Appropriate use of D-dimer testing can minimize over-utilization of venous duplex ultrasound in a contemporary high-volume hospital.

BACKGROUND: The sensitivity of d-dimer (DD) in detecting deep venous thrombosis (DVT) is remarkably high; however, many institutions send patients immediately for a venous duplex ultrasound (VDU). This study was designed to examine the appropriate utilization of DD and VDU in a high-volume hospital. METHODS: A retrospective study was conducted on consecutive patients who presented to a high-volume emergency department (ED) with lower extremity limb swelling/pain over a 30-day period, who were sent for VDU during an evaluation for DVT. VDU data were merged with electronic DD laboratory results. The enzyme-linked immunosorbent assay method was used to provide DD values and thresholds. Values above 0.60 mg/fibrinogen equivalent unit (FEU) were considered abnormal. RESULTS: We reviewed the medical records of 517 ED patients in the month of June 2013. After applying the Wells criteria, 157 patients (30.4%) were excluded because of a history of DVT or pulmonary embolism, having been screened for shortness of breath, or sent for surveillance-leaving 360 for analysis. The average age was 59.3 ± 16.5 years with more women (210, 58.3%) and the majority reported limb pain or swelling (73.9%). DD was performed on 51 patients with an average value of 3.6 ± 5.4 mg/FEU, of which 43 (84.3%) were positive. DD identified all positive and negative DVT patients (100% sensitivity and negative predictive value), but also included 40 false positives (16.7% specificity). On the other hand, 309 patients were sent directly to VDU without DD; of those, 43 (13.9%) were positive for DVT. However, 266 (86.1%) patients were negative for DVT by VDU without DD and these were deemed improper by our current study protocol. Potential charge savings were calculated as VDU for all (360 × $1000 = $360,000), DD for all (360 × $145 = $52,200), and VDU for both true and false positives (estimated to be about 25% of the cases; 90 × $1000 = $90,000); this equals a charge savings of $217,800 and would avoid unnecessary VDUs. CONCLUSIONS: Based on the results of our study, we suggest that the DD test be utilized during the initial work-up for patients with limb swelling/pain in the emergency room. Appropriate utilization of DD, as well as other clinical criteria, may limit the over-utilization and added cost of VDU, without a negative impact on patient care. The results of DD tests should be utilized to limit the number of patients sent for VDU to only those patients with a positive DD or other significant underlying concerns.