This month we continue our Perusing of the Literature. Once again, this section consists of recent articles that residents and attendings have stumbled across that have raised an eyebrow. These posts are meant to spark a discussion and do not represent a change in the standard of care (unless otherwise noted).
The Article: Bugden S, et al. Skin Glue Reduces the Failure Rate of Emergency Department-Inserted Peripheral Intravenous Catheters: A Randomized Controlled Trial. Ann Emerg Med. 2016;1-6.
The One-Liner: Skin glue, in addition to standard care securement, may reduce peripheral intravenous catheter failure rates at 48 hours for admitted patients after insertion in the ED.
Background: Frequently initiated in the ED setting, peripheral intravenous (IV) catheters may fail with inadequate fixation serving as the underlying etiology in infection, phlebitis, occlusion, or dislodgement. Failure disrupts hydration, antibiotic therapy, and analgesia for the patient, and incurs the added costs of additional supplies and staff time. In comparison to standard polyurethane dressings, medical-grade skin glue (cyanoacrylate) in addition to a dressing has been proven to be more effective in securing central venous, epidural, and peripheral arterial catheters. As peripheral IVs are administered on such a wide scale, a small increase in efficacy has the potential for great improvements in cost and patient satisfaction.
Methods: This was a single-site, 2-arm, nonblinded, randomized, controlled trial over a 5 month period. Patients were screened for inclusion after being identified as requiring hospital admission and having a patent upper limb peripheral IV. Patients were randomized to standard peripheral IV catheter securement (Tegaderm IV; Figure 1) or standard securement plus the addition of skin glue (single-use Histoacryl; Figure 2). The skin glue group received 1 drop of cyanoacrylate glue at the IV skin insertion site and 1 drop under the IV catheter hub (Video 1). The dressing was applied after allowing the glue to dry for <30 seconds. The primary outcome was peripheral IV catheter failure at 48 hours, defined by infection, phlebitis, occlusion, or dislodgement. Outcomes were assessed by RNs in person or by telephone, if no longer hospitalized. Intention to treat analysis was performed. Definitions for failure are described below:
- Infection – clinical impression of cellulitis or pus at IV site.
- Phlebitis – 2 or more symptoms of pain, redness, swelling, or palpable venous cord.
- Occlusion – inability to flush 10 mL of 0.9% saline solution or history of IV catheter removed because ‘it was not working’.
- Dislodgement – subcutaneous extravasation or history of ‘IV fell out’.
Figure 1: Standard of care – securement with transparent polyurethane dressing
Figure 2: Skin glue group – securement with 1 drop to skin insertion site and 1 drop under IV catheter hub
Video 1: Skin glue application for securement.
- 380 peripheral IVs inserted into 360 adult patients
- Some differences between standard and skin glue groups- insertion sites were different between groups, which could make big difference in dislodgement rates.
- 1° outcome: Peripheral IV failure – 17% in skin glue group, 27% in standard care group; D = -10%, 95% CI -18% to -2%, p=0.02)
- 2° outcome: Peripheral IV failure by dislodgement – 7% in skin glue group, 14% in standard care group; D = -7%, 95% CI -13% to 0%)
- Blinding impossible due to glue color
- Single site, question of external validity
- Question of generalizability as different IV insertion practices are possible (for example, almost 74% of IVs in study were inserted by physicians- can we generalize this to our hospital?)
- It is unclear if randomization really worked given differences between standard and skin glue groups. Additionally, many potential confounders were not included or reported including BMI, co-morbidities such as end-stage renal disease, or recent ED visits/ hospitalizations with IV placement.
- Is 48 hours an important time point for ED patients? Maybe future study should consider IV failure in ED.
- Outcome assessment occurred by telephone, and not direct visualization by research RN, for 209 (58%) patients discharged prior to 48 hour reassessment: This is a very significant limitation as assessing IV failure via phone could lead to underreporting of failures. Authors could have considered performing a sensitivity analysis to address this.
- Many IVs removed prior to 48 hours, therefore observed failure rates may not be accurate for 48 hours dwell time
- No cost analysis
Author: Cameron Gettel, MD PGY1
Resident Section Editor: Adam Janicki, MD PGY4
Reviewed by Tracy Madsen, MD, ScM, Assistant Professor, Department of Emergency Medicine