr Tubiana,
Dr Le Moing, and Dr Duval for their
interest in our study . They raise 3 issues
about the analysis and reporting of our
results that we would like to clarify.
The first point deals with the
proportion of patients with enterococcal
bacteremia that underwent a trans-thoracic
echocardiogram. From the 1515 patients
with enterococcal bacteremia, 388 (25.6%)
had a transthoracic echocardiogram and
183 (12.1%) a transesophageal
echocardiogram (TEE). All patients with enterococcal
endocarditis had a TEE (65 episodes). We
deeply regret that there was a typo in the
text (TEE instead of echocardiogram)
that led to this confusion, and we thank
Dr Tubiana et al for pointing it out.
Fortunately, the mistake does not affect the
validity of the message. The proportion
of patients with echocardiogram remains
low despite our efforts to implement its
systematic use. This proportion reflects
real-life practice and the fact that it is
not recommended in current guidelines
. We have not been able to find similar
information from the literature [3–5].
Second, it is true that some of the
independent predictors of infective
endocarditis (IE) which compose the NOVA score
(heart murmur, previous valve disease,
persistent bacteremia) are endocarditis
criteria. Maybe our description regarding
the selection criteria chosen for the control
patients was not clear enough. Controls
were randomly selected among patients
with enterococcal bloodstream infection
(E-BSI) and a negative TEE result that
finally were classified as not having IE.
Some Duke-Li criteria were present in
both populations (cases and controls) as
stated in the original manuscript 
page 531, table 2. We did not exclude a
control because it had heart murmur,
valve disease, or continuous bacteremia.
The potential overlap that might exist
between predictor variables and the
primary end point was overcome with the
implementation of boot-strapping
techniques that avoid overfitting. Final
variables included in the model were selected
using a backward stepwise approach and
this logistic regression model was
validated by 2 runs of 2000 bootstrap
replications. Therefore, we do not consider it
necessary to perform any further
statistical analysis. In order to further clarify this
issue, we include a table (Table 1) with
the description of the Duke criteria
present in cases and controls. If we excluded
from the 65 IE cases all those patients
who had 1 or more NOVA score factors,
none would remain as stated in the
original manuscript  page 533, figure 3.
Regarding the definition of “unknown
origin of the bacteremia,” it was not based
at all in the results of the
echocardiography. As mentioned above, all patients
with bacteremia were prospectively
evaluated by an independent investigator that
searched with the attending physician
for clinical microbiological or
radiological evidence that explained the origin
of the bacteremia. When no such source
was demonstrated the case was classified
as “unknown origin.” It is well known
Glomerulonephritis Yes 0 (0%) No 65 (100%) Osler nodes
Abbreviation: IE, infective endocarditis.
that many bacteremias with a clear origin,
such as catheter-related BSIs, are very
commonly the origin of nosocomial
endocarditis. So the classification of the
origin was not established by considering
the echocardiography results.
Finally, we concur with Dr Tubiana et al
in the need to validate the NOVA score in
larger and different populations. However,
so far, our risk prediction score (NOVA)
provides an easy to use system that could
rapidly determine which patients with
E-BSI may not require further studies to
detect IE. We can anticipate that the
NOVA score has already been validated
in an Italian cohort of patients with
enterococal bacteremia (ECCMID 2015; poster
number 2476) that will soon be published
and that a prospective trial is on its way.
We would like to take this opportunity
to invite Dr Tubiana et al and any other
group interested in endocarditis to
participate in a prospective, multicentre
validation of the NOVA score.
Potential conflicts of interest. All authors:
No potential conflicts of interest.
All authors have submitted the ICMJE Form
for Disclosure of Potential Conflicts of Interest.
Conflicts that the editors consider relevant to the
content of the manuscript have been disclosed.
Patricia Muñoz,1 Martha Kestler,1
Javier Bermejo,2 and Emilio Bouza1
1Clinical Microbiology and Infectious Diseases, and
2Department of Cardiology, Hospital General
Universitario Gregorio Marañón,
1. Bouza E , Kestler M , Beca T , et al; Grupo de Apoyo al Manejo de la Endocarditis . The NOVA score: a proposal to reduce the need for transesophageal echocardiography in patients with enterococcal bacteremia . Clin Infect Dis 2015 ; 60 : 528 - 35 .
2. Nishimura RA , Otto CM , Bonow RO , et al; ACC/AHA Task Force Members . 2014 AHA/ ACC guideline for the management of patients with valvular heart disease: executive summary a report of the American college of cardiology/American heart association task force on practice guidelines . Circulation 2014 ; 129 : 2440 - 92 .
3. Fernandez-Guerrero ML , Herrero L , Bellver M , Gadea I , Roblas RF , de Gorgolas M. Nosocomial enterococcal endocarditis: a serious hazard for hospitalized patients with enterococcal bacteraemia . J Intern Med 2002 ; 252 : 510 - 5 .
4. Anderson DJ , Murdoch DR , Sexton DJ , et al. Risk factors for infective endocarditis in patients with enterococcal bacteremia: a casecontrol study . Infection 2004 ; 32 : 72 - 7 .
5. Pinholt M , Østergaard C , Arpi M , et al; for the Danish Collaborative Bacteraemia Network (DACOBAN). Incidence, clinical characteristics and 30-day mortality of enterococcal bacteraemia in Denmark 2006-2009: a populationbased cohort study . CMI 2013 ; 20 : 145 - 51 .
Correspondence: Patricia Muñoz , MD, PhD, Department of Clinical Microbiology and Infectious Diseases, Hospital General Universitario Gregorio Marañón, Dr. Esquerdo , 46 , 28007 Madrid, Spain (pmunoz@micro .hggm.es).