National Quality Forum Performance Measures

The NQF and TJC (formerly known as the Joint Commission on Accreditation of Healthcare Organizations) have already enacted performance measures for pneumonia, heart failure, acute myocardial infarction (MI), and other conditions. Since 2005, the NQF and TJC have been collaborating to develop national consensus performance measures for the prevention and care of VTE. The VTE performance measures will apply to all medical and surgical patients and include process measures in the areas of prevention and treatment, as well as outcome measures. After pilot‐testing a range of measures for 3 years, TJC recommended 7 candidate measures in November 2007. In May 2008, the NQF endorsed 6 of these, embracing all TJC recommendations except one relating to the use and documentation of vena cava filter quality improvement (Table 2).2

The next step is for the NQF to develop a specification manual that defines which patients should be given prophylaxis using International Classification of Diseases, 9th edition (ICD‐9) codes and identifies which interventions are appropriate for each patient population. Current clinical guidelines provide important guidance for appropriate inclusion and exclusion criteria for medical and surgical prophylaxis, as well as evidence‐based recommendations for the treatment of VTE.3, 4

SCIP

The SCIP has a stated goal of reducing surgical complications by 25% by 2010.5 To accomplish this, the SCIP is targeting improvement in 4 areas: surgical‐site infection, cardiac events, postoperative pneumonia, and VTE prophylaxis. The SCIP performance measures for VTE prophylaxis in surgical patients are as follows:

• Recommended VTE prophylaxis ordered during admission; and

• Appropriate VTE prophylaxis received within 24 hours prior to surgical incision time to 24 hours after surgery end time.

After the success seen by a core group of hospitals who volunteered to participate, all Medicare‐accredited hospitals were required to submit SCIP data beginning with discharges in the first quarter of 2007 to obtain full reimbursement from the Centers for Medicare and Medicaid Services (CMS). Institutions can gauge whether they are in compliance with the SCIP VTE measures by answering a series of yes or no questions about whether prophylaxis has been ordered and received for specific patient groups and procedures. In a recent study, almost one‐half of all surgical patients at risk of VTE did not receive recommended and timely prophylaxis as specified by the SCIP performance measures.6

In addition to the 2 enacted SCIP performance measures for VTE prophylaxis, 2 outcome measures are under development. These measures address the rate at which intraoperative or postoperative pulmonary embolism (PE; SCIP VTE‐3) and deep vein thrombosis (DVT; SCIP VTE‐4) are diagnosed during the index hospitalization and within 30 days after surgery. If implemented, these measures will capture the efficacy of thromboprophylaxis.5

Other VTE Performance Initiatives

Several professional and consumer organizations are developing standards and compiling performance data for public reporting and other purposes:

• The American Medical Association Physician Consortium for Performance Improvement (PCPI) comprises more than 100 national medical specialty and state medical societies working to identify gaps in care that can be addressed with evidence‐based medicine and formal performance measures. The PCPI has endorsed a measure requiring low‐molecular‐weight heparin (LMWH), low‐dose unfractionated heparin (UFH), adjusted‐dose warfarin, fondaparinux, or mechanical prophylaxis to be given within 24 hours prior to incision time or within 24 hours after surgery end‐time for adults undergoing a procedure for which prophylaxis is indicated.7

• The Leapfrog Hospital Quality and Safety Survey hosts a searchable web‐based database that consumers can use to compare performance among participating hospitals in specific geographic regions. The Leapfrog survey includes NQF safe practices #28 (reduce occurrence of VTE) and #29 (ensure long‐term anticoagulation is effective and safe).8

• TJC National Patient Safety Goals (NPSG) target specific improvements in patient safety by providing healthcare organizations with solutions to prevalent patient safety problems. Compliance is necessary for Joint Commission accreditation, and results are reported on the Quality Check website. NPSG Goal 3 is focused on improving the safety of medications, and Goal 3E specifically addresses patient harm associated with the use of anticoagulation therapy. The 2008 NPSG goals must be implemented by January 2009.2

• The North American Thrombosis Forum (NATF), a nonprofit organization, was recently organized to address unmet needs in North America related to VTE and other thrombotic disorders. It is designed to complement existing organizations dealing with thrombosis‐related issues, with 5 major focus areas: basic translational research; clinical research; prevention and education; public policy; and advocacy. Each month, its website (http://www.natfonline.org) features several scientific papers dealing with venous and arterial thrombosis‐related issues.

• The American Venous Forum National Venous Screening Program is a national campaign designed to increase VTE awareness and promote the importance of compliance with prophylaxis protocols.9

As different organizations work to develop performance measures for VTE, conflicting standards have emerged. Although this remains a major challenge, the NQF is attempting to develop voluntary consensus standards that will harmonize VTE performance measures across all sites of care, including the acute medical, surgical, and oncology settings. Major clinical guidelines from the American College of Physicians (ACP), American College of Chest Physicians (ACCP), the American Society of Clinical Oncology (ASCO), the National Comprehensive Cancer Network (NCCN), the European Society of Cardiology (ESC), and other organizations provide data to support standardized, evidence‐based measures for VTE.

Hospital‐Level Performance Reporting

Performance results may affect an institution's ability to contract best rates with payors, obtain full reimbursement for services, and be eligible for bonus payments. For example, pay‐for‐reporting legislation from CMS provides targeted financial incentives to improve the rates at which hospitals report data on quality measures. The current legislation stipulates that hospitals must submit performance data, including data on compliance with the 2 SCIP‐VTE measures, or lose 2% of their annual CMS payment update. For a 500‐bed hospital with 80% occupancy and 50% CMS patients, failure to report data on SCIP‐VTE measures would result in an estimated annual loss of $2.6 million.10

In 2007, the first year of the CMS pay‐for‐reporting program, 93% of hospitals met the reporting goals. As penalties for nonreporting increase, an even higher compliance rate may be expected. CMS is proposing a new system that would withhold 5% of the base operating diagnosis‐related group payment from a hospital's budget; hospitals would be required to earn this back through reporting and meeting specific performance goals. Using a phase‐in system, CMS would reimburse 2.5% in the first year for pay‐for‐reporting and 2.5% for pay‐for‐performance. Ultimately, the full 5% bonus would be based on performance results.11

Performance ratings play a central role in hospital accreditation, which is critical for negotiating terms for tiered contracting arrangements with private insurers. In addition, hospital performance rankings are becoming more publicly accessible. TJC reports hospital performance in meeting the SCIP measures on its website (http://www.qualitycheck.org), and CMS will incorporate performance measures into its public reporting system, Hospital Compare (http://www.hospitalcompare.hhs.gov). Considering the widespread availability of performance ratings and the fact that payors encourage members to consider performance results when selecting their venue of care, customer choice may increasingly become a factor in a hospital's financial viability.

Physician‐Level Performance Reporting

In new quality assessment programs, physicians will also be rewarded or penalized according to their individual performance. The CMS Physician Quality Reporting Initiative (PQRI) is a claims‐based, voluntary, pay‐for‐reporting initiative targeted to Medicare providers. The PQRI program currently pays physicians 2% of total charges for covered services in exchange for voluntary reporting, and it is moving toward results‐based reimbursement.12 The 2009 PQRI Measures List describes 186 quality measures, including 2 related to VTE:13

• Quality Measure 23: Percentage of patients aged 18 years and older undergoing procedures for which VTE prophylaxis is indicated in all patients, who had an order for LMWH, low‐dose UFH, adjusted‐dose warfarin, fondaparinux, or mechanical prophylaxis to be given within 24 hours prior to incision time or within 24 hours after surgery end‐time; and

• Quality Measure 31: Percentage of patients aged 18 years and older with a diagnosis of ischemic stroke or intracranial hemorrhage who received DVT prophylaxis by end of hospital day 2.

In the PQRI, physicians report quality measures on process and patient outcomes to CMS using G‐codes or current procedural terminology (CPT)‐II codes. Approximately one‐half of the 100,000 providers who submitted quality codes during the first PQRI reporting period (July 1 to December 31, 2007) qualified for the incentive payment, totaling $36 million.12

More stringent pay‐for‐performance initiatives that hold physicians personally accountable for performance results are being developed in the private sector. For example, the Consumer‐Purchaser Disclosure Project (CPDP) is a consumer‐advocacy group that aims to improve healthcare and lower costs by holding healthcare providers publicly accountable for their quality of treatment. The CPDP has partnered with the National Committee for Quality Assurance to develop guidelines for reporting NQF performance measures.14

VTE as a Nonreimbursable Never Event

In a program that began with hospital discharges on October 1, 2008, hospitals will not receive CMS payment for 12 selected conditions that were not present on admission and were caused by medical error. These hospital‐acquired conditions (HAC), commonly known as never events, include pressure ulcers, catheter‐associated urinary tract infections, postoperative infections, and other complications. Beginning in fiscal year 2009, CMS has added hospital‐acquired VTE following hip or knee replacement surgery as a nonreimbursable never event.15 While CMS acknowledges that prophylaxis will not prevent every occurrence of DVT/PE, they feel it is a reasonably preventable HAC.15 Similar policies are expanding to state and private payor programs that require neither the patient nor the payor to reimburse the hospital for care related to reasonably preventable complications.

Brigham and Women's Hospital, Boston, MA

In 2005, Kucher et al.17 published a landmark report illustrating the benefits of an electronic alert system in increasing thromboprophylaxis and reducing VTE rates among hospitalized patients. The randomized trial identified high‐risk patients who were not receiving prophylaxis and assigned them to the intervention group, in which the treating physician was alerted to the VTE risk (n = 1255), or to the control group, in which no alert was made (n = 1251). Compared with patients in the control arm, those in the intervention arm were more than twice as likely to receive mechanical or pharmacologic prophylaxis (14.5% vs. 33.5%) and 41% less likely to develop VTE within 90 days (P < 0.001).17

In 2008, this system was evaluated in a new cohort study to determine the ongoing effectiveness of electronic alerts in a real hospital setting.18 The following steps were taken:

• Alerts were dispatched for all high‐risk cases; and

• The responsible physician for each high‐risk patient not receiving prophylaxis was issued a single alert detailing the patient's risk and encouraging the use of thromboprophylaxis

During the study period, the use of prophylaxis increased by 50% (P < 0.001). Still, nearly two‐thirds of physicians ignored the electronic alerts.18 Thus, while computer alert systems are helpful, other strategies must be employed to further improve prophylaxis rates in high‐risk medical patients.18

Roswell Park Cancer Institute, Buffalo, NY

Roswell Park Cancer Institute (RPCI), a Comprehensive Cancer Center with 24,000 active patients, initiated an institute‐wide quality improvement initiative in 2006 to improve the rates of VTE prophylaxis for all adult inpatients.19 This initiative included efforts to improve compliance with NCCN guidelines on all medical services and follow guidelines in accordance with NCCN, surgical best practices, and published standards on all surgical services. To accomplish this objective, RPCI:

• Implemented mandatory, computerized physician order entry forms;

• Promoted VTE awareness via staff education, field in‐services, and seminars; and

• Tracked compliance with manual audits of patient charts every 3 months.

When the initiative began in the fourth quarter of 2006, the rate of NCCN‐recommended VTE prophylaxis was 61% with the medical services and 86% with the surgical services. As of the second quarter of 2008, guideline compliance had increased to 90% and 100% with the medical and surgical services, respectively. Accompanying this increase in compliance was a corresponding decrease in the incidence of VTE, from 0.39% in the fourth quarter of 2006 to 0.08% in the second quarter of 2008 (P < 0.0001). The most pronounced reductions in VTE incidence were observed within the medical services and among outpatients.19

Hartford Hospital, Hartford, CT

Hartford Hospital is an 819‐bed acute‐care community hospital with 300 designated medical beds. In an effort to improve thromboprophylaxis rates among medical patients, the pharmacy, medicine, and information technology departments collaborated to develop an alert within the computerized prescriber‐order‐entry system that reminded clinicians to assess patients for VTE risk factors and the need for prophylaxis.20 When a patient met predefined criteria for VTE risk, the message was displayed until either mechanical or pharmacologic VTE prophylaxis was an active order on the patient's treatment profile (Figure 1). The program was implemented in conjunction with an extensive educational program targeting hospital staff, pharmacists, physicians, nurse practitioners, physician assistants, and nurses.20

Figure 1

Compliance with institutional prophylaxis guidelines increased from 49% to 93% following implementation (P < 0.001). Interestingly, the initiative at Hartford Hospital was able to increase the use of mechanical prophylaxis among patients with a contradiction to pharmacologic therapy from 25% prior to the program to 100% after its implementation (P < 0.001).20

Saint Elizabeth's Hospital, Collinsville, IL

In 2008, Bauer et al.21 reported the benefits of a pharmacist‐led program for VTE prevention in Saint Elizabeth's Hospital, a 278‐bed hospital with more than 13,000 admissions per year. As part of the initiative, hospital pharmacists:

• Received daily reports of all new admissions cross‐referenced with an accounting of patients currently prescribed UFH or LMWH;

• Assessed the remaining patients at risk of VTE; and

• Placed recommendations in patient charts in the form of a bold sticker alerting the physician to the patient's risk factors, their overall risk of VTE, and treatment recommendations (Figure 2).

Figure 2

The program ran 7 days per week, involved 1 pharmacist per day, and required an average of 4 hours per day. Patients in the maternity, nursery, pediatric, and psychiatric units were excluded from the program.

The program led to a significant increase in the use of VTE prophylaxis and a significant reduction in the rate of DVT (P < 0.002).21 These findings suggest that innovative programs tailored to the needs of individual institutions can dramatically increase thromboprophylaxis rates and decrease the incidence of VTE in at‐risk hospitalized patients.

National Quality Forum Performance Measures

The NQF and TJC (formerly known as the Joint Commission on Accreditation of Healthcare Organizations) have already enacted performance measures for pneumonia, heart failure, acute myocardial infarction (MI), and other conditions. Since 2005, the NQF and TJC have been collaborating to develop national consensus performance measures for the prevention and care of VTE. The VTE performance measures will apply to all medical and surgical patients and include process measures in the areas of prevention and treatment, as well as outcome measures. After pilot‐testing a range of measures for 3 years, TJC recommended 7 candidate measures in November 2007. In May 2008, the NQF endorsed 6 of these, embracing all TJC recommendations except one relating to the use and documentation of vena cava filter quality improvement (Table 2).2

The next step is for the NQF to develop a specification manual that defines which patients should be given prophylaxis using International Classification of Diseases, 9th edition (ICD‐9) codes and identifies which interventions are appropriate for each patient population. Current clinical guidelines provide important guidance for appropriate inclusion and exclusion criteria for medical and surgical prophylaxis, as well as evidence‐based recommendations for the treatment of VTE.3, 4

SCIP

The SCIP has a stated goal of reducing surgical complications by 25% by 2010.5 To accomplish this, the SCIP is targeting improvement in 4 areas: surgical‐site infection, cardiac events, postoperative pneumonia, and VTE prophylaxis. The SCIP performance measures for VTE prophylaxis in surgical patients are as follows:

• Recommended VTE prophylaxis ordered during admission; and

• Appropriate VTE prophylaxis received within 24 hours prior to surgical incision time to 24 hours after surgery end time.

After the success seen by a core group of hospitals who volunteered to participate, all Medicare‐accredited hospitals were required to submit SCIP data beginning with discharges in the first quarter of 2007 to obtain full reimbursement from the Centers for Medicare and Medicaid Services (CMS). Institutions can gauge whether they are in compliance with the SCIP VTE measures by answering a series of yes or no questions about whether prophylaxis has been ordered and received for specific patient groups and procedures. In a recent study, almost one‐half of all surgical patients at risk of VTE did not receive recommended and timely prophylaxis as specified by the SCIP performance measures.6

In addition to the 2 enacted SCIP performance measures for VTE prophylaxis, 2 outcome measures are under development. These measures address the rate at which intraoperative or postoperative pulmonary embolism (PE; SCIP VTE‐3) and deep vein thrombosis (DVT; SCIP VTE‐4) are diagnosed during the index hospitalization and within 30 days after surgery. If implemented, these measures will capture the efficacy of thromboprophylaxis.5

Other VTE Performance Initiatives

Several professional and consumer organizations are developing standards and compiling performance data for public reporting and other purposes:

• The American Medical Association Physician Consortium for Performance Improvement (PCPI) comprises more than 100 national medical specialty and state medical societies working to identify gaps in care that can be addressed with evidence‐based medicine and formal performance measures. The PCPI has endorsed a measure requiring low‐molecular‐weight heparin (LMWH), low‐dose unfractionated heparin (UFH), adjusted‐dose warfarin, fondaparinux, or mechanical prophylaxis to be given within 24 hours prior to incision time or within 24 hours after surgery end‐time for adults undergoing a procedure for which prophylaxis is indicated.7

• The Leapfrog Hospital Quality and Safety Survey hosts a searchable web‐based database that consumers can use to compare performance among participating hospitals in specific geographic regions. The Leapfrog survey includes NQF safe practices #28 (reduce occurrence of VTE) and #29 (ensure long‐term anticoagulation is effective and safe).8

• TJC National Patient Safety Goals (NPSG) target specific improvements in patient safety by providing healthcare organizations with solutions to prevalent patient safety problems. Compliance is necessary for Joint Commission accreditation, and results are reported on the Quality Check website. NPSG Goal 3 is focused on improving the safety of medications, and Goal 3E specifically addresses patient harm associated with the use of anticoagulation therapy. The 2008 NPSG goals must be implemented by January 2009.2

• The North American Thrombosis Forum (NATF), a nonprofit organization, was recently organized to address unmet needs in North America related to VTE and other thrombotic disorders. It is designed to complement existing organizations dealing with thrombosis‐related issues, with 5 major focus areas: basic translational research; clinical research; prevention and education; public policy; and advocacy. Each month, its website (http://www.natfonline.org) features several scientific papers dealing with venous and arterial thrombosis‐related issues.

• The American Venous Forum National Venous Screening Program is a national campaign designed to increase VTE awareness and promote the importance of compliance with prophylaxis protocols.9

As different organizations work to develop performance measures for VTE, conflicting standards have emerged. Although this remains a major challenge, the NQF is attempting to develop voluntary consensus standards that will harmonize VTE performance measures across all sites of care, including the acute medical, surgical, and oncology settings. Major clinical guidelines from the American College of Physicians (ACP), American College of Chest Physicians (ACCP), the American Society of Clinical Oncology (ASCO), the National Comprehensive Cancer Network (NCCN), the European Society of Cardiology (ESC), and other organizations provide data to support standardized, evidence‐based measures for VTE.

Hospital‐Level Performance Reporting

Performance results may affect an institution's ability to contract best rates with payors, obtain full reimbursement for services, and be eligible for bonus payments. For example, pay‐for‐reporting legislation from CMS provides targeted financial incentives to improve the rates at which hospitals report data on quality measures. The current legislation stipulates that hospitals must submit performance data, including data on compliance with the 2 SCIP‐VTE measures, or lose 2% of their annual CMS payment update. For a 500‐bed hospital with 80% occupancy and 50% CMS patients, failure to report data on SCIP‐VTE measures would result in an estimated annual loss of $2.6 million.10

In 2007, the first year of the CMS pay‐for‐reporting program, 93% of hospitals met the reporting goals. As penalties for nonreporting increase, an even higher compliance rate may be expected. CMS is proposing a new system that would withhold 5% of the base operating diagnosis‐related group payment from a hospital's budget; hospitals would be required to earn this back through reporting and meeting specific performance goals. Using a phase‐in system, CMS would reimburse 2.5% in the first year for pay‐for‐reporting and 2.5% for pay‐for‐performance. Ultimately, the full 5% bonus would be based on performance results.11

Performance ratings play a central role in hospital accreditation, which is critical for negotiating terms for tiered contracting arrangements with private insurers. In addition, hospital performance rankings are becoming more publicly accessible. TJC reports hospital performance in meeting the SCIP measures on its website (http://www.qualitycheck.org), and CMS will incorporate performance measures into its public reporting system, Hospital Compare (http://www.hospitalcompare.hhs.gov). Considering the widespread availability of performance ratings and the fact that payors encourage members to consider performance results when selecting their venue of care, customer choice may increasingly become a factor in a hospital's financial viability.

Physician‐Level Performance Reporting

In new quality assessment programs, physicians will also be rewarded or penalized according to their individual performance. The CMS Physician Quality Reporting Initiative (PQRI) is a claims‐based, voluntary, pay‐for‐reporting initiative targeted to Medicare providers. The PQRI program currently pays physicians 2% of total charges for covered services in exchange for voluntary reporting, and it is moving toward results‐based reimbursement.12 The 2009 PQRI Measures List describes 186 quality measures, including 2 related to VTE:13

• Quality Measure 23: Percentage of patients aged 18 years and older undergoing procedures for which VTE prophylaxis is indicated in all patients, who had an order for LMWH, low‐dose UFH, adjusted‐dose warfarin, fondaparinux, or mechanical prophylaxis to be given within 24 hours prior to incision time or within 24 hours after surgery end‐time; and

• Quality Measure 31: Percentage of patients aged 18 years and older with a diagnosis of ischemic stroke or intracranial hemorrhage who received DVT prophylaxis by end of hospital day 2.

In the PQRI, physicians report quality measures on process and patient outcomes to CMS using G‐codes or current procedural terminology (CPT)‐II codes. Approximately one‐half of the 100,000 providers who submitted quality codes during the first PQRI reporting period (July 1 to December 31, 2007) qualified for the incentive payment, totaling $36 million.12

More stringent pay‐for‐performance initiatives that hold physicians personally accountable for performance results are being developed in the private sector. For example, the Consumer‐Purchaser Disclosure Project (CPDP) is a consumer‐advocacy group that aims to improve healthcare and lower costs by holding healthcare providers publicly accountable for their quality of treatment. The CPDP has partnered with the National Committee for Quality Assurance to develop guidelines for reporting NQF performance measures.14

VTE as a Nonreimbursable Never Event

In a program that began with hospital discharges on October 1, 2008, hospitals will not receive CMS payment for 12 selected conditions that were not present on admission and were caused by medical error. These hospital‐acquired conditions (HAC), commonly known as never events, include pressure ulcers, catheter‐associated urinary tract infections, postoperative infections, and other complications. Beginning in fiscal year 2009, CMS has added hospital‐acquired VTE following hip or knee replacement surgery as a nonreimbursable never event.15 While CMS acknowledges that prophylaxis will not prevent every occurrence of DVT/PE, they feel it is a reasonably preventable HAC.15 Similar policies are expanding to state and private payor programs that require neither the patient nor the payor to reimburse the hospital for care related to reasonably preventable complications.

Brigham and Women's Hospital, Boston, MA

In 2005, Kucher et al.17 published a landmark report illustrating the benefits of an electronic alert system in increasing thromboprophylaxis and reducing VTE rates among hospitalized patients. The randomized trial identified high‐risk patients who were not receiving prophylaxis and assigned them to the intervention group, in which the treating physician was alerted to the VTE risk (n = 1255), or to the control group, in which no alert was made (n = 1251). Compared with patients in the control arm, those in the intervention arm were more than twice as likely to receive mechanical or pharmacologic prophylaxis (14.5% vs. 33.5%) and 41% less likely to develop VTE within 90 days (P < 0.001).17

In 2008, this system was evaluated in a new cohort study to determine the ongoing effectiveness of electronic alerts in a real hospital setting.18 The following steps were taken:

• Alerts were dispatched for all high‐risk cases; and

• The responsible physician for each high‐risk patient not receiving prophylaxis was issued a single alert detailing the patient's risk and encouraging the use of thromboprophylaxis

During the study period, the use of prophylaxis increased by 50% (P < 0.001). Still, nearly two‐thirds of physicians ignored the electronic alerts.18 Thus, while computer alert systems are helpful, other strategies must be employed to further improve prophylaxis rates in high‐risk medical patients.18

Roswell Park Cancer Institute, Buffalo, NY

Roswell Park Cancer Institute (RPCI), a Comprehensive Cancer Center with 24,000 active patients, initiated an institute‐wide quality improvement initiative in 2006 to improve the rates of VTE prophylaxis for all adult inpatients.19 This initiative included efforts to improve compliance with NCCN guidelines on all medical services and follow guidelines in accordance with NCCN, surgical best practices, and published standards on all surgical services. To accomplish this objective, RPCI:

• Implemented mandatory, computerized physician order entry forms;

• Promoted VTE awareness via staff education, field in‐services, and seminars; and

• Tracked compliance with manual audits of patient charts every 3 months.

When the initiative began in the fourth quarter of 2006, the rate of NCCN‐recommended VTE prophylaxis was 61% with the medical services and 86% with the surgical services. As of the second quarter of 2008, guideline compliance had increased to 90% and 100% with the medical and surgical services, respectively. Accompanying this increase in compliance was a corresponding decrease in the incidence of VTE, from 0.39% in the fourth quarter of 2006 to 0.08% in the second quarter of 2008 (P < 0.0001). The most pronounced reductions in VTE incidence were observed within the medical services and among outpatients.19

Hartford Hospital, Hartford, CT

Hartford Hospital is an 819‐bed acute‐care community hospital with 300 designated medical beds. In an effort to improve thromboprophylaxis rates among medical patients, the pharmacy, medicine, and information technology departments collaborated to develop an alert within the computerized prescriber‐order‐entry system that reminded clinicians to assess patients for VTE risk factors and the need for prophylaxis.20 When a patient met predefined criteria for VTE risk, the message was displayed until either mechanical or pharmacologic VTE prophylaxis was an active order on the patient's treatment profile (Figure 1). The program was implemented in conjunction with an extensive educational program targeting hospital staff, pharmacists, physicians, nurse practitioners, physician assistants, and nurses.20

Figure 1

Compliance with institutional prophylaxis guidelines increased from 49% to 93% following implementation (P < 0.001). Interestingly, the initiative at Hartford Hospital was able to increase the use of mechanical prophylaxis among patients with a contradiction to pharmacologic therapy from 25% prior to the program to 100% after its implementation (P < 0.001).20

Saint Elizabeth's Hospital, Collinsville, IL

In 2008, Bauer et al.21 reported the benefits of a pharmacist‐led program for VTE prevention in Saint Elizabeth's Hospital, a 278‐bed hospital with more than 13,000 admissions per year. As part of the initiative, hospital pharmacists:

• Received daily reports of all new admissions cross‐referenced with an accounting of patients currently prescribed UFH or LMWH;

• Assessed the remaining patients at risk of VTE; and

• Placed recommendations in patient charts in the form of a bold sticker alerting the physician to the patient's risk factors, their overall risk of VTE, and treatment recommendations (Figure 2).

Figure 2

The program ran 7 days per week, involved 1 pharmacist per day, and required an average of 4 hours per day. Patients in the maternity, nursery, pediatric, and psychiatric units were excluded from the program.

The program led to a significant increase in the use of VTE prophylaxis and a significant reduction in the rate of DVT (P < 0.002).21 These findings suggest that innovative programs tailored to the needs of individual institutions can dramatically increase thromboprophylaxis rates and decrease the incidence of VTE in at‐risk hospitalized patients.

National Quality Forum Performance Measures

The NQF and TJC (formerly known as the Joint Commission on Accreditation of Healthcare Organizations) have already enacted performance measures for pneumonia, heart failure, acute myocardial infarction (MI), and other conditions. Since 2005, the NQF and TJC have been collaborating to develop national consensus performance measures for the prevention and care of VTE. The VTE performance measures will apply to all medical and surgical patients and include process measures in the areas of prevention and treatment, as well as outcome measures. After pilot‐testing a range of measures for 3 years, TJC recommended 7 candidate measures in November 2007. In May 2008, the NQF endorsed 6 of these, embracing all TJC recommendations except one relating to the use and documentation of vena cava filter quality improvement (Table 2).2

The next step is for the NQF to develop a specification manual that defines which patients should be given prophylaxis using International Classification of Diseases, 9th edition (ICD‐9) codes and identifies which interventions are appropriate for each patient population. Current clinical guidelines provide important guidance for appropriate inclusion and exclusion criteria for medical and surgical prophylaxis, as well as evidence‐based recommendations for the treatment of VTE.3, 4

SCIP

The SCIP has a stated goal of reducing surgical complications by 25% by 2010.5 To accomplish this, the SCIP is targeting improvement in 4 areas: surgical‐site infection, cardiac events, postoperative pneumonia, and VTE prophylaxis. The SCIP performance measures for VTE prophylaxis in surgical patients are as follows:

• Recommended VTE prophylaxis ordered during admission; and

• Appropriate VTE prophylaxis received within 24 hours prior to surgical incision time to 24 hours after surgery end time.

After the success seen by a core group of hospitals who volunteered to participate, all Medicare‐accredited hospitals were required to submit SCIP data beginning with discharges in the first quarter of 2007 to obtain full reimbursement from the Centers for Medicare and Medicaid Services (CMS). Institutions can gauge whether they are in compliance with the SCIP VTE measures by answering a series of yes or no questions about whether prophylaxis has been ordered and received for specific patient groups and procedures. In a recent study, almost one‐half of all surgical patients at risk of VTE did not receive recommended and timely prophylaxis as specified by the SCIP performance measures.6

In addition to the 2 enacted SCIP performance measures for VTE prophylaxis, 2 outcome measures are under development. These measures address the rate at which intraoperative or postoperative pulmonary embolism (PE; SCIP VTE‐3) and deep vein thrombosis (DVT; SCIP VTE‐4) are diagnosed during the index hospitalization and within 30 days after surgery. If implemented, these measures will capture the efficacy of thromboprophylaxis.5

Other VTE Performance Initiatives

Several professional and consumer organizations are developing standards and compiling performance data for public reporting and other purposes:

• The American Medical Association Physician Consortium for Performance Improvement (PCPI) comprises more than 100 national medical specialty and state medical societies working to identify gaps in care that can be addressed with evidence‐based medicine and formal performance measures. The PCPI has endorsed a measure requiring low‐molecular‐weight heparin (LMWH), low‐dose unfractionated heparin (UFH), adjusted‐dose warfarin, fondaparinux, or mechanical prophylaxis to be given within 24 hours prior to incision time or within 24 hours after surgery end‐time for adults undergoing a procedure for which prophylaxis is indicated.7

• The Leapfrog Hospital Quality and Safety Survey hosts a searchable web‐based database that consumers can use to compare performance among participating hospitals in specific geographic regions. The Leapfrog survey includes NQF safe practices #28 (reduce occurrence of VTE) and #29 (ensure long‐term anticoagulation is effective and safe).8

• TJC National Patient Safety Goals (NPSG) target specific improvements in patient safety by providing healthcare organizations with solutions to prevalent patient safety problems. Compliance is necessary for Joint Commission accreditation, and results are reported on the Quality Check website. NPSG Goal 3 is focused on improving the safety of medications, and Goal 3E specifically addresses patient harm associated with the use of anticoagulation therapy. The 2008 NPSG goals must be implemented by January 2009.2

• The North American Thrombosis Forum (NATF), a nonprofit organization, was recently organized to address unmet needs in North America related to VTE and other thrombotic disorders. It is designed to complement existing organizations dealing with thrombosis‐related issues, with 5 major focus areas: basic translational research; clinical research; prevention and education; public policy; and advocacy. Each month, its website (http://www.natfonline.org) features several scientific papers dealing with venous and arterial thrombosis‐related issues.

• The American Venous Forum National Venous Screening Program is a national campaign designed to increase VTE awareness and promote the importance of compliance with prophylaxis protocols.9

As different organizations work to develop performance measures for VTE, conflicting standards have emerged. Although this remains a major challenge, the NQF is attempting to develop voluntary consensus standards that will harmonize VTE performance measures across all sites of care, including the acute medical, surgical, and oncology settings. Major clinical guidelines from the American College of Physicians (ACP), American College of Chest Physicians (ACCP), the American Society of Clinical Oncology (ASCO), the National Comprehensive Cancer Network (NCCN), the European Society of Cardiology (ESC), and other organizations provide data to support standardized, evidence‐based measures for VTE.

Hospital‐Level Performance Reporting

Performance results may affect an institution's ability to contract best rates with payors, obtain full reimbursement for services, and be eligible for bonus payments. For example, pay‐for‐reporting legislation from CMS provides targeted financial incentives to improve the rates at which hospitals report data on quality measures. The current legislation stipulates that hospitals must submit performance data, including data on compliance with the 2 SCIP‐VTE measures, or lose 2% of their annual CMS payment update. For a 500‐bed hospital with 80% occupancy and 50% CMS patients, failure to report data on SCIP‐VTE measures would result in an estimated annual loss of $2.6 million.10

In 2007, the first year of the CMS pay‐for‐reporting program, 93% of hospitals met the reporting goals. As penalties for nonreporting increase, an even higher compliance rate may be expected. CMS is proposing a new system that would withhold 5% of the base operating diagnosis‐related group payment from a hospital's budget; hospitals would be required to earn this back through reporting and meeting specific performance goals. Using a phase‐in system, CMS would reimburse 2.5% in the first year for pay‐for‐reporting and 2.5% for pay‐for‐performance. Ultimately, the full 5% bonus would be based on performance results.11

Performance ratings play a central role in hospital accreditation, which is critical for negotiating terms for tiered contracting arrangements with private insurers. In addition, hospital performance rankings are becoming more publicly accessible. TJC reports hospital performance in meeting the SCIP measures on its website (http://www.qualitycheck.org), and CMS will incorporate performance measures into its public reporting system, Hospital Compare (http://www.hospitalcompare.hhs.gov). Considering the widespread availability of performance ratings and the fact that payors encourage members to consider performance results when selecting their venue of care, customer choice may increasingly become a factor in a hospital's financial viability.

Physician‐Level Performance Reporting

In new quality assessment programs, physicians will also be rewarded or penalized according to their individual performance. The CMS Physician Quality Reporting Initiative (PQRI) is a claims‐based, voluntary, pay‐for‐reporting initiative targeted to Medicare providers. The PQRI program currently pays physicians 2% of total charges for covered services in exchange for voluntary reporting, and it is moving toward results‐based reimbursement.12 The 2009 PQRI Measures List describes 186 quality measures, including 2 related to VTE:13

• Quality Measure 23: Percentage of patients aged 18 years and older undergoing procedures for which VTE prophylaxis is indicated in all patients, who had an order for LMWH, low‐dose UFH, adjusted‐dose warfarin, fondaparinux, or mechanical prophylaxis to be given within 24 hours prior to incision time or within 24 hours after surgery end‐time; and

• Quality Measure 31: Percentage of patients aged 18 years and older with a diagnosis of ischemic stroke or intracranial hemorrhage who received DVT prophylaxis by end of hospital day 2.

In the PQRI, physicians report quality measures on process and patient outcomes to CMS using G‐codes or current procedural terminology (CPT)‐II codes. Approximately one‐half of the 100,000 providers who submitted quality codes during the first PQRI reporting period (July 1 to December 31, 2007) qualified for the incentive payment, totaling $36 million.12

More stringent pay‐for‐performance initiatives that hold physicians personally accountable for performance results are being developed in the private sector. For example, the Consumer‐Purchaser Disclosure Project (CPDP) is a consumer‐advocacy group that aims to improve healthcare and lower costs by holding healthcare providers publicly accountable for their quality of treatment. The CPDP has partnered with the National Committee for Quality Assurance to develop guidelines for reporting NQF performance measures.14

VTE as a Nonreimbursable Never Event

In a program that began with hospital discharges on October 1, 2008, hospitals will not receive CMS payment for 12 selected conditions that were not present on admission and were caused by medical error. These hospital‐acquired conditions (HAC), commonly known as never events, include pressure ulcers, catheter‐associated urinary tract infections, postoperative infections, and other complications. Beginning in fiscal year 2009, CMS has added hospital‐acquired VTE following hip or knee replacement surgery as a nonreimbursable never event.15 While CMS acknowledges that prophylaxis will not prevent every occurrence of DVT/PE, they feel it is a reasonably preventable HAC.15 Similar policies are expanding to state and private payor programs that require neither the patient nor the payor to reimburse the hospital for care related to reasonably preventable complications.

Brigham and Women's Hospital, Boston, MA

In 2005, Kucher et al.17 published a landmark report illustrating the benefits of an electronic alert system in increasing thromboprophylaxis and reducing VTE rates among hospitalized patients. The randomized trial identified high‐risk patients who were not receiving prophylaxis and assigned them to the intervention group, in which the treating physician was alerted to the VTE risk (n = 1255), or to the control group, in which no alert was made (n = 1251). Compared with patients in the control arm, those in the intervention arm were more than twice as likely to receive mechanical or pharmacologic prophylaxis (14.5% vs. 33.5%) and 41% less likely to develop VTE within 90 days (P < 0.001).17

In 2008, this system was evaluated in a new cohort study to determine the ongoing effectiveness of electronic alerts in a real hospital setting.18 The following steps were taken:

• Alerts were dispatched for all high‐risk cases; and

• The responsible physician for each high‐risk patient not receiving prophylaxis was issued a single alert detailing the patient's risk and encouraging the use of thromboprophylaxis

During the study period, the use of prophylaxis increased by 50% (P < 0.001). Still, nearly two‐thirds of physicians ignored the electronic alerts.18 Thus, while computer alert systems are helpful, other strategies must be employed to further improve prophylaxis rates in high‐risk medical patients.18

Roswell Park Cancer Institute, Buffalo, NY

Roswell Park Cancer Institute (RPCI), a Comprehensive Cancer Center with 24,000 active patients, initiated an institute‐wide quality improvement initiative in 2006 to improve the rates of VTE prophylaxis for all adult inpatients.19 This initiative included efforts to improve compliance with NCCN guidelines on all medical services and follow guidelines in accordance with NCCN, surgical best practices, and published standards on all surgical services. To accomplish this objective, RPCI:

• Implemented mandatory, computerized physician order entry forms;

• Promoted VTE awareness via staff education, field in‐services, and seminars; and

• Tracked compliance with manual audits of patient charts every 3 months.

When the initiative began in the fourth quarter of 2006, the rate of NCCN‐recommended VTE prophylaxis was 61% with the medical services and 86% with the surgical services. As of the second quarter of 2008, guideline compliance had increased to 90% and 100% with the medical and surgical services, respectively. Accompanying this increase in compliance was a corresponding decrease in the incidence of VTE, from 0.39% in the fourth quarter of 2006 to 0.08% in the second quarter of 2008 (P < 0.0001). The most pronounced reductions in VTE incidence were observed within the medical services and among outpatients.19

Hartford Hospital, Hartford, CT

Hartford Hospital is an 819‐bed acute‐care community hospital with 300 designated medical beds. In an effort to improve thromboprophylaxis rates among medical patients, the pharmacy, medicine, and information technology departments collaborated to develop an alert within the computerized prescriber‐order‐entry system that reminded clinicians to assess patients for VTE risk factors and the need for prophylaxis.20 When a patient met predefined criteria for VTE risk, the message was displayed until either mechanical or pharmacologic VTE prophylaxis was an active order on the patient's treatment profile (Figure 1). The program was implemented in conjunction with an extensive educational program targeting hospital staff, pharmacists, physicians, nurse practitioners, physician assistants, and nurses.20

Figure 1

Compliance with institutional prophylaxis guidelines increased from 49% to 93% following implementation (P < 0.001). Interestingly, the initiative at Hartford Hospital was able to increase the use of mechanical prophylaxis among patients with a contradiction to pharmacologic therapy from 25% prior to the program to 100% after its implementation (P < 0.001).20

Saint Elizabeth's Hospital, Collinsville, IL

In 2008, Bauer et al.21 reported the benefits of a pharmacist‐led program for VTE prevention in Saint Elizabeth's Hospital, a 278‐bed hospital with more than 13,000 admissions per year. As part of the initiative, hospital pharmacists:

• Received daily reports of all new admissions cross‐referenced with an accounting of patients currently prescribed UFH or LMWH;

• Assessed the remaining patients at risk of VTE; and

• Placed recommendations in patient charts in the form of a bold sticker alerting the physician to the patient's risk factors, their overall risk of VTE, and treatment recommendations (Figure 2).

Figure 2

The program ran 7 days per week, involved 1 pharmacist per day, and required an average of 4 hours per day. Patients in the maternity, nursery, pediatric, and psychiatric units were excluded from the program.

The program led to a significant increase in the use of VTE prophylaxis and a significant reduction in the rate of DVT (P < 0.002).21 These findings suggest that innovative programs tailored to the needs of individual institutions can dramatically increase thromboprophylaxis rates and decrease the incidence of VTE in at‐risk hospitalized patients.

Current or Recent Prior Hospitalization

The risk of VTE is elevated among hospitalized patients. The prevalence of DVT varies across hospital specialties, reaching up to 80% in major trauma, spinal cord injury, and critical care.3 The Epidemiologic International Day for the Evaluation of Patients at Risk for Venous Thromboembolism in the Acute Hospital Care Setting (ENDORSE) study evaluated the prevalence of VTE risk factors in the acute hospital care setting.12 Among the 68,183 patients enrolled, more than one‐half (52%) were judged to be at risk of VTE. In a case‐control study examining 625 patients with a first lifetime VTE, confinement to a hospital (among other risk factors) was found to be an independent and important predictor of VTE (odds ratio [OR], 8.0; 95% confidence interval [CI], 4.514.2) (Figure 1).13 Recent hospitalization is also an important risk factor for VTE, and patients who are readmitted to the hospital should be considered moderate or high risk.13

Figure 1

Age

Patient age must be considered when assessing VTE risk. VTE is predominantly a disease of older age, and age older than 75 years is an important risk factor for the condition.11 In general, patients older than 40 years have a significantly increased risk compared with younger patients, and the risk approximately doubles with each additional decade.10 Given the aging population, the prevalence of VTE and its complications are expected to increase.

Women of childbearing age experience VTE more frequently than men of the same age, due to pregnancy and exposure to contraceptive therapy.14 This risk, however, is modest compared with the risk among older patients. After age 45 years, the incidence of VTE increases markedly for both sexes, becoming more prominent in men.14 Compared with women, men also have an increased risk of recurrent VTE.15

Despite the effect of age on VTE risk, the risk among patients younger than 40 years may be underestimated because this subgroup has not been extensively studied. For reasons that are not well‐understood, the risk of VTE associated with heart failure is higher in patients younger than 40 years, and the relative risk of PE in patients with chronic obstructive pulmonary disease (COPD) is also higher in younger patients.16, 17

Cancer and Its Treatment

Cancer patients, on average, have twice the risk of VTE compared with noncancer patients.18 This risk, however, varies considerably by cancer type. According to an assessment of nearly 41 million hospitalized patients in the National Hospital Discharge Survey (NHDS), the relative risk of VTE varied from 1.02 in patients with bladder cancer to 4.34 in patients with cancer of the pancreas.18

VTE is one of the most common complications of cancer and cancer therapy, and it is the second leading cause of death among hospitalized cancer patients.19 Molecular mechanisms underlying thromboembolic events in cancer patients include tumor cell procoagulants, inflammatory cell cytokines, mediators of platelet adhesion, and tumor‐related stasis and endothelial damage.20 The clinical implications of these processes are severe. Cancer exacerbates the natural course of VTE, increasing the risk of recurrent VTE and major bleeding, and VTE worsens the prognosis of cancer, increasing the risk of death among cancer patients.

Various cancer therapiesincluding surgery, chemotherapy, hematopoietic stem cell transplantation, and even growth factor supportalso increase the risk of VTE, in part because extrinsic factors such as surgery or chemotherapy can intensify the hypercoagulable process.18, 2123 In the NHDS, cancer patients undergoing surgery had at least twice the risk of postoperative DVT and more than 3 times the risk of PE compared with noncancer patients undergoing similar procedures.18

Cancer is an independent predictor of thromboprophylaxis failure following surgery. The @RISTOS Project found that VTE was the most common cause of death among 2373 patients undergoing general, urologic, or gynecologic surgery for cancer.24 A multivariate analysis identified 5 independent risk factors for VTE after cancer surgery:

• Previous VTE (OR, 5.98; 95% CI, 2.1316.80)

• Anesthesia 2 hours (OR, 4.50; 95% CI, 1.0619.04)

• Bed rest 4 days (OR, 4.37; 95% CI, 2.457.78)

• Age 60 years (OR, 2.63; 95% CI, 1.215.71)

• Advanced‐stage cancer (OR, 2.68; 95% CI, 1.375.24)

Cardiovascular Disease

The risk of VTE is pronounced among patients with cardiovascular disease. After stroke and coronary disease, VTE is the third most common cardiovascular disorder, and PE causes more deaths each year than myocardial infarction (MI).25 Several cardiovascular diseases, including hypertension, stroke, acute MI, and heart failure, are independently associated with VTE.10, 2627 Related disorders, including diabetes and the metabolic syndrome, also increase the risk of VTE.26, 28

Congestive heart failure (CHF) is a risk factor for VTE, and the severity of illness increases risk. In the DVT‐Free Prospective Registry, 13% of patients with ultrasound‐confirmed DVT had CHF.29 In a subgroup analysis of patients of the Prophylaxis in Medical Patients with Enoxaparin (MEDENOX) study, the incidence of VTE exceeded 20% in patients with New York Heart Association (NYHA) class IV heart failure, compared with 12% in patients with NYHA class III heart failure.30 Another study found that VTE risk increases as left ventricular ejection fraction (LVEF) decreases, with an LVEF of less than 20% associated with a VTE OR of 38.3 (95% CI, 9.6152.5).31

Infectious Disease

Acute infection may increase the relative risk of VTE by as much as 50% and is associated with VTE event rates of up to 26%.11 Acute infections may be associated with acute inflammation, adverse effects on cardiac or pulmonary function, and prolonged immobilization.30, 32, 33 Human immunodeficiency virus (HIV) patients may also have an increased risk of VTE due to a circulating lupus anticoagulant and/or the presence of acute infection.34

Obesity

The 2008 ACCP guideline update recognizes obesity, for the first time, as a risk factor for VTE.3 Obesity was 1 of the 5 most frequent comorbidities found in patients with DVT in the DVT‐Free Prospective Registry.29 It increases the risk of both incident and recurrent VTE, with every 1‐point increase in body mass index (BMI) increasing the risk of recurrent VTE by 4.4% (95% CI, 1.37.6%; P < 0.001).35

Pregnancy and Puerperium

Pregnancy, particularly the postpartum period, is associated with an increased risk of VTE in women, even though the absolute risk is small.36 Still, PE is one of the leading causes of maternal death following childbirth.10 Smoking, prior VTE, and inherited thrombophilias all increase the risk of VTE in pregnant women.10 The risk begins to rise in the first trimester, and when prophylaxis is needed, it should be started early in gestation.37

Pulmonary Disease

COPD is another risk factor for the development of VTE. COPD patients who develop VTE tend to be older, hospitalized in the intensive care unit (ICU), and on mechanical ventilation.38 In the DVT‐Free Prospective Registry, 12% of patients with ultrasound‐confirmed DVT had COPD.29

Trauma and Surgery

Injury to the body tissue, via trauma or surgery, stimulates the body's clotting mechanism and increases the risk of thromboembolic complications. During the perioperative period, the circulatory system must balance a variety of assaults: an immune response to surgical stress, prolonged immobilization during surgery and recovery, vasodilation associated with general or regional anesthesia, and hypercoagulability due to venous stasis and vascular injury.39 Renal transplant recipients have an increased risk of VTE due to a chronic hypercoagulable state.40 In surgery patients, perioperative complications such as dehydration and acute infection increase the risk of VTE beyond the risk associated with the surgical procedure itself.10

VTE risk is increased approximately 13‐fold by recent major trauma or lower‐extremity injury.13 In the absence of prophylaxis, the overall risk of VTE among patients undergoing major surgery is increased nearly 22‐fold.13 After controlling for the type of surgery, additional independent risk factors for VTE within 3 months of major surgery include:41, 42

• Obesity

• Central venous catheter placement

• Malignancy

• Smoking

• Heart failure

• Previous DVT

• Prolonged immobility

• Infection

Many surgical and medical inpatients share common risk factors, and without prophylaxis, the incidence of hospital‐acquired DVT ranges from 10% to 40% for both groups.3

Inherited or Acquired Risk Factors

VTE is a multifactorial disease, and recent evidence indicates that some heritable traits may be potent risk factors for VTE.43 Approximately 35% of patients with DVT will have at least 1 of 5 traits related to an inherited blood clotting disorder:43

• Deficiencies in the anticoagulation factors protein C, protein S, or antithrombin, or

• Mutations in the factor V and prothrombin genes, resulting in Factor V Leiden and prothrombin G20210A, respectively.

Certain inherited traits and genetic polymorphisms increase the risk of VTE by interacting with clinical risk factors such as contraceptive use, pregnancy, surgery, trauma, and cancer. One recent study found that oral estrogen therapy among women with the CYP3A5*1 allele was associated with a particularly high risk of VTE.44 Although widespread screening for inherited risk factors is not currently practical, future tools may incorporate genetic polymorphisms to more precisely identify patients who would benefit from aggressive prophylaxis.

Lifestyle Factors

Lifestyle factors have a significant effect on VTE risk. Smoking increases the risk of VTE by 20% to 30%, and a sedentary lifestyle also increases the risk of VTE.26, 45 In fact, women who exercise regularly and consume alcohol in moderation have one‐half the risk of VTE as women who have a sedentary lifestyle and drink little or no alcohol.42 For both men and women, a diet high in fruits, vegetables, and fish is associated with a lower lifetime risk of VTE.46

Medications

Medications may also increase the risk of VTE. In cancer patients with anemia, for example, the use of erythropoiesis‐stimulating agents such as recombinant erythropoietin and darbepoetin was recently shown to increase the risk of VTE by 57% (95% CI, 3187%) and increase mortality risk by 10% (95% CI, 120%).23 In addition, combination hormone replacement therapy in women is associated with a higher risk of VTE compared with estrogen monotherapy, and transdermal contraceptive systems more than double the risk of VTE compared with oral contraceptives (95% CI, 1.33.8).47, 48 Recent studies have also reported an increased risk of VTE with some psychiatric drugs, including amitriptyline, clozapine, olanzapine, and risperidone.4952

Underestimation of Risk of Clotting

VTE is often clinically silent, leading some physicians to mistakenly believe that it is rare.58 In hospitalized surgical patients, for example, the incidence of thromboembolic complications during a short postoperative stay may be low. Given that many cases of symptomatic VTE occur after hospital discharge, hospitalists and surgeons may be unaware of the true incidence of DVT.59

Overestimation of the Risk of Bleeding

Physicians may also overestimate the risk of possible side effects of prophylaxis, such as major bleeding or heparin‐induced thrombocytopenia (HIT).58 Fear of excess bleeding has been cited by physicians as a leading reason for their decision to withhold thromboprophylaxis from at‐risk hospitalized patients.60 Physicians are particularly fearful of complications among elderly patients, who are less likely to receive adequate prophylaxis than younger patients with a similar risk of VTE.61 When bleeding does occur, it rarely results in death. On the other hand, PE may account for as many as 10% of hospital deaths.9

Guideline Confusion and Complexity

Discrepancies between guidelines published by different medical societies contribute to confusion in choosing a management approach. The American Academy of Orthopedic Surgeons (AAOS), for example, describes aspirin alone as a reasonable choice for VTE prophylaxis in some patients, but the ACCP guidelines advise against the use of aspirin monotherapy.58 The cumbersome nature of multiple risk‐assessment and treatment algorithms can also be problematic.61 Furthermore, certain patient subgroups, such as those with cirrhosis, severe renal failure, and epidural catheters, have been excluded from randomized controlled trials, and the management of such patients is not straightforward.

Absence of Institutional Protocols and Information Technology Support

The lack of institution‐level guidance and support can have a detrimental effect on patient care. In a 2007 survey of 127 community hospitals, the prevalence of institutional protocols related to VTE was low: only 60% had protocols to encourage prophylaxis in at‐risk patients, 54% had guidelines to assist in appropriate drug selection, and 43% had guidelines for the dosing of prophylaxis regimens.62 A lack of systems for data collection and audit has also been identified as a barrier to the implementation of prophylaxis guidelines.57 Thus, hospitals need to adopt protocols such as:3

• Written, institution‐wide thromboprophylaxis policies

• Preprinted order forms and computer decision‐support systems

• Policies specifying responsibilities for assessing VTE risk and prescribing prophylaxis

Current or Recent Prior Hospitalization

The risk of VTE is elevated among hospitalized patients. The prevalence of DVT varies across hospital specialties, reaching up to 80% in major trauma, spinal cord injury, and critical care.3 The Epidemiologic International Day for the Evaluation of Patients at Risk for Venous Thromboembolism in the Acute Hospital Care Setting (ENDORSE) study evaluated the prevalence of VTE risk factors in the acute hospital care setting.12 Among the 68,183 patients enrolled, more than one‐half (52%) were judged to be at risk of VTE. In a case‐control study examining 625 patients with a first lifetime VTE, confinement to a hospital (among other risk factors) was found to be an independent and important predictor of VTE (odds ratio [OR], 8.0; 95% confidence interval [CI], 4.514.2) (Figure 1).13 Recent hospitalization is also an important risk factor for VTE, and patients who are readmitted to the hospital should be considered moderate or high risk.13

Figure 1

Age

Patient age must be considered when assessing VTE risk. VTE is predominantly a disease of older age, and age older than 75 years is an important risk factor for the condition.11 In general, patients older than 40 years have a significantly increased risk compared with younger patients, and the risk approximately doubles with each additional decade.10 Given the aging population, the prevalence of VTE and its complications are expected to increase.

Women of childbearing age experience VTE more frequently than men of the same age, due to pregnancy and exposure to contraceptive therapy.14 This risk, however, is modest compared with the risk among older patients. After age 45 years, the incidence of VTE increases markedly for both sexes, becoming more prominent in men.14 Compared with women, men also have an increased risk of recurrent VTE.15

Despite the effect of age on VTE risk, the risk among patients younger than 40 years may be underestimated because this subgroup has not been extensively studied. For reasons that are not well‐understood, the risk of VTE associated with heart failure is higher in patients younger than 40 years, and the relative risk of PE in patients with chronic obstructive pulmonary disease (COPD) is also higher in younger patients.16, 17

Cancer and Its Treatment

Cancer patients, on average, have twice the risk of VTE compared with noncancer patients.18 This risk, however, varies considerably by cancer type. According to an assessment of nearly 41 million hospitalized patients in the National Hospital Discharge Survey (NHDS), the relative risk of VTE varied from 1.02 in patients with bladder cancer to 4.34 in patients with cancer of the pancreas.18

VTE is one of the most common complications of cancer and cancer therapy, and it is the second leading cause of death among hospitalized cancer patients.19 Molecular mechanisms underlying thromboembolic events in cancer patients include tumor cell procoagulants, inflammatory cell cytokines, mediators of platelet adhesion, and tumor‐related stasis and endothelial damage.20 The clinical implications of these processes are severe. Cancer exacerbates the natural course of VTE, increasing the risk of recurrent VTE and major bleeding, and VTE worsens the prognosis of cancer, increasing the risk of death among cancer patients.

Various cancer therapiesincluding surgery, chemotherapy, hematopoietic stem cell transplantation, and even growth factor supportalso increase the risk of VTE, in part because extrinsic factors such as surgery or chemotherapy can intensify the hypercoagulable process.18, 2123 In the NHDS, cancer patients undergoing surgery had at least twice the risk of postoperative DVT and more than 3 times the risk of PE compared with noncancer patients undergoing similar procedures.18

Cancer is an independent predictor of thromboprophylaxis failure following surgery. The @RISTOS Project found that VTE was the most common cause of death among 2373 patients undergoing general, urologic, or gynecologic surgery for cancer.24 A multivariate analysis identified 5 independent risk factors for VTE after cancer surgery:

• Previous VTE (OR, 5.98; 95% CI, 2.1316.80)

• Anesthesia 2 hours (OR, 4.50; 95% CI, 1.0619.04)

• Bed rest 4 days (OR, 4.37; 95% CI, 2.457.78)

• Age 60 years (OR, 2.63; 95% CI, 1.215.71)

• Advanced‐stage cancer (OR, 2.68; 95% CI, 1.375.24)

Cardiovascular Disease

The risk of VTE is pronounced among patients with cardiovascular disease. After stroke and coronary disease, VTE is the third most common cardiovascular disorder, and PE causes more deaths each year than myocardial infarction (MI).25 Several cardiovascular diseases, including hypertension, stroke, acute MI, and heart failure, are independently associated with VTE.10, 2627 Related disorders, including diabetes and the metabolic syndrome, also increase the risk of VTE.26, 28

Congestive heart failure (CHF) is a risk factor for VTE, and the severity of illness increases risk. In the DVT‐Free Prospective Registry, 13% of patients with ultrasound‐confirmed DVT had CHF.29 In a subgroup analysis of patients of the Prophylaxis in Medical Patients with Enoxaparin (MEDENOX) study, the incidence of VTE exceeded 20% in patients with New York Heart Association (NYHA) class IV heart failure, compared with 12% in patients with NYHA class III heart failure.30 Another study found that VTE risk increases as left ventricular ejection fraction (LVEF) decreases, with an LVEF of less than 20% associated with a VTE OR of 38.3 (95% CI, 9.6152.5).31

Infectious Disease

Acute infection may increase the relative risk of VTE by as much as 50% and is associated with VTE event rates of up to 26%.11 Acute infections may be associated with acute inflammation, adverse effects on cardiac or pulmonary function, and prolonged immobilization.30, 32, 33 Human immunodeficiency virus (HIV) patients may also have an increased risk of VTE due to a circulating lupus anticoagulant and/or the presence of acute infection.34

Obesity

The 2008 ACCP guideline update recognizes obesity, for the first time, as a risk factor for VTE.3 Obesity was 1 of the 5 most frequent comorbidities found in patients with DVT in the DVT‐Free Prospective Registry.29 It increases the risk of both incident and recurrent VTE, with every 1‐point increase in body mass index (BMI) increasing the risk of recurrent VTE by 4.4% (95% CI, 1.37.6%; P < 0.001).35

Pregnancy and Puerperium

Pregnancy, particularly the postpartum period, is associated with an increased risk of VTE in women, even though the absolute risk is small.36 Still, PE is one of the leading causes of maternal death following childbirth.10 Smoking, prior VTE, and inherited thrombophilias all increase the risk of VTE in pregnant women.10 The risk begins to rise in the first trimester, and when prophylaxis is needed, it should be started early in gestation.37

Pulmonary Disease

COPD is another risk factor for the development of VTE. COPD patients who develop VTE tend to be older, hospitalized in the intensive care unit (ICU), and on mechanical ventilation.38 In the DVT‐Free Prospective Registry, 12% of patients with ultrasound‐confirmed DVT had COPD.29

Trauma and Surgery

Injury to the body tissue, via trauma or surgery, stimulates the body's clotting mechanism and increases the risk of thromboembolic complications. During the perioperative period, the circulatory system must balance a variety of assaults: an immune response to surgical stress, prolonged immobilization during surgery and recovery, vasodilation associated with general or regional anesthesia, and hypercoagulability due to venous stasis and vascular injury.39 Renal transplant recipients have an increased risk of VTE due to a chronic hypercoagulable state.40 In surgery patients, perioperative complications such as dehydration and acute infection increase the risk of VTE beyond the risk associated with the surgical procedure itself.10

VTE risk is increased approximately 13‐fold by recent major trauma or lower‐extremity injury.13 In the absence of prophylaxis, the overall risk of VTE among patients undergoing major surgery is increased nearly 22‐fold.13 After controlling for the type of surgery, additional independent risk factors for VTE within 3 months of major surgery include:41, 42

• Obesity

• Central venous catheter placement

• Malignancy

• Smoking

• Heart failure

• Previous DVT

• Prolonged immobility

• Infection

Many surgical and medical inpatients share common risk factors, and without prophylaxis, the incidence of hospital‐acquired DVT ranges from 10% to 40% for both groups.3

Inherited or Acquired Risk Factors

VTE is a multifactorial disease, and recent evidence indicates that some heritable traits may be potent risk factors for VTE.43 Approximately 35% of patients with DVT will have at least 1 of 5 traits related to an inherited blood clotting disorder:43

• Deficiencies in the anticoagulation factors protein C, protein S, or antithrombin, or

• Mutations in the factor V and prothrombin genes, resulting in Factor V Leiden and prothrombin G20210A, respectively.

Certain inherited traits and genetic polymorphisms increase the risk of VTE by interacting with clinical risk factors such as contraceptive use, pregnancy, surgery, trauma, and cancer. One recent study found that oral estrogen therapy among women with the CYP3A5*1 allele was associated with a particularly high risk of VTE.44 Although widespread screening for inherited risk factors is not currently practical, future tools may incorporate genetic polymorphisms to more precisely identify patients who would benefit from aggressive prophylaxis.

Lifestyle Factors

Lifestyle factors have a significant effect on VTE risk. Smoking increases the risk of VTE by 20% to 30%, and a sedentary lifestyle also increases the risk of VTE.26, 45 In fact, women who exercise regularly and consume alcohol in moderation have one‐half the risk of VTE as women who have a sedentary lifestyle and drink little or no alcohol.42 For both men and women, a diet high in fruits, vegetables, and fish is associated with a lower lifetime risk of VTE.46

Medications

Medications may also increase the risk of VTE. In cancer patients with anemia, for example, the use of erythropoiesis‐stimulating agents such as recombinant erythropoietin and darbepoetin was recently shown to increase the risk of VTE by 57% (95% CI, 3187%) and increase mortality risk by 10% (95% CI, 120%).23 In addition, combination hormone replacement therapy in women is associated with a higher risk of VTE compared with estrogen monotherapy, and transdermal contraceptive systems more than double the risk of VTE compared with oral contraceptives (95% CI, 1.33.8).47, 48 Recent studies have also reported an increased risk of VTE with some psychiatric drugs, including amitriptyline, clozapine, olanzapine, and risperidone.4952

Underestimation of Risk of Clotting

VTE is often clinically silent, leading some physicians to mistakenly believe that it is rare.58 In hospitalized surgical patients, for example, the incidence of thromboembolic complications during a short postoperative stay may be low. Given that many cases of symptomatic VTE occur after hospital discharge, hospitalists and surgeons may be unaware of the true incidence of DVT.59

Overestimation of the Risk of Bleeding

Physicians may also overestimate the risk of possible side effects of prophylaxis, such as major bleeding or heparin‐induced thrombocytopenia (HIT).58 Fear of excess bleeding has been cited by physicians as a leading reason for their decision to withhold thromboprophylaxis from at‐risk hospitalized patients.60 Physicians are particularly fearful of complications among elderly patients, who are less likely to receive adequate prophylaxis than younger patients with a similar risk of VTE.61 When bleeding does occur, it rarely results in death. On the other hand, PE may account for as many as 10% of hospital deaths.9

Guideline Confusion and Complexity

Discrepancies between guidelines published by different medical societies contribute to confusion in choosing a management approach. The American Academy of Orthopedic Surgeons (AAOS), for example, describes aspirin alone as a reasonable choice for VTE prophylaxis in some patients, but the ACCP guidelines advise against the use of aspirin monotherapy.58 The cumbersome nature of multiple risk‐assessment and treatment algorithms can also be problematic.61 Furthermore, certain patient subgroups, such as those with cirrhosis, severe renal failure, and epidural catheters, have been excluded from randomized controlled trials, and the management of such patients is not straightforward.

Absence of Institutional Protocols and Information Technology Support

The lack of institution‐level guidance and support can have a detrimental effect on patient care. In a 2007 survey of 127 community hospitals, the prevalence of institutional protocols related to VTE was low: only 60% had protocols to encourage prophylaxis in at‐risk patients, 54% had guidelines to assist in appropriate drug selection, and 43% had guidelines for the dosing of prophylaxis regimens.62 A lack of systems for data collection and audit has also been identified as a barrier to the implementation of prophylaxis guidelines.57 Thus, hospitals need to adopt protocols such as:3

• Written, institution‐wide thromboprophylaxis policies

• Preprinted order forms and computer decision‐support systems

• Policies specifying responsibilities for assessing VTE risk and prescribing prophylaxis

Current or Recent Prior Hospitalization

The risk of VTE is elevated among hospitalized patients. The prevalence of DVT varies across hospital specialties, reaching up to 80% in major trauma, spinal cord injury, and critical care.3 The Epidemiologic International Day for the Evaluation of Patients at Risk for Venous Thromboembolism in the Acute Hospital Care Setting (ENDORSE) study evaluated the prevalence of VTE risk factors in the acute hospital care setting.12 Among the 68,183 patients enrolled, more than one‐half (52%) were judged to be at risk of VTE. In a case‐control study examining 625 patients with a first lifetime VTE, confinement to a hospital (among other risk factors) was found to be an independent and important predictor of VTE (odds ratio [OR], 8.0; 95% confidence interval [CI], 4.514.2) (Figure 1).13 Recent hospitalization is also an important risk factor for VTE, and patients who are readmitted to the hospital should be considered moderate or high risk.13

Figure 1

Age

Patient age must be considered when assessing VTE risk. VTE is predominantly a disease of older age, and age older than 75 years is an important risk factor for the condition.11 In general, patients older than 40 years have a significantly increased risk compared with younger patients, and the risk approximately doubles with each additional decade.10 Given the aging population, the prevalence of VTE and its complications are expected to increase.

Women of childbearing age experience VTE more frequently than men of the same age, due to pregnancy and exposure to contraceptive therapy.14 This risk, however, is modest compared with the risk among older patients. After age 45 years, the incidence of VTE increases markedly for both sexes, becoming more prominent in men.14 Compared with women, men also have an increased risk of recurrent VTE.15

Despite the effect of age on VTE risk, the risk among patients younger than 40 years may be underestimated because this subgroup has not been extensively studied. For reasons that are not well‐understood, the risk of VTE associated with heart failure is higher in patients younger than 40 years, and the relative risk of PE in patients with chronic obstructive pulmonary disease (COPD) is also higher in younger patients.16, 17

Cancer and Its Treatment

Cancer patients, on average, have twice the risk of VTE compared with noncancer patients.18 This risk, however, varies considerably by cancer type. According to an assessment of nearly 41 million hospitalized patients in the National Hospital Discharge Survey (NHDS), the relative risk of VTE varied from 1.02 in patients with bladder cancer to 4.34 in patients with cancer of the pancreas.18

VTE is one of the most common complications of cancer and cancer therapy, and it is the second leading cause of death among hospitalized cancer patients.19 Molecular mechanisms underlying thromboembolic events in cancer patients include tumor cell procoagulants, inflammatory cell cytokines, mediators of platelet adhesion, and tumor‐related stasis and endothelial damage.20 The clinical implications of these processes are severe. Cancer exacerbates the natural course of VTE, increasing the risk of recurrent VTE and major bleeding, and VTE worsens the prognosis of cancer, increasing the risk of death among cancer patients.

Various cancer therapiesincluding surgery, chemotherapy, hematopoietic stem cell transplantation, and even growth factor supportalso increase the risk of VTE, in part because extrinsic factors such as surgery or chemotherapy can intensify the hypercoagulable process.18, 2123 In the NHDS, cancer patients undergoing surgery had at least twice the risk of postoperative DVT and more than 3 times the risk of PE compared with noncancer patients undergoing similar procedures.18

Cancer is an independent predictor of thromboprophylaxis failure following surgery. The @RISTOS Project found that VTE was the most common cause of death among 2373 patients undergoing general, urologic, or gynecologic surgery for cancer.24 A multivariate analysis identified 5 independent risk factors for VTE after cancer surgery:

• Previous VTE (OR, 5.98; 95% CI, 2.1316.80)

• Anesthesia 2 hours (OR, 4.50; 95% CI, 1.0619.04)

• Bed rest 4 days (OR, 4.37; 95% CI, 2.457.78)

• Age 60 years (OR, 2.63; 95% CI, 1.215.71)

• Advanced‐stage cancer (OR, 2.68; 95% CI, 1.375.24)

Cardiovascular Disease

The risk of VTE is pronounced among patients with cardiovascular disease. After stroke and coronary disease, VTE is the third most common cardiovascular disorder, and PE causes more deaths each year than myocardial infarction (MI).25 Several cardiovascular diseases, including hypertension, stroke, acute MI, and heart failure, are independently associated with VTE.10, 2627 Related disorders, including diabetes and the metabolic syndrome, also increase the risk of VTE.26, 28

Congestive heart failure (CHF) is a risk factor for VTE, and the severity of illness increases risk. In the DVT‐Free Prospective Registry, 13% of patients with ultrasound‐confirmed DVT had CHF.29 In a subgroup analysis of patients of the Prophylaxis in Medical Patients with Enoxaparin (MEDENOX) study, the incidence of VTE exceeded 20% in patients with New York Heart Association (NYHA) class IV heart failure, compared with 12% in patients with NYHA class III heart failure.30 Another study found that VTE risk increases as left ventricular ejection fraction (LVEF) decreases, with an LVEF of less than 20% associated with a VTE OR of 38.3 (95% CI, 9.6152.5).31

Infectious Disease

Acute infection may increase the relative risk of VTE by as much as 50% and is associated with VTE event rates of up to 26%.11 Acute infections may be associated with acute inflammation, adverse effects on cardiac or pulmonary function, and prolonged immobilization.30, 32, 33 Human immunodeficiency virus (HIV) patients may also have an increased risk of VTE due to a circulating lupus anticoagulant and/or the presence of acute infection.34

Obesity

The 2008 ACCP guideline update recognizes obesity, for the first time, as a risk factor for VTE.3 Obesity was 1 of the 5 most frequent comorbidities found in patients with DVT in the DVT‐Free Prospective Registry.29 It increases the risk of both incident and recurrent VTE, with every 1‐point increase in body mass index (BMI) increasing the risk of recurrent VTE by 4.4% (95% CI, 1.37.6%; P < 0.001).35

Pregnancy and Puerperium

Pregnancy, particularly the postpartum period, is associated with an increased risk of VTE in women, even though the absolute risk is small.36 Still, PE is one of the leading causes of maternal death following childbirth.10 Smoking, prior VTE, and inherited thrombophilias all increase the risk of VTE in pregnant women.10 The risk begins to rise in the first trimester, and when prophylaxis is needed, it should be started early in gestation.37

Pulmonary Disease

COPD is another risk factor for the development of VTE. COPD patients who develop VTE tend to be older, hospitalized in the intensive care unit (ICU), and on mechanical ventilation.38 In the DVT‐Free Prospective Registry, 12% of patients with ultrasound‐confirmed DVT had COPD.29

Trauma and Surgery

Injury to the body tissue, via trauma or surgery, stimulates the body's clotting mechanism and increases the risk of thromboembolic complications. During the perioperative period, the circulatory system must balance a variety of assaults: an immune response to surgical stress, prolonged immobilization during surgery and recovery, vasodilation associated with general or regional anesthesia, and hypercoagulability due to venous stasis and vascular injury.39 Renal transplant recipients have an increased risk of VTE due to a chronic hypercoagulable state.40 In surgery patients, perioperative complications such as dehydration and acute infection increase the risk of VTE beyond the risk associated with the surgical procedure itself.10

VTE risk is increased approximately 13‐fold by recent major trauma or lower‐extremity injury.13 In the absence of prophylaxis, the overall risk of VTE among patients undergoing major surgery is increased nearly 22‐fold.13 After controlling for the type of surgery, additional independent risk factors for VTE within 3 months of major surgery include:41, 42

• Obesity

• Central venous catheter placement

• Malignancy

• Smoking

• Heart failure

• Previous DVT

• Prolonged immobility

• Infection

Many surgical and medical inpatients share common risk factors, and without prophylaxis, the incidence of hospital‐acquired DVT ranges from 10% to 40% for both groups.3

Inherited or Acquired Risk Factors

VTE is a multifactorial disease, and recent evidence indicates that some heritable traits may be potent risk factors for VTE.43 Approximately 35% of patients with DVT will have at least 1 of 5 traits related to an inherited blood clotting disorder:43

• Deficiencies in the anticoagulation factors protein C, protein S, or antithrombin, or

• Mutations in the factor V and prothrombin genes, resulting in Factor V Leiden and prothrombin G20210A, respectively.

Certain inherited traits and genetic polymorphisms increase the risk of VTE by interacting with clinical risk factors such as contraceptive use, pregnancy, surgery, trauma, and cancer. One recent study found that oral estrogen therapy among women with the CYP3A5*1 allele was associated with a particularly high risk of VTE.44 Although widespread screening for inherited risk factors is not currently practical, future tools may incorporate genetic polymorphisms to more precisely identify patients who would benefit from aggressive prophylaxis.

Lifestyle Factors

Lifestyle factors have a significant effect on VTE risk. Smoking increases the risk of VTE by 20% to 30%, and a sedentary lifestyle also increases the risk of VTE.26, 45 In fact, women who exercise regularly and consume alcohol in moderation have one‐half the risk of VTE as women who have a sedentary lifestyle and drink little or no alcohol.42 For both men and women, a diet high in fruits, vegetables, and fish is associated with a lower lifetime risk of VTE.46

Medications

Medications may also increase the risk of VTE. In cancer patients with anemia, for example, the use of erythropoiesis‐stimulating agents such as recombinant erythropoietin and darbepoetin was recently shown to increase the risk of VTE by 57% (95% CI, 3187%) and increase mortality risk by 10% (95% CI, 120%).23 In addition, combination hormone replacement therapy in women is associated with a higher risk of VTE compared with estrogen monotherapy, and transdermal contraceptive systems more than double the risk of VTE compared with oral contraceptives (95% CI, 1.33.8).47, 48 Recent studies have also reported an increased risk of VTE with some psychiatric drugs, including amitriptyline, clozapine, olanzapine, and risperidone.4952

Underestimation of Risk of Clotting

VTE is often clinically silent, leading some physicians to mistakenly believe that it is rare.58 In hospitalized surgical patients, for example, the incidence of thromboembolic complications during a short postoperative stay may be low. Given that many cases of symptomatic VTE occur after hospital discharge, hospitalists and surgeons may be unaware of the true incidence of DVT.59

Overestimation of the Risk of Bleeding

Physicians may also overestimate the risk of possible side effects of prophylaxis, such as major bleeding or heparin‐induced thrombocytopenia (HIT).58 Fear of excess bleeding has been cited by physicians as a leading reason for their decision to withhold thromboprophylaxis from at‐risk hospitalized patients.60 Physicians are particularly fearful of complications among elderly patients, who are less likely to receive adequate prophylaxis than younger patients with a similar risk of VTE.61 When bleeding does occur, it rarely results in death. On the other hand, PE may account for as many as 10% of hospital deaths.9

Guideline Confusion and Complexity

Discrepancies between guidelines published by different medical societies contribute to confusion in choosing a management approach. The American Academy of Orthopedic Surgeons (AAOS), for example, describes aspirin alone as a reasonable choice for VTE prophylaxis in some patients, but the ACCP guidelines advise against the use of aspirin monotherapy.58 The cumbersome nature of multiple risk‐assessment and treatment algorithms can also be problematic.61 Furthermore, certain patient subgroups, such as those with cirrhosis, severe renal failure, and epidural catheters, have been excluded from randomized controlled trials, and the management of such patients is not straightforward.

Absence of Institutional Protocols and Information Technology Support

The lack of institution‐level guidance and support can have a detrimental effect on patient care. In a 2007 survey of 127 community hospitals, the prevalence of institutional protocols related to VTE was low: only 60% had protocols to encourage prophylaxis in at‐risk patients, 54% had guidelines to assist in appropriate drug selection, and 43% had guidelines for the dosing of prophylaxis regimens.62 A lack of systems for data collection and audit has also been identified as a barrier to the implementation of prophylaxis guidelines.57 Thus, hospitals need to adopt protocols such as:3

• Written, institution‐wide thromboprophylaxis policies

• Preprinted order forms and computer decision‐support systems

• Policies specifying responsibilities for assessing VTE risk and prescribing prophylaxis

UFH

UFH, which is typically administered by subcutaneous injection, has the longest history as an anticoagulant in the prevention and treatment of VTE. It is an attractive option in patients with severe renal failure or those who may require a procedure in the near future. Although UFH is partially cleared by the kidney, its short half‐life can be perceived as a safety advantage in patients with severe renal impairment and an increased risk of bleeding. For most other patients, UFH holds several disadvantages compared with newer therapies, including the need for injections to be administered 3 times a day to be optimally effective, its effect on platelets, and its association with heparin‐induced thrombocytopenia (HIT).1 Given the costs of administration and potential complications, it is not less expensive than LMWHs, and it appears to be less cost‐effective.11

LMWHs

LMWHs have a higher bioavailability and longer half‐life than UFH, which translates to reliable anticoagulation levels when given subcutaneously on a weight‐based dosing schedule. Unlike UFH, LMWHs do not require laboratory tests to monitor the intensity of anticoagulation, except in special circumstances.1 The LMWHs dalteparin, enoxaparin, and tinzaparin are widely used for the prevention and treatment of VTE in the United States.

Two landmark clinical trials demonstrated the efficacy of appropriate thromboprophylaxis with LMWHs in reducing the burden of VTE in acutely ill, hospitalized medical patients. The Prophylaxis in Medical Patients with Enoxaparin (MEDENOX) trial and Prospective Evaluation of Dalteparin Efficacy for Prevention of VTE in Immobilized Patients Trial (PREVENT) demonstrated the benefits of enoxaparin and dalteparin, respectively, in reducing the risk of VTE. As shown in Table 2, thromboprophylaxis with these agents was associated with a 45% to 63% relative reduction in the risk of VTE compared with placebo.12, 13

Pentasaccharides

Fondaparinux is a synthetic factor Xa antagonist that shares many features of LMWHs, including a high bioavailability and long half‐life. Fondaparinux does not require monitoring, but it is contraindicated in patients with renal failure (CrCl < 30 mL/minute) and in patients weighing less than 50 kg.1 Although PF4 antibodies have been associated with fondaparinux administration, this drug has not, to date, been associated with HIT.14 The Arixtra for Thromboembolism Prevention in a Medical Indications Study (ARTEMIS) trial demonstrated the advantage of fondaparinux over placebo in reducing the risk of VTE (Table 2).15 The American College of Chest Physicians (ACCP) guidelines state that fondaparinux appears to be as effective and safe as LMWH.1

Vitamin K Antagonists

Vitamin K antagonists (VKAs) such as warfarin inhibit the production of prothrombin, clotting factors VII, IX, and X, and the anticoagulants protein C and protein S. Warfarin is challenging to manage because of its narrow therapeutic window, its tendency to exhibit considerable variability in dose‐response, the time required to reach target international normalized ratio (INR), its potential for interaction with diet and concomitant medications, and its need for ongoing monitoring.5 Warfarin should usually be initiated within the same 24 hours as parenteral anticoagulation, with a goal of achieving INR results between 2.0 and 3.0. An initial dose of 5 to 10 mg for the first 1 or 2 days is appropriate for most patients, and subsequent dosing should be based on INR response.5 Warfarin prophylaxis is primarily used in patients in the US undergoing orthopedic surgery, including total hip replacement and hip and knee arthroplasty.1

Future Anticoagulants

New oral agents have the potential to improve the management of patients who have a moderate to high risk of thromboembolic disease.

Duration

Although major trials support the use of short‐term prophylaxistypically 6 to 14 daysin‐hospital for acutely ill medical patients, the optimal duration of thromboprophylaxis in these patients is unclear.24 The Extended Clinical Prophylaxis in Acutely Ill Medical Patients (EXCLAIM) trial is the first randomized trial to evaluate the potential benefits of extended prophylaxis in acutely ill medical patients. In this study, 5101 hospitalized patients with varying levels of reduced mobility due to cancer, ischemic stroke, heart failure, respiratory failure, infection, and other acute medical conditions received open‐label enoxaparin 40 mg daily for a mean duration of 10 days. Patients were then randomly assigned to additional therapy with enoxaparin or placebo for a mean duration of 28 additional days.25 Preliminary findings from this trial suggest that high‐risk medical patients can benefit from extended thromboprophylaxis following hospital discharge, with significantly reduced VTE events (RR reduction, 44%; P = 0.0011). The benefits of thromboprophylaxis were apparent during the extended treatment period and persistent through 90 days.26

Guideline Recommendations

Incorrect use of thromboprophylaxis does not stem from a lack of evidence‐based recommendations. Within the past year, the ACCP, the American Society of Clinical Oncology (ASCO), and the National Comprehensive Cancer Network (NCCN) have published updated guidelines for thromboprophylaxis in hospitalized patients at risk of VTE.1, 27, 28 The 2008 ACCP guidelines include more than 700 recommendations for VTE risk assessment and management, to be implemented by a variety of physicians, including pulmonologists, cardiologists, cardiothoracic surgeons, and critical care medicine specialists.1

The ACCP guidelines organize prophylaxis recommendations on the basis of patient risk (Table 4).1 Risk assessment remains relatively subjective, however, and validated risk assessment models are not yet widely available. The prudent approach is to consider thromboprophylaxis for all hospitalized medically ill patients who do not have a specific contraindication.

Key evidence‐based recommendations regarding thromboprophylaxis for hospitalized, acutely ill patients include the following:1

• Every hospital should develop a formal strategy to addresses VTE prophylaxis;

• Aspirin alone is not recommended to prevent VTE for any patient group;

• Mechanical methods of thromboprophylaxis should be used primarily for patients who have a high bleeding risk or possibly as an adjunct to anticoagulant thromboprophylaxis;

• Thromboprophylaxis with LMWH, UFH, or fondaparinux is recommended for patients admitted to hospital with an acute medical illness (Note: fondaparinux is recommended, but not FDA‐approved, for this indication in the United States);

• On admission to the ICU, all patients should be assessed for risk of VTE, and most should receive thromboprophylaxis;

• All major trauma and all spinal‐cord injury patients should receive thromboprophylaxis.

Cancer Patients

The ASCO and NCCN guidelines endorse the use of VTE prophylaxis with anticoagulants in all hospitalized patients with active cancer or suspicion of cancer in the absence of contraindications.27, 28 The ACCP guidelines restrict this recommendation to hospitalized cancer patients who are bedridden.1 Thromboprophylaxis should continue at least through the duration of the hospital stay. Acceptable subcutaneous regimens include fondaparinux, dalteparin, or enoxaparin at the doses presented in Table 1; if UFH is chosen, the dose should be 5000 units every 8 hours.

Cancer patients who are scheduled to undergo major surgery require a different prophylaxis strategy. Even with prophylaxis, cancer patients have a 2‐fold higher risk of postoperative VTE compared with noncancer patients and more than a 3‐fold higher risk of fatal PE.30 To manage this risk, the ASCO, NCCN, and ACCP guidelines recommend extended prophylaxis in patients undergoing major cancer surgery.1, 27, 28 Specific recommendations include the following:

• All patients undergoing major surgical intervention for malignant disease should be considered for VTE prophylaxis with anticoagulants, with or without mechanical prophylaxis;

• Thromboprophylaxis should be initiated prior to the start of surgery or as early as possible following surgery;

• Mechanical interventions may supplement pharmacologic prophylaxis, especially in patients who have the highest risk;

• Prophylaxis with a LMWH should be initiated 12 to 24 hours after the surgical procedure;

• Continue prophylaxis at least 7 to 10 days postoperatively;

• Consider prolonged prophylaxis (ie, up to 4 weeks) with a LMWH for high‐risk patients (eg, patients undergoing major abdominal or pelvic surgery, those with residual malignant disease after surgery, obese patients, and patients with a history of VTE).

Routine prophylaxis with anticoagulants is not recommended for most outpatients, except for those with high‐risk factors (eg, thrombogenic chemotherapy or a central venous catheter). The strategy of restricting thromboprophylaxis to cancer outpatients with specific indications, however, may miss an opportunity to reduce VTE in this vulnerable patient population. In the PROTECHT study, 1166 ambulatory cancer patients were randomly assigned to placebo or the LMWH nadroparin for the duration of their chemotherapy. Treatment with nadroparin reduced the rate of clinical thrombosis by 47.2% compared with placebo (3.9% vs. 2.1%; P = 0.033). The risk reduction was consistent across all measured events, including DVT, PE, stroke, and visceral venous thrombosis.31

Pregnancy

Prophylaxis should be considered in pregnant women with known risk factors for VTE such as prior VTE, thrombophilia, and a history of prolonged immobility. In addition, women with a moderate to high risk of VTE associated with a cesarean section should be considered for postpartum thromboprophylaxis. For example, 1 of the following regimens may be appropriate for high‐risk women following a cesarean section:32

• UFH 5000 units subcutaneously every 12 hours until fully mobile;

• LMWH subcutaneously once daily for 5 days (such as enoxaparin 20 mg daily).

For pregnant women already receiving anticoagulant prophylaxis (eg, for hypercoagulable state, structural heart disease, or prior DVT/PE), ACCP guidelines recommend discontinuing VKAs before 6 weeks of fetal gestation to minimize the risk of birth defects and miscarriage. In general, a LMWH should be substituted for VKAs as soon as pregnancy is confirmed or prior to conception in preparation for pregnancy, as VKAs cross the placental barrier, but LMWH and UFH do not.1, 33

Renal Insufficiency

The ACCP guidelines recommend that renal function be considered when making decisions about the use and/or dose of LMWHs and fondaparinux. Because these agents are eliminated primarily via renal clearance, changes in renal function can reduce drug clearance, prolong the half‐life, and increase plasma concentrations. Consequently, the risk of treatment‐related bleeding complications is elevated in patients with renal impairment.1 Depending on the circumstances, one of the following options should be considered1:

• Avoid using an anticoagulant that bioaccumulates in the presence of renal impairment;

• Use a lower dose of the agent;

• Monitor the drug level or its anticoagulant effect.

In severe renal impairment (creatinine clearance < 30 mL/minute):710

• The prophylactic dose of enoxaparin should be adjusted to 30 mg subcutaneously once daily; no specific dosing adjustments have been recommended for dalteparin or tinzaparin;

• Fondaparinux is contraindicated.

Epidural Anesthesia

Neuraxial blockade has several advantages over systemic opioids, but the risk of spinal or epidural hematoma may be increased with the concomitant use of antithrombotic drugs. Therefore, these agents must be used cautiously in patients with neuraxial blockade.1 Guidelines from the American Society of Regional Anesthesia and Pain Medicine (ASRA) contain the following recommendations:34

• Subcutaneous UFH: No contraindication, consider delaying heparin until after block if technical difficulty is anticipated;

• LMWH: Since twice daily dosing may be associated with an increased risk of spinal hematoma, delay initiation of LMWH until at least 24 hours after surgery, regardless of anesthetic technique; for single daily dosing, administer the first dose of LMWH 6 to 8 hours postoperatively and second dose no sooner than 24 hours after the first dose;

• Warfarin: Document normal INR after discontinuation (prior to neuraxial technique); remove catheter when INR 1.5 (initiation of therapy).

Bleeding Risk

The ACCP and ASCO guidelines emphasize the importance of weighing the potential benefits of thromboprophylaxis against the potential risks of bleeding in individual patients. According to the ACCP, the overall risk of bleeding with intravenous UFH in patients with VTE is less than 3%, and thromboprophylaxis has not been shown to increase the risk of bleeding compared with placebo in major clinical trials.13, 15, 35 However, bleeding risk may increase in older patients and with higher doses of heparin. Warfarin therapy can be monitored with an INR to reduce the risk of bleeding during thromboprophylaxis.1

Anticoagulation therapy may be contraindicated in patients with certain factors and conditions that increase the risk of bleeding. These include:

• Clinically significant active or chronic bleeding;

• Recent central nervous system or spinal surgery with increased risk of bleeding;

• Thrombocytopenia (excluding HIT) or severe platelet dysfunction;

• Abnormalities associated with clotting factors.

The NCCN provides specific contraindications to anticoagulation therapy for the prevention and treatment of VTE in cancer patients.28 These include:

• Recent central nervous system bleed; intracranial, or spinal lesions at high risk of bleeding;

• Active major bleeding (> 2 units transfused in 24 hours);

• Chronic, clinically significant measurable bleeding for more than 48 hours;

• Thrombocytopenia (platelets < 50,000/L);

• Severe platelet dysfunction;

• Recent major operation with high risk of bleeding;

• Underlying coagulopathy (eg, clotting factor abnormalities or elevated prothrombin time or activated partial thromboplastin time [aPTT]);

• Spinal anesthesia or lumbar puncture;

• High risk of falls.

HIT

HIT is a serious complication that can occur as a result of exposure to heparin. It is an immune response that causes platelet activation and platelet aggregation, among other effects, and is capable of leading to severe thrombosis, amputation, or death.36 The incidence of HIT varies with subpopulations of patients and more commonly develops in patients receiving heparin in therapeutic doses. Early diagnosis (through an interpretation of clinical and laboratory information) is important to improve clinical outcomes, but difficult to achieve.36 The ACCP guidelines note that enzyme‐linked immunosorbent assay (ELISA)‐based tests for HIT are often falsely positive after surgery. As an alternative, serotonin‐release tests are more specific, although they are not as widely available.1

Substantial clinical evidence suggests that LMWH poses less of a risk of HIT than UFH. Martel et al,37 for example, conducted a meta‐analysis of 15 randomized and nonrandomized controlled trials (a total of 7287 patients) that included studies that compared prophylactic doses of UFH and LMWH and assessed postoperative or medical inpatients who received prophylaxis. The analysis revealed that the risk of HIT was 2.6% following UFH use compared with 0.2% following LMWH use.37 Despite the inclusion of UFH in the ASCO guidelines, ASCO acknowledges that a lower risk of HIT is one of the potential advantages of LMWH over UFH in cancer surgery prophylaxis.27 In addition, the recommendation to transition to outpatient therapy as soon as possible is an indirect way of stating a preference for LMWH. For cancer patients with established VTE, the recommendation is more direct: LMWH is clearly preferred over UFH for both initial and continuing antithrombotic therapy.27

UFH

UFH, which is typically administered by subcutaneous injection, has the longest history as an anticoagulant in the prevention and treatment of VTE. It is an attractive option in patients with severe renal failure or those who may require a procedure in the near future. Although UFH is partially cleared by the kidney, its short half‐life can be perceived as a safety advantage in patients with severe renal impairment and an increased risk of bleeding. For most other patients, UFH holds several disadvantages compared with newer therapies, including the need for injections to be administered 3 times a day to be optimally effective, its effect on platelets, and its association with heparin‐induced thrombocytopenia (HIT).1 Given the costs of administration and potential complications, it is not less expensive than LMWHs, and it appears to be less cost‐effective.11

LMWHs

LMWHs have a higher bioavailability and longer half‐life than UFH, which translates to reliable anticoagulation levels when given subcutaneously on a weight‐based dosing schedule. Unlike UFH, LMWHs do not require laboratory tests to monitor the intensity of anticoagulation, except in special circumstances.1 The LMWHs dalteparin, enoxaparin, and tinzaparin are widely used for the prevention and treatment of VTE in the United States.

Two landmark clinical trials demonstrated the efficacy of appropriate thromboprophylaxis with LMWHs in reducing the burden of VTE in acutely ill, hospitalized medical patients. The Prophylaxis in Medical Patients with Enoxaparin (MEDENOX) trial and Prospective Evaluation of Dalteparin Efficacy for Prevention of VTE in Immobilized Patients Trial (PREVENT) demonstrated the benefits of enoxaparin and dalteparin, respectively, in reducing the risk of VTE. As shown in Table 2, thromboprophylaxis with these agents was associated with a 45% to 63% relative reduction in the risk of VTE compared with placebo.12, 13

Pentasaccharides

Fondaparinux is a synthetic factor Xa antagonist that shares many features of LMWHs, including a high bioavailability and long half‐life. Fondaparinux does not require monitoring, but it is contraindicated in patients with renal failure (CrCl < 30 mL/minute) and in patients weighing less than 50 kg.1 Although PF4 antibodies have been associated with fondaparinux administration, this drug has not, to date, been associated with HIT.14 The Arixtra for Thromboembolism Prevention in a Medical Indications Study (ARTEMIS) trial demonstrated the advantage of fondaparinux over placebo in reducing the risk of VTE (Table 2).15 The American College of Chest Physicians (ACCP) guidelines state that fondaparinux appears to be as effective and safe as LMWH.1

Vitamin K Antagonists

Vitamin K antagonists (VKAs) such as warfarin inhibit the production of prothrombin, clotting factors VII, IX, and X, and the anticoagulants protein C and protein S. Warfarin is challenging to manage because of its narrow therapeutic window, its tendency to exhibit considerable variability in dose‐response, the time required to reach target international normalized ratio (INR), its potential for interaction with diet and concomitant medications, and its need for ongoing monitoring.5 Warfarin should usually be initiated within the same 24 hours as parenteral anticoagulation, with a goal of achieving INR results between 2.0 and 3.0. An initial dose of 5 to 10 mg for the first 1 or 2 days is appropriate for most patients, and subsequent dosing should be based on INR response.5 Warfarin prophylaxis is primarily used in patients in the US undergoing orthopedic surgery, including total hip replacement and hip and knee arthroplasty.1

Future Anticoagulants

New oral agents have the potential to improve the management of patients who have a moderate to high risk of thromboembolic disease.

Duration

Although major trials support the use of short‐term prophylaxistypically 6 to 14 daysin‐hospital for acutely ill medical patients, the optimal duration of thromboprophylaxis in these patients is unclear.24 The Extended Clinical Prophylaxis in Acutely Ill Medical Patients (EXCLAIM) trial is the first randomized trial to evaluate the potential benefits of extended prophylaxis in acutely ill medical patients. In this study, 5101 hospitalized patients with varying levels of reduced mobility due to cancer, ischemic stroke, heart failure, respiratory failure, infection, and other acute medical conditions received open‐label enoxaparin 40 mg daily for a mean duration of 10 days. Patients were then randomly assigned to additional therapy with enoxaparin or placebo for a mean duration of 28 additional days.25 Preliminary findings from this trial suggest that high‐risk medical patients can benefit from extended thromboprophylaxis following hospital discharge, with significantly reduced VTE events (RR reduction, 44%; P = 0.0011). The benefits of thromboprophylaxis were apparent during the extended treatment period and persistent through 90 days.26

Guideline Recommendations

Incorrect use of thromboprophylaxis does not stem from a lack of evidence‐based recommendations. Within the past year, the ACCP, the American Society of Clinical Oncology (ASCO), and the National Comprehensive Cancer Network (NCCN) have published updated guidelines for thromboprophylaxis in hospitalized patients at risk of VTE.1, 27, 28 The 2008 ACCP guidelines include more than 700 recommendations for VTE risk assessment and management, to be implemented by a variety of physicians, including pulmonologists, cardiologists, cardiothoracic surgeons, and critical care medicine specialists.1

The ACCP guidelines organize prophylaxis recommendations on the basis of patient risk (Table 4).1 Risk assessment remains relatively subjective, however, and validated risk assessment models are not yet widely available. The prudent approach is to consider thromboprophylaxis for all hospitalized medically ill patients who do not have a specific contraindication.

Key evidence‐based recommendations regarding thromboprophylaxis for hospitalized, acutely ill patients include the following:1

• Every hospital should develop a formal strategy to addresses VTE prophylaxis;

• Aspirin alone is not recommended to prevent VTE for any patient group;

• Mechanical methods of thromboprophylaxis should be used primarily for patients who have a high bleeding risk or possibly as an adjunct to anticoagulant thromboprophylaxis;

• Thromboprophylaxis with LMWH, UFH, or fondaparinux is recommended for patients admitted to hospital with an acute medical illness (Note: fondaparinux is recommended, but not FDA‐approved, for this indication in the United States);

• On admission to the ICU, all patients should be assessed for risk of VTE, and most should receive thromboprophylaxis;

• All major trauma and all spinal‐cord injury patients should receive thromboprophylaxis.

Cancer Patients

The ASCO and NCCN guidelines endorse the use of VTE prophylaxis with anticoagulants in all hospitalized patients with active cancer or suspicion of cancer in the absence of contraindications.27, 28 The ACCP guidelines restrict this recommendation to hospitalized cancer patients who are bedridden.1 Thromboprophylaxis should continue at least through the duration of the hospital stay. Acceptable subcutaneous regimens include fondaparinux, dalteparin, or enoxaparin at the doses presented in Table 1; if UFH is chosen, the dose should be 5000 units every 8 hours.

Cancer patients who are scheduled to undergo major surgery require a different prophylaxis strategy. Even with prophylaxis, cancer patients have a 2‐fold higher risk of postoperative VTE compared with noncancer patients and more than a 3‐fold higher risk of fatal PE.30 To manage this risk, the ASCO, NCCN, and ACCP guidelines recommend extended prophylaxis in patients undergoing major cancer surgery.1, 27, 28 Specific recommendations include the following:

• All patients undergoing major surgical intervention for malignant disease should be considered for VTE prophylaxis with anticoagulants, with or without mechanical prophylaxis;

• Thromboprophylaxis should be initiated prior to the start of surgery or as early as possible following surgery;

• Mechanical interventions may supplement pharmacologic prophylaxis, especially in patients who have the highest risk;

• Prophylaxis with a LMWH should be initiated 12 to 24 hours after the surgical procedure;

• Continue prophylaxis at least 7 to 10 days postoperatively;

• Consider prolonged prophylaxis (ie, up to 4 weeks) with a LMWH for high‐risk patients (eg, patients undergoing major abdominal or pelvic surgery, those with residual malignant disease after surgery, obese patients, and patients with a history of VTE).

Routine prophylaxis with anticoagulants is not recommended for most outpatients, except for those with high‐risk factors (eg, thrombogenic chemotherapy or a central venous catheter). The strategy of restricting thromboprophylaxis to cancer outpatients with specific indications, however, may miss an opportunity to reduce VTE in this vulnerable patient population. In the PROTECHT study, 1166 ambulatory cancer patients were randomly assigned to placebo or the LMWH nadroparin for the duration of their chemotherapy. Treatment with nadroparin reduced the rate of clinical thrombosis by 47.2% compared with placebo (3.9% vs. 2.1%; P = 0.033). The risk reduction was consistent across all measured events, including DVT, PE, stroke, and visceral venous thrombosis.31

Pregnancy

Prophylaxis should be considered in pregnant women with known risk factors for VTE such as prior VTE, thrombophilia, and a history of prolonged immobility. In addition, women with a moderate to high risk of VTE associated with a cesarean section should be considered for postpartum thromboprophylaxis. For example, 1 of the following regimens may be appropriate for high‐risk women following a cesarean section:32

• UFH 5000 units subcutaneously every 12 hours until fully mobile;

• LMWH subcutaneously once daily for 5 days (such as enoxaparin 20 mg daily).

For pregnant women already receiving anticoagulant prophylaxis (eg, for hypercoagulable state, structural heart disease, or prior DVT/PE), ACCP guidelines recommend discontinuing VKAs before 6 weeks of fetal gestation to minimize the risk of birth defects and miscarriage. In general, a LMWH should be substituted for VKAs as soon as pregnancy is confirmed or prior to conception in preparation for pregnancy, as VKAs cross the placental barrier, but LMWH and UFH do not.1, 33

Renal Insufficiency

The ACCP guidelines recommend that renal function be considered when making decisions about the use and/or dose of LMWHs and fondaparinux. Because these agents are eliminated primarily via renal clearance, changes in renal function can reduce drug clearance, prolong the half‐life, and increase plasma concentrations. Consequently, the risk of treatment‐related bleeding complications is elevated in patients with renal impairment.1 Depending on the circumstances, one of the following options should be considered1:

• Avoid using an anticoagulant that bioaccumulates in the presence of renal impairment;

• Use a lower dose of the agent;

• Monitor the drug level or its anticoagulant effect.

In severe renal impairment (creatinine clearance < 30 mL/minute):710

• The prophylactic dose of enoxaparin should be adjusted to 30 mg subcutaneously once daily; no specific dosing adjustments have been recommended for dalteparin or tinzaparin;

• Fondaparinux is contraindicated.

Epidural Anesthesia

Neuraxial blockade has several advantages over systemic opioids, but the risk of spinal or epidural hematoma may be increased with the concomitant use of antithrombotic drugs. Therefore, these agents must be used cautiously in patients with neuraxial blockade.1 Guidelines from the American Society of Regional Anesthesia and Pain Medicine (ASRA) contain the following recommendations:34

• Subcutaneous UFH: No contraindication, consider delaying heparin until after block if technical difficulty is anticipated;

• LMWH: Since twice daily dosing may be associated with an increased risk of spinal hematoma, delay initiation of LMWH until at least 24 hours after surgery, regardless of anesthetic technique; for single daily dosing, administer the first dose of LMWH 6 to 8 hours postoperatively and second dose no sooner than 24 hours after the first dose;

• Warfarin: Document normal INR after discontinuation (prior to neuraxial technique); remove catheter when INR 1.5 (initiation of therapy).

Bleeding Risk

The ACCP and ASCO guidelines emphasize the importance of weighing the potential benefits of thromboprophylaxis against the potential risks of bleeding in individual patients. According to the ACCP, the overall risk of bleeding with intravenous UFH in patients with VTE is less than 3%, and thromboprophylaxis has not been shown to increase the risk of bleeding compared with placebo in major clinical trials.13, 15, 35 However, bleeding risk may increase in older patients and with higher doses of heparin. Warfarin therapy can be monitored with an INR to reduce the risk of bleeding during thromboprophylaxis.1

Anticoagulation therapy may be contraindicated in patients with certain factors and conditions that increase the risk of bleeding. These include:

• Clinically significant active or chronic bleeding;

• Recent central nervous system or spinal surgery with increased risk of bleeding;

• Thrombocytopenia (excluding HIT) or severe platelet dysfunction;

• Abnormalities associated with clotting factors.

The NCCN provides specific contraindications to anticoagulation therapy for the prevention and treatment of VTE in cancer patients.28 These include:

• Recent central nervous system bleed; intracranial, or spinal lesions at high risk of bleeding;

• Active major bleeding (> 2 units transfused in 24 hours);

• Chronic, clinically significant measurable bleeding for more than 48 hours;

• Thrombocytopenia (platelets < 50,000/L);

• Severe platelet dysfunction;

• Recent major operation with high risk of bleeding;

• Underlying coagulopathy (eg, clotting factor abnormalities or elevated prothrombin time or activated partial thromboplastin time [aPTT]);

• Spinal anesthesia or lumbar puncture;

• High risk of falls.

HIT

HIT is a serious complication that can occur as a result of exposure to heparin. It is an immune response that causes platelet activation and platelet aggregation, among other effects, and is capable of leading to severe thrombosis, amputation, or death.36 The incidence of HIT varies with subpopulations of patients and more commonly develops in patients receiving heparin in therapeutic doses. Early diagnosis (through an interpretation of clinical and laboratory information) is important to improve clinical outcomes, but difficult to achieve.36 The ACCP guidelines note that enzyme‐linked immunosorbent assay (ELISA)‐based tests for HIT are often falsely positive after surgery. As an alternative, serotonin‐release tests are more specific, although they are not as widely available.1

Substantial clinical evidence suggests that LMWH poses less of a risk of HIT than UFH. Martel et al,37 for example, conducted a meta‐analysis of 15 randomized and nonrandomized controlled trials (a total of 7287 patients) that included studies that compared prophylactic doses of UFH and LMWH and assessed postoperative or medical inpatients who received prophylaxis. The analysis revealed that the risk of HIT was 2.6% following UFH use compared with 0.2% following LMWH use.37 Despite the inclusion of UFH in the ASCO guidelines, ASCO acknowledges that a lower risk of HIT is one of the potential advantages of LMWH over UFH in cancer surgery prophylaxis.27 In addition, the recommendation to transition to outpatient therapy as soon as possible is an indirect way of stating a preference for LMWH. For cancer patients with established VTE, the recommendation is more direct: LMWH is clearly preferred over UFH for both initial and continuing antithrombotic therapy.27

UFH

UFH, which is typically administered by subcutaneous injection, has the longest history as an anticoagulant in the prevention and treatment of VTE. It is an attractive option in patients with severe renal failure or those who may require a procedure in the near future. Although UFH is partially cleared by the kidney, its short half‐life can be perceived as a safety advantage in patients with severe renal impairment and an increased risk of bleeding. For most other patients, UFH holds several disadvantages compared with newer therapies, including the need for injections to be administered 3 times a day to be optimally effective, its effect on platelets, and its association with heparin‐induced thrombocytopenia (HIT).1 Given the costs of administration and potential complications, it is not less expensive than LMWHs, and it appears to be less cost‐effective.11

LMWHs

LMWHs have a higher bioavailability and longer half‐life than UFH, which translates to reliable anticoagulation levels when given subcutaneously on a weight‐based dosing schedule. Unlike UFH, LMWHs do not require laboratory tests to monitor the intensity of anticoagulation, except in special circumstances.1 The LMWHs dalteparin, enoxaparin, and tinzaparin are widely used for the prevention and treatment of VTE in the United States.

Two landmark clinical trials demonstrated the efficacy of appropriate thromboprophylaxis with LMWHs in reducing the burden of VTE in acutely ill, hospitalized medical patients. The Prophylaxis in Medical Patients with Enoxaparin (MEDENOX) trial and Prospective Evaluation of Dalteparin Efficacy for Prevention of VTE in Immobilized Patients Trial (PREVENT) demonstrated the benefits of enoxaparin and dalteparin, respectively, in reducing the risk of VTE. As shown in Table 2, thromboprophylaxis with these agents was associated with a 45% to 63% relative reduction in the risk of VTE compared with placebo.12, 13

Pentasaccharides

Fondaparinux is a synthetic factor Xa antagonist that shares many features of LMWHs, including a high bioavailability and long half‐life. Fondaparinux does not require monitoring, but it is contraindicated in patients with renal failure (CrCl < 30 mL/minute) and in patients weighing less than 50 kg.1 Although PF4 antibodies have been associated with fondaparinux administration, this drug has not, to date, been associated with HIT.14 The Arixtra for Thromboembolism Prevention in a Medical Indications Study (ARTEMIS) trial demonstrated the advantage of fondaparinux over placebo in reducing the risk of VTE (Table 2).15 The American College of Chest Physicians (ACCP) guidelines state that fondaparinux appears to be as effective and safe as LMWH.1

Vitamin K Antagonists

Vitamin K antagonists (VKAs) such as warfarin inhibit the production of prothrombin, clotting factors VII, IX, and X, and the anticoagulants protein C and protein S. Warfarin is challenging to manage because of its narrow therapeutic window, its tendency to exhibit considerable variability in dose‐response, the time required to reach target international normalized ratio (INR), its potential for interaction with diet and concomitant medications, and its need for ongoing monitoring.5 Warfarin should usually be initiated within the same 24 hours as parenteral anticoagulation, with a goal of achieving INR results between 2.0 and 3.0. An initial dose of 5 to 10 mg for the first 1 or 2 days is appropriate for most patients, and subsequent dosing should be based on INR response.5 Warfarin prophylaxis is primarily used in patients in the US undergoing orthopedic surgery, including total hip replacement and hip and knee arthroplasty.1

Future Anticoagulants

New oral agents have the potential to improve the management of patients who have a moderate to high risk of thromboembolic disease.

Duration

Although major trials support the use of short‐term prophylaxistypically 6 to 14 daysin‐hospital for acutely ill medical patients, the optimal duration of thromboprophylaxis in these patients is unclear.24 The Extended Clinical Prophylaxis in Acutely Ill Medical Patients (EXCLAIM) trial is the first randomized trial to evaluate the potential benefits of extended prophylaxis in acutely ill medical patients. In this study, 5101 hospitalized patients with varying levels of reduced mobility due to cancer, ischemic stroke, heart failure, respiratory failure, infection, and other acute medical conditions received open‐label enoxaparin 40 mg daily for a mean duration of 10 days. Patients were then randomly assigned to additional therapy with enoxaparin or placebo for a mean duration of 28 additional days.25 Preliminary findings from this trial suggest that high‐risk medical patients can benefit from extended thromboprophylaxis following hospital discharge, with significantly reduced VTE events (RR reduction, 44%; P = 0.0011). The benefits of thromboprophylaxis were apparent during the extended treatment period and persistent through 90 days.26

Guideline Recommendations

Incorrect use of thromboprophylaxis does not stem from a lack of evidence‐based recommendations. Within the past year, the ACCP, the American Society of Clinical Oncology (ASCO), and the National Comprehensive Cancer Network (NCCN) have published updated guidelines for thromboprophylaxis in hospitalized patients at risk of VTE.1, 27, 28 The 2008 ACCP guidelines include more than 700 recommendations for VTE risk assessment and management, to be implemented by a variety of physicians, including pulmonologists, cardiologists, cardiothoracic surgeons, and critical care medicine specialists.1

The ACCP guidelines organize prophylaxis recommendations on the basis of patient risk (Table 4).1 Risk assessment remains relatively subjective, however, and validated risk assessment models are not yet widely available. The prudent approach is to consider thromboprophylaxis for all hospitalized medically ill patients who do not have a specific contraindication.

Key evidence‐based recommendations regarding thromboprophylaxis for hospitalized, acutely ill patients include the following:1

• Every hospital should develop a formal strategy to addresses VTE prophylaxis;

• Aspirin alone is not recommended to prevent VTE for any patient group;

• Mechanical methods of thromboprophylaxis should be used primarily for patients who have a high bleeding risk or possibly as an adjunct to anticoagulant thromboprophylaxis;

• Thromboprophylaxis with LMWH, UFH, or fondaparinux is recommended for patients admitted to hospital with an acute medical illness (Note: fondaparinux is recommended, but not FDA‐approved, for this indication in the United States);

• On admission to the ICU, all patients should be assessed for risk of VTE, and most should receive thromboprophylaxis;

• All major trauma and all spinal‐cord injury patients should receive thromboprophylaxis.

Cancer Patients

The ASCO and NCCN guidelines endorse the use of VTE prophylaxis with anticoagulants in all hospitalized patients with active cancer or suspicion of cancer in the absence of contraindications.27, 28 The ACCP guidelines restrict this recommendation to hospitalized cancer patients who are bedridden.1 Thromboprophylaxis should continue at least through the duration of the hospital stay. Acceptable subcutaneous regimens include fondaparinux, dalteparin, or enoxaparin at the doses presented in Table 1; if UFH is chosen, the dose should be 5000 units every 8 hours.

Cancer patients who are scheduled to undergo major surgery require a different prophylaxis strategy. Even with prophylaxis, cancer patients have a 2‐fold higher risk of postoperative VTE compared with noncancer patients and more than a 3‐fold higher risk of fatal PE.30 To manage this risk, the ASCO, NCCN, and ACCP guidelines recommend extended prophylaxis in patients undergoing major cancer surgery.1, 27, 28 Specific recommendations include the following:

• All patients undergoing major surgical intervention for malignant disease should be considered for VTE prophylaxis with anticoagulants, with or without mechanical prophylaxis;

• Thromboprophylaxis should be initiated prior to the start of surgery or as early as possible following surgery;

• Mechanical interventions may supplement pharmacologic prophylaxis, especially in patients who have the highest risk;

• Prophylaxis with a LMWH should be initiated 12 to 24 hours after the surgical procedure;

• Continue prophylaxis at least 7 to 10 days postoperatively;

• Consider prolonged prophylaxis (ie, up to 4 weeks) with a LMWH for high‐risk patients (eg, patients undergoing major abdominal or pelvic surgery, those with residual malignant disease after surgery, obese patients, and patients with a history of VTE).

Routine prophylaxis with anticoagulants is not recommended for most outpatients, except for those with high‐risk factors (eg, thrombogenic chemotherapy or a central venous catheter). The strategy of restricting thromboprophylaxis to cancer outpatients with specific indications, however, may miss an opportunity to reduce VTE in this vulnerable patient population. In the PROTECHT study, 1166 ambulatory cancer patients were randomly assigned to placebo or the LMWH nadroparin for the duration of their chemotherapy. Treatment with nadroparin reduced the rate of clinical thrombosis by 47.2% compared with placebo (3.9% vs. 2.1%; P = 0.033). The risk reduction was consistent across all measured events, including DVT, PE, stroke, and visceral venous thrombosis.31

Pregnancy

Prophylaxis should be considered in pregnant women with known risk factors for VTE such as prior VTE, thrombophilia, and a history of prolonged immobility. In addition, women with a moderate to high risk of VTE associated with a cesarean section should be considered for postpartum thromboprophylaxis. For example, 1 of the following regimens may be appropriate for high‐risk women following a cesarean section:32

• UFH 5000 units subcutaneously every 12 hours until fully mobile;

• LMWH subcutaneously once daily for 5 days (such as enoxaparin 20 mg daily).

For pregnant women already receiving anticoagulant prophylaxis (eg, for hypercoagulable state, structural heart disease, or prior DVT/PE), ACCP guidelines recommend discontinuing VKAs before 6 weeks of fetal gestation to minimize the risk of birth defects and miscarriage. In general, a LMWH should be substituted for VKAs as soon as pregnancy is confirmed or prior to conception in preparation for pregnancy, as VKAs cross the placental barrier, but LMWH and UFH do not.1, 33

Renal Insufficiency

The ACCP guidelines recommend that renal function be considered when making decisions about the use and/or dose of LMWHs and fondaparinux. Because these agents are eliminated primarily via renal clearance, changes in renal function can reduce drug clearance, prolong the half‐life, and increase plasma concentrations. Consequently, the risk of treatment‐related bleeding complications is elevated in patients with renal impairment.1 Depending on the circumstances, one of the following options should be considered1:

• Avoid using an anticoagulant that bioaccumulates in the presence of renal impairment;

• Use a lower dose of the agent;

• Monitor the drug level or its anticoagulant effect.

In severe renal impairment (creatinine clearance < 30 mL/minute):710

• The prophylactic dose of enoxaparin should be adjusted to 30 mg subcutaneously once daily; no specific dosing adjustments have been recommended for dalteparin or tinzaparin;

• Fondaparinux is contraindicated.

Epidural Anesthesia

Neuraxial blockade has several advantages over systemic opioids, but the risk of spinal or epidural hematoma may be increased with the concomitant use of antithrombotic drugs. Therefore, these agents must be used cautiously in patients with neuraxial blockade.1 Guidelines from the American Society of Regional Anesthesia and Pain Medicine (ASRA) contain the following recommendations:34

• Subcutaneous UFH: No contraindication, consider delaying heparin until after block if technical difficulty is anticipated;

• LMWH: Since twice daily dosing may be associated with an increased risk of spinal hematoma, delay initiation of LMWH until at least 24 hours after surgery, regardless of anesthetic technique; for single daily dosing, administer the first dose of LMWH 6 to 8 hours postoperatively and second dose no sooner than 24 hours after the first dose;

• Warfarin: Document normal INR after discontinuation (prior to neuraxial technique); remove catheter when INR 1.5 (initiation of therapy).

Bleeding Risk

The ACCP and ASCO guidelines emphasize the importance of weighing the potential benefits of thromboprophylaxis against the potential risks of bleeding in individual patients. According to the ACCP, the overall risk of bleeding with intravenous UFH in patients with VTE is less than 3%, and thromboprophylaxis has not been shown to increase the risk of bleeding compared with placebo in major clinical trials.13, 15, 35 However, bleeding risk may increase in older patients and with higher doses of heparin. Warfarin therapy can be monitored with an INR to reduce the risk of bleeding during thromboprophylaxis.1

Anticoagulation therapy may be contraindicated in patients with certain factors and conditions that increase the risk of bleeding. These include:

• Clinically significant active or chronic bleeding;

• Recent central nervous system or spinal surgery with increased risk of bleeding;

• Thrombocytopenia (excluding HIT) or severe platelet dysfunction;

• Abnormalities associated with clotting factors.

The NCCN provides specific contraindications to anticoagulation therapy for the prevention and treatment of VTE in cancer patients.28 These include:

• Recent central nervous system bleed; intracranial, or spinal lesions at high risk of bleeding;

• Active major bleeding (> 2 units transfused in 24 hours);

• Chronic, clinically significant measurable bleeding for more than 48 hours;

• Thrombocytopenia (platelets < 50,000/L);

• Severe platelet dysfunction;

• Recent major operation with high risk of bleeding;

• Underlying coagulopathy (eg, clotting factor abnormalities or elevated prothrombin time or activated partial thromboplastin time [aPTT]);

• Spinal anesthesia or lumbar puncture;

• High risk of falls.

HIT

HIT is a serious complication that can occur as a result of exposure to heparin. It is an immune response that causes platelet activation and platelet aggregation, among other effects, and is capable of leading to severe thrombosis, amputation, or death.36 The incidence of HIT varies with subpopulations of patients and more commonly develops in patients receiving heparin in therapeutic doses. Early diagnosis (through an interpretation of clinical and laboratory information) is important to improve clinical outcomes, but difficult to achieve.36 The ACCP guidelines note that enzyme‐linked immunosorbent assay (ELISA)‐based tests for HIT are often falsely positive after surgery. As an alternative, serotonin‐release tests are more specific, although they are not as widely available.1

Substantial clinical evidence suggests that LMWH poses less of a risk of HIT than UFH. Martel et al,37 for example, conducted a meta‐analysis of 15 randomized and nonrandomized controlled trials (a total of 7287 patients) that included studies that compared prophylactic doses of UFH and LMWH and assessed postoperative or medical inpatients who received prophylaxis. The analysis revealed that the risk of HIT was 2.6% following UFH use compared with 0.2% following LMWH use.37 Despite the inclusion of UFH in the ASCO guidelines, ASCO acknowledges that a lower risk of HIT is one of the potential advantages of LMWH over UFH in cancer surgery prophylaxis.27 In addition, the recommendation to transition to outpatient therapy as soon as possible is an indirect way of stating a preference for LMWH. For cancer patients with established VTE, the recommendation is more direct: LMWH is clearly preferred over UFH for both initial and continuing antithrombotic therapy.27

Data Source

The Premier Perspective database is a patient‐level dataset of administrative, billing, and discharge information used for comparative analysis of clinical performance. This database contains approximately 5.5 million patient discharges per year from nearly 500 not‐for‐profit, nongovernmental, community, and teaching hospitals and health systems. Hospitals submit data to the Premier Perspective database on a monthly basis, with the data undergoing numerous quality and validation checks on entry.

Data Collection

All patient records used in this study were di‐dentified in compliance with the Health Insurance Portability and Accountability Act (HIPAA) of 1996 (http://www.hhs.gov/ocr/hipaa). Hospital and patient demographics, discharge information, principal and secondary diagnoses and procedures, and detailed resource consumption information for each discharge by day of hospitalization were extracted from the Premier Perspective database. Records related to the same hospital discharge were linked using a nonpersonal identifier assigned by the provider that prevented subject identification and the linking of identifiers to subjects. Since this study did not involve identifiable human subjects, it was exempt from Institutional Review Board overview under the Common Rule (45 CFR 46.101(b)(4)).18