Computer identification of symptomatic deep venous thrombosis associated with peripherally inserted venous catheters.

Peripherally inserted central catheters (PICCs) are considered a safe method to provide long-term antibiotic therapy, chemotherapy and nutrition support. Deep venous thrombosis (DVT) is a complication that requires early PICC removal, may extend hospitalization and can result in pulmonary embolism. PICC insertion teams strive to understand risk factors and develop methods to prevent DVTs. However, they can only manage what they can measure. At LDS Hospital, identification of PICC associated DVTs was dependent on verbal notification or manual surveillance of more than a thousand free-text vascular reports. Accurate DVT rates were not known which hindered prevention. We describe the development of a computer application (PICC-DVT monitor) to identify PICC associated DVTs each day. A one-year evaluation of the monitor by the PICC team and a review of 445 random vascular reports found a positive predictive value of 98%, sensitivity of 94%, specificity of 100% and a PICC team associated DVT rate of 2.8%.