Alerting Providers at Patient Hospital Discharge to Consider Prescribing Rivaroxaban to Reduce Venous Thromboembolism

eVTE (Electronic Venous Thromboembolism): A Cluster, Randomized, Step-wedge Type II Hybrid Study of an Alert Recommending Extended Duration Thromboprophylaxis for At-risk Discharging Medical Patients to Prevent VTE.

A new algorithm derived from only patient age and components of the complete blood count and basic metabolic panel can identify patients discharged from the hospital who may benefit from a blood thinner (called rivaroxaban) to decrease their risk of blood clots, and for whom the risk of bleeding is minimal.

The purpose of this study is to evaluate the use of a pop-up alert, which will be seen by clinicians when a discharging patient has been identified as being someone for whom the risk of blood clots is high, but for whom bleeding risk is estimated to be low.

The pop-up alert will be enabled in a sequential fashion for each group of hospitals in 1 month blocks. We will look to see if the pop-up alert changes the number of patients who receive rivaroxaban. We will also measure the outcomes of blood clots and bleeding among all discharging patients.

The goal of this prospective, cluster, randomized, type II hybrid step wedge, implementation/effectiveness study is to compare the rates of rivaroxaban prescription for extended duration thromboprophylaxis (EDT) in discharging medical patients during the baseline period when no alert informs decision-making to guide EDT, versus EDT prescription during the intervention period when an alert to the discharging clinician is delivered.

Grouped sequential hospitals will be introduced to the intervention randomly in a step wedge fashion.

Aim 1 is to assess the implementation of the alert to discharging clinicians caring for eligible hospitalized medical patients. The primary outcome for Aim 1 is the comparative rate of prescription of EDT (rivaroxaban 10 mg daily for 30 days) during the baseline period versus the intervention period among eligible patients.

Secondary outcomes for Aim 1 will capture interactions with the alert.

Aim 2 is to assess the impact of the alert on important patient clinical outcomes.

The primary efficacy outcome for Aim 2 is the composite of 90-day venous thromboembolism, non-hemorrhagic stroke, myocardial infarction and death.

The primary safety outcome for Aim 2 is 30-day major bleeding. Secondary outcomes for Aim 2 will be the net clinical benefit, defined as the primary outcome + the primary safety outcome during the baseline phase versus the intervention phase among all at risk patients, and all patients for which an alert leads to the prescription of EDT.

Additional secondary outcomes will report components of the primary efficacy and safety outcomes in various groups.

Inclusion Criteria:

• Discharging clinician must be in one of the following iCentra electronic health record (Cerner, Kansas City, MO) positions:

  • Physician, nurse practitioner, or physician assistant hospitalist
  • Physician internal medicine
  • Physician family medicine

• Patient age ≥ 18 years.

• The encounter must be inpatient.

• A signed hospital discharge order must be present.

• eVTE target population criteria (increased venous thromboembolism risk, low bleeding risk) must be met

Exclusion Criteria:

• Pregnant during encounter

• Discharge order completed by ineligible clinician type

• Exclude all cases where the patient is being actively prescribed and intended to be discharged on the following qualifying anticoagulant medications, regardless of dose form or dosing regimen (i.e., they have an active prescription for one of these medications):

  • Apixaban
  • Dabigatran
  • Dalteparin
  • Enoxaparin
  • Edoxaban
  • Betrixaban
  • Fondaparinux
  • Rivaroxaban
  • Warfarin

• Creatinine clearance <30 milliliters/minute based on last-available eligible serum creatinine value preceding discharge

  • Estimated creatine clearance based on actual body weight (preferred) ((140 - age years) * measured weight kilograms) / (72.0 * serum creatine milligrams/deciliter) (*0.85 if female)) = milliliters/minute
  • If measured body weight not available, then based on ideal body weight ((140 - age years) * ideal body weight kilograms) / (72.0 * serum creatine milligrams/deciliter) (*0.85 if female)) = milliliters/minute