The need for intravenous (IV) antimicrobial therapy is one of the most common reasons patients are admitted to hospital (Australian Commission on Safety and Quality in Health Care, 2019). Traditionally, short peripheral IV catheters (PIVCs) (<4 cm) have been the ‘default’ device chosen to meet patients’ IV needs (for periods of 1-14 days) (Chopra et al, 2015; Marsh et al, 2018a). Over half (55—73%) of short PIVCs inserted in hospitals are for antimicrobial administration, and their performance is vital for effectively treating infection (Wallis et al, 2014; Kleidon et al, 2020a; Larsen et al, 2022). However, up to one-half of short PIVCs fail during treatment, resulting in delayed antimicrobial administration and repetitive insertion procedures (Marsh et al, 2020). An estimated 7—26% of prescribed antibiotic doses are not administered, often a direct result of failed PIVCs (Wallis et al, 2014; Kleidon et al, 2020a; Larsen et al, 2022). This has serious implications for maintaining optimal antibiotic concentrations necessary for effect against infection (Roberts et al, 2012). Interruption to antimicrobial treatment has been linked to re-emergence of infectious disease (Patel et al, 2019), increased hospital stays (Rickard et al, 2012; Marsh et al, 2018b) and antimicrobial resistant organisms (Coleman et al, 2013; Patel et al, 2019; Larsen et al, 2022).
Although centrally inserted devices (eg peripherally inserted central catheters) can also be used for IV antimicrobial administration, they are associated with significant harm, particularly central line-associated bloodstream infections and thrombosis (Chopra et al, 2012; Berenholtz et al, 2014; Kleidon et al, 2021); and are therefore usually preferred for IV therapy lasting longer than 14 days (Chopra et al, 2015). Recently, longer peripheral IV devices (long PIVCs) have been introduced for short-term (1-14 days) peripherally compatible IV therapy. A long PIVC (commonly 4.5 cm to 6.4 cm in adults) is typically inserted into the forearm or upper arm with the catheter tip terminating no higher than the mid-upper arm (Qin et al, 2019). They are often inserted with imaging guidance (eg, ultrasound) but still involve a simple catheter-over-needle technique similar to that for short PIVCs, allowing them to be inserted by short-PIVC competent inserters with minimal additional training (Tuffaha et al, 2014). Compared with short PIVCs, long PIVCs allow greater catheter length within the vein, which has been strongly linked to increased device survival (Elia et al, 2012; Pandurangadu et al, 2018; Bahl et al, 2020). A prospective observational study of PIVCs inserted using ultrasound (n=86) in an emergency department found that 100% of catheters failed when less than 30% of the length was residing in the vein, compared with no failure when greater than 65% of catheter length was in the vein (Pandurangadu et al, 2018).
Although observational data appears to suggest that long PIVCs outperform short PIVCs, this has not been evaluated in an adequately powered randomised controlled trial (RCT). Therefore, the potential benefit of long PIVCs for patients requiring IV antimicrobial therapy is unknown. This is unacceptable considering IV antimicrobials are a high-volume, high-value intervention. This RCT aims to assess if long PIVCs (4.5-6.4 cm) when compared with short PIVCs optimise antimicrobial delivery by reducing PIVC failure rates.
Methods
Design
The study team will conduct a single-centre, two-arm, parallel group RCT, comparing short and long PIVCs for patients requiring IV antimicrobial therapy for 3 days or more. A pilot feasibility study (Phase 1) will be tested within an effectiveness RCT (Phase 2).
In Phase 1, protocol safety and feasibility will be assessed in an interim analysis of feasibility outcomes at 70 participants. This analysis will be presented to an independent committee appointed prior to study commencement to assess feasibility and protocol safety. If feasibility criteria are met, recruitment will continue to a fully powered study (Phase 2). The trial will be conducted in accordance with the Consolidated Standards of Reporting Trials statement (CONSORT) (Schulz et al, 2010) and has been registered with the Australian New Zealand Clinical Trials Registry (ACTRN12621001199808).
Hypotheses
Primary feasibility hypothesis (for Phase 1)
It will be feasible to conduct a definitive RCT comparing antimicrobial therapy interruption between long and short PIVCs, meeting feasibility criteria set a priori.
Primary hypothesis for fully powered RCT (Phase 2)
Patients with a long PIVC will have fewer IV antimicrobial treatment interruptions due to all-cause device failure, compared to patients receiving usual care (short PIVC).
Secondary hypotheses
Compared with patients receiving usual care (short PIVC), patients with a long PIVC will have fewer:
- Vascular access devices inserted over the course of antimicrobial therapy
- Central venous access devices (CVADs) inserted
- PIVC-related complications (insertion-related complications eg, multiple insertion attempts), bloodstream infection, local infection, phlebitis, occlusion, infiltration/extravasation, thrombosis and dislodgement).
Setting and sample
This RCT will take place in the general medical and surgical wards of a large quaternary referral hospital (>900 beds) in Queensland, Australia. Participants requiring a PIVC as part of their treatment will be screened for eligibility and be recruited as per the following inclusion criteria: 18 years or older; requiring peripherally-compatible IV antimicrobial treatment for 3 days or more, who are able to provide informed, written consent. Exclusion criteria are: patients with upper arm restrictions, limiting device placement; non-English-speaking patients without interpreter; patients receiving end-of-life care; or previous study enrolment.
Sample size and study power
For feasibility outcomes (Phase 1), sample size is not determined by statistical power to test hypotheses, rather 30-70 patients per arm are needed to test protocol feasibility and gain initial estimates of effect (Julious, 2005; Hertzog, 2008). The study team therefore plan to include 70 patients (35 patients per arm) in the interim analysis to assess the primary feasibility outcome. An independent data safety monitoring committee, consisting of content and methodological experts will be convened to perform the interim analysis. If feasibility criteria are not met on interim analysis, refinements to screening, recruitment and data collection processes will be made to improve trial efficiency. No changes in inclusion/exclusion criteria or variables collected will be made.
If the effect estimates obtained in Phase 1 require a sample size recalculation (that is, the interim analysis effect estimates are found not to be within the range of data used for initial sample size calculation), the statistician will perform that prior to progressing to Phase 2. If no sample size recalculation is required, Phase 2 will recruit to n =192 (91 patients per arm, plus 5 per arm for potential attrition), which will provide 80% power to detect significant differences between groups at P <0.05. Informing sample size calculations are local and international data reporting the incidence of IV antimicrobial interruption due to device failure of 43% for short PIVCs (Marsh et al. 2015; Marsh et al, 2018a; 2018b; 2018c; Kleidon et al, 2020b; Marsh et al, 2022) compared with 24% for long PIVCs (Bahl et al, 2020; Elia et al, 2012).
Study interventions
Participants will be randomised to receive either a short PIVC (standard care, control), less than 4 cm length eg Insyte Autoguard BC (BD, Franklin Lakes, NJ, USA) inserted into arm veins, or a long PIVC (intervention), 4.5-6.4 cm length, eg Introcan Safety Deep Access (B Braun, Melsungun, Germany) or Insyte Autoguard BC (BD, Franklin Lakes, NJ, USA) inserted in the upper arm or forearm veins. Ultrasound guidance may be used to place PIVCs as required.
Outcome measures
Primary feasibility outcome (Phase 1)
Feasibility of the protocol and processes will be established by meeting the following criteria, which are based on previous PIVC feasibility studies (Marsh et al, 2018b; Corley et al, 2023):
- Proportion of those screened who are eligible (>70%)
- Proportion of those eligible who consent (>90%)
- Proportion of those who consent who later withdraw or who are lost to follow up (≤5%)
- Proportion who adhere to the randomised protocol (>90%)
- Proportion of missing data (<5%)
- Proportion of patients satisfied with the study interventions (>80% scoring >7 on a 0-10 numerical rating scale)
- Estimates of effect to inform a feasible sample size for a larger trial.
Primary clinical outcome (Phase 2)
Disruption to antimicrobial administration from all-cause post insertion PIVC failure. A composite measure of:
- Infiltration/extravasation (infusate leaking into the interstitial tissue) (Larsen et al, 2021)
- Blockage/occlusion (PIVC will not infuse, with or without external leakage) (Marsh et al, 2018b)
- Phlebitis (defined as pain ≥2/10 on a numerical rating scale; or >2 signs/symptoms ofpain/tenderness, erythema, swelling, or palpable cord) (Larsen et al, 2021)
- Suspected or confirmed thrombosis (suspected: as assessed/suspected by the treating clinician; confirmed: ultrasound/venographic confirmed thrombosed vessel at PIVC site)
- Dislodgement (complete or partial removal of the PIVC from the vein) (Marsh et al, 2018b)
- Infection (local and laboratory-confirmed bloodstream infection) (Centers for Disease Control and Prevention, 2023).
Secondary outcome measures
- Number, type and dwell time of IV devices inserted per patient to complete the course of antimicrobials (randomised and any subsequent short PIVC, long PIVC, or CVAD)
- Overall insertion success (yes/no)
- Number of PIVCs unable to be successfully inserted
- Number of insertion attempts (needle punctures)
- Patient-reported insertion pain (0-10 numerical rating scale)
- Patient-reported satisfaction of PIVC insertion and removal (0-10 numerical rating scale)
- Staffand patient overall satisfaction with PIVC device (0-10 numerical rating scale)
- Individual PIVC complications as listed above in the primary clinical outcome
- Adverse events (eg, nerve damage, adverse skin event (rash, blister, itchiness, skin tears, adhesive residue))
- Cost analysis (in a subset of seven participants per study arm), including costs of treating PIVC-related complications and PIVC reinsertion.
Study procedures
Recruitment, randomisation, allocation concealment and blinding
A research nurse will screen wards daily to identify patients who meet all the inclusion and no exclusion criteria, recorded in a screening log. For eligible patients, the research nurse will liaise with the bedside nurse who will ask the patient’s permission to be approached for trial participation. If they agree, the research nurse will explain the trial, answering questions and provide the patient with a written information document and consent form. Patients will be given ample time to consider study participation. If the patient agrees to participate, they will complete the consent form to participate in the trial.
Using a central, web-based randomisation service (Research Electronic Data CAPture (REDCap), Vanderbilt University) (Harris et al, 2009), with allocation concealment prior to randomisation, participants will be randomised 1:1 to the short PIVC or long PIVC group, with randomly varied block sizes of 4 and 6. Each patient will be assigned a unique study identification number. Due to the nature of the intervention, blinding ofpatients/clinicians to the intervention is not possible. However, the statistician will be blinded for analysis and the microbiologist will be blinded to treatment allocation when apportioning infection outcomes.
PIVC care
All PIVCs will be inserted as per hospital policy by the research nurse experienced in PIVC insertion. PIVC length will be determined by randomisation, however insertion site and PIVC gauge will be based on patient assessment and treatment needs. After insertion and initial dressing and securement of the PIVC, all PIVCs will be maintained by clinical staff, as per hospital policy. Any additional dressings or securement products used by clinical staff will be recorded daily by the research staff. PIVC removal will be at the discretion of the clinical team using usual hospital criteria, namely PIVC complication necessitating removal, 72-hourly routine resite, and/or completion of therapy. Protocol violations are defined as randomised participants who never receive a PIVC or the randomised intervention.
Data collection
Data will be collected from the patients’ medical records and entered into the REDCap database, using a unique study identification number. Enrolled participants will be visited daily until PIVC removal by the research nurse (excluding weekends, during which data will be collected retrospectively from medical records) who will perform PIVC site assessment and collect data for primary and secondary outcomes (see Table 1). Data for feasibility outcomes will be sourced from the study screening logs (for the first three outcomes) and REDCap database (for the remaining feasibility outcomes).
Table 1. Data collected by study time point
| On enrolment | On PIVC insertion | Daily data | On PIVC removal | At 48 hours+ post-removal |
|---|---|---|---|---|
|
|
|
|
|
IV=intravenous; PIVC=peripheral intravenous catheter; VAD=vascular access device
Cost analysis substudy
In a subset of seven patients per arm, cost of consumables used for PIVC insertion will be collected. Procedural timings will be observed for three PIVC insertions per arm; and costed at ‘registered nurse’ (as per Queensland Health nursing pay scale). For PIVCs requiring replacement after device failure, reinsertion costs will be included. Costs of any PIVC-related complications (eg, bloodstream infection) will also be calculated in the sub-group.
Validity and reliability
The study team have in-built trial processes to ensure the internal validity of the study. To reduce selection bias, web-based randomisation with randomly varying block size and allocation concealment until randomisation will be used. To reduce detection bias, the infectious outcomes assessor and the data analyst will be blinded. To reduce attrition bias, an intention-to-treat analysis will be used during data analysis and all randomised patients will be reported. External validity of the study findings will be enhanced by sampling from a general medical and surgical population of a large quaternary referral hospital so that, despite this being a single-centre trial, the results are generalisable to most general ward patients requiring a PIVC as part of their care. Furthermore, clinically relevant outcome measures which are important to patients have been selected.
Statistical analysis
Trial data will be exported from the REDCap database to Stata V13.1 (Statacorp, College Station, TX, USA) for analysis. Trial feasibility outcomes will be reported descriptively and compared against the a priori acceptability limits. All randomised patients will be analysed by intention to treat. For the primary outcome, the number and proportion of devices that experience a disruption to antimicrobial administration due to all-cause post-insertion PIVC failure will be calculated per group. The between-group difference will be analysed using logistic regression with study group as the main effect. The result will be presented as an odds ratio (95% CI); P value. For secondary outcomes that are continuous, the mean and standard deviation (SD) will be calculated per group.
The between-group difference will be computed using linear regression, with study group as the main effect and the result will be presented as the mean difference (95% CI); P value. Outcomes that are count variables will be compared using Poisson regression with study group included as the main effect. Incidence rates will be presented per group, and the relative difference will be presented as an incidence rate ratio (95% CI); P value. Cost analysis will be reported descriptively. P values <0.05 will be considered significant.
Ethical considerations and dissemination
Ethics
Ethical and regulatory approvals were obtained from the Royal Brisbane and Women’s Hospital Human Research Ethics Committee (HREC) (HREC/2020/QRBW/75234) and Griffith University (Ref. 022/165). All study procedures will be performed in compliance with relevant laws and institutional guidelines. Written informed consent will be obtained from all participants prior to enrolment and randomisation. This study poses only minimal risk to study participants as both long and short PIVCs are approved for use and will be used in accordance with the manufacturer’s instructions by staff with appropriate training. Study participants will be visited daily to monitor for any adverse events (PIVC-related bloodstream infection, unplanned admission to intensive care and death), which will be reported to the HREC within the mandatory reporting timeframe. The interim analysis planned at n=70 will assess feasibility and safety data to ensure intervention safety and efficient trial processes.
Study data (paper and electronic) will be stored securely as per the Australian National Health and Medical Research Council guidelines and only accessed by the research team. All data will be deidentified at time of entering into the database. Only aggregate data will be presented in any publications or presentations
Dissemination of findings
Ensuring the efficient delivery of IV antimicrobials to treat infection will be of great interest to many stakeholders. Study results will be presented locally to key stakeholders to drive practice change. The investigators will present the findings at national and international infection prevention and vascular access specialty meetings and conferences, and a high-quality peer-reviewed journal will be targeted for publication of results. In this way, swift uptake into clinical practice guidelines related to antimicrobial delivery and vascular access device selection will be maximised.
Trial status
The trial commenced recruitment in late 2021, however this has been impacted by COVID-19 surges within the health service. Active recruitment is ongoing, and the team expect recruitment to be complete by April 2023.
Conclusion
This fully powered RCT, with in-built feasibility analysis to refine and guide recruitment and data collection processes, will compare short and long PIVCs for patients undergoing three or more days of IV antimicrobial therapy, assessing IV antimicrobial treatment interruptions, device performance and overall vascular access experience for patients completing their prescribed therapy. Short PIVCs are not meeting the needs of patients or clinicians and reliable device alternatives are urgently needed to address unacceptably high PIVC failure rates.
Key Points
- This study protocol highlights the need to test alternative longer PIVCs to prevent interruption to antimicrobial therapy, which can contribute to suboptimal antibiotic concentrations
- Short PIVCs (control) will be compared with long PIVCs in a parallel superiority randomised controlled trial, with in-built feasibility analysis
- In-built trial processes and planning will ensure the internal and external validity of the study
- The study results will be of interest to health professionals and will drive practice change as ensuring the efficient delivery of IV antimicrobials to treat infection is a patient safety priority
CPD reflective questions
- Is the study protocol feasible to compare short and long PIVCs?
- Can long PIVCs potentially decrease interruptions to intravenous antimicrobial therapy?
- Will long PIVCs potentially decrease the number of vascular access devices patients need to complete antimicrobial treatment?