Background: Clinical practice guidelines on prevention of venous thromboembolism in medical inpatients have been implemented in various settings, although few studies have assessed their impact on venous thromboembolism events.
Objective: To determine whether the implementation of a locally developed guideline is followed by changes in the rate of deep vein thrombosis.
Design: A before-and-after study consisting in two "1-day" cross-sectional studies.
Setting: Thirteen adult medical wards in a teaching hospital in France.
Subjects: All the patients hospitalized on the day of the cross-sectional study.
Intervention: A clinical guideline integrating scientific evidence and data on target medical providers' practices was developed by a local expert panel and implemented through a multifaceted intervention.
Measurements: Prevalence of deep vein thrombosis detected by systematic ultrasound examination.
Results: The study included 338 patients in the preintervention sample and 340 in the postintervention sample. The prevalence of deep vein thrombosis decreased from 9.5% (95% CI, 6.6-13.1) in the preintervention sample to 3.2% (95% CI, 1.6-5.7) in the postintervention sample (P < 0.01). The decrease in the rate of thrombosis involved all deep veins of the lower limbs and remained significant after adjustment for risk factors (adjusted odds ratio = 0.47, 95% CI, 0.32-0.70). No additional cases of pulmonary embolism or deep vein thrombosis were reported either on the day of the study or in the following 2 days.
Conclusions: Active implementation of a clinical practice guideline directed at medical providers (doctors, nurses and physical therapists) can be followed by a significant decrease in prevalence of deep vein thrombosis.