CHEST Original Research

CHEST INFECTIONS

Thrombocytosis Is a Marker of Poor

Outcome in Community-Acquired
Pneumonia
Elena Prina, MD; Miquel Ferrer, MD, PhD; Otavio T. Ranzani, MD;
Eva Polverino, MD, PhD; Catia Cillóniz, PhD; Encarnación Moreno, RN;
Josep Mensa, MD; Beatriz Montull, MD; Rosario Menéndez, MD, PhD;
Roberto Cosentini, MD; and Antoni Torres, MD, PhD

Background: Thrombocytosis, often considered a marker of normal inflammatory reaction of infec-

tions, has been recently associated with increased mortality in hospitalized patients with community-
acquired pneumonia (CAP). We assessed the characteristics and outcomes of patients with CAP
and thrombocytosis (platelet count  4 3 105/mm3) compared with thrombocytopenia (platelet
count , 105/mm3) and normal platelet count.
Methods: We prospectively analyzed 2,423 consecutive, hospitalized patients with CAP. We
excluded patients with immunosuppression, neoplasm, active TB, or hematologic disease.
Results: Fifty-three patients (2%) presented with thrombocytopenia, 204 (8%) with thrombocytosis,
and 2,166 (90%) had normal platelet counts. Patients with thrombocytosis were younger (P , .001);
those with thrombocytopenia more frequently had chronic heart and liver disease (P , .001 for
both). Patients with thrombocytosis presented more frequently with respiratory complications,
such as complicated pleural effusion and empyema (P , .001), whereas those with thrombocy-
topenia presented more often with severe sepsis (P , .001), septic shock (P 5 .009), need for inva-
sive mechanical ventilation (P , .001), and ICU admission (P 5 .011). Patients with thrombocytosis
and patients with thrombocytopenia had longer hospital stays (P 5 .004), and higher 30-day mor-
tality (P 5 .001) and readmission rates (P 5 .011) than those with normal platelet counts. Multivar-
iate analysis confirmed a significant association between thrombocytosis and 30-day mortality
(OR, 2.720; 95% CI, 1.589-4.657; P , .001). Adding thrombocytosis to the confusion, respiratory
rate, and BP plus age 65 years score slightly improved the accuracy to predict mortality (area
under the receiver operating characteristic curve increased from 0.634 to 0.654, P 5 .049).
Conclusions: Thrombocytosis in patients with CAP is associated with poor outcome, complicated
pleural effusion, and empyema. The presence of thrombocytosis in CAP should encourage ruling
out respiratory complication and could be considered for severity evaluation.
CHEST 2013; 143(3):767–775

Abbreviations: ATS 5 American Thoracic Society; AUC 5 area under the curve; CAP 5 community-acquired pneumonia;
CRB-65 5 confusion, respiratory rate, and BP plus age  65 years; IDSA 5 Infectious Disease Society of America;
TNF 5 tumor necrosis factor

Inplatelets
addition to being part of the hemostatic process,
have been increasingly recognized as an
important component of the immune response to
infection.1-3 Thrombocytopenia is a recognized sever-
ity criterion and a predictor of mortality in hospital-
ized patients with community-acquired pneumonia
(CAP)4,5 and is included in the minor severity criteria
defined by the current Infectious Disease Society of
America (IDSA)/American Thoracic Society (ATS)
guidelines to predict ICU admission.6 Conversely,
thrombocytosis has often been considered a sign of
normal inflammatory reaction, but not as a marker of
poor outcome.
Mirsaeidi et al7 showed recently in a retrospective
single-center study an association between thrombo-
cytopenia and thrombocytosis and increased mortality
in patients hospitalized with CAP. Their article, how-
ever, did not evaluate the reason for the higher mortality

journal.publications.chestnet.org CHEST / 143 / 3 / MARCH 2013 767

Downloaded From: http://journal.publications.chestnet.org/ by a University of Pittsburgh User on 05/17/2013

 rate of patients with thrombocytosis and did not assess
or compare the clinical and etiologic characteristics of
the three groups of patients with thrombocytopenia,
thrombocytosis, and normal platelet count. Further-
more, this study generated doubts on this issue8,9 that
required further research.
We hypothesized that the clinical profile, inflamma-
tory biomarkers, and mortality in patients with CAP
may be different in those presenting with thrombocy-
tosis and thrombocytopenia than in patients with a
normal platelet count. Therefore, the objective of our
study was to define these three groups of patients to
confirm the association between abnormal platelet
count and patients’ outcome.
Materials and Methods
Study Population
We prospectively assessed consecutive patients aged  14 years
hospitalized with a diagnosis of CAP at Hospital Clinic, Barcelona,
Spain, and Hospital Universitario La Fe, Valencia, Spain, from
January 2000 to October 2006. Pneumonia was defined as a new
pulmonary infiltrate on chest radiograph at admission and symptoms
and signs of lower respiratory tract infection. We excluded patients
with previous use of oral corticosteroids ( 10 mg prednisone-
equivalent per day for at least 2 weeks); other immunosuppressive
therapy; active solid or hematologic neoplasms; HIV infection;
active TB; hematologic disease involving platelets and/or leuko-
cytes, such as essential thrombocytosis or myelodysplastic syn-
drome; and hospitalization within the preceding 21 days.
This study was approved by the ethics committees of both cen-
ters (Register: 2009/5251). Patients’ identification remained anon-
ymous and informed consent was waived due to the observational
nature of the study.
Manuscript received May 15, 2012; revision accepted July 25, 2012.
Affiliations: From the Servei de Pneumologia (Drs Prina, Ferrer,
Ranzani, Polverino, Cillóniz, and Torres, and Ms Moreno), Institut
del Torax, Hospital Clinic, Institut d’Investigacions Biomèdiques
August Pi i Sunyer (IDIBAPS), Universitat de Barcelona, Spain;
Emergency Medicine Department (Drs Prina and Cosentini),
Istituto Di Ricovero e Cura a Carattere Scientifico Fondazione
Ca’ Granda, Ospedale Maggiore Policlinico, Milan, Italy; Centro
de Investigación Biomedica En Red-Enfermedades Respirato-
rias (CibeRes, CB06/06/0028) (Drs Ferrer, Polverino, Cillóniz,
Menéndez, and Torres, and Ms Moreno), Barcelona, Spain; Respi-
ratory Intensive Care Unit (Dr Ranzani), Hospital das Clínicas,
Faculdade de Medicina da Universidade de São Paulo, Brazil;
Servicio de Enfermedades Infecciosas (Dr Mensa), Hospital
Clínic, IDIBAPS, Barcelona, Spain; and Servicio de Neumologia
(Drs Montull and Menéndez), Hospital Universitario La Fe,
Valencia, Spain.
Funding/Support: This work was supported by the Centro de
Investigación Biomédica en Red-Enfermedades Respiratorias
(CibeRes CB06/06/0028)-Instituto de Salud Carlos III [Grant
2009 SGR 911], PII de infecciones respiratorias of SEPAR, and
IDIBAPS.
Correspondence to: Miquel Ferrer, MD, PhD, Unitat de Vigilancia
Intensiva Respiratoria, Servei de Pneumologia, Hospital Clínic,
Villarroel 170, 08036 Barcelona, Spain; e-mail: miferrer@clinic.ub.es
© 2013 American College of Chest Physicians. Reproduction
of this article is prohibited without written permission from the
American College of Chest Physicians. See online for more details.
DOI: 10.1378/chest.12-1235
768
Downloaded From: http://journal.publications.chestnet.org/ by a University of Pittsburgh User on 05/17/2013
Data Collection and Microbiologic Evaluation
The data collected, the severity scores, and the criteria for the
microbiologic diagnosis have been described elsewhere.10 The
laboratory findings reported were obtained within 6 h after admis-
sion to the ED.
Definitions
Thrombocytopenia and thrombocytosis were defined as plate-
let counts , 105/mm3 and  4 3 105/mm3, respectively. We defined
sepsis, severe sepsis, and septic shock according to the sepsis
guidelines.11 We calculated the Pneumonia Severity Index12 and
the confusion, respiratory rate, and BP plus age  65 years (CRB-65)
score13 at admission. The definition of severe CAP followed the
IDSA/ATS guidelines.6 We defined complications as organ dis-
eases besides the pneumonia presented at admission or appearing
during hospitalization. We defined complicated pleural effusion
and empyema according to the literature definition.14 We defined
pneumonia as the cause of death when patients died of acute
respiratory failure and/or septic shock due to the respiratory
infection. We assessed the adherence of the empirical antibiotic
treatment to the guidelines15,16 and its appropriateness to the iso-
lated pathogens.17
Determination of Cytokines, C-Reactive
Protein, and Procalcitonin
We determined serum levels of cytokines and procalcitonin in
a subgroup of 550 patients. The laboratory techniques used were
described elsewhere.18
Statistical Analysis
Categorical variables were described as frequencies and per-
centages and compared with the x2 test or Fisher exact test when
appropriate. Continuous variables were expressed as mean  SD
and compared between groups using one-way analysis of var-
iance; for data not normally distributed, these variables were
expressed as median (interquartile range) and compared using
the nonparametric Kruskal-Wallis test. All post hoc compari-
sons were made with the Bonferroni correction. Correlations
between two continuous variables were made with the Pearson
correlation.
Two different models of logistic regression were developed as
follows: First, the association between platelet count and 30-day
mortality was assessed as categorical variables, with the predefined
values, normal range being the reference category. Second, we
performed the same analysis with platelet count and leukocyte
cells as continuous variables. Continuous variables were checked
for the assumption of linearity in the logit. Single collinearity
was evaluated with the Pearson correlation among the indepen-
dent variables and multicollinearity was evaluated with the var-
iance inflation factor. The ORs and corresponding 95% CIs for
each variable were computed. The discriminative ability of the
model to predict the outcome of patients was assessed by the
area under the receiver operating characteristic curve (AUC).
Estimated AUC values were compared using the nonparamet-
ric method described by Hanley and NcNeil.19 The calibration
ability for the model was evaluated with the Hosmer-Lemeshow
goodness-of-fit statistics. To explore the role of the platelet count
in patient outcome, we retrieved the predicted 30-day mortality
for each patient from the final model and plotted it against the
respective absolute platelet count. Data were processed with
the SPSS version 18.0 (IBM). The level of significance was set at
0.05 (two-tailed).
Original Research
 Results
Patient Characteristics
During the study period, we evaluated 3,010 patients
with CAP (Fig 1), and it was possible to analyze 2,423 sub-
jects who met inclusion criteria and in whom the plate-
let count at admission was available: 53 patients (2%)
with thrombocytopenia, 204 patients (8%) with throm-
bocytosis, and 2,166 (90%) with a normal platelet count.
The general characteristics of the patients are shown
in Tables 1 and 2.
Patients with thrombocytosis were younger, had
more frequently received previous antibiotics, and had
a lower CRB-65 score at admission. Patients with throm-
bocytopenia more frequently had chronic heart and
liver disease, and heart rate, leukocyte counts, fibrin-
ogen level, and prothrombin time were all lower at
admission. Moreover, there was a weak but signifi-
cant positive correlation between platelet count and
leukocyte count (r 5 0.21, P , .001) and fibrinogen
level (r 5 0.18, P , .001).
In the overall population, 2,072 patients (86%)
received antibiotic treatment adherent to ATS guide-
lines. Among 1,060 patients with a known etiology,
947 (89%) were treated with an appropriate antimi-
crobial therapy. Adherence to the guidelines and appro-
priateness of the empirical antibiotic treatment were
not significantly different among the three groups.
Microbial Etiology
An etiologic diagnosis was established in 1,060 patients
(44%). Streptococcus pneumoniae was the most fre-
quent pathogen in all groups. Atypical agents were less
frequently identified in thrombocytopenic patients,
and there was no significant difference in the remain-
ing pathogens (Table 3).
Figure 1. Flow diagram of the selected population. CAP 5
community-acquired pneumonia.
journal.publications.chestnet.org
Downloaded From: http://journal.publications.chestnet.org/ by a University of Pittsburgh User on 05/17/2013
Complications, Length of Stay,
and Readmission Rate
Patients with thrombocytosis more frequently had
respiratory complications due to higher rate of empy-
ema and complicated pleural effusion. By contrast,
patients with thrombocytopenia presented more often
with severe CAP, severe sepsis, and septic shock, and
need for invasive mechanical ventilation and ICU
admission, with a trend toward a higher rate of car-
diovascular complications.
The length of hospital stay was significantly lon-
ger in patients with thrombocytosis and those with
thrombocytopenia compared with those with normal
platelet count; no statistical difference in length of
stay was found between the two former groups .
Among the survivors of the first hospitalization, the
30-day hospital readmission rate was higher in patients
with thrombocytopenia and those with thrombocy-
tosis, with no significant difference between them
(Table 4).
Thrombocytosis and Severity Scores
Adding thrombocytosis to the CRB-65 score slightly
improved the accuracy of this score to predict mortality
(the AUC increased from 0.634 to 0.654, P 5 .049).
Adding thrombocytosis to the pneumonia severity
index score and the IDSA/ATS minor criteria of severe
CAP did not improve the prediction of mortality and
ICU admission, respectively.
Mortality and Causes of Death
Overall, 127 patients (5%) died within 30 days of
hospital admission. The mortality rate was higher in
patients with thrombocytopenia and patients with
thrombocytosis, with no significant difference between
them (Table 4). Patients with pulmonary complications
had a higher mortality rate with respect to patients
without pulmonary complications (65.4% vs 62.7%,
P 5 .003). The multivariate analysis, using platelet
count as a categorical variable, confirmed the associ-
ation between higher platelet count and 30-day mor-
tality (Table 5). By contrast, the association of a low
platelet count with 30-day mortality did not reach
statistical significance.
To evaluate the role of continuous platelet count,
the multivariate analysis showed that the increased
risk of death was significantly associated with increas-
ing platelet count (Fig 2). The same analysis was run
for leukocyte count, without significant association
within mortality.
In all groups, the main cause of death was pneu-
monia, but through different mechanisms: Patients with
thrombocytosis died more frequently of acute respi-
ratory failure, whereas patients with thrombocytopenia
CHEST / 143 / 3 / MARCH 2013 769
 Table 1—General Characteristics of the Study Population at Admission

Characteristics Thrombocytopenia (n 5 53) Normal Platelet Count (n 5 2,166) Thrombocytosis (n 5 204) P Value
Demographic data
Age, mean ( SD), y 68 (18) 66 (18) 61 (20) , .001a
Male sex 38 (72) 1361 (63) 115 (56) .071
Previous antibiotics 11 (21) 550 (25) 67 (33) .042b
Active smoking 12 (23) 552 (25) 62 (30) .271
Active alcohol abuse 1 (2) 118 (5) 10 (5) .500
Statin therapy 3 (6) 155 (7) 8 (4) .204
Antiplatelet therapy 6 (11) 239 (11) 16 (8) .369
Comorbidities
Chronic respiratory disease 20 (38) 951 (44) 94 (46) .547
Chronic heart disease 19 (36) 406 (19) 24 (12) , .001c
Diabetes mellitus 10 (19) 414 (19) 35 (17) .789
Chronic liver disease 4 (7) 45 (2) 8 (4) .010
Chronic renal disease 5 (9) 169 (8) 11 (5) .410
Neurologic disease 8 (15) 331 (15) 27 (13) .735
Data given as No. (%) unless otherwise indicated.
aDifferences between thrombocytopenia and thrombocytosis.
bDifferences between thrombocytosis and the other groups.

cDifferences between thrombocytopenia and the other groups.

died more frequently of septic shock. The thrombo- Inflammatory Biomarkers

cytosis group did not show higher frequency of throm- IL-1 at admission was lower in patients with nor-
botic events as cause of death compared with the other mal platelet count (P 5 .048). We also observed a non-
groups (Table 6). significant trend for higher procalcitonin and tumor

Table 2—Characteristics of Pneumonia at Admission

Characteristics Thrombocytopenia (n 5 53) Normal Platelet Count (n 5 2,166) Thrombocytosis (n 5 204) P Value
Vital signs
Respiratory rate, breaths/min 25 (7) 26 (8) 26 (7) .843
Heart rate, beats/min 90 (17) 97 (19) 95 (20) .017a
Systolic BP, mm Hg 128 (27) 133 (34) 130 (24) .284
Diastolic BP, mm Hg 72 (16) 73 (14) 73 (12) .857
MAP , 70 mm Hg, No. (%) 7 (13) 122 (6) 10 (5) .056
Temperature . 38°C, No. (%) 24 (45) 732 (34) 52 (25) .010
Laboratory findings
WBC count/mm3 9,939 (6,556) 13,774 (6,386) 16,664 (7,650) , .001b
WBC count , 4,000/mm3, No. (%) 7 (13) 35 (2) 0 , .001b
WBC count . 12,000/mm3, No. (%) 17 (32) 1169 (54) 143 (70) ,.001c
C-reactive protein, mg/dL 15 (11) 16 (12) 17 (11) .470
Hematocrit, % 39 (5) 40 (5) 40 (5) .299
Prothrombin time, % 55.6 (29.5) 66.9 (30.9) 72.4 (25.3) .029c
APTT, s 37.2 (13.6) 32.1 (13.2) 32 (7.3) .120
Fibrinogen, g/L 5.4 (1.7) 7.1 (2.4) 7.8 (2.3) .003c
Albumin, mg/dL 3.2 (0.4) 3.4 (0.5) 3.3 (0.5) .237
Bilirubin, mg/dL 0.93 (0.84) 0.61 (0.49) 0.48 (0.29) , .001b
Pao2/Fio2 284 (77) 290 (72) 290 (79) .849
Paco2, mm Hg 34 (7) 36 (9) 37 (11) .239
Arterial pH 7.45 (0.08) 7.44 (0.06) 7.44 (0.07) .411
Pneumonia severity index
Risk class I-III, No. (%) 28 (53) 1053 (49) 112 (55) .204
Risk class IV-V, No. (%) 25 (47) 1111 (51) 92 (45) .204
CRB-65 score 1.96 (0.81) 1.99 (0.84) 1.82 (0.81) .018d
Data given as mean ( SD) unless otherwise indicated. APTT 5 activated partial thromboplastin time; CRB-65 5 confusion, respiratory rate, and
BP plus age  65 y score; MAP 5 mean arterial pressure.
aDifferences between thrombocytopenia and normal platelet count.

bDifferences between thrombocytopenia and the other groups.

cDifferences between thrombocytopenia and thrombocytosis.

dDifferences between thrombocytosis and normal platelet count.

770 Original Research

Downloaded From: http://journal.publications.chestnet.org/ by a University of Pittsburgh User on 05/17/2013

 Table 3—Etiologic Diagnosisa

Diagnosis Thrombocytopenia Normal Platelet Count Thrombocytosis P Value

Etiologic diagnosis 21 (40) 946 (44) 93 (46) .722
Streptococcus pneumoniae 10 (48) 483 (51) 37 (40) .113
Pseudomonas aeruginosa 0 45 (5) 6 (6) .446
Atypical 0 119 (13) 17 (18) .060
Chlamydia pneumoniae 0 38 (4) 7 (7) .172
Coxiella burnetii 0 29 (3) 2 (2) .639
Mycoplasma pneumoniae 0 54 (6) 9 (10) .154
Legionella pneumophila 2 (9) 91 (10) 10 (11) .939
Staphylococcus aureus 1 (5) 38 (4) 4 (4) .978
Haemophilus influenzae 2 (9) 71 (7) 3 (3) .285
Enterobacteriaceae 1 (5) 29 (3) 2 (2) .793
Fungi 0 12 (1) 0 .481
Respiratory viruses 3 (14) 136 (14) 17 (18) .597
Other 5 (24) 56 (6) 4 (4) …
Data given as No. (%).
aThe percentages of pathogens are related to the number of patients with etiologic diagnosis in each group.

necrosis factor (TNF)-a levels in patients with throm-
bocytopenia (Table 7).
Discussion
The main findings of our study were as follows:
(1) Thrombocytosis was observed in 8% and throm-
bocytopenia in 2% cases of a large series of consecu-
tively hospitalized patients with CAP; (2) thrombocytosis
but not thrombocytopenia is an independent marker
of poor outcome; and (3) thrombocytosis was more
frequently associated with empyema and complicated
pleural effusion, whereas patients with thrombocy-
topenia presented more often severe sepsis, septic
shock, and the need for mechanical ventilation.

Table 4—Complications and Outcome Variables

Variable Thrombocytopenia (n 5 53) Normal Platelet Count (n 5 2,166) Thrombocytosis (n 5 204) P Value

Length of hospital stay, mean ( SD), d 10 (9) 8 (8) 10 (13) .004

ICU admission 10 (19) 166 (8) 18 (9) .011a
30-d mortality 6 (11) 101 (5) 20 (10) .001c
30-d readmissionb 13 (27) 370 (18) 47 (25) .011c
Invasive mechanical ventilation 8 (15) 82 (4) 11 (5) , .001a
Noninvasive ventilation 1 (2) 64 (3) 7 (3) .832
Severe CAP 24 (45) 358 (16) 33 (16) , .001a
Complication
Respiratory complicationd 21 (40) 764 (35) 95 (47) .005
Pleural effusion 10 (19) 384 (18) 58 (28) .001
Complicated pleural effusion/empyema 0 38 (2) 18 (9) , .001
Uncomplicated pleural effusion 10 (19) 346 (16) 40 (20) .358
Multilobar pneumonia 15 (28) 463 (21) 47 (23) .425
Cavitation/abscess 0 9 (0.4) 3 (1.5) .106
Cardiovascular complication 9 (17) 191 (9) 14 (7) .069
Acute coronary syndrome 0 22 (1) 2 (1) .762
Cerebral stroke 1 (2) 8 (0.4) 1 (0.5) .231
Cardiac failure 4 (7) 77 (4) 6 (3) .265
Arrhythmia 4 (7) 84 (4) 5 (2) .218
Renal complication 14 (26) 370 (17) 31 (15) .153
Hepatic/gastrointestinal complication 3 (6) 119 (5) 15 (7) .547
Neurologic complication 11 (21) 377 (17) 29 (14) .405
Hyperglycemia 2 (4) 116 (5) 4 (2) .097
Severe sepsis 53 (100) 718 (33) 67 (33) , .001a
Septic shock 10 (19) 165 (8) 14 (7) .009a
Data given as No. (%) unless otherwise indicated. CAP 5 community-acquired pneumonia.
aDifferences between thrombocytopenia and the other groups.

bReadmission rates are calculated from the survivors from the initial episode that required hospital admission.

cDifferences between normal platelet count and the other two groups.

dOne hundred eight patients (4.5%) presented with more than one respiratory complication as follows: four (7.5%) in the thrombocytopenia group,

91 (4%) in the group. with normal platelet counts, and 13 (6%) in the thrombocytosis group (P 5 .194).

journal.publications.chestnet.org CHEST / 143 / 3 / MARCH 2013 771

Downloaded From: http://journal.publications.chestnet.org/ by a University of Pittsburgh User on 05/17/2013

 Table 5—Multivariate Logistic Regression Model In the literature, thrombocytopenia in severe CAP
for 30-d Mortality has already been studied and its role as a marker of
Variable b OR 95% CI P Value severity is well established.4,5 On the other hand, few
studies have investigated thrombocytosis in CAP.
Age 0.045 1.046 1.031-1.061 , .001
Chronic liver disease 0.896 2.449 0.966-6.209 .059
Recently, however, it has been proposed that throm-
Altered consciousness 0.673 1.961 1.282-2.999 .002 bocytosis could be a marker of poor outcome and spe-
at admission cific complications.
Pao2/Fio2 , 200 mm Hg 1.102 3.010 1.810-5.006 , .001 Mirsaeidi et al7 analyzed 500 patients with CAP, of
Septic shock at admission 1.287 3.620 2.244-5.842 , .001 whom 65 (13%) presented thrombocytosis and 27 (5%)
Platelets .001
presented thrombocytopenia. These authors reported
Normal range Reference 1 … …
Thrombocytosis 1.001 2.720 1.589-4.657 , .001 that both patients with thrombocytopenia and those
Thrombocytopenia 0.651 1.917 0.743-4.947 .179 with thrombocytosis had significantly higher mortality
and that platelet count was a better predictor of out-
come than an abnormal leukocyte count. However,
We have shown that according to platelet count, it the cause of mortality, complications, etiology, and
is possible to identify populations with different char- systemic inflammatory response were not analyzed
acteristics. Patients with thrombocytopenia were older, and the population studied included elderly patients
often had cardiac and liver diseases, and presented with cancer.
more frequently with higher acuity and severity of In pediatric patients with CAP, two studies showed
disease (ie, higher rate of severe CAP, need for ICU that thrombocytosis was associated with a poor out-
admission, need for mechanical ventilation, and sep- come and more severe and protracted disease, with
tic shock). Patients with thrombocytosis were younger, a higher probability of developing empyema.20,21 In
had a lower CRB-65 score at admission, and presented another study in adults with CAP, thrombocytosis
more often with respiratory complications (especially (platelet count . 400,000/mm3) was an independent
related to a local inflammatory reaction such as empy- risk factor for the development of complicated para-
ema and complicated pleural effusion). pneumonic effusion or empyema.22
Compared with patients with normal platelet counts, In our study, multivariate logistic regression con-
both the thrombocytopenia and thrombocytosis groups firmed the association between thrombocytosis and
presented with a worse prognosis but apparently for 30-day mortality. The adjusted risk of death was 2.7-fold
different reasons: Infection in patients with thrombo- greater in patients presenting with an elevated plate-
cytopenia patients tended to spread with the devel- let count at hospital admission relative to patients with
opment of systemic complications (ie, septic shock), a normal platelet count. Surprisingly, in contrast to
whereas in patients with thrombocytosis, infection other studies,4,23-25 we could not confirm the inde-
tended to compartmentalize with the development of pendent association between thrombocytopenia and
local respiratory complications with a subacute course mortality. Several reasons may explain this result: the
(ie, empyema/complicated pleural effusion). presence of a low number of patients with thrombo-
cytopenia, the higher weight of septic shock and hyp-
oxemia in determining mortality in these patients,
and the different characteristics of our population,
with the inclusion of patients with less severe CAP
(patients not in the ICU), compared with previous
studies. However, our results do not discard the idea
that thrombocytopenia remains a marker of serious
underlying diseases and complications that are ulti-
mately associated with death. In accordance with the
results of Mirsaeidi et al,7 we did not find an associa-
tion in our population between leukocyte count and
mortality, so an increase of platelet count appeared
to better identify patients with higher mortality risk.
A possible explanation of why patients with throm-
bocytosis had a poor outcome would be the higher
rate of empyema and complicated pleural effusion.
In the literature, these types of complications have
been associated with a longer length of hospital stay,
Figure 2. Multivariate logistic regression model to evaluate the
association between platelet count as a continuous variable and treatment failure, and higher mortality.26-30 Moreover,
mortality. as explained in the sepsis guidelines,11 besides the

772 Original Research

Downloaded From: http://journal.publications.chestnet.org/ by a University of Pittsburgh User on 05/17/2013

 Table 6—Cause of Death

Cause of Death Thrombocytopenia (n 5 6) Normal Platelet Count (n 5 101) Thrombocytosis (n 5 20) P Value
Pneumonia 5 (83) 58 (57) 16 (80) .090
Acute respiratory failure 2 (33) 41 (41) 15 (75) .015a
Septic shock 3 (50) 17 (17) 1 (5) .033b
Thrombotic event 0 10 (10) 1 (5) .576
Acute coronary syndrome 0 4 0 …
Cerebral ischemic stroke 0 5 0 …
Intestinal ischemia 0 0 1 …
Pulmonary embolism 0 1 0 …
Other causes 1 (17) 10 (15) 1 (10) .653
Unknown 0 23 (23) 2 (10) …
Data given as No. (%).
aDifferences between thrombocytosis and the other groups.
bDifferences between thrombocytopenia and the other groups.

implementation of rapid, adequate antibiotic treatment,
control of the source of infection is fundamental for
the resolution of infections. We can, therefore, hypoth-
esize that the higher mortality of patients with thrombo-
cytosis could be related to inadequate management of
these complications for different reasons, for example,
delayed diagnosis and drainage, antibiotic treatment
that is too brief, or a lack of adequate follow-up.
In our population, we did not observe a higher rate
of thrombotic/cardiovascular events in patients with
thrombocytosis, as hypothesized by Mirsaeidi et al.7
These results were in line with other studies that
showed that reactive thrombocytosis, in contrast to
primary thrombocytosis, was not associated with higher
risk of cardiovascular or thrombotic events.31-34
According to our results, we suggest that platelet
count should be monitored in patients with CAP:
Thrombocytopenia requires awareness of septic com-
plications and hemodynamic alterations, whereas in
patients with thrombocytosis, clinicians should pay
attention to local respiratory complications, such as
pleural effusion that may need specific treatment
(eg, drainage or a different follow-up). Adding throm-
bocytosis to the CRB-65 score slightly but signifi-
cantly improved the capacity of this score to recognize
patients with higher mortality risk.
This large, two-center, cohort study has some limi-
tations that should be addressed. First, we only eval-
uated data regarding platelet counts on admission,
since the main purpose of assessing platelet count
at admission was to help decide patients’ allocation
based on the prediction of outcome. Serial measure-
ments of platelet counts during hospitalization could
differentiate between a transient event and sustained
derangements in the platelet count. Second, we did
not evaluate possible differences in the functional
activity of platelets among the different groups. Third,
biomarkers and cytokines were not analyzed in all the
patients, thereby limiting our analyses.
In conclusion, thrombocytosis has been proved to
be a marker of poor outcome in CAP. Consequently,
it should be considered in the severity evaluation of

Table 7—Inflammatory Biomarkers

Biomarkers Thrombocytopenia (n 5 13) Normal Platelet Count (n 5 532) Thrombocytosis (n 5 48) P Value
Day 1
C-reactive protein, mg/dL 11 (7-24) 15 (8-24) 16 (7-23) .721
Procalcitonin, ng/mL 2.72 (0.63-10.57) 0.49 (0.18-2.27) 0.37 (0.22-1.40) .089
TNF-a, pg/mL 39 (28-66) 25 (15-44) 24.5 (14-37) .089
IL-1, pg/mL 21 (4-34) 14 (3-31) 23 (8-44) .048a
IL-6, pg/mL 101 (50-157) 83 (29-238) 78 (37-138) .625
IL-8, pg/mL 4 (0-44) 8 (2-18) 6 (1-15) .248
IL-10, pg/mL 6.5 (1-18) 5 (0-18) 10 (0-23) .522
Day 3
C-reactive protein, mg/dL 5 (2-11) 5 (2-11) 6 (2-11) .975
Procalcitonin, ng/mL 0.39 (0.09-1.32) 0.29 (0.1-0.59) 0.20 (0.10-0.41) .586
TNF-a, pg/mL 26 (9-40) 20 (12-37) 25.5 (14-40) .463
IL-1, pg/mL 7.5 (0-28) 11 (3-28) 13 (1-37) .709
IL-6, pg/mL 29 (19-55) 29 (10-73) 33.4 (8-88) .914
IL-8, pg/mL 3.5 (0-14) 4 (0-12) 6 (0-15) .793
IL-10, pg/mL 12 (1-20) 8 (2-19) 6.5 (0-22) .924
Data given as median (interquartile range). TNF 5 tumor necrosis factor.
aDifferences between thrombocytosis and normal platelet count.

journal.publications.chestnet.org CHEST / 143 / 3 / MARCH 2013 773

Downloaded From: http://journal.publications.chestnet.org/ by a University of Pittsburgh User on 05/17/2013

 patients with CAP and can help promptly identify
patients at higher risk for respiratory complications.
Acknowledgments
Author contributions: Dr Ferrer takes responsibility for the
integrity of the data and the accuracy of the data analysis.
Dr Prina: contributed to study design, analysis of data, and writing
the manuscript and served as principal author.
Dr Ferrer: contributed to the study design, analysis of data, and
writing of the manuscript.
Dr Ranzani: contributed to the study design, analysis of data, and
approved the manuscript.
Dr Polverino: contributed to the recruitment of patients, data col-
lection, and approved the manuscript.
Dr Cillóniz: contributed to the recruitment of patients, data col-
lection, and approved the manuscript.
Ms Moreno: contributed to the recruitment of patients, data col-
lection, and approved the manuscript.
Dr Mensa: contributed to the revision of the manuscript with impor-
tant intellectual content.
Dr Montull: contributed to the recruitment of patients, data col-
lection, and approved the manuscript.
Dr Menéndez: contributed to the recruitment of patients, data
collection, and approved the manuscript.
Dr Cosentini: contributed to the revision of the manuscript with
important intellectual content.
Dr Torres: contributed to the study design and writing of the
manuscript.
Financial/nonfinancial disclosures: The authors have reported
to CHEST that no potential conflicts of interest exist with any
companies/organizations whose products or services may be dis-
cussed in this article.
Role of sponsors: The sponsors had no role in the design of the
study, the collection and analysis of the data, or in the preparation
of the manuscript.
Other contributions: We thank all the clinicians who took part
in the preparation of this paper.
References
1. Yeaman MR. The role of platelets in antimicrobial host defense.
Clin Infect Dis. 1997;25(5):951-968.
2. Smyth SS, McEver RP, Weyrich AS, et al; 2009 Platelet
Colloquium Participants. Platelet functions beyond hemo-
stasis. J Thromb Haemost. 2009;7(11):1759-1766.
3. Katz JN, Kolappa KP, Becker RC. Beyond thrombosis: the ver-
satile platelet in critical illness. Chest. 2011;139(3):658-668.
4. Brogly N, Devos P, Boussekey N, Georges H, Chiche A,
Leroy O. Impact of thrombocytopenia on outcome of patients
admitted to ICU for severe community-acquired pneumonia.
J Infect. 2007;55(2):136-140.
5. Feldman C, Kallenbach JM, Levy H, et al. Community-
acquired pneumonia of diverse aetiology: prognostic features
in patients admitted to an intensive care unit and a “severity
of illness” core. Intensive Care Med. 1989;15(5):302-307.
6. Mandell LA, Wunderink RG, Anzueto A, et al; Infectious
Diseases Society of America; American Thoracic Society.
Infectious Diseases Society of America/American Thoracic Soci-
ety consensus guidelines on the management of community-
acquired pneumonia in adults. Clin Infect Dis. 2007;44(suppl 2):
S27-S72.
7. Mirsaeidi M, Peyrani P, Aliberti S, et al. Thrombocytopenia
and thrombocytosis at time of hospitalization predict mor-
tality in patients with community-acquired pneumonia. Chest.
2010;137(2):416-420.
8. Georges H, Brogly N, Olive D, Leroy O. Thrombocytosis in
patients with severe community-acquired pneumonia. Chest.
2010;138(5):1279-1280.
774
Downloaded From: http://journal.publications.chestnet.org/ by a University of Pittsburgh User on 05/17/2013
9. Cunha BA, Hage JE. Community-acquired pneumonia: diag-
nostic vs prognostic significance of the platelet count. Chest.
2011;139(5):1255-1256.
10. Cillóniz C, Ewig S, Polverino E, et al. Microbial aetiology of
community-acquired pneumonia and its relation to severity.
Thorax. 2011;66(4):340-346.
11. Dellinger RP, Levy MM, Carlet JM, et al; International
Surviving Sepsis Campaign Guidelines Committee; American
Association of Critical-Care Nurses; American College of
Chest Physicians; American College of Emergency Physicians;
Canadian Critical Care Society; European Society of Clinical
Microbiology and Infectious Diseases; European Society
of Intensive Care Medicine; European Respiratory Society;
International Sepsis Forum; Japanese Association for Acute
Medicine; Japanese Society of Intensive Care Medicine; Soci-
ety of Critical Care Medicine; Society of Hospital Medicine;
Surgical Infection Society; World Federation of Societies of
Intensive and Critical Care Medicine. Surviving Sepsis Cam-
paign: international guidelines for management of severe
sepsis and septic shock: 2008. Crit Care Med. 2008;36(1):
296-327.
12. Fine MJ, Auble TE, Yealy DM, et al. A prediction rule to iden-
tify low-risk patients with community-acquired pneumonia.
N Engl J Med. 1997;336(4):243-250.
13. Lim WS, van der Eerden MM, Laing R, et al. Defining com-
munity acquired pneumonia severity on presentation to hospital:
an international derivation and validation study. Thorax. 2003;
58(5):377-382.
14. Light RW, Girard WM, Jenkinson SG, George RB. Para-
pneumonic effusions. Am J Med. 1980;69(4):507-512.
15. Niederman MS, Bass JB Jr, Campbell GD, et al; American
Thoracic Society. Medical Section of the American Lung
Association. Guidelines for the initial management of adults
with community-acquired pneumonia: diagnosis, assessment
of severity, and initial antimicrobial therapy. Am Rev Respir
Dis. 1993;148(5):1418-1426.
16. Niederman MS, Mandell LA, Anzueto A, et al; American
Thoracic Society. Guidelines for the management of adults
with community-acquired pneumonia. Diagnosis, assessment
of severity, antimicrobial therapy, and prevention. Am J Respir
Crit Care Med. 2001;163(7):1730-1754.
17. Bodí M, Rodríguez A, Solé-Violán J, et al; Community-
Acquired Pneumonia Intensive Care Units (CAPUCI) Study
Investigators. Antibiotic prescription for community-acquired
pneumonia in the intensive care unit: impact of adherence to
Infectious Diseases Society of America guidelines on survival.
Clin Infect Dis. 2005;41(12):1709-1716.
18. Ramírez P, Ferrer M, Martí V, et al. Inflammatory biomarkers
and prediction for intensive care unit admission in severe
community-acquired pneumonia. Crit Care Med. 2011;39(10):
2211-2217.
19. Hanley JA, McNeil BJ. A method of comparing the areas
under receiver operating characteristic curves derived from
the same cases. Radiology. 1983;148(3):839-843.
20. Wolach B, Morag H, Drucker M, Sadan N. Thrombocytosis
after pneumonia with empyema and other bacterial infec-
tions in children. Pediatr Infect Dis J. 1990;9(10):718-721.
21. Dodig S, Raos M, Kovac K, et al. Thrombopoietin and inter-
leukin-6 in children with pneumonia-associated thrombocy-
tosis. Arch Med Res. 2005;36(2):124-128.
22. Chalmers JD, Singanayagam A, Murray MP, Scally C, Fawzi A,
Hill AT. Risk factors for complicated parapneumonic effusion
and empyema on presentation to hospital with community-
acquired pneumonia. Thorax. 2009;64(7):592-597.
23. Strauss R, Wehler M, Mehler K, Kreutzer D, Koebnick C,
Hahn EG. Thrombocytopenia in patients in the medical inten-
sive care unit: bleeding prevalence, transfusion requirements,
and outcome. Crit Care Med. 2002;30(8):1765-1771.
Original Research
 24. Vandijck DM, Blot SI, De Waele JJ, Hoste EA, Vandewoude
KH, Decruyenaere JM. Thrombocytopenia and outcome in
critically ill patients with bloodstream infection. Heart Lung.
2010;39(1):21-26.
25. Brun-Buisson C, Doyon F, Carlet J, et al; French ICU Group
for Severe Sepsis. Incidence, risk factors, and outcome of
severe sepsis and septic shock in adults. A multicenter pro-
spective study in intensive care units. JAMA. 1995;274(12):
968-974.
26. Ashbaugh DG. Empyema thoracis. Factors influencing mor-
bidity and mortality. Chest. 1991;99(5):1162-1165.
27. Grijalva CG, Zhu Y, Nuorti JP, Griffin MR. Emergence of
parapneumonic empyema in the USA. Thorax. 2011;66(8):
663-668.
28. Cillóniz C, Ewig S, Polverino E, et al. Pulmonary compli-
cations of pneumococcal community-acquired pneumonia:
incidence, predictors, and outcomes. Clin Microbiol Infect.
2012;18(11):1134-1142.
journal.publications.chestnet.org
Downloaded From: http://journal.publications.chestnet.org/ by a University of Pittsburgh User on 05/17/2013
29. Falguera M, Carratalà J, Bielsa S, et al. Predictive factors,
microbiology and outcome of patients with parapneumonic
effusion. Eur Respir J. 2011;38(5):1173-1179.
30. Menendez R, Torres A. Treatment failure in community-
acquired pneumonia. Chest. 2007;132(4):1348-1355.
31. van der Bom JG, Heckbert SR, Lumley T, et al. Platelet count
and the risk for thrombosis and death in the elderly. J Thromb
Haemost. 2009;7(3):399-405.
32. Zakai NA, Katz R, Jenny NS, et al. Inflammation and hemostasis
biomarkers and cardiovascular risk in the elderly: the Cardiovas-
cular Health Study. J Thromb Haemost. 2007;5(6):1128-1135.
33. Folsom AR, Rosamond WD, Shahar E, et al; The Atherosclerosis
Risk in Communities (ARIC) Study Investigators. Prospective
study of markers of hemostatic function with risk of ischemic
stroke. Circulation. 1999;100(7):736-742.
34. Meade TW, Cooper JA, Miller GJ. Platelet counts and aggre-
gation measures in the incidence of ischaemic heart disease
(IHD). Thromb Haemost. 1997;78(2):926-929.
CHEST / 143 / 3 / MARCH 2013 775