The American Journal of Surgery (2010) 199(Suppl to January 2010), S3–S10

Risk assessment as a guide for the prevention of the

many faces of venous thromboembolism
Joseph A. Caprini, M.D.

University of Chicago, Pritzker School of Medicine, Chicago, IL, USA

KEYWORDS: Abstract
Venous BACKGROUND: Approximately 900,000 cases of deep vein thrombosis and pulmonary embolism
thromboembolism; occur annually in the United States, and one-third lead to the patient’s death. A variety of surgical
Risk assessment; factors contribute to Virchow’s triad of venous stasis, vascular injury, and hypercoagulability, including
Prophylaxis; intraoperative venous distension and microvascular endothelial damage. Patients also may have indi-
Postthrombotic vidual risk factors such as a history of thromboembolism, cancer, advanced age, or a genetic trait linked
syndrome; to hypercoagulation. This article discusses recent trends in the development and validation of venous
Deep vein thrombosis; thromboembolism risk scores, including the results of a large validation study.
Pulmonary embolism DATA SOURCES: A Medline literature search was performed to identify original studies.
CONCLUSIONS: Venous thromboembolism risk scores have been developed for groups of patients
based on a few broad risk categories, but a more accurate, individualized risk score can be obtained
using a recently validated risk scoring system, which can be used to determine the type and length of
prophylaxis to administer. Further studies are under way to refine this system.
© 2010 Elsevier Inc. All rights reserved.

Venous thromboembolism (VTE), which includes deep patients discharged after abdominal surgery. The investiga-
vein thrombosis (DVT) and pulmonary embolism (PE), is a tors concluded that implementation of national VTE perfor-
serious health care problem in the United States, with an mance measures may help reduce the burden of VTE in
estimated 900,000 cases occurring annually.1 These include surgery patients.
both incident and recurrent events. VTE events continue to Approximately 300,000 deaths are attributed to PE each
be common in patients who undergo surgery, despite im- year, with one-third occurring as sudden death with no
provements in prophylaxis and perioperative and postoper- opportunity for medical intervention.1 Of the survivors,
ative surgical care.2 Postoperative VTE is the third most 15% to 18% die within 3 months of the initial event.5,6
common safety event in hospitalized patients, causing in- Another 4% develop chronic thromboembolic pulmonary
creased hospital stays, excess mortality, and higher medical hypertension within 2 years; this is a particularly high risk
costs.3 In a review of 172,320 surgical discharges, Spyro- for patients with a previous PE (odds ratio, 19.0).7 Patients
poulos et al4 found that symptomatic VTE occurred in 5% who develop pulmonary hypertension suffer significant re-
of patients who underwent orthopedic surgery and 3% of striction of activities and reduced quality of life.5
DVT, although much less likely to be fatal than PE,1 is
Dr. Caprini is on the speakers’ bureau for Covidien and sanofi-aventis; on a major clinical thrombosis that can lead to phlegmasia alba
the advisory board for ConvaTec, Covidien, Eisai, and sanofi-aventis; and a dolens (white swollen leg) or phlegmasia cerulean dolens
consultant for ConvaTec, Covidien, and sanofi-aventis.
* Corresponding author: Tel.: ⫹1-847-663-8050; fax: ⫹1-847-663-8054.
(blue swollen leg).5 Venous gangrene may set in if venous
E-mail address: jcaprini2@aol.com hypertension obstructs arterial inflow, causing occlusion of
Manuscript received July 23, 2009; revised manuscript October 1, 2009 the capillary system.5 Because up to one-half of patients

0002-9610/$ - see front matter © 2010 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjsurg.2009.10.006
S4 The American Journal of Surgery, Vol 199, No 1S, January 2010

with DVT will have an asymptomatic presentation, the suggested that it is this intraoperative venous distension that
diagnosis of DVT requires confirmatory laboratory tests, underlies the risk for DVT in patients undergoing surgery.
such as duplex ultrasound imaging or contrast phlebogra- They suggested that the venous distension is the result of
phy.5 loss of muscle tone that is caused by the muscle relaxants
A diagnosis of DVT carries the risk of additional throm- used during surgery. Muscle paralysis resulting from re-
boses. For example, patients who have had an episode of gional anesthesia also can lead to venous dilatation. These
DVT have twice the risk of new DVT compared with effects can be modified to some extent by the use of grad-
patients without a history of DVT.5 Patients with ipsilateral uated compression stockings during surgery.18 In a study of
recurrent venous thrombosis, in turn, have a 6-fold higher 40 patients undergoing surgery of the abdomen or neck, the
risk of developing postthrombotic syndrome (PTS), a com- median vein diameter in the extremity studied was 2.6 mm
plication resulting from DVT that is characterized by pain at the beginning of surgery in both the control and inter-
and leg swelling.8 An additional but rare complication of vention groups (control group, n ⫽ 20; median vein diam-
DVT is paradoxical9 stroke, which occurs when a thrombus eter, 2.6 mm; interquartile range [IQR], 2.1–3.3 mm; stock-
traverses a patent foramen ovale and enters the brain, caus- ing group, n ⫽ 20; median vein diameter, 2.6 mm; IQR,
ing an ischemic stroke.5 This in turn may lead to death or 2.1–3.7 mm). This decreased to a median vein diameter of
disability with permanent physical limitations, depending 1.6 mm (IQR, 1.3–2.8 mm) after application of a stocking,
on the location of the thrombosis. whereas vein diameter increased from 2.6 to 2.9 mm (IQR,
Patients who experience DVT often have a reduced qual- 2.3– 4.0 mm) in the control group.18
ity of life after the event, with activity restrictions because Comerota et al19 found that in patients undergoing total
of continuing symptoms.5 If leg swelling continues, the hip replacement surgery, handling of soft tissue (muscle)
patients may need elastic compression stockings to control during surgery leads to venodilation, whereas bone manip-
the swelling and reduce the risk of developing PTS.5 These ulation leads to venoconstriction. The venous dilatation that
patients also require continued anticoagulation, which car- occurs during surgery causes cracks in the endothelium,
ries with it the risk of spontaneous bleeding and the need for which provides a nidus for thrombosis as the blood coagu-
repeated blood testing.5 The use of anticoagulants also may lation system is activated. The researchers also showed that
require dietary restrictions and/or limited activity to avoid pharmacologic control of venodilation during surgery re-
injury and bleeding.5 duced postoperative DVT.19
An estimated one-fourth to one-half of patients diag- Microscopic vessel wall damage (Fig. 1),20 such as that
nosed with DVT develop PTS within 2 years.5,9,10 Risk demonstrated in patients undergoing hip and knee replace-
factors include older age, lack of adequate anticoagulation, ment surgeries, also contributes to the development of
recurrent ipsilateral thrombosis, and obesity.9 PTS should VTE.21,22 Tissue factor released from the blood vessel wall
be suspected when patients with a history of DVT experi- after injury drives thrombus formation,23 which may help
ence leg swelling, pain, or aching that increases during the explain the increased risk of VTE in patients undergoing
day but improves with bed rest.5 Patients with PTS face a surgery.
considerable reduction in quality of life,11 and up to 30% The third factor in Virchow’s triad, hypercoagulability,
have severe manifestations.12 Patients also may develop is linked to a number of factors, including certain genetic
venous ulceration, skin pigmentation, or rashes.13,14 They traits. Deficiencies of antithrombin, protein C, or protein S,
typically require multiple medical visits, extended or life- or mutations of factor V Leiden or factor II (prothrombin)
long use of compression stockings,9 and use of venoactive G20210A genes lead to hypercoagulable states.16 Although
medications.15 these genetic factors account for only a small percentage of
the total cases of VTE, more than half of all patients with
juvenile or idiopathic VTE have been identified with an
inherited thrombophilic condition.16
Risk factors for VTE
Rudolph Virchow is recognized as the first person to link
the development of VTE to the presence of at least 1 of 3 Screening for VTE
conditions: venous stasis, vascular injury, and/or hyperco-
agulability.16 Each of these factors can alter the delicate Given that VTE is the leading preventable cause of
hemostatic balance toward hypercoagulability and develop- in-hospital deaths,24 every patient should be screened before
ment of thrombosis. Several aspects of surgery can be other lesser screens are performed (bedsores, risk of falls,
linked to Virchow’s triad. nutritional evaluation, and so forth). Stated another way—
Coleridge-Smith et al17 reported in 1990 that venous every patient deserves a proper history and physical to
stasis occurs during general surgery, with veins dilating uncover any possible factors that might increase their risk of
22% to 28% in patients undergoing general anesthesia and a VTE.
surgery and up to 57% in those who also received an In 1992, the Thromboembolic Risk Factors (THRIFT)
infusion of 1 L of saline during surgery. The investigators Consensus Group identified acquired risk factors for VTE.25
J.A. Caprini VTE risk assessment S5

Figure 2 The incidence of DVT correlates with the total number

of risk factors.16 Reproduced with permission from Circulation.

VTE as those identified by THRIFT, with the addition of a

few new ones, including acute medical illness, and the
removal of smoking as a separate risk factor (Table 1).24
The incidence of VTE increases dramatically in tandem
with the number of risk factors identified in patients (Fig.
2).16,26 Most hospitalized patients have at least one risk
factor for VTE, and the most recent ACCP review of VTE
estimated that approximately 40% have 3 or more risk
Figure 1 Endothelial damage from venodilation.20 Scanning factors.24 These include fracture (hip or leg), hip or knee
electron microscope magnification, 2,500⫻; no staining applied. replacement, major general surgery, major trauma, and spi-
The venous dilatation results in cracks in the endothelium, which nal cord injury,16 as well as a history of VTE,16 thrombo-
triggers clotting as the blood comes into contact with the exposed philia,16 inflammatory bowel disease,27 postoperative infec-
collagen. Copyright Elsevier 1985. Reproduced with permission tion,24 and cancer.28 Bed rest for more than 72 hours,16,29
from The American Journal of Surgery. use of hormones,16 and impaired mobility16 are additional
risk factors.
Table 1 Selected acquired risk factors for VTE Many of these factors are not simple binary (ie, yes/no)
risks. For example, age is a significant risk factor, with the
ACCP, 200824 THRIFT, 199225 risk approximately doubling with each decade beyond age
Increasing age Increasing age 40.16,30 It is not sufficient to use a single age cut-off level to
Immobility, paresis Immobility (⬎4 d), limb paralysis define high or low risk.16 Similarly, the incidence of VTE
Previous VTE Previous VTE increases with length of surgery.31,32 In addition, Sugerman
Cancer and/or its Malignancy et al33 found higher rates of VTE in obese patients (mean
treatment Surgery (pelvis, hips, legs) body mass index, 61) who also had venous stasis syndrome;
Surgery Trauma (pelvis, hips, legs)
a simple cut-off level based on a definition of obesity would
Trauma (major or lower Obesity
not capture this increased risk. In fact, Anderson and Spen-
limbs) Varicose veins
Obesity Heart failure cer16 suggest that the association of risk of VTE and weight
Central venous Recent myocardial infarction alone is a weak one.
catheters Inflammatory bowel disease As noted earlier, hospitalized patients usually have at
Inflammatory bowel Nephrotic syndrome least 1 risk factor for VTE, and more than a third of hos-
disease Pregnancy pitalized patients have 3 risk factors or more.24 Risk factor
Nephrotic syndrome High-dose estrogen therapy weighting can be used to calculate the risk for an individual
Pregnancy and Infection patient, and the results may be used to determine several
postpartum aspects of prophylaxis, such as the length of prophylaxis
Estrogen therapy or (including out-of-hospital prophylaxis), selection of pro-
estrogen-containing
phylactic agent, timing of first dose, and the need for
oral contraceptives
combined use of physical and pharmacologic methods.
Acute medical illness
Risk assessment typically has taken 1 of 2 approaches,
group risk assessment or individual risk assessment. The
group risk assessment approach assigns patients to one of a
Sixteen years later, the most recent update of the American few broad risk categories, whereas individual risk assess-
College of Chest Physicians (ACCP) guidelines for VTE ment seeks to define risk more accurately by using individ-
prophylaxis reveals essentially the same risk factors for ualized risk scores.24 The system recommended by the 2001
S6 The American Journal of Surgery, Vol 199, No 1S, January 2010

ACCP guidelines used a group risk assessment in which the Table 2 Levels of thromboembolism risk24
type of surgery (“major” vs “minor”), age bracket, and
presence of additional risk factors were used to assign Level of risk Patient groups included
patients to 1 of 4 risk groups34; however, this was based on Low Minor surgery in mobile patients
older studies, arbitrary age cut-off levels, and inexact defi- Medical patients who are fully mobile
nitions.24 The ACCP has refined this recommendation with Moderate Most general, open gynecologic, or open
a newer one in which patients are assigned 1 of 3 VTE risk urologic surgery patients
levels based on type of surgery, patient mobility, overall Medical patients, bed rest or sick
Moderate VTE risk plus high bleeding risk
risk of bleeding, and moderate/high risk of VTE based on
High Hip or knee arthroplasty, hip fracture
the presence of additional risk factors (Table 2).24 As the surgery
investigators note, this group risk assessment approach ig- Major trauma, spinal cord injury
nores the substantial variability in patient-specific risk fac- High VTE risk plus high bleeding risk
tors, but it does take into account what they view as the
principal risk factor (surgery vs acute medical illness). This
approach is most appropriate for patients who fit the criteria
of the randomized clinical trials that were used to develop for VTE had been considered, a different decision would
the model; the investigators include a disclaimer for patient have been reached. Her risk factors included age older than
groups that have not been included in clinical trials or for 40 years, leg trauma, surgery, immobilization (she could not
types of patients who have not been tested.24 ambulate normally on crutches), leg cast, inflammatory
However, the group risk assessment approach recom- bowel disease, oral contraceptives, and obesity, a total of 8
mended by the ACCP may not be appropriate for all indi- individual factors. As discussed by Anderson and Spencer16
vidual patients.5 Out-of-hospital prophylaxis is not ad- and shown in Fig. 2, the risk of VTE in a patient with 5 or
dressed except for a few very high risk groups (major cancer more risk factors approaches 100%. Prophylaxis is, in fact,
surgery, total joint replacement).24 It may be more appro- necessary for the entire time a patient is at risk, for example,
priate to use the individual risk assessment approach to until a patient such as this one can put full weight on the leg.
identify and evaluate all possible risk factors to determine Until then, the calf muscle pump is not functioning, which
the true extent of risk for a patient.35 The ACCP guidelines, leads to venous stasis, endothelial cracks, and hypercoagu-
in fact, point out that “specific knowledge about each pa- lability because of the trapped metabolites in the leg as a
result of stasis. Providing VTE prophylaxis to this patient
tient’s risk factors for VTE” is an essential component of
for the entire 6- to 8-week period until she was walking
the decision-making process when prescribing thrombopro-
likely would have saved her life, because the fatality rate
phylaxis.24 Also, if many risk factors are present and a
owing to PE in surgical patients receiving thromboprophy-
planned procedure is based on a quality-of-life decision
laxis with unfractionated heparin or low-molecular-weight
rather than a critical medical need, the patient may come to
heparin is very low, only .15%.36 Although she still might
a different decision about whether to proceed.35 A common
have suffered a thrombosis, it would have been much less
misconception among physicians is that individual risk as-
likely to be a fatal event if she had received prophylaxis.
sessment takes longer and is more cumbersome than group
risk assessment. However, individual assessment can be
accomplished with, for example, a simple assessment form
that merely captures information from the history and phys- Clinical validation of VTE risk assessment
ical examination of the patient.
An example of the value of individual risk assessment is
tools
provided by the case of a 43-year-old woman who ruptured Several individualized VTE risk assessment models have
her Achilles tendon. She underwent a 90-minute repair of been developed and evaluated clinically. As part of a study
the tendon under general anesthesia with a 30-minute tour- of an electronic alert system, Kucher et al37 developed a risk
niquet time. She was placed in a cast and on crutches stratification procedure and followed up its use in patients
postoperatively. She was taking oral contraceptives, had a during the 90 days after hospital discharge. Eight common
body mass index greater than 30, and was on intravenous risk factors were assessed and weighted; these included
infliximab every 6 weeks for residual inflammatory bowel cancer, prior VTE, and hypercoagulability (highest risk; 3
disease. She had undergone a near-total colectomy several points each); major surgery (intermediate risk; 2 points);
years earlier. Six weeks postoperatively, she died suddenly. and several minor risk factors (advanced age, obesity, bed
Both acute and chronic pulmonary emboli and a saddle rest, hormone-replacement therapy, and oral contraceptives;
embolus were discovered at autopsy. This patient had re- 1 point each). An increased risk of VTE was defined as a
ceived no thromboprophylaxis, which agrees with the group cumulative score of 4 or more points. The researchers iden-
risk assessment approach used in the ACCP guidelines. tified a subset of patients at increased risk of VTE according
These do not recommend routine prophylaxis after Achilles to the risk stratification procedure but in whom thrombo-
tendon repair. However, if her unique individual risk factors prophylaxis had not been ordered. They randomly assigned
J.A. Caprini VTE risk assessment S7

Table 3 Anticoagulant prophylaxis safety considerations35

Factors associated with increased bleeding

Is the patient experiencing any active bleeding?
Does the patient have (or has the patient had a history
of) heparin-induced thrombocytopenia?
Is the patient’s platelet count ⬍100,000/mm3?
Is the patient taking oral anticoagulants, platelet
inhibitors (eg, nonsteroidal anti-inflammatory drugs,
clopidogrel, or salicylates)?
Is the patient’s creatinine clearance abnormal? If yes,
please insert value: ___
If the answer to any of the above questions is yes, the patient may
Figure 3 Use of a computerized reminder system reduces the not be a candidate for anticoagulant therapy, and you should consider
risk of DVT/PE.37 Adapted with permission from New England alternative prophylactic measures such as elastic stockings and/or IPC.
Journal of Medicine. Copyright © 2005 Massachusetts Medical
Society. All rights reserved.

given to patients identified as being at high risk, in this case by

1,255 of these patients to an intervention group for which an means of a risk-stratification procedure. Second, it showed that
electronic alert reminded physicians to consider thrombo- a simple computerized reminder system can significantly in-
prophylaxis during a patient’s initial hospitalization. An- crease the use of VTE prophylaxis and reduce VTE.
other 1,251 patients were assigned to a control group, for The same risk stratification used by Kucher et al37 was
whom no alert was issued. Prophylaxis was ordered for 34% evaluated by O’Shaughnessy et al38 in 27,179 patients en-
of the patients in the intervention group compared with 15% rolled in VEnous thromboembolism RegIsTrY (VERITY),
in the control group (P ⬍ .001). This simple alert produced a prospective treatment registry of patients in the United
a decrease in VTE incidence from 8% (control group) to 5% Kingdom who present to the hospital with suspected VTE.
(electronic alert group) during the 90 days after hospitaliza- They found that an increased risk score was associated with
tion (41% reduction of PE plus DVT events; hazard ratio, an approximately linear increase in the proportion of pa-
.59; 95% confidence interval, .43–.81; P ⫽ .001) (Fig. 3). tients subsequently found to have confirmed VTE. Unfor-
The study showed that VTE is reduced when prophylaxis is tunately, they also found that the Kucher et al37 model has

Figure 4 Caprini risk scoring model.40 ⴱEarlier versions of this tool have been published in 200535 and 2009.39 BMI ⫽ body mass index; COPD
⫽ chronic obstructive pulmonary disease; SVT ⫽ superficial venous thrombosis. Reproduced with permission from Annals of Surgery.
S8 The American Journal of Surgery, Vol 199, No 1S, January 2010

Table 4 Validation of risk-assessment models41

(95% confidence interval, 1.3–5.5) after correction for use
of VTE prophylaxis. The analysis by Zakai et al41 did not
Adjusted OR support 2 other models (THRIFT25 and Lutz et al44). This
OR for VTE for VTE study also identified an additional risk factor, increased
Model Cases Controls (95% CI) (95% CI)* platelet count.
Arcelus et al42 70% 56% 2.4 (1.3–4.5) 2.6 (1.3–5.5) Osborne et al45 also recommended routine VTE risk assess-
Lutz et al44 58% 54% 1.2 (.7–2.2) .9 (.5–2.0) ment using evidence-based guidelines for all patients undergo-
THRIFT26 20% 12% 1.8 (.8–4.1) 1.3 (.6–3.2) ing abdominal or pelvic surgery for cancer, and they included
CI ⫽ confidence interval; OR ⫽ odds ratio. a discussion of the Caprini model35 as modified by Bergqvist
Reproduced with permission from the Journal of Thrombosis and et al46 as one possible approach. Deheinzelin et al47 identified
Haemostasis and Wiley-Blackwell. data on risk factors for VTE and prescriptions for pharmaco-
*Adjusted for use of venous thrombosis prophylaxis (⬍1/3 of days,
1/3–2/3 of days, and ⬎2/3 of days). logic and nonpharmacologic thromboprophylaxis in 4 hospi-
tals in Sao Paulo, Brazil. They also concluded that the use of
risk assessment methods, in this instance by Caprini et al43 or
the International Union of Angiology Consensus Statement,48
helped identify patients at risk who were not receiving appro-
priate prophylaxis. The Caprini assessment tool also has been
modified for specific use in patients undergoing plastic sur-
gery49 and validated in a recent study.50 The investigators
observed a significant clinical VTE incidence in those patients
with a score of greater than 4.
Most recently, a large study conducted at the University of
Michigan was published in the Annals of Surgery on Septem-
ber 22, 2009.40 Researchers identified risk factors for 8,216
inpatients using a retrospective scoring method. Scoring was
performed by gathering available data from electronic medical
records, as well as pharmacy and billing records. Logistic
Figure 5 Correlation between risk score and proven 30-day
regression was used to calculate odds ratios for VTE within 30
VTE incidence in surgical patients.40 Reproduced with permission days after surgery for risk factors and risk level. A bivariate
from Annals of Surgery. probit model estimated the effects of risk while controlling for
adherence to prophylaxis guidelines. These patients were part
of the National Surgical Quality Improvement Program, which
poor sensitivity for those with 0 to 3 risk factors. This can included a dedicated nurse who compiled clinically evident,
be explained by the fact that only 8 factors are scored; many imaging-proven VTE events for 30 days after hospital dis-
other factors that are not measured by this approach may charge. The results of the retrospective model were compared
have been present in these patients and contributed to the with the prospective preoperative evaluations performed by
development of thrombosis. physician assistants who filled out the Caprini risk assessment
Caprini has been using a detailed individual risk assess- form for 1,470 general surgery patients admitted to the Uni-
ment model in medical and surgical patients since the late versity of Michigan Medical Center. Statistically significant
1980s and has published modified versions in 200535 and correlation between risk score and proven 30-day VTE events
2009.39 The most recent version is presented in Fig. 4.40 In was found for both models. Fig. 5 illustrates the relationship
this model, the patient fills out a health history; this is given between risk score and VTE incidence. On the basis of these
to a nurse or admitting physician who completes and scores findings, those patients with high scores may be evaluated for
the risk assessment form. Approximately 40 risk factors are postdischarge prophylaxis. The author understands that at the
listed with weights of 1 to 5 points each. The total risk factor present time this is a value judgment that needs further vali-
score then is used to group patients into 1 of 4 categories dation before a definitive recommendation can be suggested.
(low, moderate, high, and highest risk), each with a recom- The Michigan National Surgical Quality Improvement Pro-
mended prophylactic regimen.39 Additional questions ad- gram consortium is conducting another prospective study at the
dress prophylactic safety concerns about increased risks of present time in at least 5 hospitals to address this question.
bleeding (Table 3).35
Zakai et al41 evaluated 3 published VTE risk models
using a retrospective case-control study (65 cases, 123 con-
trols) from medical inpatient admissions (Table 4). They Use of risk factor score to determine
validated the model published by Arcelus et al,42 which is prophylaxis
an extension of the original Caprini et al43 assessment. This
system classified 70% of cases and 56% of controls as being A risk factor score can be calculated for any patient
at high risk for VTE at admission, with an odds ratio of 2.6 using one of the validated risk factor stratification sys-
J.A. Caprini VTE risk assessment S9

Table 5 VTE risk score as a suggested guide to thrombosis

notified that surgery should be considered carefully in light
prophylaxis* of this risk.

All moderate-risk and high-risk patients should receive UFH,

LMWH, or FXa I unless contraindicated by bleeding risk24
Scores of 2–3: IPC perioperatively and during References
hospitalization
Scores of 3–4: UFH, LMWH, FXa I, foot pump, or IPC
1. Heit JA, Cohen AT, Anderson FA Jr, et al. Estimated annual number
during hospitalization
of incident and recurrent, non-fatal and fatal venous thromboembolism
Start AC 12–24 h postoperatively
(VTE) events in the US (abstr). Blood 2005;10:910.
Scores of 5–8: AC ⫹ IPC during hospitalization and 7–10 d
2. Agnelli G. Prevention of venous thromboembolism in surgical pa-
UFH, LMWH, or FXa I tients. Circulation 2004;110(Suppl 1):IV4 –12.
Start AC 12 h preoperatively 3. Zhan C, Miller MR. Excess length of stay, charges, and mortality
Scores of ⬎8: AC ⫹ IPC during hospitalization and 30 d attributable to medical injuries during hospitalization. JAMA 2003;
UFH, LMWH, or FXa I 290:1868 –74.
AC ⫽ anticoagulation; FXa I ⫽ factor Xa inhibitor; IPC ⫽ inter- 4. Spyropoulos AC, Hussein M, Lin J, et al. Rates of symptomatic venous
mittent pneumatic compression; LMWH ⫽ low-molecular-weight hep- thromboembolism in US surgical patients: a retrospective administra-
arin; UFH ⫽ unfractionated heparin. tive database study. J Thromb Thrombolysis 2009;28:458 – 64.
*Based on the author’s analysis of the data from validated 5. Wakefield TW, Caprini J, Comerota AJ. Thromboembolic diseases.
studies.40,41,42,45,46,48 Curr Probl Surg 2008;45:844 –99.
6. Goldhaber SZ. Pulmonary embolism. N Engl J Med 1998;339:93–104.
7. Pengo V, Lensing AWA, Prins MH, et al. Incidence of chronic throm-
boembolic pulmonary hypertension after pulmonary embolism. N Engl
J Med 2004;350:2257– 64.
tems. On the basis of the relative VTE incidence shown 8. Prandoni P, Lensing AW, Cogo A, et al. The long-term clinical course
for a given score in validated studies, one can administer of acute deep venous thrombosis. Ann Intern Med 1996;125:1–7.
prophylaxis for the period that the patient is at risk. This 9. Pesavento R, Bernardi E, Concolato A, et al. Postthrombotic syn-
author’s analysis of the data from validated stud- drome. Semin Thromb Hemost 2006;32:744 –51.
ies40,41,42,45,46,48 indicates that patients with a risk score of 10. Kahn SR. The post-thrombotic syndrome: the forgotten morbidity of
deep venous thrombosis. J Thromb Thrombolysis 2006;21:41– 8.
2 to 3 should receive intermittent pneumatic compression 11. Kahn SR, Hirsch A, Shrier I. Effect of postthrombotic syndrome on
(IPC) perioperatively and during hospitalization (Table 5). health-related quality of life after deep venous thrombosis. Arch Intern
Those with high scores (⬎8) should receive anticoagulation Med 2002;162:1144 – 8.
and IPC during hospitalization and prophylaxis with unfrac- 12. Prandoni P, Lensing AWA, Cogo A, et al. The long-term clinical
tionated heparin, low-molecular-weight heparin, or a factor course of acute deep venous thrombosis. Ann Intern Med 1996;125:
1–7.
Xa inhibitor for 7 to 30 days afterward, depending on the 13. Heit JA. Venous stasis syndrome: the long-term burden of deep vein
presence of cancer, history of VTE, or ongoing risk such as thrombosis. Hosp Med 2003;64:593– 8.
infection or chemotherapy. The use of anticoagulants in 14. Bernardi E, Bagatella P, Frulla M, et al. Postthrombotic syndrome:
these high-risk individuals may not be possible for those at incidence, prevention, and management. Semin Vasc Med 2001;1:
high risk for bleeding. In those cases, IPC should be used. 71– 80.
15. Kahn SR. The Thrombosis Group of Canada guideline on the post-
It must be emphasized that this algorithm is solely the thrombotic syndrome. 2008. Available from: http://www.tigc.org/pdf/
judgment of the author and must be subjected to future post-thromboticSyndrome.pdf. Accessed: June 24, 2009.
vigorous prospective validation before definitive recom- 16. Anderson FA Jr, Spencer FA. Risk factors for venous thromboembo-
mendations can be suggested. lism. Circulation 2003;107:I9 –I16.
17. Coleridge-Smith PD, Hasty JH, Scurr JH. Venous stasis and vein
lumen changes during surgery. Br J Surg 1990;77:1055–9.
18. Coleridge Smith PD, Hasty JH, Scurr JH. Deep vein thrombosis: effect
of graduated compression stockings on distension of the deep veins of
Conclusions the calf. Br J Surg 1991;78:724 – 6.
19. Comerota AJ, Stewart GJ, Alburger PD, et al. Operative venodilation:
Physicians should perform risk assessment in all patients a previously unsuspected factor in the cause of postoperative deep vein
before any planned surgical intervention. Grouping patients thrombosis. Surgery 1989;106:301–9.
into general risk categories provides some helpful starting 20. Comerota AJ, Stewart GJ, White JV. Combined dihydroergotamine
information. However, the use of a well-designed risk as- and heparin prophylaxis of postoperative deep vein thrombosis: pro-
posed mechanism of action. Am J Surg 1985;150:39 – 44.
sessment tool that measures the total VTE risk for an indi-
21. Stamatakis JD, Kakkar VV, Sagar S, et al. Femoral vein thrombosis
vidual patient likely will prove much more helpful when and total hip replacement. BMJ 1977;2:223–5.
considering a specific patient’s need for prophylaxis and 22. Mammen EF. Pathogenesis of venous thrombosis. Chest 1992;102:
follow-up evaluation. On the basis of these results, patients 640S– 4S.
who have a positive risk profile should obtain appropriate 23. Mackman N. Role of tissue factor in hemostasis, thrombosis, and
vascular development. Arterioscler Thromb Vasc Biol 2004;24:
consultations for both prophylaxis and postsurgical fol-
1015–22.
low-up evaluation to reduce the risk of VTE. Patients who 24. Geerts WH, Bergqvist D, Pineo GF, et al. Prevention of venous
are contemplating elective quality-of-life improving proce- thromboembolism: American College of Chest Physicians Evidence-
dures but who have high scores for risk of VTE should be Based Clinical Practice Guidelines. Chest 2008;133:381S– 453S.
S10 The American Journal of Surgery, Vol 199, No 1S, January 2010

25. Thromboembolic Risk Factors (THRIFT) Consensus Group. Risk of boembolism: a VERITY registry study (abstr). Blood 2006;
and prophylaxis for venous thromboembolism in hospital patients. 108:716.
BMJ 1992;305:567–74. 39. Caprini JA. Why thromboprophylaxis fails. Vasc Dis Manage 2009;
26. Wheeler HB. Diagnosis of deep vein thrombosis. Review of clinical 6:47–51.
evaluation and impedance plethysmography. Am J Surg 1985;150:7–13. 40. Bahl V, Hu HM, Henke PK, et al. A validation study of a retrospective
27. Nguyen GC, Sam J. Rising prevalence of venous thromboembolism venous thromboembolism risk scoring method. Ann Surg 2009. Epub
and its impact on mortality among hospitalized inflammatory bowel September 22, 2009.
disease patients. Am J Gastroenterol 2008;103:2272– 80. 41. Zakai NA, Wright J, Cushman M. Risk factors for venous thrombosis
28. Kröger K, Weiland D, Ose C, et al. Risk factors for venous thrombo- in medical inpatients: validation of a thrombosis risk score. J Thromb
embolic events in cancer patients. Ann Oncol 2006;17:297–303. Haemost 2004;2:2156 – 61.
29. Kovacevich GJ, Gaich SA, Lavin JP, et al. The prevalence of throm- 42. Arcelus JI, Candocia S, Traverso CI, et al. Venous thromboembolism
boembolic events among women with extended bed rest prescribed as prophylaxis and risk assessment in medical patients. Semin Thromb
part of the treatment for premature labor or preterm premature rupture Hemost 1991;17(Suppl 3):313– 8.
of membranes. Am J Obstet Gynecol 2000;182:1089 –92. 43. Caprini JA, Arcelus JI, Reyna JJ. Effective risk stratification of sur-
30. Anderson FA Jr, Wheeler HB, Goldberg RJ, et al. A population-based gical and nonsurgical patients for venous thromboembolic disease.
perspective of the hospital incidence and case-fatality rates of deep
Semin Hematol 2001;38:12–9.
vein thrombosis and pulmonary embolism. The Worcester DVT Study.
44. Lutz L, Haas S, Hach-Wunderle V, et al. Venöse thromboembolie in
Arch Intern Med 1991;151:933– 8.
der inneren medizin: risikoeinschätzung und medikamentöse prophy-
31. Borow M, Goldson H. Postoperative venous thrombosis. Evaluation of
laxe. Med Welt 2002;53:231– 4.
five methods of treatment. Am J Surg 1981;141:245–51.
45. Osborne NH, Wakefield TW, Henke PK. Venous thromboembolism in
32. Borow M, Goldson HJ. Prevention of postoperative deep venous
cancer patients undergoing major surgery. Ann Surg Oncol 2008;15:
thrombosis and pulmonary emboli with combined modalities. Am
3567–78.
Surg 1983;49:599 – 605.
46. Bergqvist D, Caprini JA, Dotsenko O, et al. Venous thromboembolism
33. Sugerman HJ, Sugerman EL, Wolfe L, et al. Risks and benefits of
gastric bypass in morbidly obese patients with severe venous stasis and cancer. Curr Probl Surg 2007;44:157–216.
disease. Ann Surg 2001;234:41– 6. 47. Deheinzelin D, Braga AL, Martins LC, et al. Incorrect use of throm-
34. Geerts WH, Heit JA, Clagett GP, et al. Prevention of venous throm- boprophylaxis for venous thromboembolism in medical and surgical
boembolism. Chest 2001;119:132S–75S. patients: results of a multicentric, observational and cross-sectional
35. Caprini JA. Thrombosis risk assessment as a guide to quality patient study in Brazil. J Thromb Haemost 2006;4:1266 –70.
care. Dis Mon 2005;51:70 – 8. 48. Nicolaides AN, Breddin HK, Fareed J, et al. Prevention of venous
36. Haas S, Wolf H, Kakkar AK, et al. Prevention of fatal pulmonary thromboembolism. International Consensus Statement. Guidelines
embolism and mortality in surgical patients: a randomized double- compiled in accordance with the scientific evidence. Int Angiol 2001;
blind comparison of LMWH with unfractionated heparin. Thromb 20:1–37.
Haemost 2005;94:814 –9. 49. Davison SP, Venturi ML, Attinger CE, et al. Prevention of venous
37. Kucher N, Koo S, Quiroz R, et al. Electronic alerts to prevent venous thromboembolism in the plastic surgery patient. Plast Reconstr Surg
thromboembolism among hospitalized patients. N Engl J Med 2005; 2004;114:43E–51E.
352:969 –77. 50. Hatef DA, Kenkel JM, Nguyen MQ, et al. Thromboembolic risk
38. O’Shaughnessy D, Rose P, Pressley F, et al. Evaluating the perfor- assessment and the efficacy of enoxaparin prophylaxis in excisional
mance of a previously reported risk score to predict venous throm- body contouring surgery. Plast Reconstr Surg 2008;122:269 –79.