bs_bs_banner

ORIGINAL ARTICLE

Association between pneumococcal pneumonia

and venous thromboembolism in hospitalized patients:
A nationwide population-based study

YU-GUANG CHEN,1 TE-YU LIN,2 WEN-YEN HUANG,3 CHENG-LI LIN,4 MING-SHEN DAI1 AND
CHIA-HUNG KAO5,6

1
Division of Hematology/Oncology, Department of Internal Medicine, Tri-Service General Hospital, 2Division of Infectious
Diseases and Tropical Medicine, Department of Internal Medicine, Tri-Service General Hospital and 3Department of
Radiation Oncology, Tri-Service General Hospital, National Defense Medical Center, Taipei, and 4Management Office for
Health Data and 6Department of Nuclear Medicine and PET Center, China Medical University Hospital, and 5Graduate
Institute of Clinical Medical Science, China Medical University College of Medicine, Taichung, Taiwan

ABSTRACT
Background and objectives: This was a nationwide
population-based retrospective cohort study to inves-
tigate the risk of developing deep-vein thrombosis
(DVT) and pulmonary embolism (PE) in patients with a
pneumococcal pneumonia.
Methods: We analysed data from 1998 to 2010 from
the Taiwan National Health Insurance Database. The
follow-up period was extended to the end of 2011. We
identified patients with pneumococcal pneumonia and
selected a comparison cohort matched for age, sex and
diagnosis year at a ratio of one pneumococcal pneumo-
nia patient to four control patients. We analysed
the risks of DVT and PE by using Cox proportional
hazards regression models, including gender, age and
comorbidities.
Results: In total, 18 928 pneumococcal pneumonia
patients and 75 712 controls were included in the
study. The risks of developing DVT and PE were
1.78-fold (95% CI: 1.39–2.28) and 1.97-fold (95% CI:
1.43–2.72), respectively, in patients with pneumococcal
pneumonia compared to the control cohort after
adjusting for age, gender and comorbidities. The
increased risks of DVT and PE were significant in
patients who exhibited any comorbidity. The inci-
dences of DVT and PE were highest in the first 4 weeks
after pneumonia and remained slightly elevated from
13 weeks to 2 years after acute infection.
Conclusion: Pneumococcal pneumonia should be
considered a risk factor for DVT and PE, even after the
patient has recovered from the acute infection.
Correspondence: Chia-Hung Kao, Graduate Institute of Clinical
Medical Science and School of Medicine, College of Medicine,
China Medical University, No. 2, Yuh-Der Road, Taichung 404,
Taiwan. Email: d10040@mail.cmuh.org.tw
Received 15 September 2014; invited to revise 20 October and
16 November 2014; revised 3 and 23 November 2014; accepted
11 December 2014 (Associate Editor: Marcos Restrepo).
© 2015 Asian Pacific Society of Respirology
SUMMARY AT A GLANCE
Patients with pneumococcal pneumonia have a
1.78-fold and 1.97-fold risk to develop a DVT and
PE, respectively in comparison with control
cohort. The risk was particularly high in the first 4
weeks and in patients with comorbidities.
Key words: deep-vein thrombosis, nationwide population-
based study, pneumococcal pneumonia, pulmonary embolism.
Abbreviations: CI, confidence interval; DVT, deep-vein thrombo-
sis; HR, hazard ratio; ICD-9-CM, International Classification of
Disease, Ninth Revision, Clinical Modification; IRR, incidence rate
ratio; NHIRD, National Health Insurance Research Database;
PE, pulmonary embolism; VTE, venous thromboembolism
INTRODUCTION
Thrombotic events, which involve the arterial or
venous systems, are major health issues in the
general population worldwide. In developed coun-
tries, venous thromboembolism (VTE) is the third
most common cardiovascular disease following ath-
erosclerotic heart disease and stroke. In the United
States, up to 2 million cases of VTE, including deep-
vein thrombosis (DVT), calf-vein thrombosis and
pulmonary embolism (PE), are diagnosed annually.1
Compared with arterial thrombotic events, which are
characterized by pain and organ dysfunction, VTE is
frequently asymptomatic or presents with few symp-
toms; however, it is often of significant consequence.
Previous studies have identified several risk
factors for VTE including advanced age, obesity,
surgical intervention, trauma, advanced cancer,
acute medical illness and haematological or other
autoimmune disorders.2 Acute infectious processes
have been recognized as a risk factor.3 A possible
Respirology (2015)
doi: 10.1111/resp.12501
2 Y-G Chen et al.
mechanism could be because acute infection can
induce local or systemic inflammatory reactions and,
thus, affect endothelial cell function and cause subse-
quent vessel wall damage.4 In addition, infection can
cause venous vessel dilation, venous stasis and
coagulation dysfunction. Endothelial cell injury and
impairment of blood flow can result in a chronic
inflammatory reaction and thrombosis formation.
Previous large-scale, case-controlled cohort studies
on community populations have reported that acute
infectious processes, including those that occur in the
respiratory or urinary tract, are potential risk factors
for acute VTE.5 However, no analyses on the specific
pathogens involved in acute respiratory tract infec-
tion have been conducted in a community setting.
Pneumococcal pneumonia is a leading cause of
hospitalization and the most common pathogen of
community-acquired pneumonia in adults.6 The
purpose of this nationwide population-based cohort
study was to investigate the association between
pneumococcal pneumonia and VTE in a multi-ethnic
population.
METHODS
Data source
We used the database of the Taiwan National Health
Insurance (NHI) programme, a comprehensive insur-
ance programme established in 1995 that provides
coverage to 99% of the population of Taiwan (23.7
million). The Taiwan National Health Insurance
Research Database (NHIRD) contains medical claims
for all beneficiaries of hospitals and health-care
institutions in Taiwan. The Institutional Review Board
of China Medical University (CMU-REC-101-012)
approved this study based on full ethical review. We
studied inpatient claims data and used the Interna-
tional Classification of Diseases, Ninth Revision,
Clinical Modification (ICD-9-CM) to define diseases.
Study population
This was a population-based retrospective cohort
study. Using inpatient claims data, we identified
18 928 newly diagnosed pneumococcal pneumonia
(ICD-9-CM 481) patients. The diagnoses of pneumo-
coccal pneumonia and venous thromboembolism
were based on the ICD-9 codes that were assessed
and determined by specialist physicians according
to standard clinical criteria. The year of the first
pneumococcal pneumonia diagnosis served as the
diagnosis year, between 1998 and 2010. A comparison
group was formed by frequency matching four cohort
patients to one pneumococcal pneumonia patient
according to age, gender and diagnosis year. The date
of pneumococcal pneumonia diagnosis was desig-
nated as index dates to enable measuring follow-up
person-years. The follow-up person-years at the end
of 2011 were calculated for each patient until diag-
nosed with DVT (ICD-9-CM 453.8) or PE (ICD-9-CM
415.1), withdrawn from the insurance system, or cen-
sored because of death, or lost to follow-up. The ICD-
9-CM definition of PE does not include iatrogenic PE
Respirology (2015)
(ICD-9-CM 415.11). Patients with a history of DVT
or PE before index pneumonia or for whom informa-
tion was incomplete were excluded. DVT and PE
were diagnosed based on clinical manifestations,
laboratory investigations and imaging studies. We
considered certain comorbidities: atrial fibrillation
(ICD-9-CM 427.31), hypertension (ICD-9-CM 401–
405), diabetes (ICD-9-CM 250), hyperlipidemia (ICD-
9-CM 272), stroke (ICD-9-CM 430–438), heart failure
(ICD-9-CM 428), lower leg fracture or surgery (ICD-
9-CM 820, 821; 823; 81.51, 81.52, 81.53, 81.54) and all
cancers (ICD-9-CM 140-208). Taiwan launched an
NHI in 1995.
Statistical analysis
We compared the distributions of age, gender and
comorbidities of the exposure and comparison
groups by using the chi-squared test. The Student’s
t test was used to compare differences between the
two groups. The incidence of DVT and PE was esti-
mated according to age, gender and comorbidity (no/
yes) in both groups. The incidence rate ratio (IRR) of
DVT or PE in the pneumonia and comparison groups
was measured for these variables by using a Poisson
regression analysis. The multivariable Cox propor-
tional hazards regression was used to compare the
adjusted hazard ratios (HR) of differences in DVT and
PE between the two groups at a 95% confidence inter-
val (CI), and gender, age and comorbidities were
controlled for.
The SAS 9.3 statistical package (SAS Institute Inc.,
NC, USA) was used to perform statistical analyses,
and P < .05 in two-tailed tests indicated statistical sig-
nificance. The R software (R Foundation for Statistical
Computing, Vienna, Austria) was used to conduct a
Kaplan–Meier analysis to measure the cumulative
incidence of DVT and PE in both study groups, and
the log-rank test was used to assess the differences
between the two cumulative incidence curves.
RESULTS
Table 1 shows the demographic characteristics and
comorbidities of the pneumococcal pneumonia
group (n = 18 928) and the comparison group
(n = 75 712). Both groups were predominantly male
(65.7%) and over 65 years of age (56.4%). The mean
were 66.0 (standard deviation (SD): 17.8) and 65.2
(SD: 17.7), respectively. All comorbidities were more
prevalent in the pneumonia group at the baseline
(P < .0001). The incidence of DVT in the pneumonia
group was 1.24 per 1000 person-years, and compared
with the comparison group, the IRR was 1.95 (95% CI:
1.85–2.06). The adjusted HR was 1.78 (95% CI: 1.39–
2.28). The incidence of PE was 2.23 times greater in
the pneumonia group than in the comparison group
(0.78 vs 0.35 per 1000 person-years), with an adjusted
HR of 1.97 (95% CI: 1.43–2.72). Male patients with
pneumococcal pneumonia exhibited an increased
risk of developing DVT (adjusted (a)HR = 1.86, 95%
CI: 1.37, 2.53) or PE (aHR = 2.30, 95% CI: 1.56–3.40)
compared with the comparison group. The risks of
© 2015 Asian Pacific Society of Respirology
Pneumococcal pneumonia and DVT-PE 3

Table 1 Demographic characteristics and comorbidity in patients with and without pneumococcal pneumonia

Pneumococcal pneumonia

No (n = 75 712) Yes (n = 18 928)

Variables n % n % P-value†

Gender 0.99
Female 25 932 34.3 6 483 34.3
Male 49 780 65.7 12 445 65.7
Age, years 0.99
≤49 15 028 19.9 3 757 19.9
50–64 14 412 19.0 3 603 19.0
65–74 17 088 22.6 4 272 22.6
75+ 29 184 38.6 7 296 38.6
Mean ± SD† 65.2 17.7 66.0 17.8 <0.0001
Comorbidity
Atrial fibrillation 1 436 1.90 1 395 7.37 <0.0001
Hypertension 12 764 16.9 7 807 41.3 <0.0001
Diabetes 6 305 8.33 4 984 26.3 <0.0001
Hyperlipidemia 2 551 3.37 1 472 7.78 <0.0001
CVA 6 044 7.98 4 634 24.5 <0.0001
Heart failure 2 287 3.02 2 460 13.0 <0.0001
Lower leg fracture or surgery 2 463 3.25 1 249 6.60 <0.0001
Cancer 2 421 3.20 2 005 10.6 <0.0001

Chi-square test.
†
t-test.
CVA, cerebrovascular accident; SD, standard deviation.

developing DVT or PE increased with age in both
groups; however, the gradient was greater in the
pneumonia group. An association between pneu-
mococcal pneumonia and the risk of DVT or PE was
observed in patients with comorbidities (aHR = 1.78,
95% CI: 1.32–2.41 for DVT; aHR = 1.70, 95% CI: 1.16–
2.51 for PE). (Table 2).
Table 3 shows the multivariate-adjusted HR for
DVT or PE associated with pneumococcal pneumonia
and covariates. The aHR for developing DVT or
PE increased with age compared with patients
younger than 50 years of age. The effect was greater in
patients with comorbidities associated with DVT,
including cancer (aHR = 2.30, 95% CI: 1.58–3.36),
heart failure (aHR = 1.78, 95% CI: 1.21–2.63), hyper-
tension (aHR = 1.41, 95% CI: 1.08–1.82) and diabetes
(aHR = 1.35, 95% CI: 1.01–1.81). Patients with cancer,
heart failure or atrial fibrillation exhibited an
increased risk of developing PE (aHR = 2.33, 95% CI:
1.42–3.80 for cancer; aHR = 2.45, 95% CI: 1.55–3.86 for
heart failure and aHR = 2.08, 95% CI: 1.23–3.54 for
atrial fibrillation). Table 4 lists the HR of DVT or PE in
patients who had been infected with pneumococcal
pneumonia. The incidences of DVT and PE were
highest in the first 4 weeks after exposure (aHR = 11.0,
95% CI: 1.06–114.0 for DVT in the first 2 weeks;
aHR = 16.5, 95% CI: 1.75–155.8 for PE in 2 weeks to
1 month) and gradually decreased over the subse-
quent months. The incidence for PE rose slightly from
13 weeks to 1 year after acute infection (aHR = 2.77,
95% CI: 1.15–6.43). We observed a similar trend in the
DVT group, but this trend occurred 1 to 2 years after
© 2015 Asian Pacific Society of Respirology
acute infection (aHR = 1.83, 95% CI: 1.02–3.29). We
observed the patients for 14 years and used
Kaplan–Meier analysis to calculate the cumulative
incidence of DVT or PE, as shown in Figure 1. Patients
with pneumococcal pneumonia exhibited higher
incidence rates of DVT than those of the comparison
group (Fig. 1a, log-rank test: P < .0001). The cumula-
tive incidence of PE was greater in the pneumonia
group than in the comparison group (Fig. 1b, log-rank
test: P < .0001).
DISCUSSION
We observed that adults with pneumococcal pneu-
monia have 1.97-fold and 1.78-fold risks of develop-
ing PE and DVT, respectively. Our study determines
the positive correlation between pneumococcal
pneumonia and VTE, providing epidemiological evi-
dence of the relationship between these two diseases.
Our data, corrected for classical risk factor for VTE
suggests that pneumococcal pneumonia is an inde-
pendent risk factor for VTE.
Pneumococcal pneumonia caused by the Strepto-
coccus pneumoniae, a Gram-positive encapsulated
bacteria, is primary cause of community-acquired
pneumonia in children and adults. The key compo-
nents involved in the pathogenesis of pneumococcal
pneumonia are the surface capsular polysaccharides
that evade host immune responses. Virulent protein-
like pneumolysin or autolysin causes this pathogen
to invade and lyse host cells. These components
Respirology (2015)
4 Y-G Chen et al.

Table 2 Comparison of incidence and hazard ratio of DVT and PE stratified by gender and age between with and
without pneumococcal pneumonia patients

Pneumococcal pneumonia

No Yes

Variables Event PY Rate† Event PY Rate† IRR (95% CI) Adjusted HR‡ (95% CI)

DVT 297 465 580 0.64 96 77 109 1.24 1.95 (1.85, 2.06)*** 1.78 (1.39, 2.28)***
Gender
Female 106 17 069 0.62 36 31 144 1.16 1.86 (1.70, 2.03)*** 1.68 (1.11, 2.53)*
Male 191 294 900 0.65 60 45 964 1.31 2.02 (1.89, 2.15)*** 1.86 (1.37, 2.53)***
Stratify age
≤ 49 11 109 804 0.10 15 24 525 0.61 6.11 (5.47, 6.82)*** 3.60 (1.54, 8.44)**
50–64 45 101 445 0.44 15 17 780 0.84 1.90 (1.68, 2.15)*** 1.20 (0.62, 2.34)
65–74 84 116 943 0.72 23 17 289 1.33 1.85 (1.66, 2.07)*** 1.53 (0.93, 2.53)
75+ 157 13 787 1.14 43 17 515 2.45 2.15 (1.97, 2.34)*** 1.80 (1.25, 2.57)**
Comorbidity¶
No 187 377 515 0.50 21 38 802 0.54 1.09 (1.00, 1.20) 1.55 (0.98, 2.44)
Yes 110 88 065 1.25 75 38 306 1.96 1.57 (1.45, 1.69)*** 1.78 (1.32, 2.41)***
PE 162 466 069 0.35 60 77 245 0.78 2.23 (2.12, 2.36)*** 1.97 (1.43, 2.72)***
Sex
Female 62 170 810 0.36 18 31 190 0.58 1.59 (1.44, 1.75)*** 1.45 (0.82, 2.57)
Male 100 295 259 0.34 42 46 054 0.91 2.69 (2.52, 2.87)*** 2.30 (1.56, 3.40)***
Stratify age
≤ 49 6 109 821 0.05 6 24 590 0.24 4.47 (3.96, 5.04)*** 3.23 (0.94, 11.1)
50–64 14 101 549 0.14 17 17 769 0.96 6.93 (6.22, 7.75)*** 3.13 (1.37, 7.15)**
65–74 44 117 088 0.38 9 17 329 0.52 1.38 (1.22, 1.57)*** 1.14 (0.52, 2.49)
75+ 98 137 609 0.71 28 17 556 1.59 2.24 (2.05, 2.45)*** 1.80 (1.14, 2.83)*
Comorbidity¶
No 96 377 855 0.25 14 38 852 0.36 1.42 (1.30, 1.55)*** 2.18 (1.24, 3.83)**
Yes 66 88 214 0.75 46 38 392 1.20 1.60 (1.48, 1.74)*** 1.70 (1.16, 2.51)**

*P < 0.05; **P < 0.01; ***P < 0.001.

†
Per 1000 person-years.
‡
Mutually adjusted for age, gender and comorbidities in Cox proportional hazard regression.
¶
Includes atrial fibrillation, hypertension, diabetes, hyperlipidemia, CVA, heart failure, lower leg fracture or surgery and cancer.
CI, confidence interval; DVT, deep-vein thrombosis; HR, hazard ratio; IRR, incidence rate ratio; PE, pulmonary embolism;
PY, person-years.

recognize the endothelial cell through the platelet-
activating factor and invade the blood vessel and
alveolar space.7 These processes may induce an
inflammatory response and endothelial wall damage,
particularly in pulmonary alveolar cells. Severe and
persistent inflammation results in excessive fibrin
deposition and increased activation of the coagula-
tion cascade, thus inducing thromboembolism for-
mation.8 Because pneumococcal pneumonia invades
the pulmonary alveolar region and subsequently
destroys endothelial cells, these can cause an abnor-
mal coagulant cascade that affects endothelial cell
alteration and adhesion molecule expression, as
observed in Lemierre’s syndrome.9 Therefore, in
pneumococcal pneumonia patients, the pulmonary
infection/inflammatory process may induce localized
and systemic thrombotic events that explain the rela-
tively higher incidence rate of PE and DVT observed
in our study.
Although the patients with pneumococcal pneu-
monia in this study exhibited a higher prevalence of
comorbidities and coexistent conditions associated
with the development of VTE than that of the com-
Respirology (2015)
parison cohort, pneumococcal pneumonia remained
an independent risk factor for developing DVT and PE
after adjusting for comorbidities. These results are
robust according to the Cox proportional hazards
regression analysis. The IRR and HR for VTE were
higher in the younger patients with pneumococcal
pneumonia, possibly because the older age sub-
group exhibited more comorbidities and underlying
medical illness. Confounding factors, including atrial
fibrillation, metabolic syndrome and cancer, may
have reduced the effect of the pneumococcal
pneumonia.
Our study yielded findings similar to those of
Smeeth et al.5 The highest incidence rate of VTE was
observed within the first month after pneumococcal
pneumonia. Our results reveal that, after a patient
experienced an acute pneumococcal pneumonia
infection, subsequent VTE risk increased at the early
stage (0–4 weeks) as well as at the delayed stage until
1 year.
This study has several limitations. First, the
NHIRD does not contain detailed information
regarding patients’ currently used medications,
© 2015 Asian Pacific Society of Respirology
Pneumococcal pneumonia and DVT-PE 5

Table 3 Cox model with hazard ratios and 95% confidence intervals of DVT and PE associated with Pneumococcal
pneumonia and covariates

DVT PE

Variable Adjusted† HR (95% CI) Adjusted† HR (95% CI)

Age, years
≤ 49 1 (Reference) 1 (Reference)
50–64 2.45 (1.55, 3.89)*** 2.88 (1.48, 5.61)**
65–74 3.73 (2.41, 5.75)*** 4.20 (2.23, 7.91)***
75+ 5.95 (3.89, 9.09)*** 8.28 (4.50, 15.3)***
Gender (female vs male) 1.01 (0.82, 1.25) 1.00 (0.76, 1.32)
Baseline comorbidities (yes vs no)
Pneumococcal pneumonia 1.80 (1.41, 2.30)*** 2.02 (1.46, 2.78)***
Atrial fibrillation 1.13 (0.66, 1.92) 2.08 (1.23, 3.54)**
Hypertension 1.41 (1.08, 1.82)* 1.11 (0.78, 1.57)
Diabetes 1.35 (1.01, 1.81)* 0.85 (0.56, 1.31)
Hyperlipidemia 1.18 (0.77, 1.82) 1.24 (0.70, 2.19)
CVA 0.79 (0.56, 1.12) 1.08 (0.71, 1.65)
Heart failure 1.78 (1.21, 2.63)** 2.45 (1.55, 3.86)***
Lower leg fracture or surgery 1.59 (1.02, 2.46)* 1.25 (0.68, 2.32)
Cancer 2.30 (1.58, 3.36)*** 2.33 (1.42, 3.80)***

*P < 0.05, **P < 0.01, ***P < 0.001.

†
Multivariable analysis including for stratify age, gender and comorbidities of atrial fibrillation, hypertension, diabetes,
hyperlipidemia, CVA, heart failure, lower leg fracture or surgery and cancer.
CI, confidence interval; CVA, cerebrovascular accident; HR, hazard ratio; DVT, deep-vein thrombosis; PE, pulmonary embolism.

Table 4 Trends of DVT and PE event risk by stratified follow-up years

Non-pneumococcal Pneumococcal
pneumonia Cohort pneumonia Cohort

Follow time Event PY Rate† Event PY Rate† IRR (95% CI) Adjusted HR‡ (95% CI)

DVT
≤2 weeks 1 3 146 0.32 4 762 5.25 16.5 (15.4, 17.7)*** 11.0 (1.06, 114.0)*
2–4 weeks 2 3 149 0.64 0 732 0.00 — —
4–8 weeks 2 6 272 0.32 2 1 404 1.42 4.47 (4.20, 4.74)*** 3.33 (0.43, 25.6)
8–12 weeks 2 6 257 0.32 1 1 352 0.74 2.31 (2.17, 2.47)*** 0.94 (0.07, 12.2)
13–48 weeks 32 55 194 0.58 17 10 904 1.56 2.69 (2.54, 2.84)*** 1.86 (0.99, 3.50)
1–2 years 39 68 053 0.57 19 12 133 1.57 2.73 (2.58, 2.90)*** 1.83 (1.02, 3.29)*
2–3 years 53 59 890 0.88 13 9 940 1.31 1.48 (1.38, 1.59)*** 1.46 (0.77, 2.77)
PE
≤2 weeks 0 3 147 0.00 4 762 5.25 — —
2–4 weeks 1 3 149 0.32 5 732 6.83 21.5 (20.0, 23.2)*** 16.5 (1.75, 155.8)*
4–8 weeks 0 6 272 0.00 3 1 404 2.14 — —
8–12 weeks 3 6 257 0.48 2 1 352 1.48 3.09 (2.90, 3.28)*** 1.11 (0.16, 7.55)
13–48 weeks 16 55 203 0.29 10 10 905 0.92 3.16 (2.99, 3.35)*** 2.77 (1.15, 6.43)*
1–2 years 20 68 085 0.29 8 12 143 0.66 2.24 (2.11, 2.39)*** 2.15 (0.89, 5.22)
2–3 years 20 59 935 0.33 0 9 955 0.00 — —

*P < 0.05, ***P < 0.001.

†
Incidence rate, per 1000 person-years.
‡
Multiple analysis including age, gender and comorbidities.
CI, confidence interval; DVT, deep-vein thrombosis; HR, hazard ratio; IRR, incidence rate ratio; PE, pulmonary embolism; PY,
person-years.

such as hormone replacement medications, anti-
contraceptive medications, anti-coagulants or anti-
platelet agents and glucocorticosteroids. In addition,
several clinical characteristics included smoking
© 2015 Asian Pacific Society of Respirology
habits, obesity, body mass index and physical acti-
vity (whether immobilization), previous prosthetic
devices or a family history of VTE, all of which might
have been confounding factors in this study. Second,
Respirology (2015)
6 Y-G Chen et al.
Figure 1 Cumulative incidence of DVT (a) and PE (b) in patients
with pneumococcal pneumonia and comparison patients.
Without pneumococcal pneumonia; With pneumococcal
pneumonia
the serotype of S. pneumoniae is a crucial, virulent
factor involving an invasive disease burden. However,
these data are not contained in the NHIRD.10 Third, the
evidence derived from a retrospective cohort study is
typically lower in statistical quality than that derived
from randomized trials because of potential biases
related to adjustments for confounding variables.
Fourth, in our study confounding factors, such as
nephrotic syndrome, inflammatory bowel disease,
vasculitis, haematological disorders of antipho-
spholipid syndrome and heparin, induce thrombocy-
topenia, human immunodeficiency virus infection,
pregnancy, hyperviscosity syndrome and a few heredi-
tary thrombophilia were not included. Fifth, in our
cohort no severity scores were recorded, such as the
APACHE II (Acute Physiology and Chronic Health
Respirology (2015)
Evaluation II) or the pneumonia severity index.
Finally, information on management after hospita-
lization, such as central venous catheter insertion
and invasive medical devices/monitors, were not
available.
The main contribution of our study is its use of a
nationwide population-based database that contains
data on a high number of pneumococcal pneumonia
cases. The results indicate that patients with
pneumococcal pneumonia have 1.97-fold and 1.78-
fold risks of developing PE and DVT. We recommend
that physicians consider the possibility of VTE when
treating patients with pneumococcal pneumonia.
Pneumococcal pneumonia should be considered
when evaluating thromboembolism risk.
Acknowledgements
This study is supported in part by Taiwan Ministry of Health and
Welfare Clinical Trial and Research Center of Excellence
(MOHW104-TDU-B-212-113002); China Medical University Hospi-
tal, Academia Sinica Taiwan Biobank, Stroke Biosignature Project
(BM104010092); NRPB Stroke Clinical Trial Consortium (MOST
103-2325-B-039 -006); Tseng-Lien Lin Foundation, Taichung,
Taiwan; Taiwan Brain Disease Foundation, Taipei, Taiwan;
Katsuzo and Kiyo Aoshima Memorial Funds, Japan; and Health,
and welfare surcharge of tobacco products, China Medical Uni-
versity Hospital Cancer Research Center of Excellence
(MOHW104-TDU-B-212-124-002, Taiwan). The funders had no
role in study design, data collection and analysis, decision to
publish, or preparation of the manuscript. No additional external
funding received for this study.
REFERENCES
1 Greer JP. Wintrobe’s Clinical Hematology, 13th edn. Wolters
Kluwer Lippincott Williams & Wilkins Health, Philadelphia, PN,
2014.
2 Kyrle PA, Eichinger S. Deep vein thrombosis. Lancet 2005; 365:
1163–74.
3 Levine RL, LeClerc JR, Bailey JE, Monberg MJ, Sarwat S. Venous
and arterial thromboembolism in severe sepsis. Thromb.
Haemost. 2008; 99: 892–8.
4 Bhagat K, Moss R, Collier J, Vallance P. Endothelial ‘stunning’
following a brief exposure to endotoxin: a mechanism to link
infection and infarction? Cardiovasc. Res. 1996; 32: 822–9.
5 Smeeth L, Cook C, Thomas S, Hall AJ, Hubbard R, Vallance P. Risk
of deep vein thrombosis and pulmonary embolism after acute
infection in a community setting. Lancet 2006; 367: 1075–9.
6 Bartlett JG, Mundy LM. Community-acquired pneumonia. N.
Engl. J. Med. 1995; 333: 1618–24.
7 Cundell DR, Gerard NP, Gerard C, Idanpaan-Heikkila I,
Tuomanen EI. Streptococcus pneumoniae anchor to activated
human cells by the receptor for platelet-activating factor. Nature
1995; 377: 435–8.
8 Abraham E. Coagulation abnormalities in acute lung injury and
sepsis. Am. J. Respir. Cell Mol. Biol. 2000; 22: 401–4.
9 Kuppalli K, Livorsi D, Talati NJ, Osborn M. Lemierre’s syndrome
due to Fusobacterium necrophorum. Lancet Infect. Dis. 2012; 12:
808–15.
10 Hung IF, Tantawichien T, Tsai YH, Patil S, Zotomayor R. Regional
epidemiology of invasive pneumococcal disease in Asian adults:
epidemiology, disease burden, serotype distribution, and anti-
microbial resistance patterns and prevention. Int. J. Infect. Dis.
2013; 17: e364–73.
© 2015 Asian Pacific Society of Respirology