© 2009 Schattauer GmbH, Stuttgart
Blood Coagulation, Fibrinolysis and Cellular Haemostasis
In-hospital symptomatic venous thromboembolism
and antithrombotic prophylaxis in Internal Medicine
Findings from a multicenter, prospective study
Gualberto Gussoni1; Mauro Campanini2; Mauro Silingardi3; Gianluigi Scannapieco4; Antonino Mazzone5;
Giovanna Magni6; Antonella Valerio1; Ido Iori1,3; Walter Ageno7; on behalf of the GEMINI Study Group*
1FADOI Study Centre, Milan, Italy; 2Internal Medicine, Maggiore Hospital, Novara, Italy; 3Internal Medicine I, Arcispedale S. Maria Nuova,
Reggio Emilia, Italy; 4Internal Medicine, Cà Foncello Hospital, Treviso, Italy; 5Internal Medicine, Civile Hospital, Legnano, Italy; 6QBGroup SpA,
Padova, Italy; 7Department of Clinical Medicine, Insubria University, Varese, Italy
Summary
Hospitalised medical patients are at increased risk of venous
thromboembolism (VTE), but the incidence of hospitalisation-
related VTE in unselected medical inpatients has not been ex-
tensively studied, and uncertainties remain about the optimal
use of thromboprophylaxis in this setting. Aims of our prospec-
tive, observational study were to assess the prevalence of VTE
and the incidence of symptomatic, hospitalisation-related events
in a cohort of consecutive patients admitted to 27 Internal
Medicine Departments, and to evaluate clinical factors associ-
ated with the use of thromboprophylaxis. Between March and
September 2006, a total of 4,846 patients were included in the
study. Symptomatic VTE with onset of symptoms later than 48
hours after admission (”hospital-acquired” events, primary
study end-point) occurred in 26 patients (0.55%), while the
overall prevalence of VTE (including diagnosis prior to or at ad-
Keywords
Deep venous thrombosis, pulmonary embolism, prophylaxis,
management of disease
Introduction
Though an increased risk for venous thromboembolism (VTE) is
well-known in patients hospitalised for medical illnesses (1), un-
certainties remain on epidemiology, risk assessment and preven-
tion of VTE in this setting (2–5), due to heterogeneity among
available studies, and the complexity of patient population, char-
acterised by co-morbidities and patient-specific multiple risk
factors (6–8).
Correspondence to:
Gualberto Gussoni, MD
FADOI – Centro Studi
Via G.B. Bazzoni 8, Milano, Italy
Tel.: +39 02 48005140, Fax: +39 02 48027691
E-mail: gualberto_gussoni@yahoo.it
893
mission) was 3.65%. During hospital stay antithrombotic pro-
phylaxis was administered in 41.6% of patients, and in 58.7% of
Downloaded by: Lund University Libraries. Copyrighted material.
those for whom prophylaxis was recommended according to
the 2004 Guidelines of the American College of Chest Phys-
icians. The choice of administering thromboprophylaxis or not
appeared qualitatively adherent to indications from randomised
clinical trials and international guidelines, and bed rest was the
strongest determinant of the use of prophylaxis. Data from our
real-world study confirm that VTE is a relevant complication in
patients admitted to Internal Medicine Departments, and rec-
ommended tromboprophylaxis is still under-used, in particular
in some patients groups. Further efforts are needed to better
define risk profile and to optimise prophylaxis in the heterogen-
eous setting of medical inpatients.
Thromb Haemost 2009; 101: 893–901
Estimates of the prevalence of VTE in medical inpatients are
essentially based on the results of clinical trials on thrombopro-
phylaxis (4, 6, 9–16). However, these studies differ as for patients
characteristics, and their population is not necessarily represen-
tative of the real-world, due to selection criteria; moreover, re-
sults generally refer to asymptomatic instrumentally-detected
deep venous thrombosis (DVT), whose clinical relevance is still
unclear (17). Important information on the issue of VTE occur-
rence in medical patients have been recently provided (2, 7), but
*
Members of the GEMINI Study Group are listed at the end of paper.
Financial support:
The GEMINI study was supported by a research grant by GlaxoSmithKline, Italy, without
involvement in any of the study procedures.
Received: November 14, 2008
Accepted after minor revision: January 23, 2009
Prepublished online: March 11, 2009
doi:10.1160/TH08-11-0748
 Gussoni et al. Real-world incidence and prophylaxis of VTE in Internal Medicine

Table 1: Characteristics of patients at baseline. * Creatinine clear- to our knowledge no large, prospective studies are available to
ance was calculated using Cockcroft-Gault formula ** COPD, chronic address the impact of symptomatic VTE in unselected patients
obstructive pulmonary disease.
hospitalised in Internal Medicine (IM) wards.
Gender Several trials and consensus statements support the effective-
ness of VTE prophylaxis for medical inpatients (1, 18, 19). How-
Male / Female (%) 45.4 / 54.6
ever, audits from a number of countries have described under-
Age prescription of thromboprophylaxis in this setting (20), with
Mean ± SD 71.0 ± 15.9 rates ranging from less than 30% to about 60% (21–28). Unfor-
≤ 60 years (%) 21.0 tunately, little is known on determinants of choice to administer
or not prophylaxis in this complex category of patients. Further,
61–75 years (%) 30.9
the great majority of findings concerning antithrombotic pro-
76–90 years (%) 41.9 phylaxis have been collected prior to publication of the 2004 ver-
> 90 years (%) 6.2 sion of the American College of Chest Physicians (ACCP)
Body mass index Guidelines, and more recent data on these issues are therefore of
Mean ± SD 25.7 ± 5.2 potential clinical interest
< 18.5 (%) 5.6
18.5–24.9 (%) 42.5
Materials and methods
25.0–29.9 (%) 33.2 GEMINI is a prospective observational study in a cohort of uns-
≥ 30 (%) 18.7 elected consecutive patients admitted to IM Units during the
period March-September 2006. Data were collected in 27 Italian

Downloaded by: Lund University Libraries. Copyrighted material.

Creatinine clearance (ml/min) *
Departments of IM, under scientific coordination by the FADOI
Mean ± SD 66.2 ± 46.4 (Italian Federation of Internal Medicine) Study Centre.
< 30 (%) 15.1 The primary study aim was to evaluate the frequency of clini-
30–50 (%) 24.3 cally overt VTE, with onset of symptoms occurring later than 48
> 50 (%) 60.6 hours after admission to IM ("hospital-acquired VTE") (2).
Symptomatic VTE had to be objectively documented by instru-
Diseases at entry in internal medicine (%)
mental diagnostic work-up according to centre-specific pro-
Diabetes 22.3 cedures.
Cerebrovascular disease 19.6 A second objective of the study was to evaluate the attitude of
Congestive heart failure 16.5 physicians towards the use of antithrombotic prophylaxis, and
Cancer 16.4
more specifically to gather information on clinical determinants
of the choice to prescribe or not thromboprophylaxis.
COPD exacerbation ** 16.3 At hospital admission, information on medical history, prin-
Moderate – severe renal failure 8.1 cipal diagnosis and active concomitant conditions, risk factors
Peripheral vascular disease 6.9 for VTE, and previous or acute VTE events were collected. At
Moderate – severe liver 6.8 discharge data were recorded concerning occurrence of VTE,
insufficiency antithrombotic prophylaxis, and patients outcome. Prophylactic
Dementia 6.0 doses of heparin were defined as low when ≤3,400 antiXa IU/
daily for low-molecular-weight heparin (LMWH) or 5,000 U bid
Acute myocardial infarction 3.0
for unfractionated heparin (UFH) (1, 29), and as high when
Ulcer disease 2.7 >3,400 antiXa IU/daily for LMWH or 5,000 U tid for UFH. For
Connective tissue disease 1.5 patients experiencing "hospital-acquired VTE", a three-month
Inflammatory bowel disease 0.8 follow-up visit was scheduled, to evaluate clinical evolution.
The study was approved by the Ethic Committees of all par-
Acute arthritis 0.7
ticipating centres. Informed consent was obtained from all pa-
Other 55.5 tients. Data collection and management were performed by
Other characteristics (%) using a web-based system (QBGROUP SpA, Italy), allowing
Chronic bed resting 8.9 real-time control of collected data.
Previous venous 2.5
thromboembolism Statistical analysis
Recent surgery 0.9
A sample size of 4,500 patients was considered appropriate to
document a reliable estimate of the incidence of "hospital-ac-
Known thrombophilia 0.1 quired VTE" (primary end-point), by hypothesising an event rate
Hormone replacement therapy 0.1 of 0.75% (2) with 95% confidence intervals (CI) of the estimate
equal to the event rate ± 0.25%.
The association between the use of antithrombotic prophy-
laxis and the presence of risk factors for VTE in the overall study

894
 Gussoni et al. Real-world incidence and prophylaxis of VTE in Internal Medicine
Figure 1: Disposition of patients in respect to occurrence of venous thromboembolism (VTE). DVT, deep venous thrombosis; PE, pul-
monary embolism; IM, Internal Medicine.
population was evaluated by means of a multivariable logistic re-
gression analysis, excluding those patients with known or newly
diagnosed VTE ad admission. Covariates for this analysis in-
cluded age (>75 vs. ≤75), prior VTE, recent (< 3 months) major
surgery, obesity (BMI ≥30), congestive heart failure (CHF,
NYHA class II-IV), acute myocardial infarction, chronic ob-
structive pulmonary disease (COPD), inflammatory bowel dis-
ease, cancer, hemi- or paraparesis, hemi- or paraplegia, fever of
infectious origin, bed rest (chronic bedridden patients, or bed
rest > 3 days in the four weeks prior to study inclusion, or > 3
days during hospital stay). No systematic screening was sched-
uled for thrombophilia; as a consequence known thrombophilic
state was not considered as covariate due to the extremely low
frequency (<0.1%). In addition, we evaluated if an increasing
number of concomitant risk factors induced a progressively
higher use of prophylaxis.
Subsequently, by means of logistic regression analysis, we
assessed whether the administration of thromboprophylaxis was
related to those categories of patients for which the ACCP
Guidelines 2004 recommend prevention (31). A cut-off of more
than three days for bed rest, to define “patients confined to bed”,
was selected on the basis of findings in surgical setting (32), and
according to recent papers in medical inpatients (2, 4).
895
Downloaded by: Lund University Libraries. Copyrighted material.
Finally, we evaluated the role of a number of clinical factors
on the decision to not prescribe thromboprophylaxis in the sub-
group of patients with indication to prevention according to
ACCP. Covariates selected to assess this issue, by means of a spe-
cific multivariable analysis, were: age ≤75 years, bed rest ≤3
days during hospital stay, no concomitant diseases, presence of
contraindication to pharmacological prophylaxis (platelet count
< 50,000, haemoptysis / haematemesis, hepatic impairment, ac-
tive ulcer).
Odds ratios (ORs) and 95% CIs were reported with two-
tailed probability values. A p value ≤0.05 was considered statis-
tically significant.
Statistical analysis was carried-out using SAS® software,
version 9.1.3.
Results
Study population
A total of 4,846 patients were included in the study, and details
on their characteristics are specified in Table 1. The majority of
patients (54.9%) had active co-morbidities at the time of hospital
admission: in 30.4%, 16.0% and 8.6% of patients two, three or
more than three concomitant diseases were present, respectively.
 Gussoni et al. Real-world incidence and prophylaxis of VTE in Internal Medicine

Table 2: Characteristics of patients with "hospital-acquired" venous thromboembolism (VTE).

Age Gender Diseases / Number and type Prophylaxis Type of VTE Timing of VTE Outcome
symptomatology at of risk factors for and type after admission (3-month
admission VTE to IM – days follow-up)
Case # 1 79 M Diabetes, haemoptysis 2 – age, No PE 4 No recurrence
of unknown origin bed rest
Case # 2 89 F Anemia 1 – age, cancer No Proximal DVT 10 Death
Case # 3 84 M Diabetes, chronic hepa- 1 – age Yes Proximal DVT 2 No recurrence
titis, cerebrovascular LMWH HAD
disease
Case # 4 53 F Anemia No No Distal DVT 14 No recurrence
Case # 5 64 M Sepsis 3 – obesity, bed rest, No Proximal DVT + PE 9 Death
fever
Case # 6 81 F CHF 3 – age, obesity, CHF No Proximal DVT 2 No recurrence
Case # 7 90 M Cerebrovascular disease, 2 – age, No Proximal DVT 3 No recurrence
chronic renal failure bed rest
Case # 8 67 M Knee pain of No No Distal DVT 2 No recurrence
unknown origin
Case # 9 73 F Sepsis 3 – obesity, bed rest, No Distal DVT 2 No recurrence
fever
Case # 10 35 M Urinary tract infection 2 – bed rest, fever No Distal DVT 2 No recurrence

Downloaded by: Lund University Libraries. Copyrighted material.

Case # 11 85 F Arterial hypertension, 4 – age, obesity, bed Yes Proximal DVT + PE 2 No recurrence
lower limb erysipela rest, fever LMWH HAD
Case # 12 83 F Arterial hypertension, 2 – age, No PE 3 No recurrence
PAD, hemoptysis of bed rest
unknown origin
Case # 13 88 F Cancer 4 – age, Yes Proximal DVT 11 Death
bed rest, cancer, fever LMWH HAD
Case # 14 97 F COPD, hypothyroidism 4 – age, Yes Proximal DVT 5 No recurrence
bed rest, COPD, fever LMWH LAD
Case # 15 52 M AIDS 1 – fever No Proximal DVT + PE 13 No recurrence
Case # 16 75 F Rheumatoid arthritis, 3 – bed rest, fever, Yes Proximal DVT 8 No recurrence
hemiparesis, melaena hemiparesis LMWH LAD
Case # 17 67 M COPD, cerebrovascu- 3 – COPD, previous No PE 7 No recurrence
lar disease, liver insuffi- VTE, fever
ciency
Case # 18 90 F Dyspnea of unknown 2 – age, Yes Proximal DVT 10 Death
origin bed rest LMWH LAD
Case # 19 91 F Diabetes, cerebrov- 3 – age, No Proximal DVT 12 Death
ascular disease, dysp- bed rest,
nea of unknown origin fever
Case # 20 81 F Acute arthritis 3 – age, obesity, Yes Proximal DVT + PE 6 No recurrence
fever LMWH LAD
Case # 21 71 F Suspect monoclonal 1 – fever No Proximal DVT 8 Death
gammopathy, renal
failure
Case # 22 84 F Hemiparesis, anemia 3 – age, No Proximal DVT 7 No recurrence
bed rest, hemiparesis
Case # 23 75 M Diabetes, PAD, 2 – obesity, bed rest Yes Proximal DVT 16 No recurrence
asthenia LMWH HAD
Case # 24 75 F Diabetes, acute 2 – obesity, previous Yes Proximal DVT 7 No recurrence
arthritis VTE LMWH LAD
Case # 25 79 M Cancer, diabetes, 3 – age, No PE 12 Death
cerebrovascular bed rest, cancer
disease, melaena
Case # 26 72 M Cancer 2 – bed rest, cancer No Proximal bilateral 6 No recurrence
DVT
VTE, venous thromboembolism; IM, Internal Medicine; DVT, deep venous thrombosis; LMWH, low-molecular-weight heparin; HAD, high antithrombotic dose; PE, pulmonary embolism; CHF, congestive heart
failure; PAD, peripheral arterial disease; COPD, exacerbation of chronic obstructive pulmonary disease; AIDS, acquired immunodeficiency syndrome; LAD, low antithrombotic dose.

896
 Gussoni et al. Real-world incidence and prophylaxis of VTE in Internal Medicine

In 15.2% of cases, patients were transferred to IM Department Antithrombotic prophylaxis

from other Units, after a mean (± SD) stay of 4.5 ± 8.1 days
(range 0–69).
VTE events
As summarised in Figure 1, 31 patients (0.64%) were admitted to
hospital while having an already confirmed diagnosis of VTE.
Of 165 patients referring to IM with symptoms of VTE, 105 had
an instrumentally confirmed diagnosis (2.18%). Moreover, dur-
ing hospital stay, VTE occurred in 41 patients free of symptoms
at admission (0.87%). Overall, VTE was therefore registered in
177 (3.65%) patients enrolled in the study (114 DVT, 48 pulmon-
ary embolism [PE], 15 DVT+ PE).
Of these, 26 cases (0.55%) had an onset of symptoms after
more than 48 hours from admission ("hospital-acquired VTE",
primary study end-point). In Table 2 the main characteristics of
these patients are detailed. Median timing of occurrence of "hos-
pital-acquired VTE" was seven days (range 2–16 days). PE was
diagnosed in eight patients (4 isolated, 4 combined with DVT).
All DVT occurred in lower limbs (82% proximal). In the major-
ity of cases (n=17, 65.4%), patients did not receive thrombopro-
During hospital stay thromboprophylaxis was administered in
41.6% of patients. Mean duration of in-hospital prophylaxis was
11.4 ± 10.1 days, well-matching the mean length of hospital stay
(10.9 ± 10.7 days). LMWH was the most widely used agent
(31.9%, in 12.9% at low doses, in 16.0% at high doses and in
3.0% at full anticoagulant doses), followed by oral anticoagu-
lants (8.7%), UFH (0.8%), and other methods (1.0%).
Bed rest appeared as the strongest determinant of the use of
prophylaxis (Fig. 2): 32.3% of the patients had prolonged immo-
bilization, and 68.5% of them received antithrombotic prophy-
laxis. In addition, we observed that the number of concomitant
risk factors for VTE significantly influenced the choice of pre-
scribing thromboprophylaxis (OR 2.00, 95% CI 1.86–2.14, p <
0.001, at univariate analysis). This was the case for each of the
risk factors analysed; as an example, Figure 3 shows the trend in
patients with cancer.
Globally, 40.9% of the patients had indication to antithrom-
botic prophylaxis according to the 2004 guidelines of the ACCP,
and 58.4% of them received it during hospital stay. Detailed data
on the adherence to ACCP recommended strategies are reported

Downloaded by: Lund University Libraries. Copyrighted material.

phylaxis; a contraindication to pharmacological prophylaxis was in Table 3, together with the results of a multivariable analysis as-
present in five cases. No recurrent symptomatic VTE was rec- sessing the choice of adopting or not thromboprophylaxis in
orded; however, a high rate of all-cause mortality (26.9%) oc- these specific categories of patients. Moreover, by means of a
curred in these patients, at three-month follow-up. specific multivariable analysis, we observed that bed rest ≤3

2.53

Figure 2: Multivariable regression analysis to correlate known risk factors for venous thromboembolism (VTE) and prescription of
antithrombotic prophylaxis during hospital stay. Levels of statistical significance (■) are p < 0.05 for obesity and fever, and p<0.001 for age,
previous VTE, CHF, COPD, cancer, and hemi-paraparesis / hemi-paraplegia. CI, confidence interval; CHF, congestive heart failure; MI, myocardial in-
farction; COPD, chronic obstructive pulmonary disease; IBD, inflammatory bowel disease.

897
 Gussoni et al. Real-world incidence and prophylaxis of VTE in Internal Medicine
Figure 3: Percentage of antithrombotic prophylaxis adminis-
tration in cancer patients, according to the number of addi-
tional known risk factors for venous thromboembolism (VTE).
days, contraindication to pharmacological prophylaxis, and age
≤75 years were significantly related to the decision of not prescrib-
ing thromboprophylaxis despite in these patients prophylaxis
would have been recommended by ACCP guidelines (Fig. 4).
Within the group of patients without recommendation for pro-
phylaxis according to ACCP, 29.9% of them actually received it
(22.4% heparins, 6.1% oral anticoagulant, 1.4% other methods).
After hospital discharge, extended use of thromboprophyla-
xis was reported in 21.1% of patients (9.6% oral anticoagulant,
4.4% low-dose heparin, 4.0% high-dose heparin, 0.8% heparin
at full anticoagulant dose, 2.3% antiplatelet agents or physical
methods).
Contraindications to pharmacological prophylaxis because
of bleeding risk were present in 8.7% of patients in our overall
population, and in 8.0% in the subgroup of patients with indi-
cation to prophylaxis according to the ACCP.
898
Discussion
Clinical relevance of VTE in medical inpatients has been known
for a long time; however, indications for thromboprophylaxis in
this setting present substantial heterogeneity even today, hence
there is a need for up-to-date data on the epidemiology, risk pro-
file and prevention of VTE in these patients.
Results from our study confirm that VTE is a quite common
finding in medical inpatients, with an overall rate of 3.65%. If re-
lated to the high number of patients hospitalised in IM, this
prevalence accounts for a relevant burden of VTE in clinical
practice. More specifically, “hospital-acquired VTE” occurred
in 0.55% of patients. This finding seems coherent with data on
symptomatic VTE complicating medical admission, coming
from observational studies (2, 33) and randomised trials in this
setting (4, 6, 15). Moreover, this percentage is within the CIs of
the estimate of event rate we hypothesised in planning this study,
and may therefore be considered reliable from a statistical point
of view.
Audits of medical inpatients indicate that thrombosis preven-
tion is not systematically applied, or well-grounded on risk fac-
Downloaded by: Lund University Libraries. Copyrighted material.
tor assessment (20, 21). Percentages of use of prophylaxis in our
study are within the range of previously reported data, very simi-
lar to those recently published in a large international survey, as
for Western countries (34), and slightly higher than estimated for
Italy (35). In this global percentage it has to be considered that
around 30% of patients without recommendation on the basis of
guidelines by ACCP actually received prophylaxis, being per-
ceived at risk for VTE by the attending physician. On the other
side, data from GEMINI confirm that application of thrombo-
prophylaxis in at-risk medical patients is suboptimal, since more
than 40% of patients with indication to prophylaxis according to
ACCP, did not receive it, at a similar extent than previously re-
ported (26). There are a number of possible explanations for
being the implementation of thromboprophylaxis challenging in
this setting. First, medical patients typically have a number of
concomitant active conditions, and physicians may focus on
Figure 4: Potential
predictors for not
prescribing prophylaxis in
patients at high risk of
venous thromboembolism
according to the ACCP
guidelines. Percentages of
patients with and without
prophylaxis, and results of a
multivariable logistic regres-
sion analysis (* p < 0.01).
 Gussoni et al. Real-world incidence and prophylaxis of VTE in Internal Medicine

Table 3: Attitude towards antithrombotic prophylaxis in those categories for which prophylaxis was recommended by ACCP (31). ∗
The sum of percentages of the frequencies and of the different type of prophylaxis may exceed the total values due to the presence of multiple risk
factors or combined therapies in the same patient. # Results of a multivariable regression analysis performed to evaluate association between cat-
egory and prescribing of antithrombotic prophylaxis.

CHF COPD AMI Bed rest + Bed rest Bed rest + Bed rest + Bed rest +
previous VTE + cancer fever neurol. disease IBD
Frequency (%)∗ 17.0 16.6 3.1 0.9 6.3 6.9 3.6 0.2
Prophylaxis during 65.5 52.1 48.6 82.4 55.2 74.6 73.2 75.0
hospital stay (%)
Type of prophylaxis (%, absolute) ∗
OA 20.7 10.5 9.3 26.5 2.8 8.1 10.6 -
UFH 1.5 1.2 1.4 - 0.8 1.1 1.4 -
LMWH LAD 16.3 15.7 10.7 11.8 17.2 25.6 29.9 30.0
LMWH HAD 24.1 21.2 19.2 32.4 30.3 35.4 28.4 30.0
LMWH AD 3.9 2.9 7.1 11.8 4.1 3.7 5.1 15.0
Other 0.9 0.8 2.9 - - 0.7 1.4 -
OR # 3.76 1.46 1.46 4.38 1.77 4.74 5.01 3.69
(95% CI) (3.14–4.50) (1.22–1.75) (0.99–2.15) (1.72–11.16) (1.34–2.34) (3.54–6.34) (3.39–7.40) (0.66–20.73)
p value < 0.001 < 0.001 = 0.05 < 0.001 < 0.001 < 0.001 < 0.001 NS

Downloaded by: Lund University Libraries. Copyrighted material.

Prophylaxis after 36.1 24.5 20.6 67.6 24.3 35.1 42.8 28.6
discharge (%)
CHF, congestive heart failure; COPD, exacerbation of chronic obstructive pulmonary disease; AMI, acute myocardial infarction; VTE, venous thromboembolism; IBD, inflammatory bowel disease; OA, oral anti-
coagulant; UFH, unfractionated heparin; LMWH, low-molecular-weight heparin; LAD, low antithrombotic dose; HAD, high antithrombotic dose; AD, anticoagulant dose; Other, heparinoids, fondaparinux, antipla-
telets, physical methods; OR, odds ratio; NS, not significant.

treatment of the illnesses on presentation rather than on preven-
tion of potential complications. Complexity of medical inpa-
tients, and their heterogeneity as for immobilisation, concomi-
tant diseases, age etc., do not favour systematic application of
prophylaxis. Furthermore, physicians may be worried by the risk
of bleeding due to pharmacological prevention (both for under-
lying diseases and required treatments), while having perception
of low efficacy of alternative methods of thromboprophylaxis
(36–38). Our study was not designed to specifically assess the
safety of antithrombotic prophylaxis; however, neither fatal nor
other major bleeding episodes possibly related to prophylaxis
were reported in our population. Finally, risk stratification tools
to support decision-making have been proposed (39–43), but
they are poorly validated and at present not used in clinical prac-
tice.
In the GEMINI study, the majority of known risk factors for
VTE were predictors of use of thromboprophylaxis (and bed rest
appeared as the strongest determinant of prophylaxis), so docu-
menting a quite appropriate qualitative selection of patients to be
treated. As an exception to this trend, patients with cancer were
less likely to receive prophylaxis; this attitude has been pre-
viously reported (27, 44), possibly related to the absence of clear
guidelines for cancer patients (if not post-surgery or bedridden),
or physicians perception that these patients are particularly com-
plex or at high risk of bleeding.
Antithrombotic prophylaxis had not been used in 65.4% of
patients who developed "hospital-acquired VTE". This finding is
quite similar to the results from the DVT FREE trial (8). We can
not speculate on this issue, however we can not rule out that a
broader use of prophylaxis could have offered an opportunity to
significantly reduce the burden of VTE. On the other side, 9/26
patients experienced VTE in spite of applied prophylaxis. The
impact of failed rather than omitted prophylaxis has been pre-
viously reported (45), and it is well known that thromboprophy-
laxis may significantly reduce but not eliminate the risk of VTE.
However, it is possible that in a few cases low-dose heparin has
been not sufficient for an adequate protection (namely, 5 patients
received LMWH at low-antithrombotic dose, ≤3,400 antiXa IU/
daily) so claiming for a greater attention to drug regimes. Better
definition and appropriateness of prophylaxis are critical issues
in medical inpatients, who are characterised by a complex risk
profile, and frequent presence of bleeding diathesis or conditions
such as severe renal insufficiency or obesity for which concerns
have been raised on the type, dosing and monitoring of heparins
(46, 47). In this perspective, of interest could be the availability
of antithrombotics with potential to be used at a fixed, effective
dosage (4), so simplifying treatment’s choice. As a further find-
ing, non-pharmacological methods of prophylaxis, potentially
useful in patients with contraindications to drugs because of
bleeding risk, were very rarely used.
Our study, due to its design, may have some intrinsic limi-
tations. We required from the study participants that care was
performed according to institutional practices, but we can not
rule out that attitude towards thrombophylaxis was higher than
routinely done, as well as that an increased attention towards
symptoms and therefore diagnosis of VTE occurred. These two
aspects have probably compensated each other, thus making the
observed rate of "hospital-acquired VTE" clinically reliable. On
the other side, we can not rule out that figures concerning use of
prophylaxis were over-estimated if related to routine practice.

899
 Gussoni et al. Real-world incidence and prophylaxis of VTE in Internal Medicine
Furthermore, in a number of patients oral anticoagulants or anti-
platelet agents were considered by physicians as prophylaxis for
VTE, even if they were chronically and primarily used for other
purposes.
However, our study has some strengths which may make its
results valid. First, we included a high number of unselected con-
secutive medical inpatients, and therefore epidemiology of VTE
and information concerning attitude towards thromboprophyla-
xis are probably more adherent to real-world than findings from
randomised trials, where strict patients selection criteria are ap-
plied. As an example, our study included significant percentages
of patients with obesity or renal failure (Table 1), as well as pa-
tients receiving treatments active on haemostasis, and who are
usually excluded from randomised trials on thromboprophyla-
xis. Further, the prospective design, and the use of a standard-
ised, web-based, data collection, may have reduced the potential
for incomplete recording of data inherent to retrospective audits.
Finally, up to our knowledge this is the first large prospective
study in unselected medical inpatients contemporarily assessing
epidemiology of symptomatic VTE and attitude towards preven-
tion, paying particular attention to factors associated with pre-
scription or omission of thromboprophylaxis.
In conclusion, data from our real-world study confirm that
VTE is a relevant complication in patients admitted to Internal
What is known about this topic?
− Estimates of the prevalence of venous thromboembolism
(VTE) in hospitalised medical patients are essentially
based on findings from randomised clinical trials, whose
populations are not necessarily representative of the real-
world.
− Audits of medical inpatients indicate that thrombosis pre-
vention is not systematically applied, or well-grounded on
risk factor assessment. Little is known on determinants of
choice to administer or not prophylaxis in medical inpa-
tients.
What does this paper add?
− This is the first large prospective study in unselected
medical inpatients contemporarily assessing epidemiol-
ogy of symptomatic VTE and attitude towards prevention,
paying particular attention to factors associated with pre-
scription or omission of thromboprophylaxis.
− In this real-world study symptomatic VTE was a quite
common finding, being 3.65% the overall rate. More spe-
cifically, “hospital-acquired VTE” occurred in 0.55% of
patients.
− During hospital stay antithrombotic prophylaxis was
globally administered in 41.6% of patients. Less than
60% of patients for which prophylaxis was recommended
according to 2004 Guidelines by the American College of
Chest Physicians actually received it, confirming under-
use of prophylaxis, in particular in some patients group
(i.e. those with cancer).
− Bed rest appeared as the strongest determinant of prophy-
laxis.
900
Medicine Departments, and recommended thromboprophylaxis
is still under-used, in particular in some patient groups. As a gen-
eral scenario, definition of criteria for a reliable and easy-to-use
risk assessment, though a difficult goal to achieve due to pecu-
liarities of medical patients, could lead to a more systematic use
of thromboprophylaxis in this setting. Moreover, since medical
patients are often characterised by chronically active risk factors
for VTE, specific attention to prevention in outpatients, in par-
ticular after hospital discharge (48, 49), could further reduce the
burden of VTE.
Acknowledgements
We are indebted to Carlo Buniolo for project management of GEMINI; to
Erminio Bonizzoni for support in statistical analyses; to Salvatore Corrao
for contribution to study design; to Giuseppe Civardi for reviewing the
manuscript; to Irene Zaratti and Mariagrazia Riciputi for secretarial assist-
ance; and to the patients included in the study, for their trust and partici-
pation.
Abbreviations
Downloaded by: Lund University Libraries. Copyrighted material.
ACCP, American College of Chest Physicians; AD, anticoagulant dose;
BMI, body mass index; CHF, congestive heart failure; CI, confidence
interval; COPD, chronic obstructive pulmonary disease; DVT, deep ve-
nous thrombosis; HAD, high antithrombotic dose; IBD: inflammatory
bowel disease; IM, internal medicine; IU, international unit; LAD, low
antithrombotic dose; LMWH, low-molecular-weight heparin; MI, myo-
cardial infarction; NYHA, New York Heart Association; NS, not sig-
nificant; OA, oral anticoagulant; OR, odds ratio; PE, pulmonary embol-
ism; SD, standard deviation; UFH, unfractionated heparin; VTE, ve-
nous thromboembolism.
Appendix: Members of the GEMINI Study Group
A. Mazzone, F. Capelli (Internal Medicine, Hospital of Legnano); I.
Iori, M. Silingardi (Internal Medicine I, S. Maria Nuova Hospital,
Reggio Emilia); M. Mattarei, F. Pugliese (Internal Medicine, S. Maria
del Carmine Hospital, Rovereto); F. Colombo, P. Fraioli (Internal
Medicine III, Niguarda Hospital, Milano); G. Landini, L. Masotti (In-
ternal Medicine, Hospital of Cecina); A. Bulfoni, S. De Carli (Internal
Medicine, S. Maria della Misericordia Hospital, Udine); G.M. Patras-
si (Internal Medicine, Hospital of Cittadella); M. Grandi (Internal
Medicine, Hospital of Sassuolo); G. Iosa (Internal Medicine, Hospital
of Cesenatico); R. Cavaliere, S. Marengo (Internal Medicine, Maur-
iziano Hospital, Torino); A. Fontanella, P. Di Micco, D. Iannuzzo (In-
ternal Medicine, Fatebenefratelli Hospital, Napoli); R. Potì, F. Parente
(Internal Medicine, Vito Fazzi Hospital, Lecce); M. Campanini, A. Ai-
roldi (Internal Medicine, Maggiore Hospital, Novara); F. Salvati (In-
ternal Medicine, S.S. Immacolata Hospital, Guardiagrele); S. Sturbini
(Internal Medicine, Hospital of Senigallia); G. Vescovo, P. Fanton (In-
ternal Medicine, S. Bortolo Hospital, Vicenza); G. Lo Pinto, R. Poggio
(Internal Medicine, Galliera Hospital, Genova); S. Di Rosa, G. Nico-
losi (Internal Medicine, Villa Sofia-CTO Hospital, Palermo); R. Laur-
eano, G. Panigada (Internal Medicine, S.S. Cosimo e Damiana Hospi-
tal, Pescia); P. Ghiringhelli, B. Nardo (Internal Medicine, Galmarini
Hospital, Tradate); A. Sacco, G. Dentamaro (Internal Medicine, Ma-
donna delle Grazie Hospital, Matera); P. Pauletto, G. Scannapieco (In-
ternal Medicine, Cà Foncello Hospital, Treviso); C. Pedace, L. Ralli
(Internal Medicine, S. Donato Hospital, Arezzo); F. Pintus, P. Mascia
(Internal Medicine, G. Brotzu Hospital, Cagliari); R. Fariello (Internal
Medicine, Hospital of Chiari); D. Caruso, A. Margarita (Internal
Medicine, Cardarelli Hospital, Napoli); A. D’Avanzo, M. Amitrano
(Internal Medicine, S.G. Moscati Hospital, Avellino).
 Gussoni et al. Real-world incidence and prophylaxis of VTE in Internal Medicine
References
1. Geerts WH, Bergqvist D, Pineo GF, et al. Preven-
tion of venous thromboembolism: American College of
Chest Physicians Evidence-Based Clinical Practice
Guidelines (8th Ed.). Chest 2008; 133: 381S-453S.
2. Zakai NA, Wright J, Cushman M. Risk factors for
venous thrombosis in medical inpatients: validation of
a thrombosis risk score. J Thromb Haemost 2004; 2:
2156–2161.
3. Dunn AS, Brenner A, Halm EA. The magnitude of
an iatrogenic disorder: a systematic review of the inci-
dence of venous thromboembolism for general medical
inpatients. Thromb Haemost 2006; 95: 758–762.
4. Cohen AT, Davidson BL, Gallus AS, et al. Efficacy
and safety of fondaparinux for the prevention of venous
thromboembolism in older acute medical patients: ran-
domised placebo controlled trial. Br Med J 2006; 332:
325–329.
5. Lawall H, Hoffmanns W, Hoffmanns P, et al. Preva-
lence of deep vein thrombosis (DVT) in non surgical
patients at hospital admission. Thromb Haemost 2007;
98: 765–770.
6. Samama MM, Cohen AT, Darmon J-Y, et al. A com-
parison of enoxaparin with placebo for the prevention
of venous thromboembolism in acutely ill medical pa-
tients. N Engl J Med 1999; 341: 793–800.
7. Oger E, Bressolette L, Nonent M, et al. High preva-
lence of asymptomatic deep vein thrombosis on admis-
sion in a medical unit among elderly patients. The TA-
DEUS Project. Thromb Haemost 2002; 88: 592–597.
8. Goldhaber SZ, Tapson VF, DVT FREE Steering
Committee. A prospective registry of 5,451 patients
with ultrasound-confirmed deep vein thrombosis. Am J
Cardiol 2004; 93: 259–562.
9. Gallus AS, Hirsh J, Tutle RJ, et al. Small subcu-
taneous doses of heparin in prevention of venous
thrombosis. N Engl J Med 1973; 288: 545–551.
10. Belch JJ, Lowe GD, Ward AG, et al. Prevention of
deep vein thrombosis in medical patients by low-dose
heparin. Scott Med J 1981; 26: 115–117.
11. Garlund B. The Heparin Prophylaxis Study Group.
Randomised, controlled trial of low-dose heparin for
prevention of fatal pulmonary embolism in patients
with infectious diseases. Lancet 1996; 347:
1357–1361.
12. Harenberg J, Roebruck P, Heene DL. Subcutaneous
low-molecular-weight-heparin versus standard heparin
and the prevention of thromboembolism in medical in-
patients: the Heparin Study in Internal Medicine
Group. Haemostasis 1996; 26: 127–139 .
13. The International Stroke Trial (IST). A randomised
trial of aspirin, subcutaneous heparin, both, or neither
among 19435 patients with acute ischaemic stroke.
Lancet 1997; 349: 1569–1581.
14. Fraisse F, Holzapfel L, Couland JM, et al. Nadropa-
rin in the prevention of deep vein thrombosis in acute
decompensated COPD. Am J Respir Crit Care Med
2000; 161: 1109–1114.
15. Leizorovicz A, Cohen AT, Turpie AG, et al. Ran-
domized, placebo-controlled trial of dalteparin for the
prevention of venous thromboembolism in acutely ill
medical patients. Circulation 2004; 110: 874–879.
16. Diener HC, Ringelstein EB, von Kummer R, et al.
Prophylaxis of thrombotic and embolic events in acute
ischemic stroke with the low-molecular-weight-hepa-
rin certoparin: results of the PROTECT trial. Stroke
2006; 37: 139–144.
17. Lowe GDO, Sandercock PAG, Rosendaal FR. Pre-
vention of venous thromboembolism after orthopaedic
surgery: is fondaparinux an advance? Lancet 2003;
362: 504–505.
18. Dentali F, Douketis JD, Gianni M, et al. Meta-
analysis: anticoagulant prophylaxis to prevent sympto-
matic venous thromboembolism in hospitalized medi-
cal patients. Ann Intern Med 2007; 146: 278–288.
19. Själander A, Jansson J-H, Bergqvist D, et al. Effi-
cacy and safety of anticoagulant prophylaxis to prevent
venous thromboembolism in acutely ill medical inpa-
tients: a meta analysis. J Intern Med 2008; 263: 52–60.
20. Francis CW. Prophylaxis for thromboembolism in
hospitalized medical patients. N Engl J Med 2007; 356:
1438–1444.
21. Ageno W, Squizzato A, Ambrosini F, et al. Throm-
bosis prophylaxis in medical patients: a retrospective
review of clinical practice patterns. Haematologica
2002; 87: 746–750.
22. Caiafa JS, de Bastos M, Moura LK, et al. Managing
venous thromboembolism in Latin American patients:
emerging results from the Brazilian Registry. Semin
Thromb Haemost 2002; 28: 47–50.
23. Bergmann JF, Mouly S. Thromboprophylaxis in
medical patients: focus on France. Semin Thromb Hae-
most 2002; 28: 51–55.
24. Rahim SA, Panju A, Pai M, et al. Venous throm-
boembolism prophylaxis in medical inpatients: a retro-
spective chart review. Thromb Res 2003; 111:
215–219.
25. Vallano A, Arnau JM, Miralda GP, et al. Use of ve-
nous thromboprophylaxis and adherence to guideline
recommendations: a cross-sectional study. Thromb J
2004; 2: 3.
26. Chopard P, Dorffler-Melly J, Hess U, et al. Venous
thromboembolism prophylaxis in acutely ill medical
patients: definite need for improvement. J Intern Med
2005; 257: 352–357.
27. Kahn SR, Panju A, Geerts W, et al. Multicenter
evaluation of the use of venous thromboembolism pro-
phylaxis in acutely ill medical patients. Thromb Res
2007; 119: 145–155.
28. Tapson VF, Decousus H, Pini M, et al. Venous
thromboembolism prophylaxis in acutely ill hospital-
ized medical patients: findings from the International
Medical Prevention Registry on Venous Thromboem-
bolism. Chest 2007; 132: 936–945.
29. Wein L, Wein S, Haas SJ, et al. Pharmacological ve-
nous thromboembolism prophylaxis in hospitalized
medical patients. A meta-analysis of randomized con-
trolled trials. Arch Intern Med 2007; 167: 1476–1486.
30. Hirsh J, Bauer KA, Donati MB, et al. Parenteral
anticoagulants: American College of Chest Physicians
Evidence-Based Clinical Practice Guidelines (8th Ed.).
Chest 2008; 133: 141S-159S.
31. Geerts WH, Pineo GF, Heit JA, et al. Prevention of
venous thromboembolism. The Seventh ACCP Confer-
ence on Antithrombotic and Thrombolytic Therapy.
Chest 2004; 126: 338S-400S.
32. Agnelli G, Bolis G, Capussotti L, et al. A clinical-
outcome based prospective study on venous throm-
boembolism after cancer surgery – The @RISTOS pro-
ject. Ann Surg 2006; 243: 89–95.
33. Schuurman B, den Heijer M, Nijs AM. Thrombosis
prophylaxis in hospitalised medical patients: does pro-
901
phylaxis in all patients make sense? Neth J Med 2000;
56: 171–176.
34. Cohen AT, Tapson VF, Bergmann J-F, et al. Venous
thromboembolism risk and prophylaxis in the acute
hospital care setting (ENDORSE study): a multi-
national, cross-sectional study. Lancet 2008; 371:
387–394.
35. Cohen AT, Agnelli G, Anderson FA, et al. Venous
thromboembolism VTE in Europe. The number of VTE
events and associated morbidity and mortality. Thromb
Haemost 2007; 98: 756–764.
36. Alikhan R, Cohen AT, Combe S, et al. Risk factors
for venous thromboembolism in hospitalized patients
with acute medical illness. Analysis of the MEDENOX
Study. Arch Int Med 2004; 164: 963–968.
37. Kleber FX, Witt C, Vogel G, et al. Randomized
comparison of enoxaparin with unfractionated heparin
for the prevention of venous thromboembolism in
medical patients with heart failure or severe respiratory
disease. Am Heart J 2003; 145: 614–621.
38. Spyropoulos AC. Emerging strategies in the pre-
vention of venous thromboembolism in hospitalized
medical patients. Chest 2005; 128: 958–969.
39. Arcelus JJ, Candocia S, Traverso CI, et al. Venous
thromboembolism prophylaxis and risk assessment in
Downloaded by: Lund University Libraries. Copyrighted material.
medical patients. Semin Thromb Hemost 1991; 17
(Suppl 3): 313–318.
40. Thromboembolic Risk Factor THRIFT Consensus
Group. Risk of and prophylaxis for venous throm-
boembolism in hospital patients. Br Med J 1992; 305:
567–574.
41. Lutz L. Venous thromboembolism in internal medi-
cine: risk assessment and pharmaceutical prophylaxis.
Med Welt 2002; 53: 231–234.
42. Haas SK. Venous thromboembolic risk and its pre-
vention in hospitalized medical patients. Semin
Thromb Haemost 2002; 28: 577–584.
43. Cohen AT, Alikhan R, Arcelus JJ, et al. Assessment
of venous thromboembolism risk and the benefits of
thromboprophylaxis in medical patients. Thromb Hae-
most 2005; 94: 750–759.
44. Jubelirer SJ, Reyes B, Boce M, et al. Pharmacologic
thromboembolic prophylaxis for medical patients in a
tertiary care, teaching community hospital: room for
improvement. W V Med J 2008; 104: 10–14.
45. Goldhaber SZ, Dunn K, MacDougall RC. New
onset of venous thromboembolism among hospitalized
patients at Brigham and Women’s Hospital is caused
more often by prophylaxis failure than by withholding
treatment. Chest 2000; 118: 1680–1684.
46. Hirsh J, Raschke R. Heparin and Low-Molecular-
Weight Heparin. The Seventh ACCP Conference on
Antithrombotic and Thrombolytic Therapy. Chest
2004; 126: 188S-203S.
47. Mahè I, Aghassarian M, Drouet L, et al. Tinzaparin
and enoxaparin given at prophylactic doses for eight
days in medical elderly patients with impaired renal
function. A comparative pharmacokinetic study.
Thromb Haemost 2007; 97: 581–586.
48. Spencer FA, Lessard D, Emery C, et al. Venous
thromboembolism in the outpatient setting. Arch Int
Med 2007; 167: 1471–1475.
49. Goldhaber SZ. Outpatient venous thromboembol-
ism. A common but often preventable public health
threat. Arch Int Med 2007; 167: 1451–1452.