The impact of fever as a testing indication on the detection rate of deep vein thrombosis in neurocritically ill patients.

AIM

Deep vein thrombosis (DVT) is considered a possible source of non-infectious, non-central fever in the intensive care unit (ICU). In the neurocritically ill, it is unknown whether lower extremity venous Doppler ultrasonography (LEVDUS) for DVT in the setting of fever leads to a higher detection rate than the baseline detection rate of DVT in this population. The aim of this study was to compare the DVT detection rate of LEVDUS performed for the indication of fever to LEVDUS performed for other indications in a neurosciences ICU.

METHODS

Requisition forms for all LEVDUS performed in a referral neurosciences ICU were reviewed and separated into those with "fever" as the stated reason for request and those with other stated reasons. The DVT detection rate was compared between these two categories of indications.

RESULTS

Five hundred eleven LEVDUS were analyzed, of which 422 had been ordered for indications other than fever. Sixteen of these tests were positive, a detection rate of 3.8%. The remaining 89 LEVDUS had "fever" as the ordering indication. Six of these tests were positive for DVT, a detection rate of 6.7%. The likelihood of positivity of a test obtained as part of fever evaluation relative to one obtained for non-fever indications was not significantly different (OR, 1.83; 95% CI, 0.570-5.11;  = 0.246).

CONCLUSION

LEVDUS triggered by fever detected DVT at a numerically, but not statistically significantly higher rate than did LEVDUS performed for other indications in a neurocritically ill population. More rigorous investigation of this question is needed.