SEARCH

SEARCH BY CITATION

Keywords:

  • electronic reminder;
  • thromboprophylaxis;
  • venous thromboembolism

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of conflict of interests
  10. References

Summary.  Background: Compliance with venous thromboembolism (VTE) prophylaxis is poor. Objectives: We sought to determine whether a simple electronic reminder applicable to all hospitalized patients would increase prophylaxis rates and reduce VTE rates. Methods: An electronic reminder was added to the electronic medical record admission note used by all services in our hospital. Prophylaxis, VTE and bleeding rates before and after implementation were compared. Data were analyzed with sas version 9.1. Results: Among all adult medical and surgical patients admitted to our hospital during the time periods studied, 42.8% (1236/2888) before and 60.0% (1410/2350) after the reminder was added received appropriate prophylaxis as per American College of Chest Physicians (ACCP) guidelines (P < 0.001). The difference reached significance for both medical (51.0% vs. 68.9%; P < 0.001) and surgical (48.0% vs. 61.0%; P < 0.001) services. Fewer patients were diagnosed with VTE after our reminder was added (1.1% vs. 0.3%; = 0.001), and there was a trend towards fewer bleeds (1.1% vs. 0.6%; = 0.09). The presence of the reminder was an independent predictor for prophylaxis being given (odds ratio [OR] 1.92, 95% confidence interval [CI] 1.70–2.18; P < 0.001), and was independently associated with a decreased risk for VTE (OR 0.30, 95% CI 0.14–0.64; = 0.003) after adjustment for other VTE risk factors. Conclusion: Adding an electronic reminder to the admission note improved prophylaxis rates and reduced VTE rates across services. The system is easily reproducible and applicable to other facilities. The improvement obtained was modest, so additional measures will probably be needed to optimize prophylaxis rates.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of conflict of interests
  10. References

Venous thromboembolism (VTE) is a common complication of hospitalization and surgery [1]. It is the third most prevalent cardiovascular disease behind myocardial infarction and stroke, with 600 000 symptomatic cases of VTE and 300 000 VTE-related deaths per year in the USA [2]. Two-thirds of VTEs are associated with hospitalization [2,3], and VTE is felt by many to be the most common preventable cause of in-hospital death [4]. Randomized controlled trials have proven that chemical prophylaxis against VTE has a favorable risk/benefit ratio and is cost-effective, and VTE prophylaxis has been endorsed via published guidelines [1,5]. Additional studies have shown that compliance with these guidelines can reduce VTE rates and hospital costs [6,7]. On the basis of these data, the Agency for Healthcare Research and Quality (AHRQ) deemed ‘appropriate use of prophylaxis to prevent VTE in patients at risk’ the number one patient safety practice for hospitals [8].

Despite the focus on prevention, chemical prophylaxis rates range from 10% to 60% in the published literature [9–12], and hospital surveys in the USA and Europe have shown widespread underuse [13–15]. A recent publication by the Office of the Surgeon General [4] found that prophylaxis is particularly underutilized in high-risk groups [14,16–20]. When prophylaxis measures are implemented, they are often inappropriate for the admission diagnosis and patient risk level [21–23]. As a result, the deep vein thrombosis (DVT) rate in hospitalized patients has continued to climb over the past decade [24].

Automatic reminders in the electronic medical record (EMR) increase prophylaxis rates [10,21,25]. Most systems that have been studied involve integrated computer networks with programmed risk calculation that have targeted specific patient groups. Previous survey data show that providers have concerns about the acceptability, feasibility and cost associated with the implementation of computer-based reminder systems [26], and the 8th American College of Chest Physicians (ACCP) Consensus Guidelines advise against the use of calculated risk scores [1]. Therefore, we used a simple computer reminder that is applicable to all hospitalized patients, regardless of admission diagnosis or overall VTE risk. We tested to determine whether we could increase prophylaxis and decrease VTE rates both hospital-wide and in specific patient groups, and sought to verify whether prophylaxis was appropriate according to ACCP guidelines [1].

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of conflict of interests
  10. References

We based our electronic reminder on a table from a recent publication by the Society of Hospital Medicine [27], which was designed according to the ACCP 2008 consensus recommendations for the prevention of VTE [1]. The AHRQ published the reminder, and considered it to be a public domain available for others to use. Maynard et al. [28] piloted a similar system based on this table, and published their work in 2010. In contrast to the ACCP, the reminder specifies enoxaparin instead of low molecular weight heparin, because our hospital carries enoxaparin on formulary, and lists three times daily dosing of unfractionated heparin, as preferred for high-risk patients with renal failure. The reminder prompts the physician to report whether or not the patient is receiving prophylaxis, designates patients as low, medium or high risk for VTE, gives sample order choices for each patient group, and lists contraindications (Table 1). We placed the reminder in the EMR history and physical admission note, which is completed by the admitting house-staff and signed by the attending physician. The history and physical admission note cannot be saved without documenting whether or not the patient is receiving prophylaxis, but, because of software limitations, the reminder was not linked electronically to the order screen. This was not ideal, because it separated the reminder from the desired action, and physicians could acknowledge the need for prophylaxis without actually ordering it. We were able to add pre-made orders that reflected the recommendations in our reminder. The Department of Clinical Investigations at our hospital considered our project to be a part of an ongoing quality improvement initiative, and waived the need for formal institutional review board approval. The details of the project were presented to the quality improvement committee for approval, and the electronic reminder was added to the EMR at the Walter Reed Army Medical Center (WRAMC) in April 2009.

Table 1.   Eletronic reminder with risk stratification algorithm
Is the patient on DVT prophylaxis? □ Yes □ No Please refer to below then utilize VTE Prophylaxis Order Set:
Recommendations for prophylaxis
Low riskModerate riskHigh risk
 Ambulating patient with no VTE risk factors  Minor surgery in patients without additional VTE risk factors (i.e. same day surgery or OR time < 30 min and fully mobile) Not low risk or high risk Acute spinal cord injury with paresis  Multiple trauma  Gynecologic surgery for malignancy  Hip/knee arthroplasty  Hip fracture
Order:  Early ambulationOrder (1 of following) unless contraindicated or on therapeutic anticoagulation:  Enoxaparin 40 mg s.c. daily (do not use if CrCl < 30 mL min−1)  LDUH 5000 Units s.c. t.i.d.  LDUH 5000 Units s.c. b.i.d. Order optional if on pharmacologic agent or mandatory if not:  SCD*Order (1 of following) unless contraindicated or on therapeutic anticoagulation:  Enoxaparin 30 mg s.c. b.i.d. (trauma, orthopedic patients)  Enoxaparin 40 mg s.c. daily  Fondaparinux 2.5 mg s.c. daily (orthopedic patients; CI in CrCl < 30 mL min−1 or < 50 kg)  Adjusted-dose warfarin; target INR 2–3; begun night of surgery (orthopedic patients)  OR  LDUH 5000 Units s.c. t.i.d. (use ONLY if CrCl < 30 mL min−1)  AND  SCD* for spinal cord injury and trauma No pharmacologic prophylaxis because of contraindication: High bleeding risk order:  SCD*
Risk factors
Patient status  Age > 50  Obesity  Immobility  Central venous catheter  Inherited or acquired thrombophilia  Previous VTE  Surgery – moderate/major  Trauma  Acute medical illnessCardiovascular  Myocardial infarction  Congestive heart failure  Venous compression  Varicose veins/chronic stasis Gastrointestinal  Inflammatory bowel disease Pulmonary  Acute/chronic lung disease Renal  Nephrotic syndrome Hematology/oncology  Myeloproliferative disorders  Paroxysmal nocturnal hemoglobinuria  Sickle cell disease  CancerActive rheumatic disease Dehydration Gynecologic  Pregnancy and postpartum period Drugs  Selective estrogen receptor modulators  Estrogen-containing oral contraceptives or HRT
No pharmacologic agents because of contraindication
AbsoluteRelativeOther
  1. b.i.d., twice daily; CI; CrCl, creatinine clearance; DVT, deep vein thrombosis; ESA; GI, gastrointestinal; GU, genitourinary; GYN; HRT, hormone replacement therapy; INR, International Normalized Ratio; LDUH, low-dose unfractionated heparin; PLT; PT, prothrombin time; s.c., subcutaneous; SCD, sequential compression device; t.i.d., three times daily; VTE, venous thromboembolism. *Not appropriate in patient with wounds or documented/suspected VTE.

Active hemorrhage Severe trauma to head or spinal cord with hemorrhage until cleared by neurosurgery/trauma physicianHistory of intracranial hemorrhage in the last 12 months Craniotomy within 2 weeks Gastrointestinal, genitourinary hemorrhage in the last 1 month Thrombocytopenia (Platelet count < 50 000) Coagulopathy (PT > 18 s) Active intracranial lesions/neoplasms/monitoring devices Postoperative bleeding concerns Hypertensive urgency/emergency End-stage liver disease Intro-ocular surgery within 2 weeksEpidural analgesia with spinal catheter placement/removal Heparin-induced thrombocytopenia Proliferative retinopathy

To monitor prophylaxis rates, in September 2009 we began to collect relevant demographic and clinical variables on consecutive adult (> 18 years old) patients admitted to the WRAMC medical and surgical services for at least two hospital days. We included patients on the wards and in the intensive care units (ICUs). The patient was only considered to have received chemoprophylaxis if the nurse documented its administration in the EMR.

To evaluate the efficacy of our reminder, we created a control group consisting of patients admitted in the 6 months prior to institution (1 October 2008 to 31 March 2009). Again, we included consecutive adult patients admitted to a medical or surgical service for at least two hospital days, and we collected relevant demographic and clinical data. The intervention group consisted of patients admitted over a 6-month period, following a 6-month run-in, after the reminder had been added to the EMR in April 2009 (1 October 2009 to 31 March 2010). Chemoprophylaxis rates, determined by nursing administration annotated in the EMR, for the control group were compared with those for the intervention group. Patients receiving treatment-dose anticoagulation were excluded from this analysis. If the prophylaxis regimen chosen accurately reflected VTE risk category and renal function, and was administered for at least 1 day, the patient was counted as having received appropriate prophylaxis. Patients given inappropriate dosing were not counted as having received appropriate chemoprophylaxis. Dosing was labeled as inappropriate if it was not consistent with the package insert directions, was not appropriate for the patient’s risk category (i.e. twice-daily heparin in a high-risk patient with intact renal function), or did not accurately account for the patient’s renal function (i.e. 40 mg of enoxaparin daily in a patient with creatinine clearance [CrCl] < 30 mL min−1). We predefined the subgroups to be analyzed. Because patients with trauma, cancer, ICU admission, congestive heart failure (CHF), hip fracture or replacement or knee replacement are all considered to be at high risk for VTE, we wanted to assess the impact that the reminder had on these groups. Renal failure was chosen as a subgroup because recommendations for dosing change based on CrCl. We also analyzed prophylaxis rates by admitting service, and by whether the admitting service was surgical or medical.

To evaluate the effect that the reminder had on outcomes, we searched the inpatient and outpatient EMRs for new VTE and bleeding diagnoses that occurred during hospital stays and within 3 months of discharge, using ICD-9 codes [29] (Table 2). In addition, the free text fields in the summary and discharge from the EMR were also searched for events. All VTE and bleeding events were confirmed by manual chart audit to exclude events occurring prior to the hospital admission, arterial thromboembolism, and other unrelated peripheral vascular disease diagnoses. No systematic screening was performed, so evaluations for VTE were presumably initiated in response to symptoms.

Table 2.   ICD-9 codes for venous thromboembolism (VTE) and bleeding
ICD-9 Coding for VTE*
 453.42Venous embolism and thrombosis of deep vessels of distal lower extremity
 453.4Venous embolism and thrombosis of deep vessels of lower extremity
 453.40Venous embolism and thrombosis of unspecified deep vessels of lower extremity
 453.41Venous embolism and thrombosis of deep vessels of proximal lower extremity
 V12.51Venous thrombosis and embolism
 453Other venous embolism and thrombosis
 V12.51Venous thrombosis and embolism
 415.1Pulmonary embolism and infarction
 V12.51Venous thrombosis and embolism
 639.6Embolism
 415.11Iatrogenic pulmonary embolism and infarction
 673Obstetric pulmonary embolism
 415.1Pulmonary embolism and infarction
 673.8Other pulmonary embolism
 451.83Of deep veins of upper extremities
 453.9Of unspecified site
 451.11Femoral vein (deep) (superficial)
 453.42Venous embolism and thrombosis of deep vessels of distal lower extremity
ICD-9 Coding for bleeds*
 578.0Hematemesis, vomiting blood
 578.1Blood in stool
 578.9Hemorrhage of gastrointestinal tract unspecified
 459.0Hemorrhage unspecified
 430Subarachnoid hemorrhage
 431Intracerebral hemorrhage
 432.0Non-traumatic extradural hemorrhage
 432.1Subdural hemorrhage
432.9Unspecified intra-cranial hemorrhage

Statistics

Continuous variables were compared by use of the two-tailed t-test. Categorical variables were analyzed with chi-square, with the exact chi-square being used for comparisons with < 1000. Variables were selected for univariate testing on the basis of their previous identification as risk factors for VTE [1]. Multivariate analysis was used to determine predictors of proper chemoprophylaxis and VTE events, by using stepwise logistic regression. Variables were entered into the multivariate model if they showed a trend towards significance during univariate testing (defined as < 0.15). Otherwise, alpha was set to 0.05. Statistics were computed with sas 9.2 (SAS Institute Inc., Cary, NC, USA).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of conflict of interests
  10. References

Overall patient demographics

From 1 October 2008 to 31 March 2009, there were 2888 patients (control group – no reminder in the EMR) admitted to the WRAMC who met our criteria for assessment. From 1 October 2009 to 31 March 2010, there were 2350 patients (intervention group – reminder present in the EMR) who met our criteria. There were no significant differences between groups in age, gender, or the combined number of high-risk patients. The control group had more patients with cancer, and the intervention group had more operations per patient and more patients with traumatic injuries related to service in support of Operation Iraqi Freedom (OIF) or Operation Enduring Freedom (OEF) in Afghanistan (Table 3).

Table 3.   Demographics
 Entire studyControlReminderP-value
  1. OEF, Operation Enduring Freedom; OIF, Operation Iraqi Freedom. *Soldiers medically evacuated from OIF and OEF (Afghanistan) with traumatic surgical injuries. †High-risk orthopedic patients, including those with lower extremity joint replacement surgery and fracture. ‡High-risk patients include all soldiers evacuated with traumatic injuries, high-risk orthopedic patients (as defined above), cancer patients, and patients on a critical care service.

Patients, no.523828882350 
Mean Age (years +/−SD) 56.1 ± 20.955.8 ± 20.756.5 ± 11.90.15
Male, no. (%)3099 (59)1713 (59)1386 (59)0.83
Average no. of operations (+/−SD)0.60 ± 1.250.56 ± 1.290.63 ± 1.200.048
OIF/OEF*, no. (%)689 (13.2)356 (12.3)333 (14.2)0.050
Orthopedic†, no. (%)57 (1.1)29 (1.0)28 (1.2)0.61
Cancer, no. (%)770 (14.7)464 (16.1)306 (13.0)0.003
Admission related to heart failure, no. (%)271 (5.5)145 (5.4)126 (5.6)0.73
High risk‡, no. (%)1745 (33.3)951 (32.9)794 (33.8)0.51

Prophylaxis rates

The rate of appropriate pharmacologic prophylaxis for the entire population increased significantly after the electronic reminder was added (42.8% vs. 60.0%, < 0.001). There was no difference for patients with renal failure (55.9% vs. 62.5%; = 0.18) or for those who had a hip fracture or replacement, or a knee replacement (92.3% vs. 100.0%; = 0.47). Patients with CHF (61.4% vs. 78.7%; = 0.007), trauma after OIF/OEF (59.3% vs. 79.2%; < 0.001) and cancer (48.7% vs. 71.2%; < 0.001) all had an increase in prophylaxis rates. Of the patients who did not receive appropriate prophylaxis, the percentages who received inappropriate dosing, sequential compression devices (SCDs) alone or no prophylaxis are listed in Table 4. Prophylaxis rates by service are listed in Table 5.

Table 4.   Rates of prophylaxis
AnticoagulationEntire study, n (%)Control, n (%)Reminder, n (%)P-value
  1. SCD, sequential compression device.*P-value reported for chemoprophylaxis vs. no or improper chemoprophylaxis with treatment-dose patients excluded.

Chemoprophylaxis*2599 (49.6)1237 (42.8)1362 (60.0)< 0.001
Treatment dosing516 (9.9)273 (9.5)243 (10.3) 
Inappropriate dosing117 (2.2)66 (2.3)51 (2.2) 
SCDs only810 (15.5)476 (16.5)334 (14.2) 
No prophylaxis1196 (22.8)836 (28.9)360 (15.3) 
Table 5.   Rates of prophylaxis by service
 Control, no. (%)Reminder, no. (%)P-value
  1. CT, cardiothoracic; PMNR, physical medicine and rehabilitation.

Medicine services
 All medicine services697 (51)764 (68.9)< 0.001
 Internal medicine480 (55)504 (71)< 0.001
 Neurology19 (13)69 (75)< 0.001
 Oncology21 (37)40 (51)0.12
 Critical care15 (63)38 (79)0.17
 Cardiology109 (55)63 (53)0.82
 PMNR/Pain53 (76)51 (82)0.40
Surgery services
 All surgery services535 (48)586 (61)< 0.001
 Urology15 (10)17 (13)0.45
 Other surgery10 (22)13 (41)0.083
 Organ transplant14 (29)10 (40)0.43
 General surgery218 (69)241 (82)< 0.001
 CT surgery21 (48)38 (69)0.04
 Neurosurgery77 (53)78 (70)0.005
 Gynecology32 (21)48 (38)0.001
 Orthopedic surgery96 (63)115 (83)< 0.001
 Vascular surgery52 (80)26 (46)< 0.001

Event rates for VTE and bleeding

There were fewer VTEs diagnosed after the reminder was instituted than in the control group without the reminder (33 [1.1%] vs. 8 [0.34%]; = 0.001). Both DVT (12 [0.42%] vs. 3 [0.13%]; = 0.053) and pulmonary embolism (21 [0.74%] vs. 5 [0.22%]; = 0.009) rates were reduced, but the difference did not reach significance for DVT. Four of the DVTs that were diagnosed during the control period were in the upper extremity, whereas those diagnosed during the intervention period were all in the lower extremity. Bleeding rates showed a trend towards a significant decrease after the reminder was added (32 [1.1%] vs. 14 [0.6%]; = 0.09).

Univariate and multivariate analyses

To isolate the independent effect that the reminder had on prophylaxis rates, we combined the control and intervention groups and assessed all 5238 patients for univariate predictors for chemical prophylaxis. As expected, the reminder significantly increased the odds that chemical prophylaxis was given. In multivariate analysis, the reminder remained a significant predictor for prophylaxis, even after adjustment for other VTE risk factors and predictors of prophylaxis (Table 6).

Table 6.   Univariate and multivariate predictors of chemoprophylaxis
Risk factorUnivariate analysisMultivariate analysis
Odds ratio (95% CI)P-valueOdds ratio (95% CI)P-value
  1. CI, confidence interval; OEF, Operation Enduring Freedom; OIF, Operation Iraqi Freedom.

Age1.01 (1.00–1.01)< 0.0011.02 (1.01–1.02< 0.001
Number of operations1.61 (1.48–1.75)< 0.0011.63 (1.48–1.79)< 0.001
Heart failure1.76 (1.30–2.37)< 0.0011.88 (1.38–2.58)< 0.001
Renal failure1.06 (0.86–1.31)0.59
OIF/OEF1.99 (1.66–2.37)< 0.0012.12 (1.70–2.63)< 0.001
Orthopedic20.80 (5.06–85.59)< 0.00111.00 (2.62–46.09)0.001
Cancer0.98 (0.82–1.17)0.83
Reminder2.04 (1.81–2.29)< 0.0011.92 (1.70–2.18)< 0.001

A similar analysis was performed on the entire population, with occurrence of VTE as the dependent variable. The presence of the reminder significantly decreased the likelihood of VTE in multivariate analysis (Table 7).

Table 7.   Univariate and multivariate predictors of venous thromboembolism
Risk factorUnivariate analysisMultivariate analysis
Odds ratio (95% CI)P-valueOdds ratio (95% CI)P-value
  1. CI, confidence interval; OEF, Operation Enduring Freedom; OIF, Operation Iraqi Freedom.

Age0.99 (0.98–1.01)0.25
Number of operations1.22 (1.10–1.34)< 0.0011.22 (1.11–1.35)< 0.001
Heart failure0.44 (0.06–3.18)0.40
Renal failure1.19 (0.42–3.38)0.36
OIF/OEF0.91 (0.36–2.34)0.85
Orthopedic2.24 (0.30–16.59)0.43
Cancer2.22 (1.11–4.44)0.0252.21 (1.09–4.45)0.027
Reminder0.30 (0.14–0.64)0.0020.30 (0.14–0.66)0.003

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of conflict of interests
  10. References

We found that the addition of an electronic reminder to the physician history and physical admission note in the EMR improved VTE prophylaxis rates and prevented VTE. Chemical prophylaxis rates were increased for all hospitalized patients admitted to a medical or surgical service for at least 2 days. Rates were also increased for patients with CHF, combat-related surgical injuries, and cancer patients. In multivariate analysis, the reminder was an independent predictor for chemical prophylaxis and decreased the odds for VTE. The reminder was associated with a trend towards a decrease in bleeding rates.

The 8th Consensus ACCP guidelines for VTE prophylaxis recommend that every hospital adopt a formal, active strategy to address VTE prophylaxis. They advocate using computer-based decision tools, and grade the evidence supporting their use as level IA [1]. In the past, however, ACCP conference participants have questioned the feasibility of implementing a computerized reminder system, questioned the acceptability to colleagues, and worried about cost [26]. The guidelines also suggest that decisions for VTE prophylaxis should be based on the admission diagnosis, and not on a comprehensive assessment of VTE risk.

Individual studies [10,21,25,30] and a recent review of the surgical literature [31] have proved that computer-based reminders will increase VTE prophylaxis rates. These studies focused on specific patient populations and made recommendations based on a risk scoring system, as opposed to the admission diagnosis. The existing scoring systems used to assess risk have not been prospectively validated and are cumbersome to use. Maynard et al. [28] used an easier system that is applicable to all hospitalized patients, and proved that it increased prophylaxis rates and reduced VTE rates without increasing adverse events. They designed this scoring system by using feedback from clinicians at their institution. They also provided ongoing education at department meetings for the different hospital services, and gave active feedback to providers who failed to comply with recommendations.

Although VTE prophylaxis is optimized when a multifaceted approach using consensus building, multidisciplinary teams, ongoing monitoring and real-time feedback is used [28,32–34], we have proved that a technically uncomplicated reminder that is applicable to all hospitalized patients is effective for improving prophylaxis and reducing VTE rates. For hospitals using an EMR, implementing our protocol would require only a minor modification to the template for the admission note. This could mitigate past concerns over feasibility and cost. It is important to keep in mind that the overall prophylaxis rate when the reminder was in place was only 60%, which is unacceptably low. Each intervention in the multifaceted approach can provide an incremental improvement, and they should all be used together when possible.

Most of the patients who were not given appropriate prophylaxis had chemical measures omitted entirely, as opposed to receiving the wrong dose or drug. Other investigators have also found that the majority of inappropriate prophylaxis is attributable to omission [3,12,21,35]. The reminder improved appropriate prophylaxis rates by reducing the number of patients who received no prophylaxis. We believe there are two possible explanations for this. Physicians may simply forget to order prophylaxis when they are admitting their patients, and a reminder of any design that forces them to address VTE risk would improve prophylaxis rates. Alternatively, physicians may purposely omit prophylaxis because of a poor understanding of VTE risk or a perceived contraindication. If they click ‘no’ for prophylaxis but then look at the table with risk factors and contraindications (Table 1) that appears on their screen, they may then realize that prophylaxis is appropriate for their patient.

For some of the patients who received no chemical prophylaxis, it is likely that providers purposely chose to withhold anticoagulants. For example, outside of retroperitoneal lymph node dissections and cystectomy patients, the urology service does not provide chemical prophylaxis. For specific patients in other hospital settings, the provider may determine that there is a contraindication to therapy, but our data do not allow us to determine whether the provider consciously made this decision. However, if chemical prophylaxis was withheld because of bleeding risk, SCDs should have been ordered more often. Also, the fact that there were fewer VTEs without more bleeds would imply that the reminder increased the rates of clinically appropriate prophylaxis. Alternatively, the decrease in bleeding may have resulted from the decrease in VTE, which led to less need for treatment-dose anticoagulation.

In contrast to prior data, we found that overall prophylaxis rates for medical patients were slightly higher than for surgical patients. There are several possible reasons for this finding. As discussed above, specific surgical services may not use prophylaxis as often, bringing down the overall rate. In addition, 13.2% of all patients admitted during the two time periods covered were trauma patients from overseas, who often have a number of contraindications to chemical prophylaxis. These patients are randomly distributed over all surgical services, according to the nature of their injuries.

Because we considered prophylaxis to be appropriate if the correct regimen was administered for at least 1 day, we did not account for those who had appropriate prophylaxis at one point but spent a portion of their stay without it. We therefore cannot be sure whether the effect of the reminder persisted throughout the hospitalization. The reduction in the overall VTE rate would imply that it did for a significant portion of the population.

Our study has several limitations. In April 2009, when the reminder was first added to the EMR, members of our team briefed the different hospital services on the nature of our project. Our results may have been biased by the Hawthorne effect [36], although this was mitigated by the use of a 6-month roll-in period prior to data collection. There may have been other, unidentified changes in hospital procotol during the control and intervention group data collection periods that could have influenced prophylaxis, VTE or bleeding rates. Although we verified 11 VTE events, it is possible that we missed events when patients were readmitted to a hospital outside the military health system. This should have been a rare occurrence, however, as these patients are transferred to a military hospital if an extended stay is required and are followed up in a military clinic if they are discharged before transfer. Although ICD-9 codes for VTE can have low specificity, depending on where they are listed among the list of diagnoses, a manual chart audit was performed to verify accuracy [37]. Because prophylaxis was more likely during the intervention period, it is possible that patients were evaluated for VTE less often because their physician felt that the diagnosis was less likely. This may have artificially decreased the rate of VTE during this period. Finally, this was a single-center study at an academic institution that serves active duty military patients, retirees, and dependent family members. Our results cannot necessarily be generalized to other facilities that serve different populations.

In summary, in accordance with 8th ACCP Guidelines recommendations, we added a simple electronic reminder to the EMR that uses admission diagnoses to stratify the risk for VTE. It should be easy to add to most software systems, does not have to be linked with other databases within the EMR, and does not require calculation of a risk score. We tested our reminder on medical and surgical wards, and in the ICU. It increased prophylaxis rates across different services and settings, and reduced VTE rates without increasing bleeding rates for the overall population. We recommend that other hospitals adopt a similar system for VTE prophylaxis, together with the other interventions that have been proven to optimize prophylaxis rates.

Addendum

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of conflict of interests
  10. References

J. Mitchell: participated extensively in project design, data collection, statistical analysis of the data, and writing of the manuscript; J. Collen: participated in data collection and writing of the manuscript; S. Petteys: contributed to data collection and writing of the manuscript; A. Holley: participated in protocol development and submission, project design, data collection, analysis, and manuscript writing and revision. A. Holley is also the guarantor of the paper.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of conflict of interests
  10. References

We would like to acknowledge the invaluable assistance of K. Cazares, from the Department of Inpatient Pharmacy, M. Sebastianelli, who assisted in implementing our reminder in the hospital’s EMR and our daily digital data collection, and Z. Sun for assisting us in searching and cataloging our event rate data based on ICD-9 diagnostic codes.

Disclosure of conflict of interests

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of conflict of interests
  10. References

The authors state that they have no conflict of interest.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Addendum
  8. Acknowledgements
  9. Disclosure of conflict of interests
  10. References
  • 1
    Geerts WH, Bergqvist D, Pineo GF, Heit JA, Samama CM, Lassen MR, Colwell CW. Prevention of venous thromboembolism: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th Edition). Chest 2008; 133 (6 Suppl.): 381S453S.
  • 2
    Geerts WH, Pineo GF, Heit JA, Bergqvist D, Lassen MR, Colwell CW, Ray JG. Prevention of venous thromboembolism: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest 2004; 126 (3 Suppl.): 338S400S.
  • 3
    Amin AN, Stemkowski S, Lin J, Yang G. Inpatient thromboprophylaxis use in US hospitals: adherence to the seventh American College of Chest Physician’s recommendations for at-risk medical and surgical patients. J Hosp Med 2009; 4: E1521.
  • 4
    The Surgeon General’s call to action to prevent deep vein thrombosis and pulmonary embolism – 2008 at the Venous Disease Coalition’s Second Annual Meeting. Office of the Surgeon General Web site. http://www.surgeongeneral.gov/topics/deepvein/calltoaction/call-to-action-on-dvt-2008.pdf. Accessed 1 September 2009.
  • 5
    Dentali F, Douketis JD, Gianni M, Lim W, Crowther MA. Meta-analysis: anticoagulant prophylaxis to prevent symptomatic venous thromboembolism in hospitalized medical patients. Ann Intern Med 2007; 146: 27888.
  • 6
    Borah B, Henk J, Crowther MA, Selby R, Wells P. Alignment to ACCP prophylaxis guidelines and VTE outomes in THR and TKR patients. Blood 2008; 112: 70.
  • 7
    Amin A, Hussein M, Battleman D, Lin J, Stemkowski S, Merli GJ. Appropriate VTE prophylaxis is associated with lower direct medical costs. Hosp Pract (Minneap) 2010; 38: 1307.
  • 8
    Shojania KG, Duncan BW, McDonald KM, Wachter RM, Markowitz AJ. Making health care safer: a critical analysis of patient safety practices. Evid Rep Technol Assess (Summ) 2001; 43: ix, 1–668.
  • 9
    Cohen AT, Tapson VF, Bergmann JF, Goldhaber SZ, Kakkar AK, Deslandes B, Huang W, Zayaruzny M, Emery L, Anderson FA Jr. Venous thromboembolism risk and prophylaxis in the acute hospital care setting (ENDORSE study): a multinational cross-sectional study. Lancet 2008; 371: 38794.
  • 10
    Kucher N, Koo S, Quiroz R, Cooper JM, Paterno MD, Soukonnikov B, Goldhaber SZ. Electronic alerts to prevent venous thromboembolism among hospitalized patients. N Engl J Med 2005; 352: 96977.
  • 11
    Spencer FA, Lessard D, Emery C, Reed G, Goldberg RJ. Venous thromboembolism in the outpatient setting. Arch Intern Med 2007; 167: 14715.
  • 12
    Yu HT, Dylan ML, Lin J, Dubois RW. Hospitals’ compliance with prophylaxis guidelines for venous thromboembolism. Am J Health Syst Pharm 2007; 64: 6976.
  • 13
    Ageno W, Squizzato A, Ambrosini F, Dentali F, Marchesi C, Mera V, Steidl L, Venco A. Thrombosis prophylaxis in medical patients: a retrospective review of clinical practice patterns. Haematologica 2002; 87: 74650; discussion 250.
  • 14
    Goldhaber SZ, Tapson VF. A prospective registry of 5,451 patients with ultrasound-confirmed deep vein thrombosis. Am J Cardiol 2004; 93: 25962.
  • 15
    Hirsch DR, Ingenito EP, Goldhaber SZ. Prevalence of deep venous thrombosis among patients in medical intensive care. JAMA 1995; 274: 3357.
  • 16
    Anderson FA Jr, Wheeler HB, Goldberg RJ, Hosmer DW, Forcier A, Patwardhan NA. Physician practices in the prevention of venous thromboembolism. Ann Intern Med 1991; 115: 5915.
  • 17
    Kucher N, Tapson VF, Quiroz R, Mir SS, Morrison RB, McKenzie D, Goldhaber SZ. Gender differences in the administration of prophylaxis to prevent deep venous thrombosis. Thromb Haemost 2005; 93: 2848.
  • 18
    Rahim SA, Panju A, Pai M, Ginsberg J. Venous thromboembolism prophylaxis in medical inpatients: a retrospective chart review. Thromb Res 2003; 111: 21519.
  • 19
    Steier KJ, Singh G, Ullah A, Maneja J, Ha RS, Khan F. Venous thromboembolism: application and effectiveness of the American College of Chest Physicians 2001 guidelines for prophylaxis. J Am Osteopath Assoc 2006; 106: 38895.
  • 20
    Stratton MA, Anderson FA, Bussey HI, Caprini J, Comerota A, Haines ST, Hawkins DW, O’Connell MB, Smith RC, Stringer KA. Prevention of venous thromboembolism: adherence to the 1995 American College of Chest Physicians consensus guidelines for surgical patients. Arch Intern Med 2000; 160: 33440.
  • 21
    Durieux P, Nizard R, Ravaud P, Mounier N, Lepage E. A clinical decision support system for prevention of venous thromboembolism: effect on physician behavior. JAMA 2000; 283: 281621.
  • 22
    Goldhaber SZ, Dunn K, MacDougall RC. New onset of venous thromboembolism among hospitalized patients at Brigham and Women’s Hospital is caused more often by prophylaxis failure than by withholding treatment. Chest 2000; 118: 16804.
  • 23
    Kahn SR, Panju A, Geerts W, Pineo GF, Desjardins L, Turpie AG, Glezer S, Thabane L, Sebaldt RJ. Multicenter evaluation of the use of venous thromboembolism prophylaxis in acutely ill medical patients in Canada. Thromb Res 2007; 119: 14555.
  • 24
    Stein PD, Matta F, Dalen JE. Is the campaign to prevent VTE in hospitalized patients working? Chest 2011; 139: 131721.
  • 25
    Dexter PR, Perkins S, Overhage JM, Maharry K, Kohler RB, McDonald CJ. A computerized reminder system to increase the use of preventive care for hospitalized patients. N Engl J Med 2001; 345: 96570.
  • 26
    Schunemann HJ, Cook D, Grimshaw J, Liberati A, Heffner J, Tapson V, Guyatt G. Antithrombotic and thrombolytic therapy: from evidence to application: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest 2004; 126 (3 Suppl.): 688S96S.
  • 27
    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, 2008.
  • 28
    Maynard GA, Morris TA, Jenkins IH, Stone S, Lee J, Renvall M, Fink E, Schoenhaus R. Optimizing prevention of hospital-acquired venous thromboembolism (VTE): prospective validation of a VTE risk assessment model. J Hosp Med 2010; 5: 1018.
  • 29
    ICD9 Coding. http://icd9cm.chrisendres.com/ . Accessed 24 January 2011.
  • 30
    Mosen D, Elliott CG, Egger MJ, Mundorff M, Hopkins J, Patterson R, Gardner RM. The effect of a computerized reminder system on the prevention of postoperative venous thromboembolism. Chest 2004; 125: 163541.
  • 31
    Tooher R, Middleton P, Pham C, Fitridge R, Rowe S, Babidge W, Maddern G. A systematic review of strategies to improve prophylaxis for venous thromboembolism in hospitals. Ann Surg 2005; 241: 397415.
  • 32
    Anderson FA Jr, Goldhaber SZ, Tapson VF, Bergmann JF, Kakkar AK, Deslandes B, Huang W, Cohen AT. Improving practices in US hospitals to prevent venous thromboembolism: lessons from ENDORSE. Am J Med 2010; 123: 1099106.
  • 33
    Maynard G, Stein J. Designing and implementing effective venous thromboembolism prevention protocols: lessons from collaborative efforts. J Thromb Thrombolysis 2010; 29: 15966.
  • 34
    Piazza G, Goldhaber SZ. Physician alerts to prevent venous thromboembolism. J Thromb Thrombolysis 2010; 30: 16.
  • 35
    Amin A, Stemkowski S, Lin J, Yang G. Thromboprophylaxis rates in US medical centers: success or failure? J Thromb Haemost 2007; 5: 161016.
  • 36
    Franke RH, Kaul JD. The Hawthorne experiments: first statistical interpretation. Am Soc Rev 1978; 43: 62343.
  • 37
    White RH, Garcia M, Sadeghi B, Tancredi DJ, Zrelak P, Cuny J, Sama P, Gammon H, Schmaltz S, Romano PS. Evaluation of the predictive value of ICD-9-CM coded administrative data for venous thromboembolism in the United States. Thromb Res 2010; 126: 617.