Standard Versus Long Peripheral Catheters for Multiday IV Therapy: A Randomized Controlled Trial

This was an open-label RCT conducted at a tertiary pediatric center (Monash Children’s Hospital, Melbourne, Victoria, Australia) between 2018 and 2019. This study was approved by the Monash Health Human Research Ethics Committee (HREC/17/MonH/534) and prospectively registered with the Australian New Zealand Clinical Trials Registry (ACTRN12617001424392).

Children aged 1 to 17 years undergoing surgery and requiring >48 hours of postoperative IVT were randomly assigned intraoperatively to receive a PIVC or LPC. Patients were identified and recruited preoperatively on the basis of their likely pathology and postoperative treatment schedule. Written consent from a parent or guardian and, when appropriate, the child was obtained before enrollment. The following participants were excluded:

1. those aged <12 months;

2. those requiring central venous access;

3. those with connective tissue disorders;

4. those with hematologic disorders; and

5. those with allergies to multiple catheter dressing materials.

Patients were randomly assigned to receive a PIVC or LPC by weighted minimization by using dedicated software (SiMin; Institute of Child Health, London, United Kingdom). Randomization was conducted according to the following minimization criteria:

1. sex (male; female);

2. age (1–5; 6–10; 11–16);

3. weight (8–10 kg; 11–30 kg; >30 kg); and

4. elective surgery (yes; no).

All patients received an initial PIVC either in the emergency department, surgical ward, or operating room because this was necessary for the induction of anesthesia. Intraoperatively, patients were randomly assigned to PIVC or LPC after confirmation of eligibility by the surgeon. All catheters were inserted by the treating anesthesiologist, by using ultrasound guidance when necessary.

The control group received a new 2.5-cm (1-in) 22G PIVC (Introcan Safety; B. Braun, Melsungen, Germany). The intervention group received an 8-cm (3.1-in) 22G LPC (Leaderflex; Vygon GmbH & Co. KG, Aachen, Germany). All patients were analyzed on an intention-to-treat basis.

Participants were monitored until discharge from the hospital. If the index catheter failed on the ward before completion of IVT, patients would be allocated to receive oral therapy, additional PIVCs, or LPCs, as determined by the attending physician.

PIVCs were inserted by using standard nontouch technique and sterile gloves. The skin was prepared with 70% alcohol and 2% chlorhexidine solution. PIVCs were dressed with semipermeable transparent film (Tegaderm; 3M Company, Maplewood, MN).

LPCs were inserted by using the direct Seldinger technique. Veins were first accessed with a 22G PIVC to facilitate the passage of a 20-cm guidewire. The skin was prepared with a 70% alcohol and 2% chlorhexidine solution. The procedure was performed in an aseptic manner, with sterile gloves and drape. A needleless connector valve (Bionector; Vygon SA, Écouen, France) was then attached. LPCs were dressed with semipermeable transparent film (IV3000; Smith & Nephew plc, London, United Kingdom). Correct LPC position was verified by the ability to aspirate blood and freely flush the catheter with 0.9% sodium chloride solution; IVT was commenced immediately after successful insertion. A single assessor recorded the insertion time for each LPC (from skin preparation to application of dressing, including use of ultrasound). Dressings were changed weekly or more frequently if visible exudate was present. On the ward, catheters were inspected 3 times per day (once per nursing shift) and flushed with 10 mL of 0.9% sodium chloride solution, unless on continuous infusion. Viable catheters remained in situ if they were clinically indicated. All catheters were used strictly for peripherally compatible IVT. Catheters were not used for blood sampling.

The primary outcome in this study was catheter failure, defined as any complication leading to catheter removal before the completion of IVT. Catheters removed after the completion of IVT were recorded as nonfailures. Complications, including infiltration, phlebitis, dislodgement, and occlusion, were recorded. All complications were assessed by the treating nurse or lead investigator. Secondary outcomes were catheter life span, number of postoperative catheters, cost-effectiveness, patient satisfaction, and LPC insertion time. The cost of LPC insertion, including consumables and operating room costs, was calculated and compared with cost of PIVCs.

Infiltration (“tissuing”) was assessed by using the Infusion Nurses Society Infiltration Scale.¹⁵  Infiltration was graded from 1 to 4 on the basis of increasing severity.

Phlebitis was assessed by using the Visual Infusion Phlebitis Scale.¹⁶  Phlebitis was graded from 1 to 5 on the basis of increasing severity.

Catheter occlusion was determined by an inability to flush the catheter.

Patient satisfaction with the index catheter was assessed via the Catheter Satisfaction Questionnaire, which was designed for this study. Four parameters were evaluated on a 5-point Likert scale (1 = very dissatisfied, 2 = dissatisfied, 3 = neutral, 4 = satisfied, and 5 = very satisfied):

1. pain and/or discomfort;

2. anxiety;

3. impact on daily activities (eg, sleeping, bathing, eating, playing); and

4. overall satisfaction.

The questionnaire was completed immediately before discharge by parents and/or guardians or, if appropriate, the child themselves.

The primary outcome was the difference in catheter failure rate. A power calculation was performed by using data from previous observational studies in children and RCTs in adults.^2,12,14,17  We predicted a clinically relevant reduction in catheter failure from 50% for PIVCs to 15% for LPCs. A sample size of 72 participants was required to detect this difference at 90% power and 0.05 two-sided significance level.

Data are expressed as number (percentage). Parametric data are presented in terms of mean ± SD. Nonparametric data are presented as median (range). Catheter duration is reported in days.

Catheter failure and complication rates are expressed as frequencies, percentages, and relative incidence (per 1000 catheter-days), calculated by using the following formula:

All costs are expressed in US dollars and calculated in May 2020. Cost-effectiveness ratio was calculated by using the following formula:

A lower cost-effectiveness ratio was considered more cost-effective. The normality of continuous variables was determined by using the D’Agostino-Pearson normality test. The unpaired Student’s t test and Mann–Whitney U test were used to compare normally and nonnormally distributed continuous variables, respectively. The Fisher’s exact test was used to identity differences between groups of categorical variables. Kaplan-Meier analysis was used to describe catheter survival over time, and the log-rank test was used to compare curves. Statistical analysis was performed with SPSS version 23 (IBM SPSS Statistics, IBM Corporation) and GraphPad Prism version 8 (GraphPad Software, Inc, San Diego, CA). For all tests, P values <.05 were considered statistically significant.

Between February 2018 and 2019, 236 patients were assessed for eligibility, and 127 were recruited (Fig 1). A total of 55 patients were excluded from random assignment. Of these, the majority (n = 42) did not require >48 hours of IVT, whereas the remaining were excluded because the treating anesthesiologist was unfamiliar with LPCs (n = 10) or the patient required central venous access (n = 3).

In total, 72 patients were randomly assigned to PIVC (n = 36) or LPC (n = 36), all of whom received their allocated index catheter and were analyzed in the appropriate group. The procedural success rate for both catheters was 100%. Two patients, both from non-English speaking families, did not complete the satisfaction questionnaire because an interpreter was not available at the time of discharge.

Baseline characteristics, including sex, age, weight, elective status and location of pathology, were similar between the 2 groups (Table 1). Patients underwent the following surgical procedures: appendectomy for perforated appendicitis (n = 52), decortication of parapneumonic empyema (n = 8), drainage of intraabdominal abscess (n = 6), Meckel’s diverticulectomy (n = 1), ileostomy formation (n = 1), Nissen fundoplication (n = 1), resection of esophageal duplication cyst (n = 1), splenectomy (n = 1), and ileocystoplasty (n = 1). The majority of procedures (90.3%) were nonelective.

LPCs were inserted into the forearm (n = 19), antecubital fossa (n = 16), and leg (n = 1), whereas PIVCs were inserted into the hand (n = 19), antecubital fossa (n = 12), forearm (n = 4), and leg (n = 1).

LPCs were placed by 16 attending and 8 resident anesthesiologists. The median procedural time between attending physicians and residents did not differ significantly (7.0 vs 7.2 minutes; P = .30). A total of 36.1% (13 of 36) of LPCs were placed with ultrasound guidance.

The median duration of IVT for all participants was 5.1 days (1.9–16.1) and similar between groups (P = .86). LPCs remained in situ for a median of 4.7 days (1.9–8.1), compared with 3.5 days (0.3–7.8) for PIVCs (P = .01). Kaplan-Meier survival analysis revealed a clear advantage in favor of LPCs (P < .0001; Fig 2). A total of 66% of PIVCs failed before the completion of IVT, compared with 19.4% of LPCs (relative risk [RR]: 3.4 [95% confidence interval (CI): 1.8–7.1]; P = .0001). The relative incidences of PIVC and LPC failures were 187.9 and 41.0 per 1000 catheter-days, respectively. Infiltration was the most frequent complication, followed by occlusion, dislodgement, and phlebitis.

Twelve PIVCs infiltrated the surrounding tissue, compared with 1 LPC (RR: 12.0 [95% CI: 2.2–70.3]; P = .001). Occlusion was the most common complication among LPCs, resulting in the premature removal of 4 catheters. Catheter-related outcomes are displayed in Table 2.

Patients receiving LPCs were more likely to complete their course of IVT with a single catheter (RR: 2.1 [95% CI: 1.4–3.3]; P = .0006). They completed treatment with an average of 1 LPC, compared with 2 PIVCs (P = .0002).

There were 12 instances of infiltration among PIVCs, including 2 examples of grade 3 infiltration (Table 3). Both cases involved widespread upper limb edema persisting for multiple days and were treated with limb elevation. A single case of grade 2 infiltration occurred in the LPC group. There was a single case of grade 2 phlebitis in each group.

Cost data are presented in Table 4. Inserting an LPC was 2.7-fold costlier than a PIVC with a cost difference of $117 per device or $4194 across 36 patients. The largest expense was the fixed cost of using the operating room, valued at ∼$21 per minute. This expense contributed to 78% and 91% of the cost of inserting each LPC and PIVC, respectively. The cost-effectiveness ratio was 2.0 ($67.6 and 33.3%) for PIVCs, compared with 2.3 ($184.1 and 80.6%) for LPCs.

A total of 70 participants (97%) completed the Catheter Satisfaction Questionnaire at the time of discharge (Table 5). One participant from each group failed to complete the questionnaire. The median patient satisfaction was significantly higher for LPCs across all 4 parameters: pain and discomfort (3 [1–5] vs 5 [1–5]; P = .0001), anxiety (3 [1–5] vs 4 [1–5]; P = .03), impact on daily activities (3 [1–5] vs 4.5 [2–5]; P = .01) and overall satisfaction (4 [1–5] vs 5 [2–5]; P = .003).

IVT is one of the most common interventions administered to hospitalized children. Therefore, any advancements in IVT can markedly improve the quality of clinical care. This study is the first RCT investigating the use of LPCs in children. Our results demonstrated a significant reduction in the rate of catheter failure. Children receiving a PIVC were 3.4 times more likely to experience catheter failure than those receiving an LPC. In addition, LPCs had longer duration of use and reduced the median number of IV catheters per patient from 2 to 1. Parental satisfaction favored LPCs across all parameters assessed.

When PIVCs failed, they were most likely to infiltrate. Infiltration occurs because of venous erosion and migration of the catheter tip through the vein.¹⁸  It is characterized by pain, erythema, and edema as infusates enter the subcutaneous tissue, creating distress for both child and parent. In the PIVC group, there were 2 cases of grade 3 infiltration, with widespread edema persisting for multiple days. Although specific treatment was not sought, both patients required limb elevation in addition to emotional support. In contrast, LPCs were most likely to occlude. Catheter occlusion occurs when a thrombus forms in the lumen or around the tip, creating a site of obstruction.¹⁹  This could be an additional benefit of LPCs because occlusion is, typically, asymptomatic and benign.

Our results were comparable to previously published adult RCTs. We have demonstrated a reduction in catheter failure from 66.7% to 19.4% with the use of LPCs. De Prospo et al⁹  and Elia et al¹⁰  described significant reductions from 57.1% to 27.0% and 42.9% to 14.0%, respectively. De Prospo et al⁹  also reported catheters failure in relative terms; 185.9 PIVCs and 16.4 LPCs failed per 1000 catheter-days. Although their PIVCs performed almost identically to ours (185.9 vs 187.9), there was a larger difference between LPCs (41.0 vs 16.4). This is most likely due to the difference in duration of therapy. Their LPCs remained in situ for an average of 9 days, whereas ours were used for 5 days. Of note, PIVCs in our study performed worse than those in the 2 adult RCTs.^9,10  This could be attributable to the size and length of the device; both adult studies used ≥5-cm PIVCs. The 2.5-cm PIVC is the standard 22G device at our institution; it was chosen to standardize the trial because LPCs were also 22G and all catheters were inserted in superficial veins. We do acknowledge, however, that 2.5-cm PIVCs may be too short, especially in older children with deeper veins. Nonetheless, LPCs outperformed ≥5-cm PIVCs in both adult trials.^9,10  We are now routinely using 5-cm PIVCs in older children.

In several previous studies, researchers have explored LPCs as rescue catheters in patients with difficult venous access.^{8,10,13,20–24}  That is, LPCs were only considered when PIVCs were inadequate and were placed under ultrasound guidance. In contrast, we chose to investigate the utility of LPCs as primary catheters for IVT. LPCs conveyed a significant advantage over PIVCs in otherwise healthy children with normal veins. Moreover, in the presence of suitable superficial veins, ultrasound guidance was only necessary for approximately one-third of LPCs, thereby reducing the procedural time.

The majority (90.3%) of participants in our study underwent nonelective surgical procedures. This was an expected finding because most healthy children undergoing elective procedures do not require >48 hours of IVT. LPCs were particularly useful in the treatment of severe infections, such as complicated appendicitis and parapneumonic empyema, in which IVT is a critical component of medical care. Reliable IV access is, thus, crucial to the timely delivery of fluids, antibiotics, and analgesia.

Our results demonstrate that LPCs provide effective and reliable access for multiday IVT.

In our study, all catheters were inserted intraoperatively under general anesthesia; the rational for this was to allow for maximal standardization of the procedural technique for the purpose of the trial (eg, procedural environment, asepsis, and patient movement) for both devices. In our experience, LPCs are suitable for any child requiring extended peripheral IV access (eg, oncological, hematologic patients, etc) and can be placed in the ward or emergency department. At our institution, LPCs are regularly inserted by pediatricians in children who are awake, as previously demonstrated in cystic fibrosis patients.¹⁴  The direct Seldinger technique (catheter over guidewire) is a technique already familiar to many clinicians but also straightforward to learn. Furthermore, venous access for LPC insertion can be obtained without the use of the ultrasound scan; therefore, we believe that clinicians familiar with IV cannulation can accomplish the task without the need for a dedicated vascular access team. The absence of blinding is another limitation of our trial. Because of differences in the outward appearance and insertion technique of the 2 catheters, blinding was not achievable. Thus, the research team, clinical staff, and participants were aware of the allocation. We were, therefore, unable to eliminate the influence of performance and detection bias. Although this would not have had an effect on our primary outcomes, which were objective in nature, we did expect patient satisfaction to favor the novel intervention (ie, LPC). Aspects of the satisfaction questionnaire including “anxiety” and “overall satisfaction” may also reflect the patient’s emotions surrounding their procedure and not just the catheter itself.

Finally, all financial calculations were performed per the cost of items at our Australian institution, whereas this would vary between other institutions and countries. Our calculations serve as an example of how cost comparison and cost-effectiveness of IV devices can be assessed.

The results of this RCT demonstrate that LPCs should be considered the catheter of choice in all children requiring multiday IVT. LPCs lower failure rates, reduce total number of catheters, and improve patient satisfaction.

After the results of our trial, we have implemented the widespread use of LPCs at our institution. There are many anesthesiologists, emergency physicians, and pediatricians (including trainees) who are now proficient in LPC insertion.

We thank the following colleagues from the Department of Anesthesia at Monash Health for the support provided during the trial: Drs Damian Castanelli, Rachel Chapman, Joanne Ee, Tracy Jackson, Cassandra Lang, Paul McCallum, Sangeetha Murthi, Samuel Sha, Joseph Speekman, and Valerie Taylor.

CI

confidence interval

IV

intravenous

IVT

intravenous therapy

LPC

long peripheral catheter

PIVC

peripheral intravenous catheter

RCT

randomized controlled trial

RR

relative risk