The Association of Vasopressor Administration through a Midline Catheter with Catheter-related Complications.

Little is known about the safety of infusing vasopressors through a midline catheter. To evaluate safety outcomes after vasopressor administration through a midline. We conducted a cohort study of adults admitted to 39 hospitals in Michigan (December 2017-March 2022) who received vasopressors while either a midline or peripherally inserted central catheter (PICC) was in place. Patients receiving vasopressors through a midline were compared with those receiving vasopressors through a PICC and, separately, to those with midlines in place but who received vasopressors through a different catheter. We used descriptive statistics to characterize and compare cohort characteristics. Multivariable mixed effects logistic regression models were fit to determine the association between vasopressor administration through a midline with outcomes, primarily catheter-related complications (bloodstream infection, superficial thrombophlebitis, exit site infection, or catheter occlusion). Our cohort included 287 patients with midlines through which vasopressors were administered, 1,660 with PICCs through which vasopressors were administered, and 884 patients with midlines who received vasopressors through a separate catheter. Age (median [interquartile range]: 68.7 [58.6-75.7], 66.6 [57.1-75.0], and 67.6 [58.7-75.8] yr) and gender (percentage female: 50.5%, 47.3%, and 43.8%) were similar in all groups. The frequency of catheter-related complications was lower in patients with midlines used for vasopressors than PICCs used for vasopressors (5.2% vs. 13.4%;  < 0.001) but similar to midlines with vasopressor administration through a different device (5.2% vs. 6.3%;  = 0.49). After adjustment, administration of vasopressors through a midline was not associated with catheter-related complications compared with PICCs with vasopressors (adjusted odds ratios [aOR], 0.65 [95% confidence interval, 0.31-1.33];  = 0.23) or midlines with vasopressors elsewhere (aOR, 0.85 [0.46-1.58];  = 0.59). Midlines used for vasopressors were associated with greater risk of systemic thromboembolism (vs. PICCs with vasopressors: aOR, 2.69 [1.31-5.49];  = 0.008; vs. midlines with vasopressors elsewhere: aOR, 2.42 [1.29-4.54];  = 0.008) but not thromboses restricted to the ipsilateral upper extremity (vs. PICCs with vasopressors: aOR, 2.35 [0.83-6.63];  = 0.10; model did not converge for vs. midlines with vasopressors elsewhere). We found no significant association of vasopressor administration through a midline with catheter-related complications. However, we identified increased odds of systemic (but not ipsilateral upper extremity) venous thromboembolism warranting further evaluation.