Impact of a Clinical Decision-Support Tool on Venous Thromboembolism Prophylaxis in Acutely Ill Medical Patients

Abstract

Background: Adherence to guideline-based venous thromboembolism (VTE) prophylaxis recommendations is often inadequate. Effective improvement strategies are needed. Objective: The purpose of this quality improvement initiative was to increase use of the facility’s preferred pharmacologic VTE prophylaxis, reduce unnecessary VTE prophylaxis use, and reduce use of pharmacologic VTE prophylaxis in high bleeding risk patients, in accordance with guideline-based recommendations. Methods: Clinical pharmacists spearheaded the development and implementation of a clinical decision-support tool (CDST) integrated within a Veterans Health Administration electronic health record (EHR). The CDST focused on VTE prophylaxis in acutely ill medical patients and guided prescribers to guideline-based recommendations. Following review and approval, the CDST underwent activation in the EHR. A subsequent intervention occurred, year 2 post-intervention, which embedded this CDST into the EHR admission process and admission menus. A drug message was added in the EHR to alert prescribers that low-molecular-weight heparin was the preferred agent. Measures were evaluated pre-intervention, year 1 post-intervention, and year 2 post-intervention. Results: After intervention, there were statistically significant increases in the proportion of patients receiving the facility’s preferred pharmacologic VTE prophylaxis agent, enoxaparin, and a statistically significant decrease in the proportion of unwarranted VTE prophylaxis. The proportion of inappropriate pharmacologic VTE prophylaxis in high bleeding risk patients decreased, but this result did not reach statistical significance. Conclusion: The improvements observed suggest the beneficial role of CDSTs integrated into the EHR to increase adherence to guideline-based VTE prophylaxis recommendations.

Background

Hospitalizations due to acute medical illnesses are associated with risk for venous thromboembolism (VTE). VTE is a substantial cause of morbidity and mortality throughout the world.^1-7 Given the associated patient safety implications, a variety of national programs and initiatives have been created to increase awareness and appropriate VTE prophylaxis.^8-13 As a considerable portion of VTE events occur in patients admitted to medical services, the American College of Chest Physicians (ACCP) has guidelines that delineate recommendations for thromboprophylaxis of acutely ill hospitalized medical patients.¹ Separate guidance exists for surgical patients and patients with stroke, trauma, and spinal cord injury.

Risk for VTE and necessity of prophylaxis has been derived from a risk assessment model known as the Padua Prediction Score.¹⁴ Guidelines infer that the Padua Prediction Score is the best tool available for VTE risk assessment in nonsurgical hospitalized patients.¹ Risk factors for bleeding in hospitalized medical patients are derived from a multinational observational study in which bleeding incidence was analyzed to identify independent in-hospital general bleeding risk factors.^15,16 Guidelines recommend the use of these bleeding risk factors to assess bleeding scores for patients prescribed VTE prophylaxis.^1,16

Pharmacologic VTE prophylaxis options include low-molecular-weight heparin (LMWH), low-dose unfractionated heparin (LDUH) twice daily, LDUH thrice daily, or fondaparinux.¹ Direct oral anticoagulants also contain labeled indications for VTE prophylaxis in acutely ill hospitalized medical patients.^17,18 In place of pharmacologic VTE prophylaxis, mechanical VTE prophylaxis (graduated compression stockings or intermittent pneumatic compression devices) is recommended in high-risk VTE patients at high risk of bleeding.¹

Globally, there is limited adherence to guideline-based VTE prophylaxis that can have patient safety consequences.¹ Complicating these safety outcomes is the potential preventability of the occurrence with appropriate prophylaxis. In the United States, pulmonary embolism events are reported to be the most preventable cause of death in hospitalized patients.¹⁹ Effective interventions to improve concordance with guidelines include educational interventions, risk assessment documentation, prescriber alerts, review and feedback, or a combination of these strategies.^1,20-22

Despite the existence of these strategies, inappropriate VTE prophylaxis persists.^1,20-24 Questions also remain as to which design for intervention is best to use given the variability among studies. Incorporating approaches from a variety of studies may be most beneficial to health care systems in devising system-wide interventions to address this priority. Hence, development and integration of a novel clinical decision-support tool (CDST) within the electronic health record (EHR) may optimize adherence with guideline-based VTE prophylaxis recommendations and potentially improve health care outcomes.

Objective

North Florida/South Georgia Veterans Health System clinical pharmacists spearheaded the development and implementation of a CDST integrated within the Veterans Health Administration EHR to guide prescribers to guideline-based recommendations. The purpose of the CDST was to improve adherence to guideline-based VTE prophylaxis in acutely ill medical patients. Specifically, the CDST aimed to reduce unwarranted VTE prophylaxis, reduce use of pharmacologic VTE prophylaxis in patients with high bleeding risk, and increase use of the facility’s preferred pharmacologic VTE prophylaxis. The purpose of this study was to evaluate the effectiveness of the CDST.

Methods

North Florida/South Georgia Veterans Health System is a large multicenter Veterans health system containing 2 hospitals, 3 outpatient clinics, and 9 community-based outpatient clinics. The main, tertiary care medical center was the site of investigation. The intervention entailed developing a CDST aligning with ACCP VTE prophylaxis guidelines and facility Hematology Service guidance for acutely ill medical patients. A stepwise algorithm was created that included a therapeutic anticoagulation assessment, VTE risk assessment, bleeding risk assessment, and a kidney function assessment. The CDST would then guide prescribers to the guideline-based recommendation of no prophylaxis needed, mechanical prophylaxis, or pharmacologic prophylaxis. Combination mechanical and pharmacologic prophylaxis is not recommended in these ACCP VTE prophylaxis guidelines and was therefore not part of the CDST. The algorithm was created and embedded electronically as a CDST within the EHR by the facility’s Clinical Applications Coordinator team. Specifically, the CDST was built as an electronic clinical pathway. It would be accessed as an order from the electronic ordering menu within the EHR. After selecting the order, each screen of the clinical pathway would then display one page at a time with a new question (Figure 1). Following each question and the prescriber’s self-tally of risk scores that were displayed on relevant screens (the tool did not have functionality to calculate the total score itself), the prescriber clicked “yes” versus “no” or “less than” versus “greater than” to prompt the pathway onto its next step and question. Depending on the answers selected, the electronic clinical pathway ended as an electronic notification on the screen of “no prophylaxis needed,” an electronic prepopulated nursing text order for mechanical prophylaxis that the prescriber signed to activate in the EHR, or an electronic prepopulated medication order for enoxaparin or LDUH that the prescriber signed to activate in the EHR. On review and approval by the facility’s Hematology Service, the CDST was approved formally through the local Pharmacy and Therapeutics Committee. Within the CDST, enoxaparin was classified as the preferred pharmacologic VTE prophylaxis modality barring any contraindications or kidney dysfunction (defined as a creatinine clearance <30 mL/minute utilizing Cockcroft-Gault formula); LDUH twice daily was designated as the preferred pharmacologic VTE prophylaxis modality in patients with kidney dysfunction. LDUH thrice daily was removed from all standard ordering menus, although it was still available for prescribers to manually order based on their clinical discretion. Following activation of the CDST in the EHR, it was presented and trialed on one of the facility’s teaching medicine teams followed by notification and expansion to all medicine teams.

Figure 1.

Clinical decision-support tool.

ACCP, American College of Chest Physicians; GCS, graduated compression stockings; Q12H, every 12 hours; Q24H, every 24 hours; VTE, venous thromboembolism.

Derived from ACCP Guidelines, facility Hematology Service guidance, and local facility preferred therapies.

A subsequent intervention was made, year 2 post-intervention, which required this CDST to be embedded into the EHR admission process and admission menus that all prescribers would be alerted to use. Additionally, a drug message was added and displayed in the EHR to alert prescribers manually ordering pharmacologic VTE prophylaxis that LMWH was the preferred agent.

To assess the impact of these interventions, a cross-sectional study was conducted. The before period included a single day in January 2014. Post-intervention data were collected on a single day after the interventions occurred. The year 1 post-intervention period included a single day in October 2016, and the year 2 post-intervention period included a single day in September 2017. The 1-day timeframe was selected to capture all hospitalized medical patients on a selected day so that appropriateness of guideline-recommended VTE prophylaxis could be evaluated on that day of their hospitalization. As patient factors can change during hospitalization, a single day cross-section was deemed most relevant to capture appropriateness of prescribing. The amount of time between post-intervention periods was dictated based on timeframes of facility review, approval, trial, and concurrence.

Patients were included if they were acutely ill hospitalized medical patients. Patients were excluded if they were admitted to non–internal medicine services (ie, neurology, psychiatry, surgery, or critical care), if they had stroke or spinal cord injury, or if they were receiving therapeutic anticoagulation. Trauma patients were automatically excluded as the hospital does not have a trauma unit. In the pre-intervention group, patients’ VTE risk factors were based solely on ACCP VTE prophylaxis guidelines (Padua Prediction Score). On implementation of the CDST, the facility’s preference was to exclude all heart failure patients from this CDST given heart failure guideline–specific VTE prophylaxis recommendations.²⁵ Therefore, all heart failure patients were excluded from post-intervention analyses (Figure 2). Additionally, in the pre-intervention group, bleeding risks factors were based solely on ACCP VTE prophylaxis guidelines; however, the facility’s preference was to exclude the following bleeding risk factors: critical care unit admission, central venous catheter, rheumatic disease, current cancer, and male sex. Therefore, these bleeding risk factors were excluded from post-intervention analyses (Figure 2).

Figure 2.

Venous thromboembolism (VTE) and bleeding risk factors.

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

Derived from ACCP Guidelines and facility Hematology Service guidance. Italicized bleeding risk factors are minor bleeding risk factors.

Proportion of unwarranted VTE prophylaxis, high bleeding risk patients inappropriately receiving pharmacologic VTE prophylaxis, and patients receiving the preferred pharmacologic agent were calculated for the composite groups before and after intervention. Unwarranted VTE prophylaxis was defined as receipt of VTE prophylaxis, either mechanical or pharmacological, in the absence of a risk score necessitating its use. High bleeding risk was defined as having one or more major bleeding risk factors or 2 or more minor bleeding risk factors (Figure 2). These proportions were compared through statistical analysis with Fisher’s exact test or χ² test as appropriate for nominal data. Inferential statistical analysis was conducted using the GraphPad QuickCalcs (GraphPad Software); an α = .05 was set for statistical significance. This project was conducted in accordance with institutional procedures for quality improvement projects.

Results

Eighty-seven patients were included in the pre-intervention group, 57 patients were included in the year 1 post-intervention group, and 50 patients were included in the year 2 post-intervention group. The proportion of unwarranted VTE prophylaxis was calculated as the number of patients who did not warrant VTE prophylaxis out of the number of total patients receiving VTE prophylaxis. This proportion decreased from 59.5% (47 of 79 patients) pre-intervention to 56.5% (26 of 46 patients) year 1 post-intervention (P = .89) and 35.7% (15 of 42 patients) year 2 post-intervention (P = .02; for the comparison to the pre-intervention group).

The proportion of high bleeding risk patients inappropriately receiving pharmacologic VTE prophylaxis was calculated as the number of patients who were considered high bleeding risk who received pharmacologic VTE prophylaxis out of the total number of patients considered high bleeding risk. This proportion decreased from 69.1% (38 of 55 patients) pre-intervention to 60.0% (6 of 10 patients) year 1 post-intervention (P = .72) and 25.0% (1 of 4 patients) year 2 post-intervention (P = .11; for the comparison to the pre-intervention group). Results are summarized in Table 1.

Table 1.

Appropriateness of Venous Thromboembolism (VTE) Prophylaxis.

Outcomes	Pre-intervention	Year 1 post-intervention	P a	Year 2 post-intervention	P a
Unwarranted VTE prophylaxis	47/79 (59.5%)	26/46 (56.5%)	.89	15/42 (35.7%)	.02
High bleeding risk receiving pharmacologic VTE prophylaxis	38/55 (69.1%)	6/10 (60.0%)	.72	1/4 (25.0%)	.11

^aP value comparisons are to the pre-intervention group.

The proportion of patients receiving the preferred pharmacologic agent, enoxaparin, was calculated as the number of patients receiving the preferred agent out of the total number of patients receiving any form of pharmacologic VTE prophylaxis. This proportion increased from 3.3% (2 of 61 patients) pre-intervention to 41.5% (17 of 41 patients) year 1 post-intervention (P < .01) and 52.6% (20 of 38 patients) year 2 post-intervention (P < .01; for the comparison to the pre-intervention group). The proportions of patients receiving LDUH twice daily and thrice daily were also calculated as the number of patients receiving these doses out of the total number of patients receiving any form of pharmacologic VTE prophylaxis. LDUH twice daily proportions increased from 3.3% (2 of 61 patients) pre-intervention to 17.1% (7 of 41 patients) year 1 post-intervention (P = .03) and 23.7% (9 of 38 patients) year 2 post-intervention (P < .01; for the comparison to the pre-intervention group). LDUH thrice daily proportions decreased from 93.4% (57 of 61 patients) pre-intervention to 41.5% (17 of 41 patients) year 1 post-intervention (P < .01) and 23.7% (9 of 38 patients) year 2 post-intervention (P < .01; for the comparison to the pre-intervention group). Results are summarized in Table 2.

Table 2.

Pharmacologic Venous Thromboembolism (VTE) Prophylaxis Prescribed.

Pharmacologic Agent	Pre-intervention	Year 1 Post-intervention	P a	Year 2 post-intervention	P a
Enoxaparin 40 mg subcutaneous Q24H	2/61 (3.3%)	17/41 (41.5%)	<.01	20/38 (52.6%)	<.01
Heparin 5000 units subcutaneous Q12H	2/61 (3.3%)	7/41 (17.1%)	.03	9/38 (23.7%)	<.01
Heparin 5000 units subcutaneous Q8H	57/61 (93.4%)	17/41 (41.5%)	<.01	9/38 (23.7%)	<.01

Abbreviations: VTE, venous thromboembolism; Q8H, every 8 hours; Q12H, every 12 hours; Q24H, every 24.

^aP value comparisons are to the pre-intervention group.

Discussion

Significant decreases in the proportion of unwarranted VTE prophylaxis and significant increases in use of the preferred pharmacological agent were observed following the intervention. The proportion of high bleeding risk patients inappropriately receiving pharmacologic VTE prophylaxis was numerically, but not significantly, decreased in the post-intervention groups. These results suggest that integration of a CDST in the EHR was associated with prescribing VTE prophylaxis more appropriately.

While improvements were seen, continued efforts are needed to ensure that adherence to guideline-based VTE prophylaxis ensues. Methods for sustaining the improvements seen include mandating use of the CDST, ongoing staff education, and routine reevaluation of adherence to guideline-based recommendations. Also, it may be important to analyze the rate of warranted VTE prophylaxis and potentially improving provision of prophylaxis to patients who may have not received it, as this was not assessed.

An increase in the proportion of patients receiving the preferred pharmacologic agent, enoxaparin, was observed in the post-intervention composite groups. LMWH and LDUH have similar efficacy in VTE reduction.¹ Given the results of the Avoid-Heparin Initiative, increase in LMWH use may have positive safety implications by reducing heparin-induced thrombocytopenia.²⁶ Due to the less frequent intervals for injections of enoxaparin, fewer materials needed (eg, pre-filled syringe) in VTE prophylaxis dosing, and lower drug acquisition costs, this may also have positive implications for humanistic and economic outcomes. Additionally, LDUH twice daily and thrice daily have similar efficacy.¹ Therefore, increasing use of twice daily dosing may be favorable for similar reasons. A significant decrease in LDUH thrice daily and a significant increase in LDUH twice daily were observed in this study.

Although there was a significant increase in enoxaparin prescribing from pre-intervention, almost half of patients were still not receiving the preferred pharmacologic agent on year 2 post-intervention analysis. This may be because patients with kidney dysfunction or contraindications then appropriately received LDUH prophylaxis, however, this distinction in appropriateness was not analyzed among each pharmacologic modality. While definitions vary, results of this study showed similar use of appropriate prophylaxis as has been seen with other reported interventions.²² Consistent with other studies, the magnitude of change with these interventions and why larger changes have not been observed are known issues and require further research.²² Although there is heterogeneity in studies evaluating these interventions, mandatory tools may provide the highest rates of attainment with measured outcomes. However, even these mandatory tools do not result in full adherence.^21,24 When implementing such tools and alerts at a health care system, developing a feasible system that does not hinder workflow and that is associated with the greatest magnitude of change and benefit is crucial to consider. Requiring documentation by prescribers of rationale for nonadherence to developed tools may also adjudicate any ambiguity as to appropriateness.

There are limitations to this evaluation that warrant discussion. Inherent in observational study designs are limitations to its internal validity including potential for selection bias, information bias, and confounding factors. No specific adjustments or corrections for these biases were employed. The single-center study sample may limit the external validity of the results. Another limitation is the adjustment of risk assessments following pre-intervention review analyses. As the facility’s preference was to exclude all heart failure patients from the VTE risk assessment and Padua Prediction Score given heart failure guideline–specific VTE prophylaxis recommendations, heart failure patients were omitted from all post-intervention analyses. This may have led to a higher observed rate of unwarranted VTE prophylaxis in the pre-intervention group. Additionally, as the facility’s preference was to exclude certain bleeding risk factors from the bleeding risk assessment, these were excluded from all post-intervention analyses. This may have led to higher observed rates of high bleeding risk patients inappropriately receiving pharmacologic VTE prophylaxis in the pre-intervention group. The appropriateness of mechanical prophylaxis in high bleeding risk patients was not evaluated.

Finally, a limitation exists with accounting for clinical discretion of the prescriber. Prescribers were not mandated to follow the CDST’s recommendations and were able to prescribe based on their clinical judgement and what was deemed as the most appropriate therapy. Although patient charts were reviewed for notation by the prescriber when their clinical determination differed from the VTE prophylaxis CDST’s recommendation, the rationale or discussion may not have been documented in the EHR to delineate this decision. Although this CDST did not require these comments, it may be useful to add a designated spot for prescriber justification and documentation when clinical discretion warrants deviating from guideline-based recommendations.

The measures evaluated in this quality improvement initiative provide promising results, but it should be noted that these are surrogate outcomes related to the endpoints they represent. Of note, in a systematic review of studies evaluating interventions for implementation of VTE prophylaxis, it was cited that more studies should report the clinically relevant outcome of appropriate prophylaxis as many studies report outcomes as rates of prophylaxis.²² This quality improvement initiative did report rates of unwarranted prophylaxis that may directly translate to improved safety outcomes such as reduction in bleeding. Given the nature of VTE diagnosis and the length of time for detection, the clinical outcome of VTE was not included in this quality improvement initiative that focused solely on outcomes during a single-day review. Evaluating clinical, humanistic, and economic outcomes would be useful to determine key health care endpoints that would be affected from improved adherence to guideline-based recommendations. When mandatory electronic risk assessments are required to be used, resultant clinical outcomes such as hospital-associated thrombosis, have been reported to decrease.²¹ When pre-intervention VTE prophylaxis rates are low, mandatory risk assessments are also shown to reduce hospital-associated thrombosis and improve use of VTE prophylaxis.^27-30 Additionally, computer-based clinical decision support systems have shown to reduce inpatient VTE costs, and electronic alert systems have shown to reduce VTE incidence with resultant net cost savings.²²

Conclusion

This quality improvement initiative provides support of a CDST created to improve adherence with guideline-based VTE prophylaxis recommendations. Improvements including decreased use of unwarranted VTE prophylaxis and increased use of the preferred pharmacologic agent suggest the beneficial role of a CDST integrated within the EHR. Modalities to improve guideline-based VTE prophylaxis should continue to be evaluated. As evidence evolves and guidelines change, this will need to be revisited to ensure tools are continuously updated.