Safety of a DVT Chemoprophylaxis Protocol Following Traumatic Brain Injury: A Single Center Quality Improvement Initiative

Abstract

Background

Venous thromboembolism (VTE) is a complication that affects approximately 30 % of moderate and severe traumatic brain injury (TBI) patients when pharmacologic prophylaxis is not used. Following TBI, specifically in the case of contusions, the safety and efficacy of pharmacologic thromboembolism prophylaxis (PTP) has been studied only in small sample sizes. In this study, we attempt to assess the safety and efficacy of a PTP protocol for TBI patients, as a quality improvement (QI) initiative, in the neuroscience intensive care unit (NSICU).

Methods

Between January 1st and December 31st, 2009, consecutive patients discharged from the University of Wisconsin NSICU after >a 48 h minimum stay were evaluated as part of a QI project. A protocol for the initiation of PTP was designed and implemented for NSICU patients. The protocol did not vary based on type of intracranial injury. The rate of VTE was reported as was heparin-induced thrombocytopenia and PTP-related expansion of intracranial hemorrhage (IH) requiring reoperation. The number of patients receiving PTP and the timing of therapy were tracked. Patients were excluded for persistent coagulopathy, other organ system bleeding (such as the gastrointestinal tract), or pregnancy. Faculty could opt out of the protocol without reason. Using the same criteria, patients discharged during the preceding 6 months, from July 1st to December 31st, 2008, were evaluated as controls as the PTP protocol was not in effect during this time.

Results

During the control period, there were 48 head trauma admissions who met the inclusion criteria. In 22 patients (45.8 %), PTP was initiated at an average of 4.9 ± 5.4 days after admission. During the protocol period, there were 87 head trauma admissions taken from 1,143 total NSICU stays who met criteria. In 63 patients (72.4 %), the care team in the NSICU successfully initiated PTP, at an average of 3.4 ± 2.8 days after admission. All 87 trauma patients were analyzed, and the rate of clinically significant deep venous thrombosis (DVT) was 6.9 % (6 of 87). Three protocol patients (3.45 %) went to the operating room for surgery after the initiation of PTP; none of these patients had a measurable change in hemorrhage size on head CT. The change in percentage of patients receiving PTP was significantly increased by the protocol (p < 0.0001); while the average days to first PTP dose trended down with institution of the protocol, this change was not statistically significant.

Conclusion

A PTP protocol in the NSICU is useful in controlling the number of complications from DVT and pulmonary embolism while avoiding additional IH. This protocol, based on a published body of literature, allowed for VTE rates similar to published rates, while having no PTP-related hemorrhage expansion. The protocol significantly changed physician behavior, increasing the percentage of patients receiving PTP during their hospitalization; whether long-term patient outcomes are affected is a potential goal for future study.