Objective

To compare the effectiveness of usual care versus educational interventions for nurses and patients on nonadministration and patient refusal of anticoagulants for venous thromboembolism (VTE) prevention among hospitalized patient.

Study Design

Design Element	Description
Design	Nurse educational intervention: randomized controlled trial Patient educational intervention: controlled pre- versus postintervention comparison
Population	Nurse educational intervention: 933 internal medicine and surgical nurses Patient educational intervention: 19,652 patient visits in which the patient was prescribed at least 1 dose of anticoagulant for VTE prevention
Interventions/ Comparators	Nurse educational intervention: static versus interactive educational intervention Patient educational intervention: usual care versus 3-component patient-education bundle delivered in response to a real-time alert
Outcomes	Primary: overall proportion of anticoagulants not administered Secondary: proportion of anticoagulants not administered due to refusal by patients or family
Timeframe	12-month follow-up for primary outcome

This study included two components:

• A double-blind cluster randomized controlled trial of two web-based modules—static versus interactive—to educate nurses about the harms of VTE, the benefits and risks of anticoagulants, and strategies for communicating with patients about anticoagulants
• A controlled pre- versus postintervention comparison of a patient-education bundle delivered when a patient missed a prescribed dose of anticoagulants

The primary outcome for both study components was the proportion of anticoagulants not administered. The secondary outcome for both study components was the proportion of anticoagulants not administered due to refusal by patients or family.

The nurse education study included 933 permanent nursing staff from the departments of internal medicine and surgery at a hospital in Baltimore, Maryland. The research team cluster randomized nurses by floor to receive one of two online trainings on anticoagulants: a static PowerPoint presentation with a voiceover to present concepts, or a dynamic module with positive reinforcement and corrective feedback. The research team monitored anticoagulant administration practices using electronic health records over one year, stratified into time periods: baseline (preintervention) and postintervention.

The patient-education study included hospitalized patients on 16 adult nonintensive care nursing floors of the hospital. There were 4 intervention floors and 12 control floors. Patients were close to evenly split between men and women in the intervention and control groups. Almost half were white and a little less than half were African American. In response to a real-time alert, patients on intervention floors who missed a dose of anticoagulants received at least one component of an educational bundle, according to their preference. The patient-education bundle included a one-on-one conversation with a health-educator nurse, a 2-page paper handout, and a 10-minute video. Patients on control floors received usual care. The research team compared anticoagulants not administered pre- and postintervention.

Patients and caregivers, some of whom had experience with VTE, and other stakeholders helped create the educational materials.

Results

Nurse educational intervention.

• Medication not administered. Among nurses in either education arm, the proportion of nonadministration was significantly lower following nurse education (12.4% versus 11.1%, conditional odds ratio (OR) 0.87, 95% confidence interval (CI), 0.80-0.95, p=0.002). The difference between the dynamic and static arms was not statistically significant.
• Medication not administered because of patient refusal. Following nurse education, there was no change overall or in either education arm in the proportion of doses that patients or family refused.

Patient educational intervention.

• Medication not administered. The odds of missing doses of anticoagulants declined by 43% on intervention floors (OR 0.57, 95% CI, 0.48-0.67). There was no change on control floors (OR 0.98, 95% CI, 0.91-1.07).
• Medication not administered because of patient refusal. The odds of missing doses of anticoagulants due to patient or family refusal decreased by 47% on intervention floors (OR 0.53, 95% CI, 0.43-0.65). There was no change in patient or family refusal of doses on control floors (OR 0.98, 95% CI, 0.89-1.08).

Limitations

The study took place at only one hospital, so results may not be generalizable to other hospitals. For the patient-education component, not all patients who missed doses took part in the study; the hospital discharged some patients before the intervention began, and others were unable or unwilling to meet with a health-educator nurse.

Conclusions and Relevance

This study found that both nurse and patient education reduced nonadministration of anticoagulants. The study did not find evidence that one type of nurse education is better than another, because both reduced the proportion of missed doses among patients.

Future Research Needs

Future research could examine optimal strategies for training nurses or the effect of nurse- or patient-education interventions on VTE events. Future research could also focus on implementing the education interventions in hospitals in other locations.