Is deep vein thrombosis surveillance warranted in high-risk trauma patients?

BACKGROUND

Deep vein thrombosis (DVT) has been reported to occur in 20% to 40% of high-risk trauma patients if no prophylaxis is used. The purpose of this study was to determine the incidence of DVT and utility of a screening program in a high-risk group of trauma patients for whom routine DVT prophylaxis was utilized.

PATIENTS AND METHODS

Of 3,154 trauma admissions over a 20-month period, 343 patients (10.9%) identified as high risk based on established criteria (prolonged bed rest, Glasgow coma score (GCS) of 7, spinal injury, lower extremity or pelvic fracture) were placed on a prospective surveillance protocol using color-flow duplex scanning and received thromboembolic prophylaxis.

RESULTS

Twenty-three thromboembolic complications occurred, including 20 DVTs (5.8%) and 3 pulmonary emboli ([PE] 1%). Univariate analysis showed that the risk of DVT was related to age (52.6 + 19.9 years versus 38.1 + 18.5; P = 0.001), a longer hospital stay (31.4 versus 17.8 days; P = 0.001), or the presence of spinal fracture (12.6% versus 3.5%; P = 0.01). Discriminant function analysis revealed that length of stay, intensive care unit days, age, and GCS allowed correct classification of those who did not develop DVT in 97% of cases but was only correct in 15% of cases in predicting those who would develop DVT. Injury severity score (ISS) was not predictive in this multivariate analysis. Seventeen (85%) DVTs were unsuspected clinically. Study patients received an average of 3.5 studies at an overall charge of $313,330 to detect 17 clinically unsuspected DVTs (5%). This represents about 5% of the total bed charges for these patients, or $18,000 per DVT.

CONCLUSIONS

These results suggest that standard use of DVT prophylaxis in a high-risk trauma population leads to a low incidence of DVT and that a screening protocol is effective in detecting unsuspected DVTs. Use of a surveillance protocol, however, may reduce but will not eliminate the incidence of pulmonary emboli in this patient population.