Effects of a standardized emergency department protocol on after-hours use of venous duplex ultrasound.

OBJECTIVE: Vascular laboratory (VL) venous duplex ultrasound is the "gold standard" for diagnosis of lower extremity deep venous thrombosis (DVT), which is linked to many morbid conditions. Decreasing night and weekend use of VL services in the emergency department (ED) represents a potentially viable means of reducing costs as skilled personnel must remain on call and receive a wage premium when activated. We investigated the effects of workflow changes that required ED providers to use a computerized decision-making tool, integrated into the electronic medical record, to calculate a Wells score for each patient considered for an after-hours venous duplex ultrasound study for suspected DVT.

METHODS: The rate of VL use and study positivity before and after implementation of the decision-making tool were examined in addition to measures of ED throughput, rate of concomitant pulmonary embolism, disposition of examined patients from the ED, observed thrombus distribution in duplex ultrasound studies positive for DVT, and calculated personnel costs of after-hours VL use.

RESULTS: A total of 391 after-hours, ED-initiated venous duplex ultrasound studies were obtained during the 4-year study period (n = 213 before intervention, n = 178 after intervention; P = .12). Whereas the period immediately after the start of the intervention saw a decrease in VL use, this was not sustained. Studies performed after the intervention were not more likely to be positive for acute DVT (12.2% vs 18%; P = .1179). The average Wells score was 2.8 (range, 0-6). VL personnel were called in 347 times during the 4-year period, with a total cost of $14,643.40. Nurse-ordered studies were significantly more likely to be positive, with 22% revealing acute DVT compared with 12% for physician-ordered studies (P = .042). The intervention resulted in significant improvements in ED throughput, with time between triage and study request falling from 226 minutes to 165 minutes (P < .001). Observed thrombus distribution revealed involvement of the most proximal external iliac system in a minority of cases (11%), whereas most thrombi (89%) were limited to the femoropopliteal, calf, and superficial venous systems.

CONCLUSIONS: A requirement for ED providers to document a Wells score before obtaining an after-hours venous duplex ultrasound study resulted in only a transient decrease in VL use but improved ED throughput. Studies ordered by nurses were significantly more likely to be positive, possibly as a result of consistent protocol adherence compared with the physicians. Future studies may warrant investigation into this provider variance.