Finding the Middle Ground for Intravenous Therapy
Vineet Chopra
DOI: https://doi.org/10.1001/jamanetworkopen.2023.55724
2024-02-14
JAMA Network Open
Abstract:With shifts in care delivery from inpatient to outpatient settings, a growing need for reliable devices that enable intravenous (IV) therapies outside the hospital has emerged. 1 Peripherally inserted central catheters (PICCs) have longed served as the go-to device for this purpose. PICCs offer many advantages in this regard, eg, their insertion in the hospital or home setting is straightforward and convenient; they provide durable venous access for both infusion and phlebotomy; patients and families can be taught to care for them with relative ease; and home-based teams or skilled nursing facilities are well versed in managing these devices. 2 However, accumulating data suggest that use of PICCs is not without risks, including morbid complications such as catheter-related bloodstream infection (CRBSI) and deep vein thrombosis (DVT). 3 ,4 Data also suggest that PICCs are often used inappropriately, potentially exposing patients to risk without clear benefit. While guidelines that seek to inform health care practitioners when to use PICCs are available, 5 ,6 a paucity of head-to-head studies between device types limits the utility of these recommendations. Midline catheters were introduced at the turn of the century to provide middle-of-the-ground therapy. 7 Unlike PICCs, which terminate in the deep vessels of the chest, midline catheters terminate just below the axilla in peripheral veins of the arm. In theory, therefore, their use may not be associated with the same risks as those from PICCs, as they are not a central venous catheter. However, observational studies have reported mixed results. For example, in a large cohort study, Swaminathan et al 8 found that midline catheters were associated with reduced prevalence of occlusion (2.1% vs 7.0%; P < .001) and bloodstream infection (0.4% vs 1.6%; P < .001) compared with PICCs. However, when examining time-to-event models, the risk of DVT was lower in patients who received PICCs compared with midline catheters (hazard ratio, 0.53; 95% CI, 0.38-0.74). 8 Other observational studies have reported conflicting findings, some suggesting no difference in the risk of complications whereas others suggest the converse. 9 ,10 Most recently, a systematic review reported higher rates of thrombosis and lower rates of infection among patients who received midline catheters compared with PICCs. 11 Because almost all of these studies use observational designs with inherent limitations, when to use a midline catheter—especially when considering a PICC—remains debatable. In this edition of JAMA Network Open , Thomsen and colleagues 12 advance science by reporting results from a randomized clinical trial comparing midline catheters with PICCs. These investigators randomized 304 patients in Denmark in a 1:1 ratio to receive either a midline catheter or PICC for therapy ranging from 5 to 28 days. The most common indication for device placement was antibiotic therapy, and the median dwell time was 10 and 11 days for midline catheters and PICCs, respectively. The authors report 1 CRBSI and 2 symptomatic DVTs in the PICC group, whereas those randomized to midline catheters experienced none of these events. However, midline catheter recipients more often experienced minor complications, such as occlusion, leaking, accidental removal, and infiltration, compared with those who received PICCs (13.2% vs 5.3%, respectively). Consequently, there were significantly more premature removals (ie, device discontinuation prior to completion of therapy) in the midline group vs the PICC group (13.2% vs 6.6%; P = .045). In post hoc analyses stratified by dwell time, however, no differences in rates of minor complications or premature discontinuation were observed in midline catheters that dwelled 15 days or less vs PICCs. It was only among midlines with a dwell time of 16 days or longer that a greater rate of premature removals and complications was observed. Although important, these findings have limitations. First, we do not precisely know which infusions were delivered or classified as peripherally compatible in the study. This gap is important because debate about which drugs are vesicants, irritants, cytotoxic, and safe for administration via peripheral veins remains. 13 Second, how many patients left the hospital with their assigned device is unknown; thus, it is difficult to extrapolate these data to the home setting or care outside of the hospital, where midline catheters have an important role. Third, contemporary practices, such as use of tissue plasminogen activator for declotting, electrocardiography for catheter tip ascertainment in PICCs, ascertainment of catheter-to-vein ratio, or midline tip position, were not performed during device placement. Absence of these p -Abstract Truncated-
medicine, general & internal