Inserting a peripheral vascular access device (PVAD) promptly to administer drugs and fluids is crucial for the treatment of patients in emergency departments (ED). Failing to secure reliable vascular access could contribute to ED crowding, as well as delays in patient treatments and diagnostic testing (Witting, 2012). Despite PVAD insertion being one of the most common tasks for ED clinicians, the task remains challenging when considering one in three patients present with difficult venous access (DIVA; Rippey et al, 2016; Whalen et al, 2017). Patients with DIVA are often individuals who are overweight, without visible or palpable veins, IV drug users, receiving recent chemotherapy, or have medical conditions (eg, sickle cell disease; Fields et al, 2014; Rippey et al, 2016).
First PVAD insertion failure in DIVA patients is as high as 58% (Farrell et al, 2017; Kleidon et al, 2019; Larsen, 2010; Reigart et al, 2012). Repeated PVAD insertion not only causes patient distress, pain and dissatisfaction (Fields, Piela, & Ku, 2014), but also increases staff workload and organizational costs (Tuffaha et al, 2014). Failure to secure a PVAD in ED patients who are unstable or critically unwell can lead to a more invasive procedure, such as insertion of a central vascular access device (CVAD) (Schoenfeld et al, 2011). To increase insertion success, ultrasound is widely utilized to assess vein location and characteristics during PVAD insertion, to avoid ‘blind puncture’ (Bahl et al, 2016; van Loon et al, 2018). Findings from a systematic review suggest that ultrasound-guided insertion has a higher success rate compared with landmark insertion (81% versus 70%; van Loon et al, 2018).
Traditionally, short PVADs (length <4 cm) are the first line device choice by clinicians, including for DIVA patients. Research demonstrates that long PVADs (4.5 to 6.4 cm) allow access to deeper veins and a larger proportion of catheters to reside within the vein. This significantly improves PVAD stability and increases the longevity of the device dwell (Bahl et al, 2019). Indeed, as there are more choices of various types of PVADs on the market, clinicians are becoming more conscious of selecting the right devices for the challenging DIVA population, to enhance first insertion success.
New PVAD devices using advanced technology are promising in enhancing first-time insertion success. An example is a guidewire (GW) PVAD with a retractable coiled tip, designed to reduce damage to the endothelial layer of blood vessels. However, current evidence about its effectiveness in increasing first-insertion success and reducing complications is mixed. Results of a randomized controlled trial (RCT) showed using a long GW-PVAD was associated with higher first-insertion success (89% GWPVAD versus 47% standard PVAD), increased dwell time (4.4 days GW-PVAD versus 1.5 days standard PVAD) and fewer complications (8% GW-PVAD versus 52% standard PVAD) compared with standard PVADs in a medical ward (Idemoto et al, 2014). Similarly, two observational studies found that using GW-PVAD is associated with a lower insertion failure rate and better patient satisfaction in ED settings (Chiricolo et al, 2015; Raio et al, 2018). In contrast, a small RCT (n=93) conducted in prehospital settings compared insertion time and found a small difference (42 seconds longer, P < 0.0001) to insert GW-PVAD compared with standard PVADs (Jin et al, 2018). Furthermore, a single-site RCT conducted in an interventional radiology outpatient population found no difference between the GW-PVAD and standard PVADs in the first-time insertion success (77% GW-PVAD versus 82% standard PVAD; Chick et al, 2017). Previous research has been conducted as either small single-site RCTs (Chick et al, 2017; Idemoto et al, 2014) or observation studies (Chiricolo et al, 2015; Raio et al, 2018). Rigorous evidence from a large multi-site RCT is needed to determine the effectiveness of long GW-PVADs in an ED setting. This study aims to determine whether a GW-PVAD, compared to a standard PVAD, is clinically and cost-effective in the ED setting.
Methods and analysis
Design
A prospective, two-arm, multi-site RCT will be conducted to compare long GW-PVADs (intervention) with standard care PVADs (control). The design and reporting of the trial is informed by the CONSORT guideline (Moher et al, 2010). The trial was prospectively registered with the Australian New Zealand Clinical Trial Registry ACTRN12622000299707).
Hypotheses
Primary hypothesis
- Patients with long GW-PVADs will have a higher firstinsertion success rate, compared with patients receiving standard care PVADs.
Secondary hypothesis
- Compared with patients receiving standard care PVADs, patients with long GW-PVADs will have lower all-cause PVAD failure and longer dwell-time.
Setting and participants
This RCT will be undertaken in two metropolitan secondary level hospital EDs in Queensland, Australia. All patients presenting to the EDs who are ≥18 years old requiring a PVAD and identified as DIVA patients will be eligible for trial entry (Table 1).
Table 1. Inclusion and exclusion criteria
| Inclusion criteria | Exclusion criteria |
|---|---|
|
|
Study interventions
Participants will be randomized to either the standard care PVAD (control) arm or the GW-PVAD (intervention) arm.
- Control: either a short PVAD (<4 cm length, eg BD Insyte™ Autoguard™ BC, Introcan Safety®3 catheter, B Braun) or long PVAD catheter (4.5–6.4 cm length, eg, BD Insyte™ Autoguard™ BC, Introcan Safety®Deep Access, B Braun), inserted by a registered nurse or doctor with established PVAD insertion skills, with or without ultrasound guidance as per clinical judgement. The choice of long or short will be based on clinician judgement
- Intervention arm: a long GW-PVAD, inserted in the upper arm or forearm vein, (5.8 cm length, AccuCath Ace™ Intravascular Cannula, BD, Salt lake City, UT, USA), by a registered nurse or doctor competent to insert GWPVAD, with ultrasound guidance.
Outcomes
Primary outcome
- First-time insertion success: on the first-insertion attempt there is a visible presence of blood at the PVAD hub after piercing through the skin into the vein, and a successful flush of normal saline into the vein is achieved (Carr et al, 2016).
Secondary outcomes
- Number of PVAD insertion attempts (needle punctures to insert PVAD)
- Time to insert PVAD (time from randomization to successful insertion)
- PVAD failure rate: any causes of ‘failure’ (unintentional cessation of PVAD function before treatment is complete), a composite of infiltration or extravasation (Doellman et al, 2009), blockage or occlusion (with/without leakage; Helm et al, 2015), dislodgement (partial/complete), infection (laboratory-confirmed local or bloodstream infection as per NHSN 2021; National Healthcare Safety Network, 2021), phlebitis (2 or more of pain/tenderness, erythema and/or swelling >1 cm, palpable cord and/or vein streak >2 cm; Rickard et al, 2018), and thrombosis (confirmed by ultrasound)
PVAD dwell-time (time from PVAD insertion to removal, in hours)
- Patient reported pain from the insertion procedure (0–10 Verbal Rating Scale)
- Serious adverse events (eg ICU admission)
- Individual complications (Marsh et al, 2020): extravasation or infiltration, occlusion or blockage (with or without leakage), venous thrombosis, catheter dislodgement (complete or partial), phlebitis (vessel wall irritation), or infection (laboratory-confirmed bloodstream infection NHSN 2021 criteria: positive blood culture from the peripheral vein, clinical signs of infection with no other apparent source except the PVAD, and localized PVAD site infection without bloodstream infection; NHSN CVS-VASC criteria)
- Subsequent vascular devices required to complete required treatment (until discharge, insertion of CVAD)
- Satisfaction: patient satisfaction (10-point Likert scale) and staff satisfaction (10-point Likert scale) will be collected after PVAD insertion
- Cost data will be collected in both arms for the cost-effectiveness analysis. This sub-study will focus on collecting direct variable costs (includes cost and number of products used and staff for PVAD insertion; costs of treating any complications, including PVAD reinsertions) in a subset of 30 per arm, assuming fixed costs (eg buildings, overhead) are similar in both arms (Tuffaha et al, 2014).
Sample size
Assuming a 40% first-insertion failure rate in DIVA patients based on previous studies (Farrell et al, 2017; Kleidon et al, 2019; Larsen, 2010; Reigart et al, 2012), we hypothesize an absolute reduction of 15% with the novel catheters (Chiricolo et al, 2015; Idemoto et al, 2014). With two-sided alpha of 0.05 and 90% power, 203 participants in each group (406 total) are required. We do not expect significant attrition or missing outcome data given the study design.
Recruitment, randomization, allocation concealment, and blinding
Hospital-based research nurses (ReNs) will screen the EDs' patients who meet the inclusion criteria on a daily basis (Monday to Friday). These patients will be approached by the ReN, who will explain the trial and give ample opportunity for the patients to ask questions. The participants who sign the consent form will be randomized 1:1 to the GW-PVAD and standard care PVAD groups with varied block sizes of 4 and 6 using a central randomization service. Allocation concealment will be maintained.
Blinding of ReNs, clinicians, and patients to treatment allocation is not possible due to the type of intervention. However, the infectious disease physician who allocates infection outcomes and the statistician will be blinded to treatment allocation.
Data collection
Dedicated ReNs will collect and enter de-identified data into the Research Electronic Data Capture (REDCap, Vanderbilt) (Harris et al, 2009). All patients will be followed by ReNs daily until 48 hours after removal of their PVADs. Clinical data will be collected from inserters, patients, treating clinicians, clinical notes and pathology records. All data will be collected by using a unique study identification number. Quality checks for allocation integrity and 100% data validation for the initial five patients, consent forms, primary outcomes, and a random 5% other data for all patients will be undertaken by the principal investigator.
At study enrolment/insertion
Data including demographic variables (eg age, gender), clinical factors (eg comorbidities, diagnosis), catheter/vessel variables (eg catheter size/length, insertion site, catheterized vein condition, used for blood collection or not, number of PVAD insertion attempts, time to insertion between first needle stick to catheter secured by dressing, need for rescue inserter), number of concurrent PVADs, patient satisfaction score, and staff satisfaction score will be collected by ReNs.
Daily check
ReNs will perform daily checks from Monday to Friday to assess patients for site complications. Weekend daily data will be completed based on the information in patients' clinical notes; an attempt to contact the relevant nurse will be made where missing data exists.
Within 24 hours of PVAD removal
Data regarding PVAD removal time, removal reason, insertion site assessment, completion of treatment, and needs for other vascular access device(s) will be collected.
48 hours post PVAD removal
Microbiological results from blood cultures, catheter tips, or insertion site swabs collected, as per the treating clinical team, will be allocated to our infection outcomes by an independent blinded infectious disease physician (not part of the research team). Additional information about patient outcomes (eg discharge, dead) before 48 hours will also be collected.
Data collection for cost analysis
ReNs will collect direct cost (eg staff time, materials) of a convenience sample of 60 participants in both control and treatment groups (30 in each group). ReNs will observe the length of each insertion time, the number of PVAD insertion attempts, and the materials during insertion (eg PVAD/GW-PVAD, dressing pack, sterile probe for ultrasound guided PVAD insertions, dressings, or securements used to secure the PVAD). The unit costs of required materials will be collected from the most recent Queensland Health purchase prices. Labour costs will be calculated using directly measured time for procedures and estimated hourly salary for Queensland Health staff. Likewise, the total cost of materials will be calculated by multiplying the number of resources used and their unit costs. Costs associated with complications will be estimated using both observed data from the study and literature synthesis. The cost of complications from literature (Tuffaha et al, 2014) will be used, because with the relatively small sample size of 30 and that the risk of complication is low, data from the sample with complications may not be sufficient to generate reliable statistics for the costs of complications.
Validity and reliability
A range of strategies was utilized in the design phase to minimize bias and enhance the internal validity of the trial, including using a randomization website, varied block sizes, allocation concealment, blinding during data analysis, and intention-to-treat analysis method. Sampling from two metropolitan EDs ensures the external validity of the trial and the findings will be generalizable to other metropolitan EDs. Additionally, ReNs and a vascular access specialist will perform inter-rater reliability testing of 5% daily site inspections to assess outcome assessment reliability.
Statistical analysis
Data will be transferred from the REDCap database to Stata (StataCorp, College Station, TX, USA) for analysis. Data will be analyzed according to the intention-to-treat (ITT) principle. A statistical significance level will be set at P ≤0.05. The primary outcome, first-insertion success, will be assessed using logistic regression. Secondary outcomes measured with interval data will be assessed using linear regression analyses, while outcomes with count data will be analyzed using Poisson regression models. Time-to-event data will be analyzed using Cox proportional hazards methods and displayed using Kaplan-Meier curves. Missing data will be modelled for best- and worst-case outcomes and the sensitivity of results to missing data will be investigated using multiple imputation or pattern-mixture methods. Sensitivity analysis will be performed to observe any difference in outcomes related to inserters' skill levels, as well as comparing long GW-PVAD with various types of standard care PVADs.
Cost-effectiveness analysis
A decision tree model will be used to conduct the cost-effectiveness of the trial. Particularly, costs and selected outcomes of the control and intervention groups will be compared to generate an incremental cost-effectiveness ratio (ICER). Statistical properties of costs, outcomes, and the ICER will be generated using Monte Carlo simulations. The estimated ICER will be compared to the willingness to pay (WTP) for the selected outcome to determine the cost-effectiveness of the trial. One-way and probabilistic sensitivity analyses will be conducted to determine the robustness of cost-effectiveness results to changes in factors affecting costs and outcomes. The WTP for the primary outcome will be estimated using two options. First, the WTP is proxied by observed costs associated with successful catheter insertion using the usual care approach. This includes costs of staff time and materials required for successful insertion. Second, literature synthesis and consultation with expert clinicians in the research team will be conducted to ensure the robustness of the findings.
Trial status
Recruitment is scheduled to commence in June 2022. Data collection is anticipated to be completed within six months.
Discussion
Establishing vascular access for DIVA patients is an ongoing challenge for ED clinicians. This RCT is the first trial investigating whether a novel GW-PVAD makes a difference in the first-insertion success rate and PVAD complications in an ED setting. This study will generate new evidence to fill the current knowledge gap and assist ED clinicians to make safe, evidence-based decisions in managing this challenging clinical task.
Strengths and limitations
- This study is a multi-site, fully powered RCT that will provide rigorous evidence regarding the clinical and cost-effectiveness of an innovative GW-PVAD
- A range of strategies was utilized in the planning to ensure the rigour of the design including internal and external validity and inter-rater reliability
- Blinding will not be possible due to the nature of the intervention. However, independent infectious disease physicians, and the statistician, will be blinded to treatment allocation
- The study will be conducted in a single region in Australia with an adult cohort, which might restrict the generalizability of the findings.
Ethics and dissemination
The trial has obtained ethical approval from Metro South Ethics Committee (HREC/2022/QMS/82264). The trial will be conducted in compliance with the Australian Government National Health and Medical Research Council (NHMRC) National Statement on Ethical Conduct in Human Research (National Health and Medical Research Council, 2018). Written consent will be obtained from all participants before randomization. Participants can withdraw from the study at any stage. Patient confidentiality will be maintained as no individual patient data will be presented in publications and conference presentations. Serious adverse events (eg death, unplanned admission to intensive care unit) will be reported to the ethics committee. The findings of the trial will be published in a peer-reviewed journal.