Reviewer's report N°1
Specific comments
The title should state that the scope of the study is to compared newer and older
quinolones.
The title was changed to: “Severe Pneumococcal Pneumonia: Impact of New Quinolones on
Prognosis.”
Abstract: the objective of the analysis is to compare newer and older quinolones,
rather than determining risk factors for mortality. For the latter objective the
sample size would largely be inadequate.
We changed “objectives” as wished: To evaluate the outcome of patients with severe
pneumococcal CAP, focusing on the impact of new rather than old fluoroquinolones
combined with β-lactam in the empirical antimicrobial treatment.
Abstract: “three independent factors associated with survival in ICU” rather than
outcome.
Correction performed
Abstract: conclusions, levofloxacine is associated with lower mortality, rather
than being more effective.
Correction performed
In the introduction there is no mention on the ERS guidelines (Woodhead M.,
ERJ 2005) in which the recommended treatment regimen for CAP is a
beta-lactam plus a macrolide OR a quinolone (levofloxacine, moxifloxacine).
Please consider and comment.
This reference was added in the manuscript as wished.
Methods: amoxicillin and cefotaxime doses of 50 mg/kg/d are at the lower edge
of the recommended range. Can the author specify that the mean administered
doses were similar in Group A and Group B ?
French guidelines during the study period recommended these doses. Observed doses in both
groups were similar.
As specified in the methods the survival curves have been drafted and compared
for each patients in each group. Does figure 1 show survival in group A
compared to group B ? If so, is it correct to label the figure as a comparison of
the effect of beta-lactam plus old quinolones vs. beta-lactams plus newer
quinolones ?
We do not understand your comment as the label of Figure 1 is: 15-day survival curves in
patients treated with β-lactam combined with levofloxacin versus β-lactam combined with
ofloxacin or ciprofloxacin
Discussion, first sentence: the main finding of the study is that newer quinolones
are associated with improved survival rather than being more effective on severe
pneumococcal CAP.
We agree with your comment and the sentence was changed as wished.
Reviewer's report N°2
Reviewer's report:
MAJOR COMPULSORY REVISIONS:
1. Material and Methods. Patients. (page 4) Definition of CAP: “We excluded patients coming
from nursing homes or hospitalized within 30 days”: if the authors are going to exclude
Health care-associated CAP, 90 days should be used (Ref. Niederman MS. Hospitalacquired pneumonia, health care-associated pneumonia, ventilator-associated pneumonia,
and ventilator-associated tracheobronchitis: definitions and challenges in trial design. Clin
Infect Dis. 2010 Aug 1;51 Suppl 1:S12-7.).
We agree with your comments. According to our database, no patient was hospitalized within
90 days prior to developing pneumonia. Consequently we changed 30 days for 90 days in the
revised manuscript.
2. Appropriate drug dosage (page 5): please refer to international literature or reference book
your definition of “appropriate”, in particular regarding pro kilo dosage of beta-lactams and
levofloxacin dosage (500 mg/12h is an “evidence based” dosage in clinical practice to treat
severe infections, however current CAP guidelines reported 750 mg OD). Please discuss
your choices.
These points were clarified in the manuscript as followed: Appropriate drug dosages were
defined in the French recommendations [1991, 1999, 2006] as: amoxicillin > 50 mg/kg/d,
cefotaxime > 50mg/kg/d, ceftriaxone > 20 mg/kg/d, piperacillin > 200 mg/kg/d, ofloxacin =
200 mg/12h, ciprofloxacin = 400 mg/12h, levofloxacin = 500 mg/12h. These drug dosages for
β-lactam, ofloxacin and ciprofloxacin were unchanged during the study period. Thus, doses
used in both groups were similar.
3.Table1:
a. Please specify that age was measured in years.
Correction performed
b. Please show also percentages for male sex and PSI 4/5.
Correction performed
c. Please report in table 1 the #-lactams used in the two groups and the p value
for between group comparison. A difference in #-lactams used could influence
the results.
The number and percentage of patients treated by cephalosporins is now reported in Table 2
4. Apparently in group B a higher proportion of patients received a third generation
cephalosporin as #-lactam. Is there a significant between group difference? Theoretically an
in vivo higher activity of cephalosporin in comparison to amoxicillin-clavulanic acid could
contribute to the differences observed in mortality. The authors should comment in the
discussion if this difference could influence the results.
In our study, the use of cephalosporin is more frequent in group B than in group A (p=0.01).
This point was added in Table 2. However, as reported in Table 3, the use of cephalosporin
rather than amoxicillin has no impact on prognosis. Such a result is not surprising as, to the
best of our knowledge, no clinical study demonstrates a third generation cephalosporin is
superior to amoxicillin to treat pneumococcal CAP due to non penicillin-resistant S.
pneumoniae as soon as drug dosage is correct. However, this point was added in the
limitations of our study.
5. Was the antibiotic therapy modified after results of cultures? In how many patients and
which antibiotic were used? This could be reported to better interpret the results.
The goal of our study was to compare the empirical drug regimens. So, we did not focus on
the antibiotics instituted after microbiological documentation. In our unit, we apply a strict deescalation strategy. This was not modified during the study period.
6. The phrase “all underlying diseases...in this analysis” appears unclear.
According to table 3 Chronic heart failure appeared associated to mortality, while
cephalosporin and bacteremia were not: please clarify. Moreover, mechanical
ventilation is not reported in table 3: please report.
The sentence was clarified. Mechanical ventilation was reported as MV on D1 in Table 3.
7. Table 3: Please show also percentages for male sex, PSI 4/5, patients in
group A.
Correction performed
8. Why sepsis-related complications and HA-LRT superinfections were not
included in the multivariate model? Their development could be strictly related to
the empirical antibiotic therapy used. If authors believe that these complications
could be mainly related to factors other than antibiotic therapy (for example
invasive procedures), why did they not excluded patients with these
complications from the analysis?
As our precise goal was in fact to compare old vs. new quinolones (this point was asked by
reviewer N°1), we focused on prognostic factors present on ICU admission or within the first
48 hours of ICU stay. These points are now clearly announced in the chapter methods (Data
collection and statistical analysis). Moreover, we think that complications occurring during
the ICU stay are more an explanation of death than a cause of death. Consequently, it would
limit the interest of including these variables in the prognostic analysis.
9. Discussion “Finally, some important prognostic parameters such as the time
elapsed between admission and the first dose of antibiotic were not taken into
account in our study.”: Why? It is a fundamental information to be inserted in
statistical analysis. If not available, it should be discussed more accurately.
We agree with your comments. This point is now more clearly discussed as follows: Before
2006, we did not have computerized data charts thus, exact time of admission and antibiotics
admission, particularly for patients transferred from other departments/hospitals cannot be
obtained.
MINOR ESSENTIAL REVISIONS:
10. Results: “we identified 70 patients treated by a #-lactam combined with a
fluoroquinolone, including 53 men and 17 women. The mean age was 63.8 ±
16.8 yrs. S. pneumoniae was identified in blood cultures in 25 patients. For 18
patients infection was polymicrobial.” AND “Thirty-eight patients were classified
as Group A. #-lactams combined with ofloxacin (n=33) or ciprofloxacin (n=5)
were amoxicillin ± clavulanic acid (n=16), a third generation cephalosporin
(n=20), and piperacillin-tazobactam (n=2). Thirty-two patients were classified as
Group B. #-lactams combined with levofloxacin were amoxicillin ± clavulanic acid
(n=5), a third generation cephalosporin (n=26), and piperacillin-tazobactam
(n=1)”: please report percentages.
Percentages were added.
11. Global revisions of English language is desirable: e.g. risks factors "of"
mortality, treated "by", but the difference "is" not statistically significant, etc..).
Reviewer's report N°3.
Major Revisions
The study has several methodological problems.
Methodological issue
Aim and design
There is a discrepancy among the aim of the study, the design of the study and the main
findings.
- Aim:”to analyze risk factors for mortality in patients treated by a a combination of B-lactam
plus fluoroquinolone for severe pneumonia.” -->based on this aim, readers would like to see
definitions/aOR of factors associated with mortality in this group of patients.
We do not understand your comments, since all variables were defined in the text and we
provided AOR in the chapter results.
- Main findings: “Levofloxacin combined with b-lactam is more effective than ofloxacin or
ciprofloxacin combined with a b-lactam in severe pneumococcal CAP.” -->This speculation
cannot be derived from the design of the study applied and the data presented.
We modified the sentence and now, we concluded, as wished by reviewer N°1 “ that, when
combined a β-lactam, levofloxacin is associated with lower mortality than ofloxacin or
ciprofloxacin in severe pneumococcal community-acquired pneumonia.”
- Design: Dr Olive and coworkers designed a retrospective observational study during 13
years. During this long period different techniques have been developed that could have
impacted the outcome analyzed as correctly mentioned by the authors. Based on the
particular objective presented in the introduction, authors could think about different designs,
such as a nested case-control study or a therapeutic observational retrospective study.
For us, the reported study is “a therapeutic observational retrospective study.
- Study population: It would be interesting to have the number of patients with CAP admitted
to the ICU during the long study period in order to have an idea regarding the inception
cohort. It should be noticed that the study period was composed by 13 years with 83 severe
pneumococcal pneumonia (6-7 patients/year) enrolled. This should be explained by the
authors who should also provide information regarding the study setting (which king of ICU?
Where? Hospital beds? and so on).
These points were added in the revised manuscript (number of beds in ICU, kind of ICU, and
number of CAP treated in our ICU during the study period).
Definitions / criteria:
- Authors should better clarify the definition of severe CAP: patients admitted to ICU? PSI
risk class 4?? My suggestion would be to include all the patients admitted to ICU or those
with a PSI risk class of 5 according to the previous literature.
We chose to include patients both admitted in ICU and exhibiting a PSI risk class 4 or 5. We
excluded patients in class 2 or 3, even if they were admitted in ICU. According to IDSA/ATS
Guidelines published in 2007, “objective criteria to identify patients for ICU admission
included…PSI severity class IV or V” Moreover, our choice appears logical since among the
21 patients in class 4, 12 were mechanically ventilated, 5 had non invasive ventilation, and 4
exhibited septic shock.
- Is not completely clear to me the reason why the authors have decided to exclude penicillinresistant strain of SP. Please, clarify.
In group A, the only effective antibiotic against S.pneumoniae was the beta-lactam, we
limited the analysis to penicillin susceptible strains as it allowed us to evaluate the added
benefit of a fluoroquinolone.
- Authors should give references for the appropriate drug dosages given in the manuscript.
References were added.
- It would be also necessary to give the exact definition for septic shock that is included in the
multivariable analysis in the results section but not mentioned in the Materials & Methods
section.
Shock was defined on page 5 with a precise reference. Shock was considered as a septic
shock, because patients had concomitant bacterial pneumonia.
- The definition for time to clinical stability is well standardized in literature (see papers
published by Halm, Menendez or Julio Ramirez). Authors could use criteria for clinical
stability already validated in literature.
Criteria proposed in the literature to define clinical stability are the followings “A variable
was considered stable if all measurements in the 24-hour period met stability criteria—a
convention used by other investigators. Stable values for vital signs were selected prior to
analysis based on the literature and common clinical practice. The stability cut point for HR
was at least 100 beats/min; SBP, more than 90 mm Hg; and RR, 24 breaths/min or less.
Oxygenation was considered stable if the oxygen saturation was 90% or greater or the PaO 2
was 60 mm Hg or greater and a patient was not receiving mechanical ventilation or
supplemental oxygen by face mask.” [Halm et al JAMA 1998]. Of course, such a definition
is not applicable for ICU patients receiving mechanical ventilation. We used in this paper
definitions used in a previous paper published by our team (Chest 1999; 116: 157)
Outcome
I strongly suggest to focuse the analysis just on one outcome. The outcome mentioned on
pag 9 that was associated with septic schock, age, and initial treatment is not clear to me.
What kind of mortality was used in this case? Mortality on day 15, day 28 or ICU discharge?
Furthermore, what is the rationale in analyzing only mortality on day 15, day 28 and at the
time of ICU discharge? Why did not the authors chose only the mortality at ICU discharge as
the sole outcome?
We agree with you that our results could be confusing. We added in the text and in Tables 3
and 4 that outcome was evaluated at ICU discharge. We chose to also report mortality data
evaluated on D15, because in severe infectious diseases the antimicrobial treatment influences
only the initial course of disease (i.e., in VAP, adequate antimicrobial treatment improves the
cure rate but has no significant impact on ICU mortality). In addition, we suppressed data
about mortality on D28.
Speculation
As data are presented, authors cannot declare that “when severe CAP causative agent in
SP, a combination levofloxacin plus a beta lactam is more effective than a combination
ofloxacin or ciprofloxacin plus beta lactam”. This sentence in the Discussion section cannot
be accepted.
We agree with your remark. As wished by the reviewer N°1, this sentence was replaced by
the following: “levofloxacin combined with a β-lactam is associated with improved survival
in comparison with ofloxacin or ciprofloxacin combined with a β-lactam”