Improving Practices in US Hospitals to Prevent Venous ...

CLINICAL RESEARCH STUDY

Improving Practices in US Hospitals to Prevent Venous Thromboembolism: Lessons from ENDORSE Frederick A. Anderson, Jr, PhD,a Samuel Z. Goldhaber, MD,b Victor F. Tapson, MD,c Jean-Francois Bergmann, MD,d Ajay K. Kakkar, MD,e Bruno Deslandes, MD,f Wei Huang, MS,a Alexander T. Cohen MD,g the ENDORSE Investigators* a Center for Outcomes Research, University of Massachusetts Medical School, Worcester; bBrigham and Women’s Hospital, Harvard Medical School, Boston, Mass; cDuke University Medical Center, Durham, NC; dHôpital Lariboisière, University Paris 7, Paris, France; eThrombosis Research Institute and Queen Mary University of London, United Kingdom; fsanofi-aventis, Paris, France; g King’s College Hospital, London, United Kingdom.

ABSTRACT BACKGROUND: Venous thromboembolism prophylaxis is suboptimal in the US despite long-standing evidencebased recommendations. The aim of this subset analysis of the Epidemiologic International Day for the Evaluation of Patients at Risk for Venous Thromboembolism in the Acute Hospital Care Setting (ENDORSE) study was to identify characteristics of hospitals with high guideline-recommended prophylaxis use. METHODS: Between September and November 2006, charts from eligible patients aged ⱖ40 years with an acute medical illness or age ⱖ18 years and undergoing a surgical procedure were reviewed from randomly selected US acute-care hospitals. Hospitals were ranked based on the proportion of at-risk patients who received American College of Chest Physicians–recommended types of prophylaxis. Hospital characteristics were compared to determine factors related to more frequent prophylaxis use. Hospitals were followed up 1 year after the chart audit. RESULTS: Overall, 9257 patients were evaluated from 81 hospitals. Appropriate types of prophylaxis were prescribed to more at-risk patients in hospitals in the highest quartile compared with the lowest quartile of prophylaxis use (74% vs 36%). All quartiles had a similar percentage of at-risk patients (61%-65%). Significantly more hospitals in the highest quartile had residency training programs (43% vs 5%), a larger median number of beds (277 vs 140), and had adopted hospital-wide prophylaxis protocols (76% vs 40%). In the follow-up survey, more hospitals overall had adopted hospital-wide written guidelines for venous thromboembolism prevention. CONCLUSIONS: These findings support the value of hospital-wide protocols and local audits for VTE prevention, as recommended by several national quality-of-care groups. © 2010 Elsevier Inc. All rights reserved. • The American Journal of Medicine (2010) 123, 1099-1106 KEYWORDS: Guideline adherence; Prophylaxis; Venous thromboembolism

*A full list of the ENDORSE Investigators is available in Appendix I, which appears in the online version of this article. Funding: This study was sponsored by sanofi-aventis. Editorial support for this publication was funded by sanofi-aventis US, Inc. Conflict of Interest: Dr. J. F. Bergmann receives consultancy fees from AstraZeneca, sanofi-aventis, and GlaxoSmithKline; Dr. A. T. Cohen has received consultancy fees and clinical research funding from AstraZeneca, Bayer, Boehringer-Ingelheim, BMS, Daiichi, GlaxoSmithKline, Johnson & Johnson, Mitsubishi Pharma, Organon, Pfizer, sanofi-aventis, Schering Plough, and Takeda; Dr. V. F. Tapson has received consultancy fees and clinical research funding from Bayer and sanofi-aventis; Dr. S. Z. Goldhaber has received funding from Bayer, Boehringer-Ingelheim, BMS, Eisai, and sanofi-aventis for consultancy and clinical research; Dr. A. K. Kakkar reports receiving consultancy fees, research funding support, and honoraria from sanofi-aventis; Dr. B. Deslandes is an employee of sanofi0002-9343/$ -see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.amjmed.2010.07.011

aventis; Drs. W. Huang and F. A. Anderson, Jr. are employees of the Center for Outcomes Research, which receives research grants from sanofiaventis and The Medicines Company. Dr. F. A. Anderson, Jr. has also received speaker honoraria and consulting contracts from sanofi-aventis, The Medicines Company, Millennium Pharmaceuticals, GlaxoSmithKline, and Johnson & Johnson. Authorship: The authors had full access to the data for this paper and all data analysis was performed by the authors. This article was written by the authors. Editorial support for the final version of this article, comprising language editing, content checking, formatting, and referencing was provided by Daniel Bridges, PhD, and funded by sanofi-aventis US, Inc. Requests for Reprints should be addressed to Frederick A. Anderson, PhD, Center for Outcomes Research, University of Massachusetts Medical School, One Innovation Drive, Suite 110, Worcester, MA 01605. E-mail address: [email protected]

1100

The American Journal of Medicine, Vol 123, No 12, December 2010

Venous thromboembolism is a frequent complication METHODS during and after hospitalization for acute medical illness Procedures and Patients. Between September and Noand surgery. Recent epidemiologic studies estimate that vember 2006, data were obtained for inpatients in 81 approximately one-third of acute care hospital patients randomly selected acute care hospitals in the United are at risk for developing venous thromboembolism.1 The States, participating in the cross-sectional ENDORSE surgeon general has cited venous survey.4 Hospital wards were elthromboembolism as the most igible if they were predomicommon preventable cause of innately occupied by acutely ill CLINICAL SIGNIFICANCE hospital death.2 Prophylaxis can medical patients, or surgical pareduce the incidence of venous tients requiring general or epi● This study highlights the current underthromboembolism substantially. dural anesthesia lasting at least use of guideline-recommended prophyOrganizations such as the Amer45 minutes (eg, general medical, laxis in hospitalized medical and surgical ican College of Chest Physicians respiratory, cardiac, general surpatients at risk of venous thromboembo(ACCP) have developed evigical, and orthopedic wards). All lism in the US. dence-based recommendations eligible wards within enrolled regarding its appropriate use.3 ● Findings from this study suggest that hospitals were included in the Despite the existence of guidethe adoption of hospital-wide thrombostudy. Patients aged ⱖ40 years lines for evidence-based pracin eligible medical wards or prophylaxis protocols, coupled with petice, recent studies of real-world those aged ⱖ18 years in eligible riodic local audits of practice, could populations demonstrate that vesurgical wards were screened; substantially improve current prophynous thromboembolism prophypatients admitted for the treatlaxis practices. laxis implementation remains ment of venous thromboembo● Improving adherence with best-practice suboptimal.4-8 The Epidemiolism were excluded. Patient logic International Day for the guidelines is urgently required to recharts were reviewed using recEvaluation of Patients at Risk for duce the substantial clinical and ecoommendations from the 2004 Venous Thromboembolism in ACCP guidelines to evaluate venomic burden associated with venous the Acute Hospital Care Setting nous thromboembolism risk and thromboembolism in the US. (ENDORSE) study assessed the prophylaxis use (Appendix II, onproportion of at-risk patients line).11 Only the type of prophylaxis who received effective types of (and not the dose or duration) was prophylaxis in 358 hospitals in 32 countries during considered when assessing whether prophylaxis was compliant 2006.4 Only two-thirds of at-risk surgical patients and with ACCP guideline recommendations. fewer than half of at-risk medical patients received any form of prophylaxis. Furthermore, appropriate types of Selection of Participating Hospitals. The American Hosprophylaxis, according to the 2004 ACCP guidelines, pital Association (AHA) list of acute care hospitals was were received by 59% of surgical and 40% of medical used to select 50 hospitals at random in each of 8 US census patients.4 Analyses of large US managed-care databases regions, for a total of 400 possible candidate hospitals.12 have found that the majority of patients prescribed veHospitals were considered eligible for enrollment if they nous thromboembolism prophylaxis do not receive contained more than 50 beds, admitted patients for the ACCP guideline-recommended prophylaxis.6,7 treatment of medical illnesses and exacerbations of chronic The US National Quality Forum (NQF) has recognized diseases, and scheduled routine major surgical procedures. the need to optimize the prevention and care of venous After contacting eligible hospitals from the list, 100 agreed thromboembolism. In collaboration with the Joint Comto participate, but 13 hospitals failed to obtain institutional mission, the NQF recently developed standard perforapproval, and an additional 6 failed to collect patient data. mance measures in order to improve venous thromboemTherefore, 81 US hospitals that had completed the ENbolism risk assessment, prophylaxis, diagnosis, and DORSE protocol were included in this analysis. Based on treatment.9 Resources for improving venous thromboemthe AHA database,12 the key organizational characteristics bolism prophylaxis practices, such as the Agency for of the hospitals that chose not to participate in this study, Healthcare Research and Quality’s “Preventing Hospitalincluding number of licensed beds and academic affiliation, Acquired Venous Thromboembolism” publication, can were similar to those of hospitals that completed the study. help increase the understanding of factors which impact Ethical committee approval to undertake the survey was on prophylaxis practices, and thus may drive hospitalobtained by the study lead investigator at participating hoswide improvements.10 The objective of this subanalysis pitals. The requirement for patient consent was waived by of the ENDORSE study was to compare prophylaxis all 81 US hospital ethics committees. practices in representative US hospitals and to identify characteristics that are associated with better delivery of Data Collection. Data were collected from a review of venous thromboembolism prevention. hospital charts on standard case report forms by trained data

Anderson et al Table 1

Improving Thromboprophylaxis Practices in US Hospitals

1101

Risk Factors for VTE by Hospital Quartile According to Prescription of ACCP-Recommended Prophylaxis

No. (%) No. of hospitals Patients in surgical wards At ACCP-defined VTE risk Received VTE prophylaxis Median no. of VTE risk factors Major orthopedic surgery Abdominal/thoracic surgery Major urologic/gynecologic surgery Other major surgery Patients in medical wards At ACCP-defined VTE risk Received VTE prophylaxis Median no. of VTE risk factors Admitted to ICU Immobile ⬎24 hours Active cancer

Quartile 1 Lowest

Quartile 2

Quartile 3

Quartile 4 Highest

P Value for Linear Trend

20 396 (32) 335 (85) 151 (45) 3.6

20 915 (41) 688 (75) 425 (62) 4.1

20 1055 (44) 864 (82) 594 (69) 4.2

21 1695 (50) 1278 (75) 1074 (84) 4.1

⬍.001 .009 ⬍.001 .009

51 (15) 87 (26) 38 (11)

101 (15) 186 (27) 61 (9)

120 (14) 226 (26) 63 (7)

199 (16) 326 (26) 86 (7)

.312 .182 .002

124 (37) 850 (68) 454 (53) 131 (29) 3.7

300 1298 665 274 4.0

190 (42) 143 (32) 38 (8)

(44) (59) (51) (41)

382 1356 694 340 4.3

278 (42) 202 (30) 53 (8)

(44) (56) (51) (49)

288 (42) 248 (36) 39 (6)

579 1692 907 547 4.1

(45) (50) (54) (60)

.001 ⬍.001 .65 ⬍.001 .002

398 (44) 375 (41) 145 (16)

.42 ⬍.001 ⬍.001

ACCP ⫽ American College of Chest Physicians; ICU ⫽ intensive care unit; VTE ⫽ venous thromboembolism.

abstractors, which included physicians, nurses, and other hospital staff. Each ward was assessed on a single, prespecified day to complete collection of patient data. For inclusion, every patient must have been an inpatient in an eligible ward on the day it was surveyed. In this analysis, the 81 US hospitals that completed the ENDORSE study were ranked based on the proportion of at-risk patients who received effective types of prophylaxis for venous thromboembolism according to the 2004 ACCP recommendations. Hospitals were divided into 4 groups of 20 or 21 hospitals based on level of prophylaxis use. Various selfreported characteristics (eg, hospital type, number of beds, and the presence of a hospital venous thromboembolism prevention protocol) were recorded for hospitals in each quartile.

Approximately 1 year after distribution of the ENDORSE audit findings to participating hospitals, a supplementary follow-up survey was sent to all participating hospitals to obtain further information regarding changes in hospital characteristics that might affect prophylaxis practices, including adoption of a hospital-wide venous thromboembolism prevention protocol, if any, and whether hospitals reported their performance on prevention of venous thromboembolism to an outside agency. Statistical Analysis. Quantitative data was summarized as median (interquartile range). Categorical data was summarized as the number and percentage of the population. Summary data were reported globally and by quartile of pro-

Table 2 Prophylaxis Practices and Characteristics of Hospitals Ranked into Quartiles According to Prescription of ACCPRecommended Prophylaxis No. (%) No. of hospitals Hospitals with teaching programs Hospitals with a formal VTE prophylaxis protocol at audit Median no. of hospital beds Eligible patients distribution (No. ⫽ 9257) Patients at VTE risk Patients prescribed ACCP-recommended prophylaxis among patients at VTE risk Patients prescribed any prophylaxis among patients at VTE risk

Quartile 1 Lowest 20

Quartile 2 20

Quartile 3

Quartile 4 Highest

P-Value for Linear Trend

1 (5) 8 (40)

1 (5) 14 (70)

20 3 (15) 14 (70)

21 9 (43) 16 (76)

140 1246 (14) 789 (63) 282 (36)

242 2213 (24) 1353 (61) 699 (52)

278 2411 (26) 1558 (65) 934 (60)

277 3387 (37) 2185 (65) 1621 (74)

.061 ⬍.001

392 (50)

929 (69)

1117 (72)

1857 (85)

⬍.001

ACCP ⫽ American College of Chest Physicians; VTE ⫽ venous thromboembolism.

⬍.001 .001 .002 –

1102


Figure 1 Proportion of hospitals with a formal venous thromboembolism prophylaxis protocol by quartile of hospitalwide prophylaxis use in at-risk patients. VTE ⫽ venous thromboembolism.

phylaxis use. Use of ACCP-recommended types of prophylaxis in medical and surgical patients at risk of venous thromboembolism was assessed using the chi-squared test for linear trend. SAS version 9.1 (Cary, NC) was used for all statistical analyses.

RESULTS Hospital and Patient Characteristics. The 81 participating hospitals were located in 37 US states and included 14 teaching hospitals and 67 community-based hospitals. Of the 23,554 licensed beds in the 81 participating hospitals, 9666 beds were in ineligible wards and 2489 of the beds in eligible wards were unoccupied on the day of audit. An additional 2142 patients were ineligible for the study, resulting in 9257 evaluable patients. Almost two-thirds of evaluable patients (64%) were at ACCP-defined risk of venous thromboembolism, and 60% of these patients received ACCP-recommended types of prophylaxis. For medical patients (n ⫽ 5196), 52% of patients were at risk of venous thromboembolism, and of these, 48% were prescribed appropriate types of prophylaxis (Table 1). For surgical patients (n ⫽ 4061), 78% of patients were at risk of venous thromboembolism, and 71% of these received appropriate types of prophylaxis (Table 1). When hospitals were ranked based on their use of ACCP-recommended types of prophylaxis, the proportion of at-risk patients who received appropriate prophylaxis was 36% in the 20 hospitals in the lowest quartile compared with 74% in 21 hospitals in the highest quartile (P ⬍.001 for trend; Table 2). The large difference in the prescribing of appropriate prophylaxis between hospitals in the lowest and highest quartiles occurred despite a similar risk profile for venous thromboembolism; 61 to 65% of patients were atrisk of venous thromboembolism across the quartiles (P ⫽ .061; Table 2). An additional 11% to 17% of patients in each quartile received a type of prophylaxis judged in-

adequate by the ACCP guidelines (Table 2). When hospital characteristics were evaluated, significantly more hospitals in the highest quartile were teaching hospitals compared with the lowest quartile (9 of 21 [43%] vs 1 of 20 [5%]; P ⫽ .001). In addition, significantly more hospitals in the highest quartile had a formal, hospital-wide venous thromboembolism prophylaxis protocol compared with the lowest quartile (16 of 21 [76%] vs 8 of 20 [40%]; P ⬍.001; Figure 1). Hospital size also affected venous thromboembolism prophylaxis practices, with the median number of beds significantly greater in the highest quartile than the lowest quartile (277 vs 140; P ⬍ .002). The percentage of patients in surgical wards increased from 32% in the lowest quartile to 50% in the highest quartile group (P ⬍.001; Table 1). Separate examination of the medical and surgical at-risk populations demonstrated a correlation between the use of ACCP-recommended types of prophylaxis in at-risk medical compared with at-risk surgical patients, r ⫽ 0.42 (P ⬍.001; Figure 2). A separate analysis of hospital rank based on prophylaxis use in medical patients demonstrated similar findings. Significantly more hospitals in the highest quartile of prophylaxis use in at-risk medical patients were teaching hospitals compared with the lowest quartile (6 of 20 [30%] vs 0 of 19 [0%]; P ⫽ .001). In addition, significantly more hospitals in the highest quartile had a formal, hospital-wide venous thromboembolism prophylaxis protocol compared with the lowest quartile (15 of 20 [76%] vs 11 of 19 [40%]; P ⬍.001). Hospital size also affected prophylaxis practices, with the median number of beds significantly greater in the highest quartile than the lowest quartile (254 vs 213; P ⬍.01).The type of prophylaxis received by patients in the different quartiles was also compared between patients at ACCP-defined risk and those at low risk of venous thromboembolism (Table 3). A greater proportion of at-risk patients in the highest quartile received anticoagulants compared with at-risk patients in the lowest quartile (63.6% vs 35.1%). A larger proportion of at-risk patients, both those at ACCP-defined risk and those at low risk of venous throm-

Figure 2 Prophylaxis use in at-risk medical and surgical patients by quartile.

Anderson et al


1103

Table 3 Types of VTE Prophylaxis Received by At-Risk Patients as Defined by the ACCP Guidelines versus Those at Low Risk of VTE in Hospitals, Ranked by Quartiles According to Prescription of ACCP-Recommended Prophylaxis No. (%) No. of hospitals Eligible patients Patients at ACCP-defined VTE risk Anticoagulants Intermittent pneumatic compression alone Graduated compression stockings alone Acetylsalicylic acid alone No prophylaxis Patients at low VTE risk Anticoagulants Intermittent pneumatic compression alone Graduated compression stockings alone Acetylsalicylic acid alone No prophylaxis

Quartile 1 Lowest

Quartile 2

Quartile 3

Quartile 4 Highest

20 1246 789 (63.3) 277 (35.1) 73 (9.3)

20 2213 1353 (61.1) 594 (43.9) 234 (17.3)

20 2411 1558 (64.6) 691 (44.4) 364 (23.4)

21 3387 2185 (64.5) 1389 (63.6) 428 (19.6)

32 (4.1) 10 (1.3) 397 (50.3) 394 85 (21.6) 10 (2.5)

93 (6.9) 8 (0.6) 424 (31.3) 648 187 (28.9) 57 (8.8)

50 (3.2) 12 (0.8) 441 (28.3) 682 158 (23.2) 78 (11.4)

33 7 328 798 404 96

12 (3.1) 4 (1.0) 283 (71.8)

29 (4.5) 5 (0.8) 370 (57.1)

20 (2.9) 7 (1.0) 419 (61.4)

16 (2.0) 7 (0.9) 275 (34.5)


⬍.001

(1.5) (0.3) (15.0) (50.6) (12.0)

⬍.001

ACCP ⫽ American College of Chest Physicians; VTE ⫽ venous thromboembolism.

boembolism, received no prophylaxis in the lowest quartile compared with the highest quartile (Table 3). Follow-Up Survey. Participating hospitals were sent a supplementary follow-up survey 1 year after the original medical record audit, with 79 of the 81 participating hospitals returning the form (Table 4; Appendix III, online). The number of hospitals that had a venous thromboembolism prophylaxis protocol according to the follow-up survey increased in all quartiles, but with persistently greater use in the highest quartile compared with the lowest quartile (Figure 3). Furthermore, the number of hospitals in the highest quartile with a venous thromboembolism prophylaxis protocol based on ACCP guidelines increased: 11 of 21 in the original audit vs 15 of 21 in the 12-month follow-up survey. Only 10 hospitals in the lowest quartile stated that they reported on performance compared with 16 or 17 hospitals in the other quartiles (Table 4).

DISCUSSION In this real-world study of US medical and surgical patients at risk of venous thromboembolism, prescribing rates for guideline-recommended prophylaxis were low. Appropriate prophylaxis was received by 36% of at-risk patients in the lowest quartile compared with 74% of patients in the highest-performing quartile of hospitals. These results are consistent with those obtained in the multinational analysis of the ENDORSE study, where 40% of medical and 59% of surgical patients received guideline-recommended prophylaxis.4 The results of the present analysis are also consistent with studies that used large managed-care databases6,7 and registries5,8 to assess prophylaxis practices in the United States. In the International Medical Prevention Registry on Venous Thromboembolism, prophylaxis was only administered to 61% of the acutely ill medical patients in the United States (n ⫽ 15,156) who should have received prophylaxis

Table 4 Characteristics of Hospitals Ranked into Quartiles According to Prescription of ACCP-Recommended Prophylaxis Based on Findings from the Supplementary Follow-Up Survey No. (%) No. of hospitals Returned the survey With a formal VTE prophylaxis protocol With a VTE prophylaxis protocol based on ACCP guidelines Report performance

Quartile 1 Lowest

Quartile 2

Quartile 3

Quartile 4 Highest

Total


20 19 (95.0) 14 (73.7*) 11 (78.6†)

20 20 (100) 17 (85.0*) 14 (82.4†)

20 19 (95.0) 17 (89.5*) 15 (88.2†)

21 21 (100) 19 (90.5*) 15 (79.0†)

81 79 (97.5) 67 (84.8*) 55 (82.1†)

— .136 .76

10 (52.6*)

16 (80.0*)

17 (89.5*)

16 (76.2*)

59 (75.6*)

.040

ACCP ⫽ American College of Chest Physicians; VTE ⫽ venous thromboembolism. *As a percentage of hospitals which returned the survey. †As a percentage of hospitals with a VTE prophylaxis protocol.

1104


Figure 3 After distribution of the findings of the ENDORSE audit to participating hospitals, a post-audit survey demonstrated increases in adoption of hospital-wide protocols for venous thromboembolism prevention. VTE ⫽ venous thromboembolism.

based on the ACCP guidelines.5 As a result of its examination of patient characteristics and prophylaxis practices in randomly selected US acute care hospitals, ENDORSE provides more generalizable findings compared with previous audits that were based on selected groups of hospitals or patients. Moreover, this subanalysis is unique in providing information about hospital characteristics that may be associated with superior prophylaxis practices. Factors significantly associated with higher prophylaxis use included hospital teaching status, size, adoption of hospital-wide venous thromboembolism prophylaxis protocols, and a larger percentage of at-risk surgical versus medical patients. Although teaching status and hospital size are difficult characteristics to modify, implementation of an institution-wide prophylaxis policy represents a relatively simple and inexpensive strategy with which hospitals can improve venous thromboembolism prevention. Indeed, both the ACCP consensus statement3 and the NQF13 recommend “that every healthcare facility should have a written policy that is evidence-based and that drives continuous quality improvement related to venous thromboembolism risk assessment, prophylaxis, diagnosis, and treatment.” The NQF developed its statement of policy, preferred practices, and performance measures in recognition of the need to reduce the US venous thromboembolism burden by improving prevention, and also the care of patients with venous thromboembolism. In a 2003 study of US venous thromboembolism disease burden, more than 12 million of the 38 million patients (31%) discharged from US acute hospitals were estimated to be at risk of venous thromboembolism.1 Furthermore, venous thromboembolism is associated with significant mortality and morbidity, and considerable economic burden.14,15 In a large, retrospective analysis, total annual healthcare costs of a venous thromboembolism event ranged from $7594 to $16,644, depending on the type of event and whether venous thromboembolism was a primary or secondary diagnosis.14 An additional push toward improved venous thromboembolism prophylaxis patterns in US hospitals is being provided by the ongoing implementation of hospital perfor-

mance reporting. For example, both the Joint Commission and the Surgical Care Improvement Project have recently developed performance measures which aim to improve inhospital prophylaxis use in at-risk patients.9,16 Other initiatives also may help to improve upon current practices. For example, in a study of surgical and medical patients, the use of a computer alert linked to prophylaxis guidelines significantly increased orders for prophylaxis, from 15% to 34% (P ⬍.001) and significantly reduced the risk of symptomatic imaging-confirmed venous thromboembolism by 41% compared with the control group where no alert was issued (P ⫽ .001).17 Integrated quality improvement initiatives incorporating several active strategies may be more effective at improving venous thromboembolism prophylaxis practices compared with a single strategy.18,19 A study that combined regular physician education, a decision support tool, and a mechanism for audit and feedback found significant improvement in the proportion of medically ill patients receiving appropriate prophylaxis, from 43% to 85% after 18 months (P ⬍.01 vs baseline).18 The relative independence of thromboprophylaxis practices observed in medical and surgical wards (r ⫽ 0.42) suggests that evidence of good practices in at-risk surgical patients does not necessarily guarantee good performance in at-risk medical patients. Consequently, physician and nurse education, design of practice guidelines, and audit of physician performance should cover medical and surgical patients separately. Study limitations include a lack of data on dose or duration of prophylaxis, due to the cross-sectional design. Thus, the ACCP definition of appropriate prophylaxis did not take into account whether the patient received prophylaxis for the guideline-recommended duration, which may mean that some of the prophylaxis included in this study did not meet all criteria for best practice. There also may be possible limitations based on the use of patient charts rather than interviews, which could have led to some inaccuracies and failure to capture the most recent information. However, the approach used obviated the need for patient consent, thereby increasing the representative nature of the patient sample. Strengths of this study include the participation of 81 randomly selected US acute care hospitals, which increases the likelihood that these hospitals are representative of US hospitals as a whole. In addition, the hospital charts of all patients hospitalized in eligible wards were reviewed, providing a broad view of patient risk factors and prophylaxis practice. When recruiting hospitals, we contacted the Director of Quality Assurance (QA) to assess interest in participating in the ENDORSE study. Anecdotal comments from QA directors who declined to participate showed that many felt that their staff was too busy to take on new projects. A small number of QA directors (⬍10) stated that they had recently completed a self-assessment audit of local practices in the prevention of venous thromboembolism and were not interested in beginning a new audit. Although we found no significant differences between hospital practices by geo-

Anderson et al


graphic region, our sample of 81 hospitals was inadequate to detect modest differences across the 8 US Census regions included in this study. Also, our sample of hospitals was too small to allow robust multivariable modeling of our findings. Physicians who were aware of the study may have altered their prescribing behavior; however, the data still show a large gap between ideal and real-world practices despite the theoretical possibility that physician awareness might have led to increased use of prophylaxis. To try to alleviate this issue, prior to the ENDORSE chart audit, we asked the lead investigator at each study hospital not to inform physicians that a study was being conducted in their hospital. After chart audits were completed, we mailed a confidential summary report (which included local findings and benchmarks) to all 81 participating US hospitals. The lead investigator at each hospital (typically the director of QA) was encouraged to share ENDORSE findings at grand rounds or similar venues. While the significant association between the existence of local practice guidelines and prophylaxis practices demonstrated by our chart audit (Figure 1) suggested that feedback of local audit findings may be associated with significantly better local prophylaxis practices, we did not have the resources to perform a follow-up chart audit to confirm an improvement in prophylaxis practice following distribution of audit reports. Nonetheless, after distribution of the ENDORSE audit findings, the uptake of practice guidelines by many of the study hospitals was encouraging (Figure 3). Questions that remain to be answered include: are guidelines developed from a systematic review of the literature, such as those published by the ACCP, suitable for individual patient decision making? Can better prediction models be derived from large observational databases? Are improvements in prophylaxis practice maintained over time? Do hospitals that report their performance to the Surgical Care Improvement Project16 perform better? What is the frequency of inadequate prophylaxis administration (eg, wrong dose, inadequate duration)? Given the decreasing availability of autopsy as a gold standard to document fatal pulmonary embolism, can the real-world benefit of prophylaxis be assessed from overall mortality rates? Given the cost and time required for this type of manual chart audit, these questions may never be answered; however, the new US Federal mandate to establish interoperable Electronic Health Records may soon allow examination of these questions in a more robust, timely, and cost-effective manner.20

CONCLUSIONS This study highlights the current underuse of guideline-recommended prophylaxis in hospitalized medical and surgical patients at risk of venous thromboembolism in the United States. Findings from this study suggest that the adoption of hospitalwide thromboprophylaxis protocols, coupled with periodic local audits of practice, could substantially improve current prophylaxis practices. Improving adherence with best-practice guidelines is urgently required to reduce the substantial clinical

1105

and economic burden associated with venous thromboembolism in the United States.

ACKNOWLEDGMENTS The authors would like to thank the physicians and study coordinators participating in the ENDORSE study; the staff at the Center for Outcomes Research; and Olivia Wu, PhD, Division of Cardiovascular & Medical Sciences, University of Glasgow, United Kingdom. The ENDORSE study was conducted by the Center for Outcomes Research, University of Massachusetts Medical School, with financial support from sanofi-aventis. All data analyses were conducted by Center for Outcomes Research statisticians under the guidance of the authors. Editorial support for the final version of this article, comprising language editing, content checking, formatting, and referencing was provided by Daniel Bridges, PhD, and funded by sanofi-aventis US, Inc. ENDORSE is overseen by an independent medical advisory board of clinicians. Further information about the registry can be found at http://www.outcomes.org/endorse.

References 1. Anderson FA Jr, Zayaruzny M, Heit JA, et al. Estimated annual numbers of US acute-care hospital patients at risk for venous thromboembolism. Am J Hematol. 2007;82:777-782. 2. US Department of Health and Human Services. The Surgeon General’s Call to Action to Prevent Deep Vein Thrombosis and Pulmonary Embolism, 2008. Available at: http://www.surgeongeneral.gov/topics/ deepvein/index.html. Accessed January 18, 2010. 3. Geerts WH, Bergqvist D, Pineo GF, et al. Prevention of venous thromboembolism: American College of Chest Physicians EvidenceBased Clinical Practice Guidelines (8th Edition). Chest. 2008;133(6 Suppl):381S-453S. 4. Cohen AT, Tapson VF, Bergmann JF, et al; ENDORSE Investigators. Venous thromboembolism risk and prophylaxis in the acute hospital care setting (ENDORSE study): a multinational cross-sectional study. Lancet. 2008;371:387-394. 5. Tapson VF, Decousus H, Pini M, et al; IMPROVE Investigators. Venous thromboembolism prophylaxis in acutely ill hospitalized medical patients: findings from the International Medical Prevention Registry on Venous Thromboembolism. Chest. 2007;132:936-945. 6. Amin A, Stemkowski S, Lin J, Yang G. Thromboprophylaxis rates in US medical centers: success or failure? J Thromb Haemost. 2007;5: 1610-1616. 7. Amin AN, Stemkowski S, Lin J, Yang G. Preventing venous thromboembolism in US hospitals: are surgical patients receiving appropriate prophylaxis? Thromb Haemost. 2008;99:796-797. 8. Friedman RJ, Gallus AS, Cushner FD, et al; Global Orthopaedic Registry Investigators. Physician compliance with guidelines for deepvein thrombosis prevention in total hip and knee arthroplasty. Curr Med Res Opin. 2008;24:87-97. 9. The Joint Commission. Performance Measurement Initiatives. Venous Thromboembolism (VTE) Core Measure Set. Last updated 4/2009. Available at: http://www.jointcommission.org/PerformanceMeasure ment/PerformanceMeasurement/VTE.htm. Accessed January 18, 2010. 10. Maynard G, Stein J. Preventing Hospital-Acquired Venous Thromboembolism: A Guide for Effective Quality Improvement. Prepared by the Society of Hospital Medicine. AHRQ Publication No. 08-0075. Rockville, MD: Agency for Healthcare Research and Quality; August 2008. Available at: http://www.ahrq.gov/qual/vtguide/vtguide.pdf. Accessed January 18, 2010.

1106


11. Geerts WH, Pineo GF, Heit JA, et al. Prevention of venous thromboembolism: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest. 2004;126(3 Suppl):338S-400S. 12. American Hospital Association Annual Survey Database for Fiscal Year 2004. Published by Health Forum, An American Hospital Association Company, Chicago, IL, 2006. 13. National Quality Forum. Executive Summary National Voluntary Consensus Standards for Prevention and Care of Venous Thromboembolism: Policy, Preferred Practices, and Initial Performance Measures. Available at: http://www.qualityforum.org/Publications/2006/12/ Executive_Summary_for_VTE_Consensus_Report.aspx. Accessed January 18, 2010. 14. Spyropoulos AC, Lin J. Direct medical costs of venous thromboembolism and subsequent hospital readmission rates: an administrative claims analysis from 30 managed care organizations. J Manag Care Pharm. 2007;13:475-486.

15. MacDougall DA, Feliu AL, Boccuzzi SJ, Lin J. Economic burden of deep-vein thrombosis, pulmonary embolism, and post-thrombotic syndrome. Am J Health Syst Pharm. 2006;63(20 Suppl 6):S5-S15. 16. QualityNet. Surgical Care Improvement Project. Venous thromboembolism. Available at: http://www.qualitynet.org/dcs/ContentServer?c⫽ MQParents&pagename⫽Medqic%2FContent%2FParentShellTemplate &cid⫽1137448680860&parentName⫽Topic. Accessed January 18, 2010. 17. Kucher N, Koo S, Quiroz R, et al. Electronic alerts to prevent venous thromboembolism among hospitalized patients. N Engl J Med. 2005; 352:969-977. 18. Cohn SL, Adekile A, Mahabir V. Improved use of thromboprophylaxis for deep vein thrombosis following an educational intervention. J Hosp Med. 2006;1:331-338. 19. Michota FA. Bridging the gap between evidence and practice in venous thromboembolism prophylaxis: the quality improvement process. J Gen Intern Med. 2007;22:1762-1770. 20. Blumenthal D. Launching HITECH. N Engl J Med. 2010;362:382-385.

Anderson et al Appendix I


1106.e1

List of Participating US Hospitals and Investigators

ENDORSE is overseen by a medical advisory board of clinicians. The names of the participating hospitals and study coordinators are listed below. Further information about ENDORSE can be found at http://www.outcomes.org/endorse. Principal Investigator – Victor Tapson, MD, Duke University (Durham, NC) Investigators and coordinators representing 81 participating USA hospitals: Fred Anderson, PhD, University Of Massachusetts Memorial Medical Center (Worcester, Mass) Anthony Antonacci, MD, Christ Hospital (Jersey City, NJ) Monica Aragon, Alta Vista Regional Hospital (Las Vegas, NM) Gennfer Baker, CNS, RN, Cullman Regional Medical Center (Cullman, Ala) Richard Beaver, MD, VP of Quality, Sewickley Valley Hospital and The Medical Center, Beaver (Sewickley, Pa) William Becker, VAMC, Fargo (Fargo, ND) Cynthia Beckett, PhD, RNC, Flagstaff Medical Center (Flagstaff, Ariz) Susan Black, Harbor UCLA Medical Center (Torrance, Calif) Carla Bomben, MPA, San Joaquin General Hospital (Stockton, Calif) Judith Bonomi, RN, Rush-Copley Medical Center (Aurora, Ill) Carla Borchardt, RN, Avera McKennan (Sioux Falls, SD) Marianne Brassil, RN, New York University Hospital System (New York, NY) Linda Brauner, Tampa General Hospital (Tampa, Fla) Michael Briggs, MD, Innovis Health (Fargo, ND) Joseph Bruckman, Floyd Memorial Hospital and Health Services (New Albany, Ind) Marlene Brunswick, Blanchard Valley Regional Health Center (Findlay, Ohio) Lisa Buck, RN, Wadley Regional Medical Center (Texarkana, Tex) Janny Byrnes, New York University Hospital System (New York, NY) Philip Caushaj, Western Pennsylvania Hospital (Pittsburgh, Pa) Alan Chausow, El Camino Hospital (Mountain View, Calif) Danielle Christopulos, Ogden Regional Medical Center (Ogden, Utah) Paula Cica, UPMC Horizon (Greenville, Pa) Jean Cody, MS, RN, Southwestern Vermont Medical Center (Bennington, Vt) Bruce Condit, MD, Central Maine Medical Center (Lewiston, Me) Ruth Crawley, Fort Sanders Regional Medical Center (Knoxville, Tenn) Sherill Cronin, PhD, Jewish Hospital (Louisville, Ky) Barbara Cysyk, RN, Upper Chesapeake Health System (Bel Air, Md) Cathy Daly, RN, Mercy Hospital and Medical Center (Chicago, Ill) Sharon Darkovitch, RN, Lake Hospital System, Inc (Painesville, Ohio) Catherine DeLuna, RN, Tampa General Hospital (Tampa, Fla) Jamie Deneau, APRN, Rush-Copley Medical Center (Aurora, Ill) Linda Dettmann, VA Puget Sound Health Care System (Seattle, Wash) Anne Diefendorf, MS, RD, LD, Concord Hospital (Concord, NH) LeAnn Doddridge, Floyd Memorial Hospital and Health Services (New Albany, Ind) Irene Downs, New York University Hospitals System (New York, NY) Steven Driggers, MD, Holy Family Memorial (Manitowoc, Wis) Bonnie Durick, Mercy Medical Center, Oregon (Roseburg, Ore) Paula Ellis, RN, MSN, St. Francis Health Center (Topeka, Kan) Jaeda Elvenia, RN, Wilcox Memorial Hospital (Lihue, Hawaii) David Feinbloom, MD, Beth Israel Deaconess Medical Center (Boston, Mass) Mark Feldman, MD, Presbyterian Hospital of Dallas (Dallas, Tex) Lori Fink, St. Mary’s Hospital and Medical Center (Grand Jct., Colo) Kelly Jo Finnerty, Sewickley Valley Hospital and The Medical Center, Beaver (Beaver, Pa) Marian Fischer, RN, MS, F. F. Thompson Hospital (Canandaigua, NY) James Froehlich, MD, MPH, University of Michigan Health System (Ann Arbor, Mich) Christopher Frost, MD, Methodist Medical Center of Oak Ridge (Oak Ridge, Tenn) Donna Fulmer, F. F. Thompson Hospital (Canandaigua, NY) Brenda Gaffey, McAllen Medical Center (McAllen, Tex) Joseph Gaffey, RN, McAllen Medical Center (McAllen, Tex) Karen Gagnon, Central Maine Medical Center (Lewiston, Me) Ruth Galster, RN, Mercy Medical Center, Oregon (Roseburg, Ore) Richard Geldmeier, Watauga Medical Center (Boone, NC) Joyce George, MPH, RN, St. John Medical Center (Tulsa, Okla)

1106.e2 Appendix I

The American Journal of Medicine, Vol 123, No 12, December 2010 Continued

Vincent Glielmi, Ephrata Community Hospital (Ephrata, Pa) James Goodwin, MD, Harrington Memorial Hospital (Southbridge, Mass) Anita Gottlieb, RN, St. Joseph’s Mercy Health Center (Hot Springs, Ariz) Bruce Gray, DO, Greenville Hospital System - Greenville Memorial (Greenville, SC) Darrell Harrington, MD, Harbor-UCLA Medical Center (Torrance, Calif) Edwina Henry, RN, Medical Center at Terrell (Terrell, Tex) Mary Herman, St. Elizabeth Hospital (Appleton, Wis) Cecilia Hernandez, MD, Sutter Medical Center Sacramento (SMG) (SMH) (Sacramento, Calif) David Hill, Waterbury Hospital Health Center (Waterbury, Conn) Chris Hunter, El Camino Hospital (Mountain View, Calif) Melissa Hutchinson, VA Puget Sound Health Care System (Seattle, Wash) Juanita Johnson, RS, MSN, Albemarle Hospital (Elizabeth City, NC) Roberta Johnson, RN, John C. Lincoln Hospital NM (Phoenix, Ariz) Shandrea Johnson, BS, CCRC, Presbyterian Hospital of Dallas (Dallas, Tex) Sheela Kapre, MD, San Joaquin General Hospital (Stockton, Calif) Robyn Kedzie, RN, Midland Memorial Hospital (Midland, Tex) Harvey Kowaloff, MD, Jordan Hospital Inc. (Plymouth, Mass) John Krodel, MD, Norman Regional Hospital (Norman, Okla) Francis LaMonica, RN, Mercy Hospital and Medical Center (Chicago, Ill) James Lehman, MD, Genesis Medical Center - Davenport (Davenport, Iowa) Florence Lynch, Christ Hospital (Jersey City, NJ) Marlene Mach, Harrington Memorial Hospital (Southbridge, Mass) Coletta Manning, Methodist Medical Center of Oak Ridge (Oak Ridge, Tenn) Terri Marney, RN, Plains Regional Medical Center (Clovis, NM) Ingeborg Marshall, Gulf Health Hospitals, Inc., DBA Thomas Hospital (Fairhope, Ala) Mary McDaniels, RN, Catawba Valley Medical Center (Hickory, NC) Mary McLaughlin, Mercy Hospital and Medical Center (Chicago, Ill) Wendy McMahon, Methodist Medical Center of Oak Ridge (Oak Ridge, Tenn) Sherron Meeks, RN, Midland Memorial Hospital (Midland, Tex) Robert Middleton, III, Southwest Mississippi Regional Medical Center (McComb, Miss) Phillip Mitchell, Fort Sanders Regional Medical Center (Knoxville, Tenn) Christine Murphy, RN, Jordan Hospital Inc. (Plymouth, Mass) Allyson Navyac, University of Michigan Health System (Ann Arbor, Mich) Carrie Nelson, MD, Rush-Copley Medical Center (Aurora, Ill) Susan Netherland, Morehead Memorial Hospital (Eden, NC) Kathy New, Three Rivers Medical Center (Louisa, Ky) Chris Norris, PharmD, Fort Sanders Regional Medical Center (Knoxville, Tenn) Maureen Ogden, RN, MHA, Tampa General Hospital (Tampa, Fla) Roseann Owens, Mary Black Memorial Hospital (Spartanburg, SC) Keith Packard, Alta Vista Regional Hospital (Las Vegas, NM) Carol Parise, Sutter Medical Center Sacramento (SMG) (SMH) (Sacramento, Calif) Susan Pickle, RN, McLeod Regional Medical Center (Florence, SC) Marlys Pieske, RN, Jamestown Hospital (Jamestown, ND) Lorie Poole, PharmD, Thomasville Medical Center (Thomasville, NC) Barbara Pritchard, RN, Innovis Health (Fargo, ND) Martha Radford, MD, New York University Hospitals System (New York, NY) Suman Rathbun, VAMC, Oklahoma City (Oklahoma City, Okla) Penny Razo-Mosier, LPN, CCRP, VAMC, Oklahoma City (Oklahoma City, Okla) Dierdre Reynolds, RN, Southwest Mississippi Regional Medical Center (McComb, Miss) Cathy Roberts, RN, MS, St. Mary’s Hospital and Medical Center (Grand Jct., Colo) Margaret Robinson, Sr., VP/CCO. Midland Memorial Hospital (Midland, Tex) David Rollins, MD, Lake Hospital System, Inc (Willoughby, Ohio) Donna Rubinate, St. Anne’s Hospital (Fall River, Mass) JoAnn Rushenberg, St. John Medical Center (Tulsa, Okla) Sherie Scarnati, Western Pennsylvania Hospital (Pittsburgh, Pa) Cheryl Schmidt, RN, St. Elizabeth Hospital (Appleton, Wis)

Anderson et al Appendix I


Continued

Julia Scott, RN, Northern Nevada Medical Center (Sparks, Nev) Dianna Shie, Genesis Medical Center - Davenport (Davenport, Iowa) Debbie Shook, Watauga Medical Center (Boone, NC) Sandra Shutt, MS, Blanchard Valley Regional Health Center (Findlay, Ohio) Lindsay Siewert, BSN, RN, Jewish Hospital (Jeffersonville, Ind) Betty Slone, RN, CPHQ, Three Rivers Medical Center (Louisa, Ky) Karren Smedley, St. Joseph’s Mercy Health Center (Hot Springs, Ark) Darin Smith, PharmD, Norman Regional Hospital (Norman, Okla) Alex Spyropoulos, MD, Lovelace Sandia Health Systems (Albuquerque, NM) Karen Steffen, RN, MSN, Thomas Jefferson University Hospital (Philadelphia, Pa) Seshan Subramanian, MD, Mercy Hospital and Medical Center (Chicago, Ill) Jacqueline Sullivan, PhD, RN, Thomas Jefferson University Hospital (Philadelphia, Pa) Debbie Tallman, RN, CRC, Lovelace Sandia Health Systems (Albuquerque, NM) Carman Timmerman, EdD, Rapid City Regional Hospital (Rapid City, SD) George Twaroski, Holy Family Memorial (Manitowoc, Wis) Carolyn Vizzard, Jeanes Hospital (Philadelphia, Pa) Wanda Watson, RN, Rapid City Regional Hospital (Rapid City, SD) Kelly Welker, Ogden Regional Medical Center (Ogden, Utah) Connie Wheeler, RN, BSN, Jordan Valley Hospital (West Jordan, Utah) Patti Whims, Greenville Hospital System - Greenville Memorial (Greenville, SC) Diane White, Exeter Hospital (Exeter, NH) Nancy White, Mary Black Memorial Hospital (Spartanburg, SC) Alva Whitehead, MD, McLeod Regional Medical Center (Florence, SC) Claire Wilkie, MPH, RN, CNAA, CPHQ, Watauga Medical Center (Boone, NC) Patricia Windham, Ephrata Community Hospital (Ephrata, Pa) Joyce Wright, MSW, Armstrong County Memorial Hospital (Kittanning, Pa) Robert Wroblewski, VAMC-Fargo (Fargo, ND) Dereck Young, PharmD, North Mississippi Medical Center (Tupelo, Miss)

1106.e3

1106.e4


Appendix II Algorithm employed to identify patients at ACCP-defined risk for acute venous thromboembolism and ACCPrecommended types of prophylaxis according to risk-category (Based on ACCP 2004 recommendations)11 Surgical Patients at Risk for Venous Thromboembolism Level of Risk

Type of Surgery Specified in ACCP Recommendation

ENDORSE Variables

ACCP-Recommended Types of Prophylaxis

Low molecular weight heparin or Vitamin-K antagonist Low molecular weight heparin or Fondaparinux or Vitamin-K antagonist or Intermittent pneumatic compression Low molecular weight heparin or Fondaparinux or Vitamin-K antagonist or Intermittent pneumatic compression/graduated compression stockings plus low-dose unfractionated heparin For patients at high risk for bleeding, intermittent pneumatic compression alone (with or without graduated compression stockings) Low molecular weight heparin or Fondaparinux Age ⱖ40 years and History of venous or Vitamin-K antagonist or Intermittent thromboembolism and Active cancer or pneumatic compression and/or Graduated Surgery for cancer compression stockings, plus low-dose Note: ENDORSE did not capture spinal cord unfractionated heparin injury Surgical patients not previously captured under Low-dose unfractionated heparin or Low High Surgery in patients ⬎60 molecular weight heparin or Intermittent any of the higher risk groups (above) years or age 40–60 years pneumatic compression Age ⬎60 years or Age 40-60 years plus any with additional risk factors of the following: history of venous (ie, prior venous thromboembolism or surgery for cancer or thromboembolism, cancer, current active malignancy or cancer molecular therapy in hospital hypercoagulability) Moderate Surgery in patients aged 40- Surgical patients not previously captured under Low-dose unfractionated heparin or Low molecular weight heparin or Intermittent any of the higher risk groups (above) 60 years with no pneumatic compression or Graduated Age 40-60 years or age 18-39 years plus additional risk factors or compression stockings minor surgery in patients another risk factor (other than surgery) with additional risk factors Low Minor surgery in patients Surgical patients not previously captured under No specific prophylaxis aged ⬍40 years with no any of the higher risk groups (above) additional risk factors Age 18-39 years with no additional risk factors present Note: Most low risk surgical cases were excluded from ENDORSE per protocol based on the length of anesthesia Highest

Hip or knee arthroplasty; hip Total hip arthroplasty fracture surgery; major trauma; spinal cord injury Total knee arthroplasty or surgery in patients with multiple risk factors (age Hip fracture surgery or Admitted with major ⱖ40 years, cancer, prior multiple trauma venous thromboembolism)

Patients at Medical Risk for Venous Thromboembolism (ACCP 2004 Definition) Medical Condition/Illness Cited in ACCP 2004 Recommendations

ENDORSE Variables Matching ACCP Logic

ACCP Recommended Types of Prophylaxis

6.0.1. In acutely ill medical patients who have been admitted to the hospital with congestive heart failure or severe respiratory disease

Acute heart failure (current admission) or acute non-infectious respiratory disease (current admission) or pulmonary infection

Low-dose unfractionated heparin or Low molecular weight heparin or intermittent pneumatic compression

Anderson et al Appendix II


1106.e5

Continued

Patients at Medical Risk for Venous Thromboembolism (ACCP 2004 Definition) Medical Condition/Illness Cited in ACCP 2004 Recommendations

ENDORSE Variables Matching ACCP Logic

ACCP Recommended Types of Prophylaxis

Or who are confined to bed and have one or more additional risk factors, including active cancer, previous venous thromboembolism, sepsis, acute neurological disease, or inflammatory bowel disease, we recommend prophylaxis with low dose unfractionated heparin (grade 1A) or low molecular weight heparin (grade 1A) 8.3. For ICU patients who are at moderate risk for venous thromboembolism (eg, medically ill or postoperative patients), we recommend using low-dose unfractionated heparin or low molecular weight heparin prophylaxis (grade 1A)

Complete immobilization or immobile with bathroom privileges plus one or more of the following (current admission): history of venous thromboembolism, active cancer (including current active malignancy or cancer therapy), infection (non-respiratory ischemic), or hemorrhagic stroke, rheumatologic or inflammatory disease

Low-dose unfractionated heparin or low molecular weight heparin For patients at high risk for bleeding, intermittent pneumatic compression and/or graduated compression stockings alone

Admitted to ICU/CCU (during current admission) or mechanical ventilation

Low-dose unfractionated heparin or low molecular weight heparin For patients at high risk for bleeding, intermittent pneumatic compression and/or graduated compression stockings alone

ACCP ⫽ American College of Chest Physicians; CCU ⫽ coronary care unit; ENDORSE ⫽ Epidemiologic International Day for the Evaluation of Patients at Risk for Venous Thromboembolism in the Acute Hospital Care Setting; ICU ⫽ intensive care unit.

1106.e6 Appendix III

The American Journal of Medicine, Vol 123, No 12, December 2010 Follow-up Study

ENDORSE SUPPLEMENTAL SURVEY: HOSPITAL EFFORTS TO PREVENT VENOUS THROMBOEMBOLISM Request for supplemental hospital data from investigators participating in the ENDORSE Study Site ID: ___________________________79/81 (97.5%) rate of survey return Site Name: __________________________________________ Site Respondent: ____________________________________________ Site Coordinator: __________________________________Telephone ________________________ Lead Investigator: __________________________________Telephone ___________________________________ 1. Does your hospital have a protocol for prevention of venous thromboembolism? □ Yes 67 (84.8%) □ No 12 (15.2%) (If no, skip to Question 2) a. Does it include: i. Risk assessment guidelines □ Yes 63/67 (94.0%) □ No 4/67 (6.0%) ii. A prophylaxis protocol □ Yes 66/67 (98.1%) □ No 1/67 (1.9%) b. Year produced/adopted? Year range ⫽ 1998-2007 (Before 2002: 13/63 [9.7%]) □ Unknown 5/67 (7.5%) c. Which patient groups do the venous thromboembolism protocols apply to? (Check all that apply) □ All acute care in-patients 56/67 (83.6%) □ Orthopedics 36/67 (53.7%) □ Trauma 20/67 (29.9%) □ General surgery patients 36/67 (53.7%) □ General medical patients 33/67 (49.3%) □ Intensive care units 32/67 (47.8%) □ OB/GYN 22/67 (32.8%) □ Neurology 27/67 (40.3%) □ Other 7/67 (10.5%): Specify _______________________________ Rehab, all other, cardiac/telemetry, oncology-extended stay outpatient d. Who was responsible for developing your venous thromboembolism prevention protocol? (Check all that apply) i. A Group □ Hospital quality assurance committee 18/67 (26.9%) □ Pharmacy 24/67 (35.8%) □ Anticoagulation Task Force 13/67 (19.4%) □ Nursing 10/67 (14.9%) □ Ad Hoc hospital group 31/67 (46.3%): Specify_______________________________ ii. An Individual □ Chief of staff 3/67 (4.5%) □ Patient safety officer 3/67 (4.5%) □ Hematologist/vascular specialist 3/67 (4.5%) □ Cardiologist 1/67 (1.5%) □ Vascular surgeon 5/67 (7.5%) □ Other 19/67 (28.4%): Specify_______________________ e. What your protocol based on? □ American College of Chest Physicians Consensus 55/67 (82.1%) □ Other published guidelines 11/67 (16.4%): Specify_______________________ □ Local consensus guidelines 6/67 (9.0%) □ Other 1/67 (1.5%): Specify _______________________ Northwestern Uni. VTE Risk Factor Assessment f. How was the protocol disseminated? (Check all that apply) □ Physician grand rounds 27/67 (40.3%) □ Nursing grand rounds 19/67 (28.4%) □ Departmental staff meetings 60/67 (89.6%) □ In-service presentations 48/67 (71.6%) □ Email formal policy 16/67 (23.9%) g. Was there training for staff? □ Yes 52/67 (77.6%) □ No 15/67 (22.4%) If yes, check all that apply. □ Residents 19/52 (36.5%) □ Medical students 6/52 (11.5%) □ Attendings 40/52 (76.9%) □ Nurses 45/52 (86.5%) What type? (Web based?) In-service, staff meetings, grand-round

Anderson et al Appendix III


1106.e7

continued

h. Has it been reviewed or updated? □ Yes 32/67 (47.8%) (Year range ⫽ 2003-2007; In 2007: 22/30 [73.3%]) □ No/unknown 27/67 (40.3%) 2. Do you know of a champion(s) for venous thromboembolism prevention at your hospital? May be person other than noted in question 1D. □ Yes 63/78 (80.8%) □ No 15/78 (19.2%) (If no, skip to Question 3) If yes, what is their specialty? □ Hematology 5/63 (7.9%) □ Vascular Surgery 7/63 (11.1%) □ Cardiology 1/63 (1.6%) □ Intensivist 12/63 (19.1%) □ Pharmacy 7/63 (11.1%) □ Nursing 7/63 (11.1%) □ Patient safety officer 2/63 (3.2%) □ Other 22/63 (34.9%): Specify Internal medicine, family medicine 3. Which methods or resources does your hospital use to promote or support the assessment and prophylaxis of venous thromboembolism? (Check all that apply) □ Clinical care pathways 32/79 (40.5%) □ Patient information leaflets 8/79 (10.1%) □ MD educational programs 39/79 (49.4%) □ Service-specific reminders 16/79 (20.3%) □ Standing orders 69/79 (87.3%) □ Personal compensation incentives 0 □ Personal performance review 10/79 (12.7%) □ Dedicated staff focused on venous thromboembolism prevention 14/79 (17.7%) □ Dedicated budget allocated to venous thromboembolism prevention in at risk populations ⴝ 0 □ Unit-wide and/or department-wide clinical improvement program aimed at ⴝ 39/79 (49.4%) i. Assessing risk for venous thromboembolism □ Yes 36/39 (92.3%) □ No 3/39 (7.7%) ii. Providing effective prophylaxis □ Yes 38/39 (97.4%) □ No 1/39 (2.6%) Is your hospital involved with regional or national programs to prevent venous thromboembolism? □ National Consensus Standards for Prevention and Care of venous thromboembolism pilot project 4/79 (5.1%) □ Regional coalition 4/79 (5.1%) □ Other 17/79 (21.5%): Specify SCIP, IHI, Oryx Has your hospital conducted chart audits (other than ENDORSE) on venous thromboembolism prophylaxis? □ Yes 55/79 (69.6%) Which department(s) is responsible for performing the audit? Quality Which departments receive the audit results? All clinical departments Date of most recent audit year (range ⴝ 2006-2008; in 2007: 39/45 [86.7%]) □ Unknown 7/55 (12.7%) Did the audit lead to changes in policies? □ Yes 35/55 (63.6%) □ No 20/55 (36.4%) If yes, did it: □ Lead to the creation of protocols for venous thromboembolism prevention 17/35 (48.6%) □ Or other action? 21/35 (60.0%) Specify Change in protocol □ Ongoing (routinely scheduled) 0 □ No 24/79 (30.4%) If no, are there plans to institute audits? □ Yes 12/24 (50%) □ No 12/24 (50%) 4. Does your hospital report rates of venous thromboembolism prophylaxis data to any national or regional QA organization? □ Yes 59/78 (75.6%) □ No 19/78 (24.4%) If yes, (Check all that apply) □ Leapfrog Group Hospital Quality and Safety Survey 11/59 (18.6%) □ University Health System Consortium DVT/PE Benchmarking Project 4/59 (6.8%) □ Surgical Care Improvement Project (SCIP) 51/59 (86.4%) □ Institute for Healthcare Improvement (IHI) 10/59 (17.0%) □ Quality Net 19/59 (32.2%)

1106.e8 Appendix III

The American Journal of Medicine, Vol 123, No 12, December 2010 continued

□ State specific QA organization 4/59 (6.8%) □ Other 8/59 (13.6%), Specify___________________________ 5. Comments: ___________________________________________________ Interviewer: Date: Nov 15, 2007 – March 24, 2008