P023 - Association of Vasopressor Administration Through a Midline with Complications in Critically Ill Patients

David Paje, MD, MPH - Michigan Medicine
Scott Flanders, MD - Michigan Medicine
Tanima Basu, MA, MS - Michigan Medicine
Elizabeth McLaughlin, MS, RN - Michigan Medicine
Megan O'Malley, PhD - Michigan Medicine
Jennifer Horowitz, MA - Michigan Medicine
Lama Hsaiky, PharmD - Beaumont Health
Vineet Chopra, MD, MSc - University of Colorado Anschutz Medical Campus
Hallie Prescott, MD. MSc - Michigan Medicine

Purpose: : Traditionally, vasopressors were believed to only be safe to deliver through central venous catheters. Using data from hospitals in the Michigan Hospital Medicine Safety Consortium (HMS), we quantified the impact of administering vasopressors through midline catheters.

Methods: : From December 2017 to March 2022, data were abstracted from medical records of adult medicine patients who had a peripherally inserted central catheter (PICC) or midline inserted during hospitalization and received vasopressors. Our primary outcome was a composite of catheter-related complications including infection, venous thromboembolism (VTE), catheter occlusion, superficial thrombophlebitis, and exit site infection. We constructed two mixed effects multivariable regression models to evaluate whether complications were associated with having a midline used for vasopressor delivery compared to: (1) having a PICC used for vasopressor delivery and (2) having a midline not used for vasopressor delivery (vasopressors given through an alternate catheter). Both models included adjustments for demographics (age, sex, and race), patient characteristics (body mass index, current/former smoker, history of VTE, peripheral vascular disease, cancer, diabetes, and kidney failure/dialysis) and line characteristics (insertion vein, lumens, gauge, and line duration) as well as hospital-specific intercepts to control for clustering.

Results: : Among 7,473 patients with a PICC, 1,660 (22.2%) received vasopressors via PICC. Among 4,865 patients with a midline, 287 (5.9%) received vasopressors via midline and 844 (17.3%) received vasopressors through another catheter. Catheter-related complications were found in 13.4% (222/1,660) of patients receiving vasopressors via PICC, 5.2% (15/287) of patients receiving vasopressors via midline, and 6.3% (56/884) of patients with midlines, but receiving vasopressors through an alternate catheter. There was no association of receipt of vasopressors via midline with the composite of catheter-related complications (versus PICC used for vasopressors: aOR 0.62 [95% CI: 0.29-1.33], p = 0.21; versus midline not used for vasopressors: 0.85 [0.46-1.58], p = 0.59). However, receipt of vasopressors through midlines was associated with a higher odds of experiencing VTE (vs PICC: 3.24 [1.42-7.38], p = 0.007; vs midlines not used for vasopressors: 2.47 [1.25, 4.88], p = 0.01).

Limitations:: Study limitations include the observational and retrospective nature of the study, risk of residual confounding, and the inclusion of only hospitals from Michigan.

Conclusions: : While rates of catheter-related complications vary across catheter type, the delivery of vasopressors through a midline does not appear to be associated with catheter-related complications in aggregate. Confidence intervals are wide; however, an association was found with VTE; as such, there might need to be further investigation among patient subgroups.