Determining the magnitude of surveillance bias in the assessment of lower extremity deep venous thrombosis: A prospective observational study of two centers

Steven R. Shackford*, Mark D. Cipolle, Jayraan Badiee, Danielle L. Mosby, M. Margaret Knudson, Paul R. Lewis, Victoria S. McDonald, Erik J. Olson, Kimberly A. Thompson, Jan Michael Van Gent, Ashley L. Zander

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

32 Scopus citations


BACKGROUND: Venous thromboembolism (VTE) remains a significant cause of morbidity and mortality in trauma. Controversy exists regarding the use of lower extremity duplex ultrasound screening and surveillance (LEDUS). Advocates cite earlier diagnosis and treatment of deep venous thrombosis (DVT) to prevent clot propagation and pulmonary embolism (PE). Opponents argue that LEDUS identifies more DVT (surveillance bias) but does not reduce the incidence of PE.We sought to determine the magnitude of surveillance bias associated with LEDUS and test the hypothesis that LEDUS does not decrease the incidence of PE after injury. METHODS: We compared data from two Level 1 trauma centers: Scripps Mercy Hospital, which used serial LEDUS, and Christiana Care Health System, which used LEDUS only for symptomatic patients. Beginning in 2013, both centers prospectively collected data on demographics, injury severity, and VTE risk for patients admitted for more than 48 hours. Both centers used mechanical and pharmacologic prophylaxis based on VTE risk assessment. RESULTS: Scripps Mercy treated 772 patients and Christiana Care treated 454 patients with similar injury severity and VTE risk. The incidence of PE was 0.4% at both centers. The odds of a DVT diagnosis were 5.3 times higher (odds ratio, 5.3; 95% confidence interval, 2.5-12.9; p < 0.0001) for patients admitted to Scripps Mercy than for patients admitted to Christiana Care. Of the 80 patients who developed DVT, PE, or both, 99% received prophylaxis before the event. Among those who received pharmacologic prophylaxis, the VTE rates between the two centers were not statistically significantly different (Scripps Mercy, 11% vs. Christiana Care, 3%; p = 0.06). CONCLUSION: The odds of a diagnosis of DVT are increased significantly when a program of LEDUS is used in trauma patients. Neither pharmacologic prophylaxis nor mechanical prophylaxis is completely effective in preventing VTE in trauma patients. VTE should not be considered a "never event" in this cohort.

Original languageEnglish
Pages (from-to)734-741
Number of pages8
JournalJournal of Trauma and Acute Care Surgery
Issue number5
StatePublished - 17 Feb 2016
Externally publishedYes


  • Duplex surveillance
  • Surveillance bias
  • Trauma
  • Venous thromboembolism


Dive into the research topics of 'Determining the magnitude of surveillance bias in the assessment of lower extremity deep venous thrombosis: A prospective observational study of two centers'. Together they form a unique fingerprint.

Cite this