O2092: Tissue Plasminogen Activator Administration for Restoration of Central Line Patency is Associated with Subsequent Hospital-Acquired Venous Thromboembolism: A Single Center Cohort Study

Background: Risk prediction models for hospital-acquired venous thromboembolism (HA-VTE) among critically ill children have identified central venous catheter (CVC) presence as a prothrombotic risk factor. Tissue plasminogen activator (tPA) is a fibrinolytic agent used to restore line patency in cases of intraluminal thrombosis and its use may infer the subsequent risk of symptomatic HA-VTE. We sought to determine the association between tPA administration for restoration of line patency and future HA-VTE risk (i.e., cumulative frequency) among critically ill children with a CVC.

Methods: We performed a single center retrospective cohort study including critically ill children < 18 years of age hospitalized in a quaternary pediatric referral center with a CVC excluding those with HA-VTE prior to CVC insertion. Each CVC was assessed independently. The primary exposure variable was tPA for restoration of line patency. The primary outcome was radiographically-confirmed HA-VTE including limb/neck deep venous thrombosis and pulmonary embolism. Demographics, anthropometrics, Pediatric Risk of Mortality-3-Probability of Mortality (PRISM-3-POM), CVC features, length of stay, and mortality rates were summated. Comparative statistics were employed to assess for differences by cohorts defined by tPA exposure and subsequent HA-VTE. An adjusted logistic regression was derived to determine if tPA was associated with HA-VTE. All statistics were two-sided, type I error was set at 0.05, and no missing data were imputed.

Results: Of 790 patient encounters with 1,332 unique CVCs, 522 (39.2%) catheters required tPA for restoration of line patency. The types of CVCs included 623 (46.8%) peripherally inserted central catheters (PICCs), 368 (27.6%) percutaneous CVCs, 193 (14.5%) tunneled CVCs, and 148 (11.1) percutaneous hemodialysis catheters. Those that required tPA (as compared to not) were more frequently PICCs (50.2% versus 44.6%), percutaneous CVCs (30.8% versus 25.6%), had longer median CVC duration (9.1 [interquartile range, IQR:4.1-15.9] versus 4.1 (IQR:1.8-8.] days), and had greater HA-VTE rates (11.3% versus 4.9%) (all p< 0.001, see Table 1). A total of 99 (7.4%) HA-VTE were identified during the study period at a median of 6.2 (IQR:2-14.6) days following CVC placement (see Table 2). Those with a HA-VTE (as compared to not) were more frequently noted in a percutaneous CVCs (43.4% versus 26.4%, p=0.003), had longer median CVC durations (8.1 [IQR:4.5-18.2] versus 5.6 [IQR:2.2-11] days, p< 0.001), and had greater frequency of tPA exposure (59.6% versus 37.6%, p< 0.001). In an adjusted logistic model, tPA exposure was independently associated with HA-VTE (adjusted odds ratio: 2.27, 95% confidence interval: 1.46-3.51, p< 0.001).

Conclusion: In this single-center report, the rate of HA-VTE among critically ill children with CVCs was 7.4%. Prior exposure to tPA for restoration of line patency was independently associated with subsequent HA-VTE development in this population. While prospective validation is required, clinicians should consider CVC lumen occlusion in their assessment of VTE risk.