PO15 | Venous thromboembolism risk associated with midline catheters and peripherally inserted central catheters: a systematic review and meta-analysis

D. Santagata¹, F. Subri¹, F. Dentali^1,2, M.P. Donadini¹, I. Giarretta² | ¹Department of Medicine and Surgery, Research Center on Thromboembolic Disorders and Antithrombotic Therapies, University of Insubria, Varese and Como; ²Department of Internal Medicine, Ospedale di Circolo and Fondazione Macchi, Varese, Italy

Background and Aims: Peripherally inserted central catheters (PICCs) and midline catheters (MCs) are essential for administering intravenous therapies. PICCs, which extend from peripheral veins to the central venous system, are effective for high-volume infusions but are associated with significant complications, including venous thromboembolism (VTE) and catheter-related bloodstream infections. In contrast, MCs (terminating in peripheral veins) could present a different thrombotic risk, although current evidence remains conflicting. This systematic review and meta-analysis was conducted to compare VTE rates between these two devices and clarify their safety profiles.

Methods: A systematic literature search was performed across PubMed, Embase, Scopus, Web of Science, and The Cochrane Library following the PRISMA 2020 guidelines. Studies were eligible if they reported VTE outcomes (deep vein thrombosis, superficial venous thrombosis, pulmonary embolism) for both MCs and PICCs in adult patients, with a minimum of 15 catheters per type. Risk of bias was assessed using the Cochrane RoB2 tool for randomized controlled trials and the ROBINS-E tool for non-randomized studies. Subgroup and sensitivity analyses were also performed to explore potential sources of variability.

Results: After screening 2878 records, 15 studies were included, encompassing 11,496 MCs and 13,344 PICCs. Among these, 2 were randomized controlled trials and 13 were cohort studies (12 retrospective and 1 prospective). Thrombotic events were observed in 468 MC cases (4.1%) and in 549 PICC cases (4.1%). The random-effects model revealed no significant difference in the risk of venous thromboembolism between MCs and PICCs (OR 1.06; 95% CI 0.81–1.38) (Figure 1). These findings remained consistent after excluding high risk-of-bias studies and in the leave-one-out sensitivity analysis. Subgroup analyses based on ICU setting, placement year, and hospitalization status yielded similar results to the overall analysis. However, when examining placement methods, there was an increased VTE risk for MCs compared to PICCs ECG-guided techniques and echo-guided techniques were used for placement (OR 1.57; 95% CI 1.25–1.97). Moreover, when the analysis has been restricted to catheter related thrombosis (CRT) events only (available in seven studies) we reported a non significant increased risk for MCs compared to PICCs (OR 1.30; 95% 0.88 – 1.82). Overall, the quality of the included studies was moderate to low regarding our outcome of interest.

Conclusions: In summary, the overall rates of VTE were similar between MCs and PICCs across most analyzed sub populations. However, when advanced placement techniques were utilized, MCs were associated with an increased risk of VTE. Further high-quality studies are warranted to strengthen these findings.