EHR-based decision support system helps prevent VTE in hospitalized patients
Click Here to Manage Email Alerts
CHICAGO — In hospitalized medically ill patients, an electronic health record-based decision support system reduced major thromboembolic events compared with usual care, researchers reported.
Use of the system was also associated with higher rates of appropriate thromboprophylaxis compared with usual care.
Alex C. Spyropoulos, MD, professor at the Zucker School of Medicine at Hofstra/Northwell and the Institute of Health System Science at the Feinstein Institutes for Medical Research in Manhasset, New York, and colleagues designed the cluster-randomized IMPROVE-DD trial to determine whether an EHR-based clinical decision support system, which informed clinicians of each patient’s IMPROVE-DD venous thromboembolism risk score, would increase the use of appropriate thromboprophylaxis and reduce major thromboembolic events compared with usual care in hospitalized medically ill patients. The results were presented at the American Heart Association Scientific Sessions.
Increasing thromboprophylaxis
“A big proportion of venous thromboembolism in the community is due to hospitalization for recent medical illness,” Spyropoulos told Healio. “By far, more severe and fatal VTE events occur in nonsurgical medically ill patients. There are approximately 8 million hospitalized medically ill patients in the United States every year. Thromboprophylaxis for this heterogeneous, difficult population is misused or underused, especially in the post-discharge period. Data show only 4% of hospitalized medically ill patients receive any kind of post-discharge prophylaxis. We know that about one in four hospitalized medically ill patients are at high risk of thrombosis, not just in the hospital but after discharge.”
Existing systems to improve VTE prophylaxis in this population are either inefficient or very expensive, he said.
“Our institute developed a novel universal clinical decision support platform that was able to integrate clinical decision support into any electronic health record using an EHR-agnostic capability,” Spyropoulos told Healio. “This tool was effective in altering clinician behavior, which was not easy. We garnered that technology and incorporated a validated VTE risk model called IMPROVE-DD. I call it the CHA2DS2-VASc score for medically ill patients. We hoped that using this tool would increase rates of appropriate thromboprophylaxis as a marker of change in clinician behavior. The home run would be to reduce hard outcomes, which are major thromboembolic events.”
All patients were older than 60 years and were hospitalized for one of five reasons: exacerbation of congestive HF, acute respiratory insufficiency, acute infectious disease including COVID-19, acute inflammatory disease including rheumatic disease, and acute stroke with or without paralysis, Spyropoulos said during a presentation.
The IMPROVE-DD VTE score, developed and validated over 10 years, covers a scale of 0 to 12, and designates patients with a score of 0 or 1 as low risk, those with a score of 2 or 3 as moderate risk (who benefit from appropriate in-hospital prophylaxis) and those with a score of 4 to 12 as high risk (who benefit from appropriate in-hospital and post-discharge prophylaxis), Spyropoulos said. It is compiled as follows: 3 points for previous VTE; 2 points for known thrombophilia, current lower-limb paralysis, cancer, or D-dimer at least two times the upper limit of normal; and 1 point for immobilization for at least 1 day, an ICU or coronary care unit stay, or being older than 60 years.
He said the appropriate thromboprophylaxis strategies are, for an inpatient with IMPROVE-DD VTE score of 2 or more, subcutaneous unfractionated or low-molecular-weight heparin, and for a post-discharge patient with IMPROVE-DD VTE score of 4 or more, rivaroxaban (Xarelto, Janssen/Bayer) 10 mg once daily for 30 days.
For the trial, two hospitals were randomly assigned to the usual care and two hospitals were randomly assigned to the EHR-based decision support system. There were 5,249 patients (mean age, 75 years; 48% men) in the intervention group and 5,450 patients (mean age, 74 years; 46% men) in the control group, enough to power the trial for hard clinical endpoints, Spyropoulos told Healio.
“We used the tool at admission, which is the normal clinical workflow, but we also created a brand-new workflow that doesn’t exist in U.S. hospitals, which is the tool being brought up at discharge as well, during the medication reconciliation process,” he said in an interview. “Using the tool took all of about 10 seconds, so it was not a difficult thing for the clinicians to do. The tool calculated a score and a 3-month risk, and when the user recorded those results, both the IMPROVE-DD score and an appropriate thromboprophylaxis recommendation were displayed in the EHR. It forced the providers to do something about it.”
The primary outcome was the rate of appropriate in-hospital and post-discharge thromboprophylaxis in moderate-risk and high-risk patients. Secondary outcomes included rates of VTE, arterial thromboembolism, VTE plus arterial thromboembolism, major bleeding, all-cause mortality, death/all-cause readmission and death/VTE readmission, all at 30 days after discharge.
‘Big numbers’
The in-hospital primary outcome occurred in 80.1% of patients in the intervention group and 72.5% of patients in the control group (OR = 1.52; 95% CI, 1.31-1.67; P < .001), whereas the post-discharge primary outcome occurred in 13.6% of patients with IMPROVE-DD VTE score of 4 or more in the intervention group and 7.5% of those patients in the control group (OR = 1.93; 95% CI, 1.6-2.33; P < .001), Spyropoulos said.
“The tool did exactly what it was supposed to do,” he told Healio. “Six percent to 8% absolute increases in prophylaxis are big numbers and have consequences in terms of outcomes and quality improvement metrics.”
Of note, the 72.5% in-hospital prophylaxis rate in the control group was “very high,” Spyropoulos said, noting that it far exceeded the rate in nonacademic hospitals, which is about 55%.
“The thinking is, there should be a greater benefit in nonacademic hospital settings,” he told Healio.
At 30 days after discharge, compared with the control group, the intervention group had lower rates of VTE (OR = 0.8; 95% CI, 0.64-1; P = .048), arterial thromboembolism (OR = 0.35; 95% CI, 0.19-0.67; P < .001) and VTE/arterial thromboembolism (OR = 0.71; 95% CI, 0.58-0.88; P = .002), but a higher rate of all-cause mortality (OR = 1.32; 95% CI, 1.15-1.53; P < .001), according to the researchers. There were no significant differences between the groups in major bleeding, death/all-cause readmission and death/VTE readmission.
“By increasing in-hospital and at-discharge thromboprophylaxis, the tool was able to reduce major thromboembolism,” Spyropoulos told Healio. “And we didn’t pay a price for this in terms of bleeding.”
The tool should be freely available in the library of tools of the Agency for Healthcare Research and Quality, which gave a grant to the informatics and clinical team of Spyropoulos’ group at the Feinstein Institutes for Medical Research to develop it, he said, noting the IMPROVE-DD risk model will soon be available for those who want to use it individually on most apps that physicians use.
Collapse
Spyropoulos AC, et al. LBS.07: Preventing Limb Loss Through Vascular Interventions and Venous Therapies. Presented at: American Heart Association Scientific Sessions; Nov. 5-7, 2022; Chicago (hybrid meeting).
Read more about
Click Here to Manage Email Alerts