Peripherally inserted central catheter design and material for reducing catheter failure and complications

Abstract

Background

Peripherally inserted central catheters (PICCs) facilitate diagnostic and therapeutic interventions in health care. PICCs can fail due to infective and non‐infective complications, which PICC materials and design may contribute to, leading to negative sequelae for patients and healthcare systems.

Objectives

To assess the effectiveness of PICC material and design in reducing catheter failure and complications.

Search methods

The University of Queensland and Cochrane Vascular Information Specialist searched the Cochrane Vascular Specialised Register, CENTRAL, MEDLINE, Embase, and CINAHL databases and the WHO ICTRP and ClinicalTrials.gov trials registers to 16 May 2023. We aimed to identify other potentially eligible trials or ancillary publications by searching the reference lists of retrieved included trials, as well as relevant systematic reviews, meta‐analyses, and health technology assessment reports. We contacted experts in the field to ascertain additional relevant information.

Selection criteria

We included randomised controlled trials (RCTs) evaluating PICC design and materials.

Data collection and analysis

We used standard Cochrane methods. Our primary outcomes were venous thromboembolism (VTE), PICC‐associated bloodstream infection (BSI), occlusion, and all‐cause mortality. Secondary outcomes were catheter failure, PICC‐related BSI, catheter breakage, PICC dwell time, and safety endpoints. We assessed the certainty of evidence using GRADE.

Main results

We included 12 RCTs involving approximately 2913 participants (one multi‐arm study). All studies except one had a high risk of bias in one or more risk of bias domain.

Integrated valve technology compared to no valve technology for peripherally inserted central catheter design

Integrated valve technology may make little or no difference to VTE risk when compared with PICCs with no valve (risk ratio (RR) 0.71, 95% confidence interval (CI) 0.19 to 2.63; I² = 0%; 3 studies; 437 participants; low certainty evidence). We are uncertain whether integrated valve technology reduces PICC‐associated BSI risk, as the certainty of the evidence is very low (RR 0.20, 95% CI 0.01 to 4.00; I² = not applicable; 2 studies (no events in 1 study); 257 participants). Integrated valve technology may make little or no difference to occlusion risk when compared with PICCs with no valve (RR 0.86, 95% CI 0.53 to 1.38; I² = 0%; 5 studies; 900 participants; low certainty evidence). We are uncertain whether use of integrated valve technology reduces all‐cause mortality risk, as the certainty of evidence is very low (RR 0.85, 95% CI 0.44 to 1.64; I² = 0%; 2 studies; 473 participants).

Integrated valve technology may make little or no difference to catheter failure risk when compared with PICCs with no valve (RR 0.80, 95% CI 0.62 to 1.03; I² = 0%; 4 studies; 720 participants; low certainty evidence). We are uncertain whether integrated‐valve technology reduces PICC‐related BSI risk (RR 0.51, 95% CI 0.19 to 1.32; I² = not applicable; 2 studies (no events in 1 study); 542 participants) or catheter breakage, as the certainty of evidence is very low (RR 1.05, 95% CI 0.22 to 5.06; I² = 20%; 4 studies; 799 participants).

Anti‐thrombogenic surface modification compared to no anti‐thrombogenic surface modification for peripherally inserted central catheter design

We are uncertain whether use of anti‐thrombogenic surface modified catheters reduces risk of VTE (RR 0.67, 95% CI 0.13 to 3.54; I² = 15%; 2 studies; 257 participants) or PICC‐associated BSI, as the certainty of evidence is very low (RR 0.20, 95% CI 0.01 to 4.00; I² = not applicable; 2 studies (no events in 1 study); 257 participants). We are uncertain whether use of anti‐thrombogenic surface modified catheters reduces occlusion (RR 0.69, 95% CI 0.04 to 11.22; I² = 70%; 2 studies; 257 participants) or all‐cause mortality risk, as the certainty of evidence is very low (RR 0.49, 95% CI 0.05 to 5.26; I² = not applicable; 1 study; 111 participants).

Use of anti‐thrombogenic surface modified catheters may make little or no difference to risk of catheter failure (RR 0.76, 95% CI 0.37 to 1.54; I² = 46%; 2 studies; 257 participants; low certainty evidence). No PICC‐related BSIs were reported in one study (111 participants). As such, we are uncertain whether use of anti‐thrombogenic surface modified catheters reduces PICC‐related BSI risk (RR not estimable; I² = not applicable; very low certainty evidence). We are uncertain whether use of anti‐thrombogenic surface modified catheters reduces the risk of catheter breakage, as the certainty of evidence is very low (RR 0.15, 95% CI 0.01 to 2.79; I² = not applicable; 2 studies (no events in 1 study); 257 participants).

Antimicrobial impregnation compared to non‐antimicrobial impregnation for peripherally inserted central catheter design

We are uncertain whether use of antimicrobial‐impregnated catheters reduces VTE risk (RR 0.54, 95% CI 0.05 to 5.88; I² = not applicable; 1 study; 167 participants) or PICC‐associated BSI risk, as the certainty of evidence is very low (RR 2.17, 95% CI 0.20 to 23.53; I² = not applicable; 1 study; 167 participants). Antimicrobial‐impregnated catheters probably make little or no difference to occlusion risk (RR 1.00, 95% CI 0.57 to 1.74; I² = 0%; 2 studies; 1025 participants; moderate certainty evidence) or all‐cause mortality (RR 1.12, 95% CI 0.71 to 1.75; I² = 0%; 2 studies; 1082 participants; moderate certainty evidence).

Antimicrobial‐impregnated catheters may make little or no difference to risk of catheter failure (RR 1.04, 95% CI 0.82 to 1.30; I² = not applicable; 1 study; 221 participants; low certainty evidence). Antimicrobial‐impregnated catheters probably make little or no difference to PICC‐related BSI risk (RR 1.05, 95% CI 0.71 to 1.55; I² = not applicable; 2 studies (no events in 1 study); 1082 participants; moderate certainty evidence). Antimicrobial‐impregnated catheters may make little or no difference to risk of catheter breakage (RR 0.86, 95% CI 0.19 to 3.83; I² = not applicable; 1 study; 804 participants; low certainty evidence).

Authors' conclusions

There is limited high‐quality RCT evidence available to inform clinician decision‐making for PICC materials and design. Limitations of the current evidence include small sample sizes, infrequent events, and risk of bias. There may be little to no difference in the risk of VTE, PICC‐associated BSI, occlusion, or mortality across PICC materials and designs. Further rigorous RCTs are needed to reduce uncertainty.

Plain language summary

What are the benefits and risks of catheter material and design innovations (thin plastic tubes that deliver blood and medicine to the veins) to reduce complications and infections that lead to premature catheter failure?

Key messages

Peripherally inserted central catheters (PICCs) are associated with serious complications, a high failure rate, and negative aftereffects for patients and healthcare systems.

There is a lack of strong evidence to evaluate the benefits and risks of PICC design and material (e.g. silicone versus polyurethane) to prevent catheter complications, infection, and failure.

Future research in this area should focus on the effectiveness of surface‐modified catheter material (anti‐clot barrier), antimicrobial coating (anti‐infection barrier), and new developments in catheter designs (e.g. polyurethane catheters with pressure valves) to reduce catheter failure, as well as looking at any unwanted effects of these treatments.

What is a peripherally inserted central catheter, and what is it used for?

A PICC is a tube that is inserted in a peripheral vein with the catheter tip advanced to a central vein in adults and children requiring medical treatment such as fluid, blood products, or medication. These invasive medical devices are typically inserted in patients requiring therapy that is harmful to smaller, peripheral veins and/or for those needing therapy for more than seven days.

PICCs are commonly complicated by infection, catheter blockage, and breakage. These complications lead to device failure, removal, and interruption of medical treatment.

PICC failures necessitating PICC reinsertions are associated with increased complications and increased procedural complexity due to vascular anatomical changes. In children, subsequent PICC placement is associated with an increased risk of deep vein thrombosis.

What did we want to find out?

We wanted to know if different PICC material (what the PICC is made from) or design (e.g. valve or no valve) reduces the occurrence of PICC complications and device failure.

What did we do?

We searched medical databases up to 16 May 2023 for randomised controlled trials (a type of study where participants are randomly assigned to one of two or more treatment groups) that compared different PICC material and designs in patients of any age.

What did we find?

We found 12 trials (including approximately 2913 participants) of different catheter material and designs used to prevent catheter complications and infections. Overall, there was not enough information to determine what catheter material or design should be used to prevent catheter complications such as blood clots and catheter infections.

There may be a slightly reduced risk of the catheter breaking or failing when an open‐ended tip is used. Use of a proximal valve may also slightly reduce the risk of the catheter breaking or failing. More research is needed with larger groups of people to confirm whether different PICC materials and designs are effective at preventing complications and infections. Patients should talk to their doctor about the different catheter materials and designs available, and which one is best for them based on their individual needs and risks.

What are the limitations of this evidence?

We are not confident in the evidence because most trials were small (fewer than 300 people) with few complication events. There is a need for trials with many hundreds or even thousands of people included to find out if these catheter materials or designs help prevent complications and infection. Further, methods used to randomly assign participants to treatment groups and blinding of outcome assessors to the treatment received were not well documented and should be improved in future trials.

How up‐to‐date is the evidence?

The evidence is current to 16 May 2023.

Summary of findings

Summary of findings 1. No valve versus integrated valve technology for peripherally inserted central catheter design.

No valve versus integrated valve technology for peripherally inserted central catheter design
Patient or population: patient of any age Setting: hospital Intervention: integrated valve technology Comparison: no valve
Outcomes	Anticipated absolute effects (95% CI)*		Relative effect (95% CI)	No. of participants (studies)	Certainty of evidence (GRADE)
	Risk with no valve	Risk with integrated valve technology
Venous thromboembolism (follow‐up: insertion to 2 months)	27 per 1000	19 per 1000 (5 to 70)	RR 0.71 (0.19 to 2.63)	437 (3 RCTs)	⨁⨁◯◯ Low a
PICC‐associated BSI (follow‐up: 2 days to 2 months)	16 per 1000	3 per 1000 (0 to 62)	RR 0.20 (0.01 to 4.00)	257 (2 RCTs)	⨁◯◯◯ Very low b
Occlusion (follow‐up: insertion to 2 months)	59 per 1000	51 per 1000 (31 to 82)	RR 0.86 (0.53 to 1.38)	900 (5 RCTs)	⨁⨁◯◯ Low c
All‐cause mortality (follow‐up: insertion to 2 months)	76 per 1000	65 per 1000 (33 to 125)	RR 0.85 (0.44 to 1.64)	473 (2 RCTs)	⨁◯◯◯ Very low d
Catheter failure (follow‐up: insertion to 2 months)	275 per 1000	220 per 1000 (171 to 284)	RR 0.80 (0.62 to 1.03)	720 (4 RCTs)	⨁⨁◯◯ Low c
PICC‐related BSI (follow‐up: insertion to 2 months)	50 per 1000	25 per 1000 (9 to 66)	RR 0.51 (0.19 to 1.32)	542 (2 RCTs)	⨁◯◯◯ Very low e
Catheter breakage (follow‐up: insertion to 2 months)	14 per 1000	14 per 1000 (3 to 68)	RR 1.05 (0.22 to 5.06)	799 (4 RCTs)	⨁◯◯◯ Very low f
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). BSI: bloodstream infection; CI: confidence interval; PICC: peripherally inserted central catheter; RCT: randomised controlled trial; RR: risk ratio
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

^aWe downgraded a total of two levels due to concerns related to risk of bias (uncertain allocation concealment and sequence generation, one study) and imprecision (few events).
^bWe downgraded three levels due to concerns related to imprecision (few events, small sample size, and wide CIs).
^cWe downgraded a total of two levels due to concerns related to risk of bias (uncertain allocation concealment and sequence generation, one study) and imprecision (small study sizes (early study termination)).
^dWe downgraded a total of two levels due to concerns related to risk of bias (uncertain allocation concealment and sequence generation, one study) and imprecision (few events and small study sizes (early study termination)).
^eWe downgraded three levels due to concerns related to risk of bias (uncertain allocation concealment, sequence generation, and blinding of outcome assessor) and imprecision (few events and small sample sizes (early study termination)).
^fWe downgraded a total of three levels due to concerns related to risk of bias (uncertain allocation concealment and sequence generation, one study) and imprecision (few events, small study sample sizes, and wide CIs).

Summary of findings 2. Catheters without anti‐thrombogenic surface modification versus anti‐thrombogenic surface‐modified catheters for peripherally inserted central catheter design.

Catheters without anti‐thrombogenic surface modification versus anti‐thrombogenic surface‐modified catheters for peripherally inserted central catheter design
Patient or population: patient of any age Setting: hospital Intervention: anti‐thrombogenic surface modification Comparison: no anti‐thrombogenic surface modification
Outcomes	Anticipated absolute effects (95% CI)*		Relative effect (95% CI)	No. of participants (studies)	Certainty of evidence (GRADE)
	Risk with no modification	Risk with anti‐thrombogenic surface modification
Venous thromboembolism (follow‐up: insertion to 2 months)	39 per 1000	26 per 1000 (5 to 137)	RR 0.67 (0.13 to 3.54)	257 (2 RCTs)	⨁◯◯◯ Very low a
PICC‐associated BSI (follow‐up: 2 days to 2 months)	16 per 1000	3 per 1000 (0 to 62)	RR 0.20 (0.01 to 4.00)	257 (2 RCTs)	⨁◯◯◯ Very low a
Occlusion (follow‐up: insertion to 2 months)	54 per 1000	37 per 1000 (2 to 609)	RR 0.69 (0.04 to 11.22)	257 (2 RCTs)	⨁◯◯◯ Very low a
All‐cause mortality (follow‐up: insertion to 2 months)	36 per 1000	18 per 1000 (2 to 191)	RR 0.49 (0.05 to 5.26)	111 (1 RCT)	⨁◯◯◯ Very low a
Catheter failure (follow‐up: insertion to 2 months)	217 per 1000	165 per 1000 (80 to 334)	RR 0.76 (0.37 to 1.54)	257 (2 RCTs)	⨁⨁◯◯ Low b
PICC‐related BSI (follow‐up: insertion to 2 months)	No events were reported in either study group.		‐	111 (1 RCT)	⨁◯◯◯ Very low c
Catheter breakage (follow‐up: insertion to 2 months)	23 per 1000	3 per 1000 (0 to 65)	RR 0.15 (0.01 to 2.79)	257 (2 RCTs)	⨁◯◯◯ Very low a
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). BSI: bloodstream infection; CI: confidence interval; PICC: peripherally inserted central catheter; RCT: randomised controlled trial; RR: risk ratio
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

^aWe downgraded a total of three levels due to concerns related to imprecision (few events, small sample size, and wide CIs).
^bWe downgraded a total of two levels due to concerns related to imprecision (few events and small sample size).
^cWe downgraded a total of three levels due to concerns related to imprecision (no events, small sample size, and wide CIs).

Summary of findings 3. No antimicrobial‐impregnated catheters versus antimicrobial‐impregnated catheters for peripherally inserted central catheter design.

No antimicrobial‐impregnated catheters versus antimicrobial‐impregnated catheters for peripherally inserted central catheter design
Patient or population: patient of any age Setting: hospital Intervention: antimicrobial impregnation Comparison: no antimicrobial impregnation
Outcomes	Anticipated absolute effects (95% CI)*		Relative effect (95% CI)	No. of participants (studies)	Certainty of evidence (GRADE)	Comments
	Risk with no modification	Risk with anti‐thrombogenic surface modification
Venous thromboembolism (follow‐up: insertion to 2 months)	23 per 1000	12 per 1000 (1 to 135)	RR 0.54 (0.05 to 5.88)	167 (1 RCT)	⨁◯◯◯ Very low a
PICC‐associated BSI (follow‐up: 2 days to 2 months)	11 per 1000	25 per 1000 (0 to 0)	RR 2.17 (0.20 to 23.53)	167 (1 RCT)	⨁◯◯◯ Very low a	1 additional study investigated this outcome, but incomplete reporting precluded pooling of data.
Occlusion (follow‐up: insertion to 2 months)	46 per 1000	46 per 1000 (26 to 80)	RR 1.00 (0.57 to 1.74)	1025 (2 RCTs)	⨁⨁⨁◯ Moderate b
All‐cause mortality (follow‐up: insertion to 2 months)	60 per 1000	68 per 1000 (43 to 106)	RR 1.12 (0.71 to 1.75)	1082 (2 RCTs)	⨁⨁⨁◯ Moderate b
Catheter failure (follow‐up: insertion to 2 months)	560 per 1000	583 per 1000 (459 to 728)	RR 1.04 (0.82 to 1.30)	221 (1 RCT)	⨁⨁◯◯ Low c
PICC‐related BSI (follow‐up: 2 days to 2 months)	80 per 1000	84 per 1000 (57 to 125)	RR 1.05 (0.71 to 1.55)	1082 (2 RCTs)	⨁⨁⨁◯ Moderate b
Catheter breakage (follow‐up: insertion to 2 months)	9 per 1000	8 per 1000 (2 to 36)	RR 0.86 (0.19 to 3.83)	804 (1 RCT)	⨁⨁◯◯ Low d
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). BSI: bloodstream infection; CI: confidence interval; PICC: peripherally inserted central catheter; RCT: randomised controlled trial; RR: risk ratio
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

^aWe downgraded a total of three levels due to concerns related to imprecision (few events, small sample size (early study termination), and wide CIs).
^bWe downgraded one level due to concerns related to risk of bias (uncertain sequence generation).
^cWe downgraded one level due to concerns related to risk of bias (uncertain sequence generation) and one level due to concerns related to imprecision (small sample size).
^dWe downgraded a total of two levels due to concerns related to imprecision (few events and wide CIs).

Background

Peripherally inserted central catheters (PICCs) are routinely inserted in adults and children who require intermediate intravascular therapy such as total parenteral nutrition (Russell 2014; Ullman 2017). PICCs are long (50 cm to 60 cm), flexible catheters usually constructed of polyurethane or silicone material. Typically inserted in the basilic, brachial, or cephalic veins of the upper arm, the tip of the PICC terminates in a central vessel providing natural haemodilution of irritant infusates, such as chemotherapy. PICC use has increased in recent decades due to perceived advantages in comparison to central venous catheters (CVCs), such as cost (bedside placement by non‐medical staff) and reduced complication profile during insertion (Bertoglio 2016; Chopra 2013a; Johansson 2013). PICCs support therapeutic interventions for as long as the treatment is required. A recent US study of hospitalised patients across 52 centres reported a median PICC dwell of 8 to 56 days (Chopra 2022). PICCs are associated with greater patient‐reported satisfaction in adults requiring central venous access (Periard 2008). They have also been demonstrated to be a cost‐effective intervention, with the average cost of a PICC insertion estimated at USD 690 per patient (Periard 2008), compared to approximately USD 1500 for other central venous access devices (Di Carlo 2012). However, despite these perceived benefits, PICC complications are common, with 30% of PICCs failing prior to the completion of therapy (Shen 2009; Ullman 2015).

Description of the condition

PICC failure can occur due to infectious (catheter‐associated blood stream infection (CABSI)), mechanical (occlusion, dislodgement, or breakage), or vascular (deep venous thrombosis (DVT)) complications (Abedin 2008; Yamamoto 2002; Yap 2006). A systematic review in paediatric patients found PICCs had the second‐highest failure rate of all central venous access devices after haemodialysis catheters (incidence rate 12.4; 95% confidence interval (CI) 10.0 to 14.9) (Ullman 2015). In adults, a meta‐analysis of more than 29,000 patients demonstrated that PICCs were associated with a higher risk of DVT than CVCs (odds ratio 2.55, 95% CI 1.54 to 4.23, P < 0.001), with a weighted frequency of PICC‐related DVT of 6.7% (95% CI 4.7 to 8.6) (Chopra 2013b). PICC‐related DVTs were more prevalent in vulnerable diagnostic groups such as intensive care patients or those with an oncology diagnosis (Chopra 2013b). PICC‐related thrombus is associated with increased morbidity and mortality (Chopra 2012). In paediatrics, a systematic review of 74 cohort studies found that PICCs had the highest incidence of catheter occlusion or blockage per 1000 catheter days (Ullman 2015). PICC‐related DVTs affect around 7% of children, with recent pilot trial data demonstrating that DVTs were the primary reason for PICC failure (Kleidon 2018). However, the true incidence of PICC‐related DVTs is likely to be higher due to the presence of unscreened and asymptomatic DVTs.

Infection is another serious complication associated with central venous access and PICC placement. PICC‐related CABSI affects more than 5% of hospitalised adults (Chopra 2013c), and 8% of hospitalised children (Ullman 2015), with Staphylococcus aureus and S epidermidis the most commonly isolated pathogens (Ullman 2015). A cohort study of hospitalised adults (966 PICCs; 26,887 catheter days) found that CABSI was associated with the use of multi‐lumen PICCs and longer length of hospitalisation (Chopra 2014). CABSI particularly impacts vulnerable patients such as neonates or oncology patients (Chopra 2013c; Shalabi 2015), and is associated with an almost three‐fold increase in hospital mortality (Ziegler 2015). PICC failure due to CABSI is estimated to cost the US healthcare system between USD 11,000 and USD 69,000 per episode (Warren 2006; Wilson 2014).

Description of the intervention

It is purported that innovation in PICC material and design might reduce the incidence of PICC complications and associated failure. The interventions under consideration are new PICC material (power injectable polyurethane, and anti‐thrombogenic or antimicrobial surface modifications) and design (integrated valve technology). Desirable properties of PICC material and design include:

• soft, flexible, and robust catheter material for patient comfort and ease of insertion, reducing procedural risk such as vessel trauma;

• mechanical stability of the outer lumen to withstand rapid injection pressures, while maximising the internal lumen to enable reliable infusion and aspiration;

• low adherence of blood components, biofilm formation, and microbial colonisation; and

• cost‐efficiency.

How the intervention might work

The primary goal of central venous access with a PICC is to facilitate the reliable delivery of infusates over a prolonged period, for both in‐ and outpatient settings. Choice of PICC material and design can contribute to preventing or reducing device failure and subsequent reinsertion procedures. Early PICCs were predominately manufactured using silicone‐based materials, which were considered soft and 'stretchable' for increased patient comfort and ease of insertion (Gallieni 2008). Due to the softer, less stable nature of silicone, more plastic was required in the outer lumen to provide the requisite stability. The unavoidable trade‐off to this was a narrower, internal lumen with lower injectable pressures (50 pounds per square inch (psi)). The consequence of this was an increased risk of fracture and dislodgement (Poli 2016). Subsequent PICCs were composed of a stronger polyurethane material which tolerated higher pressures (> 100 psi) with only a small decrease in catheter flexibility. However, these first‐generation polyurethane PICCs were associated with an increased risk of phlebitis and vessel trauma due to the rigidity of the material (Seckold 2015). Newer polyurethane PICCs are composed of material blends, which soften with body temperature and facilitate injection pressures of up to 300 psi (power injectable), while maintaining catheter workability and resilience (May 2015).

Surface‐modified PICCs are becoming increasingly available in clinical practice. Current approaches for introducing anti‐thrombogenic properties into PICCs include the use of hydrophilic, hydrophobic, or biological surfaces, and the addition of drugs (Ullman 2018). Hydrophilic surfaces ‐ or grafted surface polymers ‐ decrease protein reabsorption through a water‐soluble surface layer; conversely, hydrophobic polymers rapidly absorb proteins and have the ability to 'repel' water mass. Biological grafting entails coating the surface of the PICC with a protein that may reduce the development of thrombosis such as albumin (Freitas 2003). In vitro studies have demonstrated anti‐thrombogenic PICC surface modifications that inhibit platelet activation, and adherence of blood components to the catheter wall, with the potential to lower the risk of catheter‐associated venous thrombosis (Kleidon 2018). Finally, catheter coating or impregnation using drugs such as chlorhexidine may inhibit bacterial cell growth and division within the PICC. This is particularly important in CABSI prevention strategies, where detached microbial cells from the biofilm can re‐infect the blood, leading to organism resistance to antibiotics. Antimicrobial PICCs are impregnated with an antibiotic (e.g. minocycline plus rifampicin) or antiseptic (chlorhexidine‐silver sulfadiazine). Antimicrobial impregnation in CVCs has been demonstrated to reduce the risk of CABSI significantly by preventing intraluminal colonisation (Lai 2016). It is proposed that PICCs with modified materials provide protective properties, while not adversely impacting catheter mechanical properties (Mermel 2001; Raad 2009).

In addition to PICC material modification, innovation in PICC design, such as valved technology, has been initiated to reduce the occurrence of PICC‐related complications such as occlusion. Valved PICCs incorporate a valve either at the distal tip in a closed‐end catheter, or proximally, at the catheter hub. The valve opens with infusion or aspiration pressure (Hoffer 2001), and closes during pressure fluctuations, creating a closed system and reducing the potential for blood reflux into the catheter (Pittiruti 2014). It is suggested that valve technology might provide clinical improvement through reducing the risk of thrombus formation, occlusion, and subsequently device fracture (Hoffer 2001; Kleidon 2018). With the number and variability in technologically modified PICCs, clinicians need to know the clinical and cost‐effectiveness of these catheter modifications, and which innovations reduce PICC‐associated complications and failure.

Why it is important to do this review

PICCs can be associated with serious complications, a high failure rate, and negative sequelae for patients and healthcare systems. PICC failures necessitating PICC reinsertions are associated with increased complications and increased procedural complexity due to vascular anatomical changes (Yang 2012). In children, subsequent PICC placement is associated with an increased risk of DVT. In a prospective cohort study, Gnannt 2018 describes an almost six‐fold increase in the risk of developing a symptomatic DVT compared to first PICC placement (95% CI 2.25 to 16.04). With such significant patient safety outcomes, choosing the most appropriate PICC for the patient and healthcare system is important. Knowledge of the benefits and harms of various PICC materials and designs will likely contribute to these important catheter selection decisions, and subsequent reduction in healthcare‐related costs and PICC complications.

Objectives

To assess the effectiveness of PICC material and design in reducing catheter failure and complications.

Methods

Criteria for considering studies for this review

Types of studies

We included randomised controlled trials (RCTs) that evaluated PICC design and materials. In order to minimise potential bias, we did not include cluster‐ or cross‐over RCTs or quasi‐experimental studies (Higgins 2021).

Types of participants

We included participants of any age, in any setting (inpatient or outpatient), who required a PICC.

Types of interventions

We included trials comparing one type of catheter material or catheter design to standard PICC (without design or material modification), or with any other modification. Modification of PICC material and design could include, but was not limited to, the following.

Material:

• power‐injectable polyurethane;

• polyurethane;

• surface‐modified polyurethane;

• chemical‐bonded PICCs;

• medication‐impregnated PICCs;

• antiseptic‐coated PICCs.

Design:

• valve technology, including distal or proximal valves.

We planned to investigate the following main comparisons:

• anti‐thrombogenic surface‐modified catheters versus catheters without surface modification;

• antimicrobial‐impregnated catheters versus non‐impregnated catheters;

• catheters with integrated‐valve technology versus catheters without valve technology; and

• power‐injectable polyurethane catheters versus silicone catheters.

Types of outcome measures

Follow‐up time for all outcomes was until PICC removal, except for CRBSI, which could be identified at 48 hours post‐PICC removal (Mermel 2009).

Primary outcomes

• Venous thromboembolism, defined as either:

  • development of symptomatic thrombosed vessel (partial or complete) at the PICC site, diagnosed via ultrasound (Frey 2006; Yamamoto 2002); or

  • symptomatic DVT as described by the trial investigator.

• PICC‐associated bloodstream infection (CABSI; Mermel 2009), as defined by one of the following criteria:

  • primary bacteraemia or fungaemia with at least one positive blood culture from a peripheral vein with no other identifiable source for the bloodstream infection (BSI) other than the PICC, plus one of:

    • a positive semiquantitative (> 15 colony‐forming units (cfu)); or

    • a quantitative (> 10³ cfu) device culture, with the same organism (species and antibiogram) isolated from the PICC and blood;

  • two blood cultures (one from the PICC hub and one from a peripheral vein), both of which meet the PICC‐related BSI criteria for quantitative blood cultures (three‐fold greater colony count of growth for the same organism as from the peripheral blood), or differential time to positivity (DTP; growth of the same microbe from hub drawn blood at least two hours before growth from the peripheral blood);

  • two quantitative blood cultures of samples obtained through two catheter lumens in which the colony count for the blood sample drawn through one lumen is at least three‐fold greater than the colony count for the blood sample from the second lumen; or

  • laboratory‐confirmed BSI (LCBSI) that is not secondary to an infection at another body site (excluding mucosal barrier injury LCBSI), with PICC in place for more than two calendar days on the day of the BSI (i.e. the day of PICC placement is Day 1) and the PICC in place on the date of the event or the day before, when all elements of LCBSI were first present together (Horan 2008).

• Occlusion: complete blockage of the PICC lumen(s) including fibrin sheath and medication precipitate. This includes aspiration and infusion occlusion and occlusions that resolve with a fibrinolytic such as tissue plasminogen activator (Goossens 2016), and intraluminal thrombosis as described by the trial investigator.

• All‐cause mortality.

Secondary outcomes

• Catheter failure: cessation of catheter function prior to the completion of necessary therapy.

• Incidence of PICC‐related BSI: laboratory confirmed with matched organism from blood culture and catheter tip culture.

• Catheter breakage: visible split in PICC material diagnosed by leakage or radiographic evidence of infiltration or extravasation from a portion of the PICC into tissue.

• PICC dwell time: hours from insertion until removal.

• Other safety endpoints: adverse effects including any local or systemic allergic reactions to chemical or drugs used to coat, bond, or impregnate PICCs.

Search methods for identification of studies

Electronic searches

The University of Queensland and Cochrane Vascular Information Specialist conducted systematic searches of the following databases for RCTs without language, publication year, or publication status restrictions. The latest searches were run on 16 May 2023.

• Cochrane Vascular Specialised Register via the Cochrane Register of Studies (CRS‐Web) (16 May 2023)

• Cochrane Central Register of Controlled Trials (CENTRAL; 2023, Issue 5) via the Cochrane Register of Studies Online (CRSO)

• MEDLINE (Ovid MEDLINE Epub Ahead of Print, In‐Process & Other Non‐Indexed Citations, Ovid MEDLINE Daily and Ovid MEDLINE) (1946 to 16 May 2023)

• Embase Ovid (1980 to 2023 week 20)

• CINAHL EBSCO (Cumulative Index to Nursing and Allied Health Literature) (1937 to 16 May 2023)

We developed search strategies for other databases from the search strategy designed for MEDLINE. Where appropriate, these were combined with adaptations of the Cochrane Highly Sensitive Search Strategy for identifying RCTs and controlled clinical trials (Lefebvre 2023). Search strategies for the major databases are provided in Appendix 1.

We also searched the following trials registries.

• World Health Organization International Clinical Trials Registry Platform (trialsearch.who.int/); up to 16 May 2023

• ClinicalTrials.gov (clinicaltrials.gov/); up to 16 May 2023

Searching other resources

We aimed to identify other potentially eligible trials or ancillary publications by searching the reference lists of retrieved included trials, as well as relevant systematic reviews, meta‐analyses, and health technology assessment reports. We contacted experts in the field to ascertain additional information relevant to the review.

Data collection and analysis

We carried out data collection and analysis according to methods stated in the published protocol (Schults 2019), which were based on the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2021).

Selection of studies

Two review authors (JAS, KC) independently screened the titles and abstracts of retrieved studies for relevance. We retrieved the full‐text versions of all potentially eligible studies. The same two review authors independently screened the full‐text papers for eligibility, using the predefined inclusion and exclusion criteria to select eligible studies. Any discrepancies between review authors were resolved through discussion and consensus with a third review author (AJU). To facilitate transparency in reporting, a PRISMA flowchart is provided (Liberati 2009).

Data extraction and management

We extracted data from eligible studies using a data extraction sheet. The following data were extracted: study characteristics, information relating to the risk of bias, primary and secondary outcome measures, and outcome data. One review author (KC) entered data into RevMan (RevMan 2024), and two review authors (JAS, EY) independently cross‐checked the data for accuracy and agreement. Any discrepancies were resolved through discussion and consensus with an additional review author (AJU). For studies with more than one report, we extracted the maximum data, but did not duplicate data in analyses.

Assessment of risk of bias in included studies

Two review authors (JAS, KC) independently assessed risk of bias in the included studies using Cochrane's RoB 1 tool (Higgins 2011). Risk of bias assessment is based on the following domains:

• adequacy of sequence generation;

• adequacy of allocation concealment;

• blinding;

• incomplete outcome data;

• selective reporting; and

• other sources of bias.

We assigned a judgement of low, high, or unclear risk of bias for each domain. Any discrepancies were resolved through discussion and consensus. A risk of bias table is provided for each eligible study.

Measures of treatment effect

The primary analysis involved pair‐wise comparison of treatment effect between PICC material and design types, using the predefined outcomes. For dichotomous outcomes, we calculated effect measures using risk ratios (RR) and 95% confidence intervals (CI). For continuous outcomes, we calculated effect measures using mean difference (MD) and 95% CI. For outcomes presented as rate‐per‐time period, we planned to perform inverse variance analysis using rate ratios (RaR) and standard errors (SE). We planned to perform analysis using the standardised mean difference (SMD) if studies used different measurement scales.

Unit of analysis issues

The unit of analysis was based on the predefined included study design (RCT) and was expected to be per participant. We planned to include studies that defined the unit of analysis as PICC and perform a sensitivity analysis. For multi‐arm studies (with more than two intervention groups), we planned to (i) to omit groups that are not relevant to the comparison being made or ii) combine multiple groups that were eligible as the experimental or comparator intervention to create a single pair‐wise comparison.

Dealing with missing data

We contacted investigators or study sponsors to obtain numerical outcome data where possible.

Assessment of heterogeneity

We considered methodological, statistical, and clinical heterogeneity of the included studies. We investigated statistical heterogeneity using a combination of methods, including visually inspecting the forest plot and the associated Chi² statistic (using an alpha level 0.10 to determine statistical significance). To assess the impact of heterogeneity among trials (variation in effect estimates not due to chance) (Higgins 2021), we interpreted the I² statistic using the Higgins‐Thompson method (where low heterogeneity, moderate heterogeneity, and high heterogeneity can loosely be equated to 25%, 50%, and 75%, respectively) (Higgins 2003). Due to anticipated clinical heterogeneity, we used random‐effects models (Higgins 2003). We explored clinical variation across trials using descriptive statistics to summarise participant characteristics, sample size, intervention, and outcome measures.

Assessment of reporting biases

We assessed reporting bias using funnel plots if 10 or more trials met the review inclusion criteria. We planned to report each outcome separately. We contacted study authors if further clarification of outcomes reported in methods versus outcomes reported in results was required.

Data synthesis

Methods of synthesising study data depended on trial quality, outcomes, design, and heterogeneity. Initially, we used qualitative synthesis to summarise study results. We then entered data into RevMan (RevMan 2024). We undertook a meta‐analysis where more than one study applied the same intervention and measured the same outcome. Due to expected clinical heterogeneity, we used a random‐effects model for all analyses. Where pooled analyses were not possible, we reported the results of the individual studies narratively. We performed analyses using intention‐to‐treat data where possible.

Subgroup analysis and investigation of heterogeneity

We planned to undertake the following subgroup analyses for the primary outcomes in the case of sufficient data:

• paediatric participants (less than 18 years) versus adult participants (18 years or over);

• participants diagnosed with oncology or haematology pathology versus other participants;

• participants in the intensive care unit versus participants in other settings;

• inpatient versus outpatient settings; and

• participants receiving lipid and parenteral nutrition (PN) versus participants not receiving lipid and PN.

In addition to the main pair‐wise analysis, we planned to investigate the following comparisons if data were available:

• PICCs with a proximal valve versus PICCs with a distal valve;

• PICCs with one type of anti‐thrombotic coating versus all other anti‐thrombotic coated PICCs;

• Chlorhexidine gluconate (CHG)‐impregnated PICCs versus all other antimicrobial‐impregnated PICCs;

• minocycline‐impregnated PICCs versus all other antimicrobial‐impregnated PICCs;

• rifampicin‐impregnated PICCs versus all other antimicrobial‐impregnated PICCs;

• PICCs with two or more modifications (e.g. anti‐thrombogenic material with valve technology, antimicrobial‐impregnated with valve technology, impregnation, valve technology) versus PICCs without modification.

Data were insufficient to facilitate preplanned subgroup analyses with the exception of age (paediatric versus adult). Further, data were insufficient to undertake comparisons of 'clustered' interventions due to the variety of PICC technology and the aim of each design or modification innovation.

Sensitivity analysis

We planned to conduct the following prespecified sensitivity analyses in the case of sufficient data:

• excluding studies that defined PICC as the unit of analysis; and

• excluding studies with a high risk of bias; we planned to only include studies assessed as having a low risk of bias in all key domains, namely adequate sequence generation, adequate allocation concealment, and blinding of outcome assessor for estimates of treatment effect. However, high or unclear risk of bias for sequence generation, allocation concealment, and outcome assessor blinding was evident across studies, therefore this preplanned analysis was not undertaken for any comparison.

In addition to these preplanned analyses, we conducted a post hoc sensitivity analysis for the primary outcomes excluding studies that were terminated early. Four studies were terminated early due to safety (blood sample haemolysis (Johnston 2012), Groshong catheter recall (Miyagaki 2012), PICC rupture (Pittiruti 2014), catheter fracture (Hoffer 2001); see Characteristics of included studies); one for failure to recruit (Storey 2016); and one for multiple reasons (Ong 2010).

Summary of findings and assessment of the certainty of the evidence

We have presented the main results of the review in summary of findings tables (GRADEpro GDT), which provide key information concerning the quality, clinical importance, and context of the evidence, the magnitude of the effects of the interventions examined, and the sum of available data for the main outcomes (GRADE Working Group; Schünemann 2023). These tables also include an overall grading of the evidence relating to main review outcomes using the GRADE approach, which defines the certainty of a body of evidence regarding the extent to which one can be confident that an estimate of effect or association is close to the true quantity of specific interest. We assessed the certainty of the body of evidence for each outcome based on the five GRADE considerations (study limitations, consistency of effect, imprecision, indirectness, and publication bias) as high, moderate, low, or very low (GRADE Handbook 2013; Higgins 2023).

We planned to present the following summary of findings tables for the outcomes that are most clinically relevant to healthcare professionals and consumers.

• Anti‐thrombogenic surface‐modified catheters versus catheters without surface modification

• Antimicrobial‐impregnated catheters versus non‐impregnated catheters

• Catheters with integrated‐valve technology versus catheters without valve technology

We also planned to present a summary of findings table for power‐injectable polyurethane catheters versus silicone catheters; however, this was precluded by insufficient studies.

We planned to include the following outcomes in the summary of findings tables.

• Venous thromboembolism

• PICC‐associated bloodstream infection

• Occlusion

• All‐cause mortality

• Catheter failure

• PICC‐related BSI

• Catheter breakage

Two review authors (JAS and KC) independently evaluated the certainty of the evidence according to the GRADE approach (GRADE Handbook 2013), with any differences in evaluation resolved by consulting a third review author (AJU). For the risk of bias domain, we downgraded the certainty of the evidence if a study was evaluated as being at high risk of bias (except for blinding of participants and personnel). For the imprecision domain, we downgraded the certainty of the evidence when there were few events or no events, and if the CI crossed the line of no effect or was wide. For continuous outcomes and as per GRADE guidance, we considered downgrading for imprecision with sample sizes less than 400.

Results

Description of studies

The characteristics of included studies are outlined in Table 4 (see also Characteristics of included studies). See Excluded studies for details of excluded studies. The flow of studies included in the review is outlined in Figure 1 (see also Characteristics of studies awaiting classification; Characteristics of ongoing studies).

1. Study characteristics.

Study with publication year	Patient population	Unit of randomisation	Control	Intervention	Primary outcomes	Secondary outcomes
Gavin 2020	Single centre ‐ adults	Per participant	Arrow International ‐ power‐injectable with external clamp ‐ polyurethane	Navilyst Medical BioFlo PICC ‐ Endexo power‐ injectable with PASV proximal valve technology ‐ polyurethane	Venous thromboembolism PICC‐associated BSI Occlusion All‐cause mortality	Catheter failure Catheter breakage PICC dwell time Systemic allergic reaction
Gilbert 2019a	Multicentre ‐ neonates	Per participant	Vygon Premicath PICC ‐ standard ‐ polyurethane	Vygon Premistar PICC ‐ miconazole‐ and rifampicin‐impregnated PICC ‐ polyurethane	Venous thromboembolism All‐cause mortality	Catheter breakage PICC dwell time
Hoffer 1999b	Single centre ‐ adults	Per PICC	Cook ‐ clamped non‐valved	Catheter Innovations Clampless valved polyurethane PICC (proximal PASV)	Occlusion	Catheter failure PICC‐related BSI Catheter breakage All‐cause mortality PICC dwell time
Hoffer 2001a	Single centre ‐ adults	Per PICC	Bard Groshong PICC ‐ distal‐valved ‐ silicone	Catheter Innovations ‐ proximal valve with PASV technology ‐ silicone	Occlusion	Catheter failure PICC‐related BSI Catheter breakage All‐cause mortality PICC dwell time
Itkin 2014	Single centre ‐ adults	Per participant	Teleflex ‐ non‐tapered	Bard ‐ reverse tapered	Venous thromboembolism Occlusion All‐cause mortality	PICC dwell time
Johnston 2012	Single centre ‐ adults ‐ ICU	Per participant	Cook Turbo‐Flo PICC ‐ non‐valved, open‐ended ‐ polyurethane ‐ non‐tapered	1. Bard Groshong PICC ‐ distal valve, distal slide splits ‐ non‐tapered ‐ silicone 2. Navilyst Medical Vaxcel PICC ‐ proximal valve with PASV technology ‐ polyurethane ‐ open‐ended tip ‐ reverse tapered	Occlusion	Catheter failure PICC dwell time
Kleidon 2018	Single centre ‐ paediatrics ‐ medical/surgical wards	Per participant	Cook Turbo‐Ject PICC ‐ power‐injectable, external clamp ‐ polyurethane	Navilyst Medical BioFlo PICC ‐ Endexo and PASV proximal valve technology ‐ polyurethane	Venous thromboembolism PICC‐associated BSI Occlusion	Catheter failure Catheter breakage PICC dwell time
Miyagaki 2012	Single centre ‐ adults ‐ oncology	Per participant	PI Catheter, Covidien ‐ open‐end tip ‐ polyurethane	Bard Groshong ‐ distal valve side slits ‐ silicone	Venous thromboembolism Occlusion	Catheter failure Catheter breakage PICC dwell time
Ong 2010a	Single centre ‐ adults ‐ all wards/outpatients	Per PICC	Bard Groshong PICC ‐ distal valve ‐ silicone	Navilyst Medical Vaxcel PICC ‐ proximal valve with PASV valve technology ‐ polyurethane	Occlusion	Catheter failure PICC‐related BSI Catheter breakage PICC dwell time
Pittiruti 2014	Single centre ‐ adults ‐ oncology outpatients	Per participant	Medcomp ProPICC ‐ open‐ended, power‐injectable with no valve ‐ polyurethane	1. Bard PowerPICC Solo 2 ‐ open‐ended, reverse‐tapered, power‐injectable, proximal valves ‐ polyurethane 2. Navilyst Medical Xcela PICC open‐ended, power‐injectable, proximal valve with PASV valve technology ‐ polyurethane	Venous thromboembolism Occlusion	PICC‐related BSI PICC dwell time
Sheretz 1997	Single centre ‐ adults ‐ medical/surgical wards	Per participant	Vialon ‐ standard ‐ polyurethane	Vialon ‐ chlorhexidine‐coated ‐ polyurethane	N/A	Catheter failure PICC‐related BSI All‐cause mortality PICC dwell time
Storey 2016	Single centre ‐ adults ‐ ICU, oncology, cardiothoracic wards	Per participant	Power‐injected PICC	Chlorhexidine‐impregnated, power‐injected PICC	Venous thromboembolism PICC‐associated BSI	PICC dwell time

Abbreviations: BSI: bloodstream infection; ICU: intensive care unit; N/A: not applicable; PICC: peripherally inserted central catheter

^aStudy authors contacted, unpublished data provided and included in the review.
^b Study authors contacted, no response received.

1.

Study flow diagram.

Results of the search

We identified 5622 records from the electronic searches and 1 additional record through other sources. Following removal of duplicates (n = 85), 5537 abstracts remained. We excluded 5493 references based on title and abstract review. We retrieved the full‐text reports of 44 articles, of which 25 were excluded with reasons provided (Moher 2009). Reasons for exclusion were study design (not an RCT) (n = 16) and did not examine PICC material or design (n = 9). A further two studies are awaiting classification and five studies are ongoing. We included 12 articles in the final review.

Included studies

For further details, see Characteristics of included studies.

We included 12 studies involving approximately 2913 participants. We attempted to contact study authors to obtain unpublished data, but received no reply (Hoffer 1999; Hoffer 2001; Ong 2010).

Design

The included studies were RCTs. One study was multicentre (Gilbert 2019), and two studies were multi‐arm (Johnston 2012; Pittiruti 2014). Gavin 2020 and Kleidon 2018 were identified as pilot trials. In nine studies participants were the unit of randomisation (Gavin 2020; Gilbert 2019; Itkin 2014; Johnston 2012; Kleidon 2018; Miyagaki 2012; Pittiruti 2014; Sheretz 1997; Storey 2016), while in three trials randomisation took place at the level of the PICC (multiple PICCs per participant were randomised) (Hoffer 1999; Hoffer 2001; Ong 2010).

Country

The included studies were conducted in the United Kingdom (n = 2; Gilbert 2019; Johnston 2012), the United States (n = 5; Hoffer 1999; Hoffer 2001; Itkin 2014; Sheretz 1997; Storey 2016), Australia (n = 2; Gavin 2020; Kleidon 2018), Japan (Miyagaki 2012), Singapore (Ong 2010), and Italy (Pittiruti 2014).

Sample size

Study sample size ranged from 26 participants in Miyagaki 2012 to 861 participants in Gilbert 2019. Per‐participant data were not available for all trials, therefore a mean sample size calculation was not possible.

Participants and setting

Most studies examined adult cohorts, with the exception of one trial that included neonates (Gilbert 2019) and one that included children (Kleidon 2018). Studies were conducted in a variety of settings. Miyagaki 2012 and Pittiruti 2014 investigated patients in the oncology unit; Johnston 2012 examined an intensive care unit (ICU) cohort. Gavin 2020 sampled adult medical/surgical wards, while Kleidon 2018 included paediatric patients from general medical and surgical wards. Storey 2016 included patients from oncology, ICU, and cardiothoracic wards, and Gilbert 2019 examined neonatal patients. Most PICCs were inserted in the department of medical imaging, with patients returning to a variety of home wards (Hoffer 1999; Hoffer 2001; Itkin 2014; Ong 2010; Sheretz 1997).

Characteristics of interventions

Comparisons

The included studies examined the following material and design interventions.

• No valve versus integrated valve technology (Gavin 2020; Hoffer 1999; Johnston 2012; Kleidon 2018; Pittiruti 2014)

• Distal valve technology versus proximal valve technology (Hoffer 2001; Johnston 2012; Ong 2010)

• Distal valve technology versus proximal valve technology with PASV (Johnston 2012; Ong 2010)

• Proximal valve technology versus proximal valve technology with PASV (Pittiruti 2014)

• Non‐tapered catheter versus tapered catheter (Itkin 2014; Pittiruti 2014)

• Silicone catheter versus polyurethane catheter (Johnston 2012; Miyagaki 2012; Ong 2010)

• Closed versus open‐end tip (Hoffer 2001; Johnston 2012; Miyagaki 2012; Ong 2010)

• Catheters without surface modification versus anti‐thrombogenic surface‐modified catheters (Gavin 2020; Kleidon 2018)

• Non‐antimicrobial‐impregnated catheters versus antimicrobial‐impregnated catheters (Gilbert 2019; Sheretz 1997; Storey 2016)

Primary outcome measures

Venous thromboembolism as defined in the methods was investigated by seven studies (Gavin 2020; Gilbert 2019; Itkin 2014; Kleidon 2018; Miyagaki 2012; Pittiruti 2014; Storey 2016). PICC‐associated bloodstream infection was investigated by three investigators according to the definition stated in Centers for Disease Control and Prevention (CDC) guidelines (Gavin 2020; Kleidon 2018; Storey 2016).

Occlusion was evaluated in 11 studies (Gavin 2020; Gilbert 2019; Hoffer 1999; Hoffer 2001; Itkin 2014; Johnston 2012; Kleidon 2018; Miyagaki 2012; Ong 2010; Pittiruti 2014; Sheretz 1997). All‐cause mortality was investigated in three studies as a prespecified outcome (Gavin 2020; Gilbert 2019; Itkin 2014). A further three trials reported all‐cause mortality within their results (Hoffer 1999; Hoffer 2001; Sheretz 1997).

Secondary outcome measures

Catheter failure was investigated in eight studies (Gavin 2020; Hoffer 1999; Hoffer 2001; Johnston 2012; Kleidon 2018; Miyagaki 2012; Ong 2010; Sheretz 1997). Outcome definitions were not provided in six studies (Hoffer 1999; Hoffer 2001; Johnston 2012; Miyagaki 2012; Ong 2010; Sheretz 1997). Incidence of PICC‐related BSI was reported in five studies (Hoffer 1999; Hoffer 2001; Ong 2010; Pittiruti 2014; Sheretz 1997)

Catheter breakage was investigated in six studies as a primary or secondary outcome according to the definition in the protocol (Gavin 2020; Gilbert 2019; Hoffer 1999; Hoffer 2001; Kleidon 2018; Pittiruti 2014). Miyagaki 2012 and Ong 2010 also reported incidence of catheter breakage without record of the definition used.

PICC dwell time was reported in 10 studies (Gavin 2020; Gilbert 2019; Hoffer 1999; Hoffer 2001; Johnston 2012; Kleidon 2018; Miyagaki 2012; Ong 2010; Sheretz 1997; Storey 2016).

Skin reaction (other safety endpoints) was investigated by Gavin 2020.

Excluded studies

For further details, see Characteristics of excluded studies.

We excluded a total of 25 studies on full‐text review. Many of these studies were excluded because they failed to meet one or more of the inclusion criteria in this review. In summary, 15 studies were not an RCT, and 9 studies did not examine PICC material or design.

Studies awaiting classification and ongoing studies

We identified two studies awaiting classification and five ongoing studies.

Risk of bias in included studies

Risk of bias summaries are provided in Figure 2 and Figure 3. All included studies had unclear or high risk of bias for one or more domains. Given the nature of the intervention, participants and caregivers were generally not blinded, therefore almost all studies were rated at high or unclear risk of bias for this domain, except for one study that was a double‐blind trial (Sheretz 1997).

2.

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

3.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Allocation

Generation of random allocation sequence

Six of the trial investigators reported that random allocation was achieved via a computer‐based sequence generator (Gavin 2020; Gilbert 2019; Hoffer 2001; Itkin 2014; Kleidon 2018; Miyagaki 2012). Ong 2010 and Storey 2016 both utilised the assistance of a statistician or third party to achieve the random allocation sequence. In the remaining studies (Hoffer 1999; Johnston 2012; Pittiruti 2014; Sheretz 1997), it was unclear how the sequence was generated.

Allocation concealment

Three studies used sealed envelopes for allocation concealment (Ong 2010; Sheretz 1997; Storey 2016). Five trials used a central telephone or computer‐based service (Gavin 2020; Gilbert 2019; Itkin 2014; Kleidon 2018; Miyagaki 2012). The remaining four studies did not address allocation concealment and were assessed as at unclear risk of bias (Hoffer 1999; Hoffer 2001; Johnston 2012; Pittiruti 2014).

Blinding

Participant and treatment providers

Only one trial was able to achieve participant and caregiver blinding; this was achieved due to the local treatment of the catheter with the study material, making both control and coated catheters identical in appearance (Sheretz 1997). Eight studies reported inadequate blinding of participants and treatment providers mainly due to the nature of the intervention and were therefore judged as at high risk of bias (Gavin 2020; Gilbert 2019; Hoffer 1999; Hoffer 2001; Itkin 2014; Kleidon 2018; Ong 2010; Storey 2016). Three studies provided no statement regarding blinding and were thus judged to be at high risk of bias (Johnston 2012; Miyagaki 2012; Pittiruti 2014).

Outcome assessors

Five studies described adequate blinding outcome assessors and were therefore judged to be at low risk of bias for this domain (Gavin 2020; Gilbert 2019; Itkin 2014; Kleidon 2018; Sheretz 1997). The remaining studies did not comment on the blinding of outcome assessors and were judged as at unclear risk of bias (Hoffer 1999; Hoffer 2001; Johnston 2012; Miyagaki 2012; Ong 2010; Pittiruti 2014; Storey 2016).

Incomplete outcome data

Six studies were terminated early due to slow recruitment (Storey 2016), recall of product (Miyagaki 2012), haemolysis of blood samples (Johnston 2012), and fractures in the lines (Hoffer 2001; Pittiruti 2014); the remaining study did not provide an explanation for not reaching sample size calculation (Ong 2010). We assessed all six studies as at high risk of bias as results were not sufficiently powered.

Three studies provided data for all participants enrolled in the study and were thus judged to be at low risk of bias (Gavin 2020; Gilbert 2019; Kleidon 2018). Three studies provided explanations for participant attrition following randomisation (Hoffer 1999; Itkin 2014; Sheretz 1997); one study reported a 17% loss to follow‐up as PICC sites were not scanned on removal (Itkin 2014). We therefore judged these studies to be at low risk of attrition bias, with similar numbers of participants lost in all arms of the study and valid reasons for attrition provided.

Selective reporting

We judged all included studies to be at low or unclear risk reporting bias. Study protocols were available for three trials, and reporting followed pre‐planned analyses (Gavin 2020; Gilbert 2019; Kleidon 2018).

Other potential sources of bias

Four studies reported receiving funding from industry sponsors (Gavin 2020; Itkin 2014; Kleidon 2018; Miyagaki 2012). In three studies, the extent of industry involvement was clearly stated, and the study was deemed to be at low risk of bias. We assessed one study to be at high risk of bias for this domain due to industry funding involvement in study design and data collection (Itkin 2014). Gilbert 2019 reported funding from research grants with no influence on the outcome of the project and was thus assessed as at low risk of bias.

We judged six studies to be at unclear risk of other bias (Hoffer 1999; Hoffer 2001; Johnston 2012; Ong 2010; Sheretz 1997; Storey 2016). We assessed one study as at high risk of other bias, as no sample size calculation was provided (Pittiruti 2014).

Effects of interventions

See: Table 1; Table 2; Table 3

See: Table 1; Table 2; Table 3.

Comparison 1: No valve versus integrated valve technology

Five studies compared integrated valve technology with no valve technology (Gavin 2020; Hoffer 1999; Johnston 2012; Kleidon 2018; Pittiruti 2014). See Table 1 Integrated valve technology compared to no valve technology for PICC design.

Primary outcomes

Venous thromboembolism (VTE)

Three studies reported VTE (Gavin 2020; Kleidon 2018; Pittiruti 2014). Meta‐analysis demonstrated that integrated valve technology may make little or no difference to VTE risk when compared with PICCs with no valve (risk ratio (RR) 0.71, 95% confidence interval (CI) 0.19 to 2.63; I² = 0%; 3 studies; 437 participants; low certainty evidence; Analysis 1.1). We downgraded the certainty of evidence to low due to risk of bias (uncertain allocation concealment and sequence generation, one study) and imprecision (few events). In a post hoc sensitivity analysis, there was no change in the interpretation of results when excluding one study considered to have potential unit of analysis issues (Pittiruti 2014) (RR 0.67, 95% CI 0.13 to 3.54; Analysis 1.3).

1.1. Analysis.

Comparison 1: No valve versus integrated valve technology, Outcome 1: Venous thromboembolism

1.3. Analysis.

Comparison 1: No valve versus integrated valve technology, Outcome 3: Venous thromboembolism (sensitivity analysis)

PICC‐associated bloodstream infection (BSI)

Two trials reported PICC‐associated BSI (Gavin 2020; Kleidon 2018). Kleidon 2018 reported no events. We are uncertain whether integrated valve technology reduces PICC‐associated BSI risk as the certainty of the evidence is very low (RR 0.20, 95% CI 0.01 to 4.00; 257 participants; Analysis 1.4), downgraded due to imprecision (few events, small sample size, and wide CIs).

1.4. Analysis.

Comparison 1: No valve versus integrated valve technology, Outcome 4: PICC‐associated BSI

Occlusion

Five trials reported occlusion (Gavin 2020; Hoffer 1999; Johnston 2012; Kleidon 2018; Pittiruti 2014). Integrated valve technology may make little or no difference to occlusion risk when compared with PICCs with no valve (RR 0.86, 95% CI 0.53 to 1.38; I² = 0%; 5 studies; 900 participants; low certainty evidence; Analysis 1.5). We downgraded the certainty of evidence to low due to imprecision (small sample size (early study termination)) and risk of bias (uncertain allocation concealment and sequence generation).

1.5. Analysis.

Comparison 1: No valve versus integrated valve technology, Outcome 5: Occlusion

There remained no evidence of a difference after applying sensitivity analysis and removing one study that performed randomisation at the PICC level (Hoffer 1999) (RR 0.87, 95% CI 0.51 to 1.49; Analysis 1.7). In a post hoc sensitivity analysis, there was no change in the interpretation of results when we excluded a single study that was terminated early (RR 0.71, 95% CI 0.18 to 2.91; Analysis 1.7).

1.7. Analysis.

Comparison 1: No valve versus integrated valve technology, Outcome 7: Occlusion (sensitivity analyses)

All‐cause mortality

Two trials reported all‐cause mortality (Gavin 2020; Hoffer 1999). Thirty‐three deaths were reported. We are uncertain whether use of integrated valve technology reduces all‐cause mortality risk as the certainty of evidence is very low (RR 0.85, 95% CI 0.44 to 1.64; I² = 0%; 2 studies; 473 participants; Analysis 1.8). We downgraded the certainty of evidence to very low due to imprecision (few events and small sample size (early study termination)) and risk of bias (uncertain allocation concealment and sequence generation).

1.8. Analysis.

Comparison 1: No valve versus integrated valve technology, Outcome 8: All‐cause mortality

There remained no evidence of a difference after applying sensitivity analysis and removing one study that performed randomisation at the PICC level and was terminated early (Hoffer 1999) (RR 0.49, 95% CI 0.05 to 5.26).

Secondary outcomes

Catheter failure

Four trials examined catheter failure (Gavin 2020; Hoffer 1999; Johnston 2012; Kleidon 2018). Integrated valve technology may make little or no difference to catheter failure risk when compared with PICCs with no valve (RR 0.80, 95% CI 0.62 to 1.03; I² = 0%; 4 studies; 720 participants; low certainty evidence; Analysis 1.9). We downgraded the certainty of evidence to low due to imprecision (small sample size (early study termination)) and risk of bias due to uncertain allocation concealment and sequence generation.

1.9. Analysis.

Comparison 1: No valve versus integrated valve technology, Outcome 9: Catheter failure

There remained no evidence of a difference after applying sensitivity analysis and removing one study that performed randomisation at the PICC level and was terminated early (Hoffer 1999) (RR 0.76, 95% CI 0.46 to 1.27; 3 studies; Analysis 1.10).

1.10. Analysis.

Comparison 1: No valve versus integrated valve technology, Outcome 10: Catheter failure (sensitivity analyses)

PICC‐related BSI

Two trials reported this outcome (Hoffer 1999; Pittiruti 2014). Pittiruti 2014 reported no events, precluding sensitivity analysis. We are uncertain whether integrated valve technology reduces PICC‐related BSI risk as the certainty of the evidence is very low (RR 0.51, 95% CI 0.19 to 1.32; 542 participants; Analysis 1.11), downgraded due to risk of bias (uncertain allocation concealment, sequence generation, and blinding of outcome assessor) and imprecision (few events and small sample size (early study termination)).

1.11. Analysis.

Comparison 1: No valve versus integrated valve technology, Outcome 11: Incidence of PICC‐related BSI

Catheter breakage

Four studies examined catheter breakage (Gavin 2020; Hoffer 1999; Kleidon 2018; Pittiruti 2014). Gavin 2020 reported no cases. We are uncertain whether integrated valve technology reduces the risk of catheter breakage as the certainty of evidence is very low (RR 1.05, 95% CI 0.22 to 5.06; I² = 20%; 4 studies; 799 participants; Analysis 1.12), downgraded for imprecision (few events, small sample sizes, wide CIs) and risk of bias (uncertain allocation concealment and sequence generation).

1.12. Analysis.

Comparison 1: No valve versus integrated valve technology, Outcome 12: Catheter breakage

There remained no evidence of a difference after applying a sensitivity analysis and removing one study that performed randomisation at the PICC level (Hoffer 1999) (RR 0.71, 95% CI 0.03 to 15.65; I² = 55%; 3 studies; 437 participants; Analysis 1.13). In a post hoc sensitivity analysis, there was no change in the interpretation of results when we excluded studies that were terminated early (RR 0.64, 95% CI 0.07 to 5.84; Analysis 1.13).

1.13. Analysis.

Comparison 1: No valve versus integrated valve technology, Outcome 13: Catheter breakage (sensitivity analyses)

PICC dwell time

Four studies reported catheter dwell, two as mean days (Hoffer 1999; Johnston 2012), and two as median days (Gavin 2020; Kleidon 2018), thereby precluding the pooling of data. Kleidon 2018 reported PICC dwell of 12.9 days (interquartile range (IQR) 9.0 to 14.1) for non‐valved PICCs and 13.8 days (10.0 to 17.3) for catheters with integrated valve technology. Gavin 2020 reported PICC dwell of 8 days (IQR 5 to 15) and 12 days (IQR 5 to 21) for non‐valved and valved PICCs, respectively.

We included data from two studies in the meta‐analysis (Hoffer 1999; Johnston 2012). We are uncertain whether integrated valve technology improves PICC dwell time when compared with PICCs with no valve as the certainty of evidence is very low (mean difference (MD) −0.90, 95% −4.39 to 2.59; I² = 0%; 2 studies; 463 participants; Analysis 1.14), downgraded due to risk of bias (uncertain allocation concealment and sequence generation) and imprecision (small sample size and wide CIs). There remained no evidence of a difference after applying a sensitivity analysis and removing one study that performed randomisation at the PICC level (Hoffer 1999) (MD −0.48, 95% CI −4.58 to 3.62; 1 study; 111 participants).

1.14. Analysis.

Comparison 1: No valve versus integrated valve technology, Outcome 14: PICC dwell time

Other safety endpoints ‐ skin reaction

One trial (111 participants) examined local skin reaction and reported one case (Gavin 2020). We are uncertain whether integrated valve technology reduces the risk of skin reaction as the certainty of the evidence is very low (RR 2.95, 95% CI 0.12 to 70.82; Analysis 1.15), downgraded due to imprecision (few events, small sample size, and wide CIs).

1.15. Analysis.

Comparison 1: No valve versus integrated valve technology, Outcome 15: Skin reaction

Subgroup analysis: age (paediatric versus adult)

One study included paediatric patients only (Kleidon 2018). We used these data in meta‐analyses.

VTE

One study included only children (Kleidon 2018). This study showed no clear impact of the intervention on VTE risk in children (RR 0.41, 95% CI 0.08 to 2.05; 146 participants). Two studies included only adult patients (Gavin 2020; Pittiruti 2014). Meta‐analysis showed no clear impact of intervention on VTE risk (RR 2.09, 95% CI 0.22 to 19.81; I²= 0%; 2 studies; 291 participants; Analysis 1.2).

1.2. Analysis.

Comparison 1: No valve versus integrated valve technology, Outcome 2: Venous thromboembolism (subgroup analysis ‐ age)

PICC‐associated BSI

Kleidon 2018 reported no events, thereby precluding subgroup analyses.

Occlusion

One study included only children (Kleidon 2018). This study showed no clear impact of the intervention on occlusion risk in children (RR 0.21, 95% CI 0.02 to 1.72; 146 participants). Four studies included only adult patients (Gavin 2020; Hoffer 1999; Johnston 2012; Pittiruti 2014). Meta‐analysis showed no clear impact of intervention on occlusion risk (RR 0.92, 95% CI 0.57 to 1.50; 754 participants; Analysis 1.6).

1.6. Analysis.

Comparison 1: No valve versus integrated valve technology, Outcome 6: Occlusion (subgroup analyses ‐ age)

All‐cause mortality

No data were reported for this intervention.

Comparison 2: Distal valve technology versus proximal valve technology

Three studies compared distal valve technology versus proximal valve technology (Hoffer 2001; Johnston 2012; Ong 2010). No study reported VTE or other safety endpoints (skin reaction).

Primary outcomes

VTE

No data were reported for this intervention.

PICC‐associated BSI

One study examined this outcome and reported seven PICC‐associated BSIs (Ong 2010). We are uncertain whether proximal valve technology reduces PICC‐associated BSI risk as the certainty of evidence is very low (RR 0.39, 95% CI 0.08 to 1.99; 392 participants; Analysis 2.1), downgraded due to risk of bias (uncertain outcome assessor blinding) and imprecision (few events and small sample size (early study termination)).

2.1. Analysis.

Comparison 2: Distal valve technology versus proximal valve technology, Outcome 1: PICC‐associated BSI

Occlusion

Three trials reported PICC occlusion (Hoffer 2001; Johnston 2012; Ong 2010). A total of 67 PICCs (12%) developed occlusions over the study period. We are uncertain whether proximal valve technology reduces occlusion risk as the certainty of evidence is very low (RR 0.79, 95% CI 0.51 to 1.22; I² = 0%; 3 studies; 560 participants; Analysis 2.2), downgraded due to risk of bias (uncertain allocation concealment and sequence generation) and imprecision (small study sample size (all studies were terminated early)). There remained no evidence of a difference after applying a sensitivity analysis and removing two studies that performed randomisation at the PICC level (Hoffer 2001; Ong 2010) (RR 0.71, 95% CI 0.35 to 1.44; 67 participants).

2.2. Analysis.

Comparison 2: Distal valve technology versus proximal valve technology, Outcome 2: Occlusion

All‐cause mortality

One trial investigated this outcome and reported seven events (Hoffer 2001). We are uncertain whether proximal valve technology reduces the risk of all‐cause mortality when compared with distal valve technology use as the certainty of evidence is very low (RR 1.23, 95% CI 0.29 to 5.22; 100 participants; Analysis 2.3), downgraded due to risk of bias (uncertain allocation concealment) and imprecision (small sample size (early study termination) and few events).

2.3. Analysis.

Comparison 2: Distal valve technology versus proximal valve technology, Outcome 3: All‐cause mortality

Secondary outcomes

Catheter failure

Two trials reported catheter failure (Hoffer 2001; Ong 2010). The pooled findings showed that when compared with distal valve technology, use of proximal valve technology may slightly reduce the risk of catheter failure (RR 0.54, 95% CI 0.42 to 0.70; I²= 0%; 492 participants; Analysis 2.4). We downgraded the certainty of evidence to low due to risk of bias (unclear allocation concealment) and imprecision (early study termination).

2.4. Analysis.

Comparison 2: Distal valve technology versus proximal valve technology, Outcome 4: Catheter failure

PICC‐related BSI

One trial reported this outcome (Hoffer 2001). We are uncertain whether proximal valve technology reduces the risk of PICC‐related BSI as the certainty of evidence is very low (RR 0.31, 95% CI 0.01 to 7.39; 100 participants; Analysis 2.5), downgraded due to risk of bias (uncertain allocation concealment) and imprecision (small sample size and few events).

2.5. Analysis.

Comparison 2: Distal valve technology versus proximal valve technology, Outcome 5: PICC‐related BSI

Catheter breakage

Two trials reported this outcome (Hoffer 2001; Ong 2010). The pooled findings showed that when compared with distal valve technology, use of proximal valve technology may slightly reduce the risk of catheter breakage (RR 0.21, 95% CI 0.05 to 0.84; I²= 0%; 492 participants; Analysis 2.6). We downgraded the certainty of evidence to low due to risk of bias (unclear allocation concealment) and imprecision (early study termination).

2.6. Analysis.

Comparison 2: Distal valve technology versus proximal valve technology, Outcome 6: Catheter breakage

PICC dwell time

Three studies reported catheter dwell, one as mean days (Johnston 2012), and two as median days (Hoffer 2001; Ong 2010), thereby precluding the pooling of data. Hoffer 2001 (100 participants) reported PICC dwell of 37.5 days (IQR 1 to 306) in PICCs with distal valve technology and 35.1 days (IQR 1 to 211) in PICCs with proximal valve technology. Ong 2010 reported a median dwell of 23.3 days (IQR 1 to 168) in PICCs with a distal valve and 27.8 days (IQR 2 to 245) in PICCs with a proximal valve.

We included data from one study in the meta‐analysis. We are uncertain whether proximal valve technology improves PICC dwell when compared with distal valve technology as the certainty of evidence is very low (MD −4.00, 95% CI −8.20 to 0.20; 1 study; 67 participants; Analysis 2.7), downgraded due to risk of bias (uncertain allocation concealment and sequence generation), imprecision (small sample size (early study termination)), and inconsistency (wide CIs).

2.7. Analysis.

Comparison 2: Distal valve technology versus proximal valve technology, Outcome 7: PICC dwell time

Other safety endpoints ‐ skin reaction

No data were reported for this intervention.

Subgroup analysis: age (paediatric versus adult)

No data were reported for this intervention.

Comparison 3: Distal valve technology versus proximal valve technology with PASV

Two studies compared distal valve technology versus proximal valve technology with PASV (Johnston 2012; Ong 2010). No study reported VTE, all‐cause mortality, PICC‐related BSI, or other safety endpoints (skin reaction).

Primary outcomes

VTE

No data were reported for this intervention.

PICC‐associated BSI

One trial examined this outcome (Ong 2010). We are uncertain whether proximal valve technology with PASV reduces PICC‐associated BSI riskas the certainty of the evidence is very low (RR 0.39, 95% CI 0.08 to 2.00; 392 participants; Analysis 3.1), downgraded for imprecision (few events and small sample size (early study termination)) and risk of bias (uncertain outcome assessor blinding).

3.1. Analysis.

Comparison 3: Distal valve technology versus proximal valve technology with PASV, Outcome 1: PICC‐associated BSI

Occlusion

Two trials reported this outcome (Johnston 2012; Ong 2010). We are uncertain whether proximal valve technology with PASV reduces occlusion riskas the certainty of the evidence is very low (RR 0.77, 95% CI 0.49 to 1.20; I² = 0%; 459 participants; Analysis 3.2), downgraded for imprecision (few events and small sample size (early study termination)) and risk of bias (unclear risk of bias for sequence generation and allocation concealment).

3.2. Analysis.

Comparison 3: Distal valve technology versus proximal valve technology with PASV, Outcome 2: Occlusion

All‐cause mortality

No data were reported for this intervention.

Secondary outcomes

Catheter failure

One trial reported this outcome (Ong 2010). We are uncertain whether use of proximal valve technology with PASV reduces the risk of catheter failure as the certainty of the evidence is very low (RR 0.56, 95% CI 0.42 to 0.73; 392 participants; Analysis 3.3), downgraded for imprecision (few events and small sample size (early study termination)).

3.3. Analysis.

Comparison 3: Distal valve technology versus proximal valve technology with PASV, Outcome 3: Catheter failure

PICC‐related BSI

No data were reported for this intervention.

Catheter breakage

One trial reported this outcome (Ong 2010; 392 PICCs). We are uncertain whether proximal valve technology with PASV reduces the risk of catheter breakage as the certainty of the evidence is very low (RR 0.28, 95% CI 0.06 to 1.33; 392 participants; Analysis 3.4), downgraded for imprecision (few events and small sample size (early study termination)).

3.4. Analysis.

Comparison 3: Distal valve technology versus proximal valve technology with PASV, Outcome 4: Catheter breakage

PICC dwell time

Two trials reported PICC dwell time, one as mean days (Johnston 2012), and one as median days (Ong 2010), thereby precluding the pooling of data. Ong 2010 reported a median dwell of 23.3 days (IQR 1 to 168) in PICCs with a distal valve and 27.8 days (IQR 2 to 245) in PICCs with proximal valve technology with PASV.

We included data from one study in the meta‐analysis (Johnston 2012). We are uncertain whether proximal valve technology with PASV improves PICC dwell time as the certainty of evidence is very low (MD −4.00, 95% CI −8.20 to 0.20; 1 study; 67 participants; Analysis 3.5), downgraded for imprecision (few events and small sample size (early study termination)) and risk of bias (unclear risk of bias for sequence generation and allocation concealment).

3.5. Analysis.

Comparison 3: Distal valve technology versus proximal valve technology with PASV, Outcome 5: PICC dwell time

Other safety endpoints ‐ skin reaction

No data were reported for this intervention.

Subgroup analysis: age (paediatric versus adult)

No data were reported for this intervention.

Comparison 4: Proximal valve technology versus proximal valve technology with PASV

One study compared proximal valve technology versus proximal valve technology with PASV (Pittiruti 2014). No study reported PICC‐associated BSI, all‐cause mortality, PICC dwell time, or other safety endpoints (skin reaction) related to proximal valve versus proximal valve technology with PASV.

Primary outcomes

VTE

One trial examined this outcome and reported one event (Pittiruti 2014). We are uncertain whether proximal valve technology with PASV reduces VTE riskas the certainty of evidence is very low (RR 3.05, 95% CI 0.13 to 73.40; 121 participants; Analysis 4.1), downgraded due to risk of bias (unclear allocation concealment), imprecision (small sample size (early study termination) and few events), and inconsistency (wide CIs).

4.1. Analysis.

Comparison 4: Proximal valve technology versus proximal valve technology with PASV, Outcome 1: Venous thromboembolism

PICC‐associated BSI

No data were reported for this intervention.

Occlusion

One trial investigated this outcome and reported five cases (Pittiruti 2014). We are uncertain whether proximal valve technology with PASV reduces occlusion riskas the certainty of the evidence is very low (RR 0.25, 95% CI 0.03 to 2.21; 121 participants; Analysis 4.2), downgraded due to risk of bias (unclear allocation concealment), imprecision (small sample size (early study termination) and few events), and inconsistency (wide CIs).

4.2. Analysis.

Comparison 4: Proximal valve technology versus proximal valve technology with PASV, Outcome 2: Occlusion

All‐cause mortality

No data were reported for this intervention.

Secondary outcomes

Catheter failure

No data were reported for this intervention.

PICC‐related BSI

Pittiruti 2014 (121 participants) investigated this outcome but reported no events. We are uncertain whether proximal valve technology with PASV reduces the risk of PICC‐related BSI as the certainty of evidence is very low, downgraded due to risk of bias (unclear allocation concealment) and imprecision (small sample size (early study termination) and few events).

Catheter breakage

One trial investigated this outcome and reported three events (Pittiruti 2014). We are uncertain whether proximal valve technology with PASV reduces the risk of catheter breakage as the certainty of the evidence is very low (RR 0.15, 95% CI 0.01 to 2.75; 121 participants; Analysis 4.3), downgraded due to risk of bias (unclear allocation concealment), imprecision (small sample size (early study termination) and few events), and inconsistency (wide CIs).

4.3. Analysis.

Comparison 4: Proximal valve technology versus proximal valve technology with PASV, Outcome 3: Catheter breakage

PICC dwell time

No data were reported for this intervention.

Other safety endpoints ‐ skin reaction

No data were reported for this intervention.

Subgroup analysis: age (paediatric versus adult)

No data were reported for this intervention.

Comparison 5: Non‐tapered catheter versus tapered catheter

Three studies compared tapered catheters with non‐tapered catheters (Itkin 2014; Johnston 2012; Pittiruti 2014). No study reported PICC‐associated BSI or other safety endpoints (skin reaction).

Primary outcomes

VTE

Two studies examined this outcome and reported 14 cases of VTE (Itkin 2014; Pittiruti 2014). We are uncertain whether use of tapered catheters reduces VTE risk as the certainty of the evidence is very low (RR 0.82, 95% CI 0.30 to 2.26; I² = 0%; 511 participants; Analysis 5.1), downgraded for imprecision (few events and early study termination), risk of bias (unclear allocation concealment), and inconsistency (wide CIs).

5.1. Analysis.

Comparison 5: Non‐tapered catheter versus tapered catheter, Outcome 1: Venous thromboembolism

In a post hoc sensitivity analysis, there was no change in the interpretation of results when we excluded one study that was terminated early (Pittiruti 2014) (RR 0.84, 95% CI 0.29 to 2.45; 331 participants).

PICC‐associated BSI

No data were reported for this intervention.

Occlusion

Three trials reported this outcome (Itkin 2014; Johnston 2012; Pittiruti 2014). We are uncertain whether use of tapered catheters reduces occlusion risk as the certainty of the evidence is very low (RR 1.08, 95% CI 0.64 to 1.83; I² = 38%; 613 participants; Analysis 5.2), downgraded due to risk of bias (unclear allocation concealment and sequence generation) and imprecision (small sample size (early study termination)).

5.2. Analysis.

Comparison 5: Non‐tapered catheter versus tapered catheter, Outcome 2: Occlusion

In a post hoc sensitivity analysis, there was no change in the interpretation of results when we excluded two studies that were terminated early (Johnston 2012; Pittiruti 2014) (RR 1.21, 95% CI 0.75 to 1.96; 332 participants).

All‐cause mortality

One trial examined this outcome and reported 14 deaths over the study period (Itkin 2014). Use of tapered catheters may make little or no difference to all‐cause mortality risk (RR 1.30, 95% CI 0.46 to 3.67; 332 participants; Analysis 5.3; low certainty evidence). We downgraded the certainty of evidence for imprecision (few events and small sample size) and inconsistency (wide CIs).

5.3. Analysis.

Comparison 5: Non‐tapered catheter versus tapered catheter, Outcome 3: All‐cause mortality

Secondary outcomes

Catheter failure

No data were reported for this intervention.

PICC‐related BSI

Pittiruti 2014 (180 participants) compared tapered with non‐tapered catheters for this outcome but reported no cases. We downgraded the certainty of evidence to very low due to risk of bias (unclear allocation concealment and sequence generation) and imprecision (small sample size and few events (early study termination)).

Catheter breakage

One trial compared tapered with non‐tapered catheters for catheter breakage (Pittiruti 2014). We are uncertain whether use of tapered catheters reduces the risk of catheter breakage as the certainty of the evidence is very low (RR 13.55, 95% CI 0.71 to 258.15; 180 participants; Analysis 5.4), downgraded due to risk of bias (unclear allocation concealment and sequence generation), imprecision (small sample size and few events (early study termination)), and inconsistency (wide CIs).

5.4. Analysis.

Comparison 5: Non‐tapered catheter versus tapered catheter, Outcome 4: Catheter breakage

PICC dwell time

One study reported this outcome (Johnston 2012). We are uncertain whether use of tapered catheters improves PICC dwell as the certainty of the evidence is very low (MD −0.48, 95% CI −4.58 to 3.62; 101 participants; Analysis 5.5), downgraded due to risk of bias (uncertain allocation concealment and sequence generation), imprecision (small sample size (early study termination)), and inconsistency (wide CIs).

5.5. Analysis.

Comparison 5: Non‐tapered catheter versus tapered catheter, Outcome 5: PICC dwell time

Other safety endpoints ‐ skin reaction

No data were reported for this intervention.

Subgroup analysis: age (paediatric versus adult)

No data were reported for this intervention.

Comparison 6: Silicone catheter versus polyurethane catheter

Three studies compared silicone catheter material with polyurethane catheter material (Johnston 2012; Miyagaki 2012; Ong 2010). No study reported all‐cause mortality, PICC‐related BSI, or other safety endpoints (skin reaction).

Primary outcomes

VTE

Miyagaki 2012 examined this outcome but reported no cases in 25 participants. We are uncertain whether use of polyurethane catheters reduces VTE risk as the certainty of evidence is very low, downgraded due to risk of bias (unclear allocation concealment and sequence generation) and imprecision (small sample size ‐ early study termination).

PICC‐associated BSI

One study reported this outcome (Ong 2010). We are uncertain whether use of polyurethane catheters reduces the risk of PICC‐associated BSI as the certainty of the evidence is very low (RR 0.39, 95% CI 0.08 to 2.00; 392 participants; Analysis 6.1), downgraded due to risk of bias (uncertain outcome assessor blinding) and imprecision (small sample size (early study termination) and few events).

6.1. Analysis.

Comparison 6: Silicone versus polyurethane catheter, Outcome 1: PICC‐associated BSI

Occlusion

Three trials reported this outcome (Johnston 2012; Miyagaki 2012; Ong 2010). We are uncertain whether use of polyurethane catheters reduces occlusion risk as the certainty of the evidence is very low (RR 0.86, 95% CI 0.58 to 1.27; I² = 0%; 518 participants; Analysis 6.2), downgraded due to risk of bias (unclear allocation concealment and sequence generation) and imprecision (small sample size (early study termination)).

6.2. Analysis.

Comparison 6: Silicone versus polyurethane catheter, Outcome 2: Occlusion

There remained no evidence of a difference after applying a sensitivity analysis and removing one study that performed randomisation at the PICC level (Ong 2010) (RR 0.92, 95% CI 0.50 to 1.67; I² = 2%; 126 participants; Analysis 6.3). In a post hoc sensitivity analysis, there was no change in effect when we excluded two studies that were terminated early (Johnston 2012; Miyagaki 2012) (RR 0.81, 95% CI 0.46 to 1.44).

6.3. Analysis.

Comparison 6: Silicone versus polyurethane catheter, Outcome 3: Occlusion ‐ sensitivity analysis

All‐cause mortality

No data were reported for this intervention.

Secondary outcomes

Catheter failure

Three trials reported this outcome (Johnston 2012; Miyagaki 2012; Ong 2010). We are uncertain whether use of polyurethane catheters reduces the risk of catheter failure as the certainty of evidence is very low (RR 0.56, 95% CI 0.42 to 0.73; I² = 0%; 518 participants; Analysis 6.4), downgraded due to risk of bias (unclear allocation concealment and sequence generation) and imprecision (all studies were terminated early). The possible slight benefit was lost after applying sensitivity analysis by removing one study that performed randomisation at the PICC level (Ong 2010) (RR 0.44, 95% CI 0.08 to 2.50; I² = 0%; 2 studies; 126 participants; Analysis 6.5).

6.4. Analysis.

Comparison 6: Silicone versus polyurethane catheter, Outcome 4: Catheter failure

6.5. Analysis.

Comparison 6: Silicone versus polyurethane catheter, Outcome 5: Catheter failure ‐ sensitivity analysis

PICC‐related BSI

No data were reported for this intervention.

Catheter breakage

Two trials examined this outcome (Miyagaki 2012; Ong 2010). Miyagaki 2012 reported no events, thereby precluding sensitivity analysis. Use of polyurethane catheters may make little or no difference to catheter breakage risk (RR 0.28, 95% CI 0.06 to 1.33; 417 participants; low certainty evidence; Analysis 6.6). We downgraded the certainty of evidence for imprecision (few events and small sample size).

6.6. Analysis.

Comparison 6: Silicone versus polyurethane catheter, Outcome 6: Catheter breakage

PICC dwell time

Three trials reported PICC dwell time, two as median dwell time, and one as mean dwell, precluding the pooling of data (Johnston 2012; Miyagaki 2012; Ong 2010). Ong 2010 reported a median dwell of 23.3 days (IQR 1 to 168) in silicone PICCs and 27.8 days (IQR 2 to 245) in polyurethane PICCs. Miyagaki 2012 reported a median dwell of 16 days (IQR 9 to 39) in silicone PICCs and 16 days (IQR 5 to 52) in polyurethane PICCs.

We are uncertain whether use of polyurethane catheters improves PICC dwell as the certainty of the evidence is very low (MD −2.74, 95% CI −6.74 to 1.26; 1 study; 101 participants; Analysis 6.7) (Johnston 2012), downgraded due to risk of bias (uncertain allocation concealment and sequence generation), imprecision (small sample size (all studies were terminated early)), and inconsistency (wide CIs).

6.7. Analysis.

Comparison 6: Silicone versus polyurethane catheter, Outcome 7: PICC dwell time

Other safety endpoints ‐ skin reaction

No data were reported for this intervention.

Subgroup analysis: age (paediatric versus adult)

No data were reported for this intervention.

Comparison 7: Closed‐ versus open‐end tip

Four studies compared PICCs with closed‐end PICCs (distal side split) with PICCs with an open‐end tip (Hoffer 2001; Johnston 2012; Miyagaki 2012; Ong 2010). No study reported other safety endpoints (skin reaction).

Primary outcomes

VTE

Miyagaki 2012 examined this outcome but reported no cases in 25 participants. We are uncertain whether use of the open‐ended PICCs reduces VTE risk as the certainty of evidence is very low, downgraded due to risk of bias (unclear allocation concealment and sequence generation) and imprecision (small sample size ‐ early study termination).

PICC‐associated BSI

One trial reported this outcome (Ong 2010; 392 PICCs). We are uncertain whether use of PICCs with open‐ended tips reduces the risk of PICC‐associated BSI as the certainty of evidence is very low (RR 0.28, 95% CI 0.06 to 1.33; 392 participants; Analysis 7.1), downgraded due to risk of bias (uncertain outcome assessor blinding), imprecision (few events and small sample size (early study termination)), and inconsistency (wide CIs).

7.1. Analysis.

Comparison 7: Closed‐ versus open‐end tip, Outcome 1: PICC‐associated BSI

Occlusion

Four trials compared PICCs with closed‐end tip PICCs with an open‐end tip for occlusion (Hoffer 2001; Johnston 2012; Miyagaki 2012; Ong 2010).

We are uncertain whether use of PICCs with open‐ended tips reduces occlusion risk as the certainty of evidence is very low (RR 0.88, 95% CI 0.59 to 1.29; I² = 0%; 618 participants; Analysis 7.2), downgraded due to risk of bias (unclear allocation concealment and sequence generation) and imprecision (all studies were terminated early). There remained no evidence of a difference after applying a sensitivity analysis and removing two studies that performed randomisation at the PICC level (Hoffer 2001; Ong 2010) (RR 0.92, 95% CI 0.50 to 1.67; I² = 2%; 126 participants; Analysis 7.3).

7.2. Analysis.

Comparison 7: Closed‐ versus open‐end tip, Outcome 2: Occlusion

7.3. Analysis.

Comparison 7: Closed‐ versus open‐end tip, Outcome 3: Occlusion ‐ sensitivity analysis

All‐cause mortality

One trial investigated this outcome (Hoffer 2001; 100 PICCs). We are uncertain whether use of PICCs with open‐ended tips reduces all‐cause mortality risk as the certainty of evidence is very low (RR 1.23, 95% CI 0.29 to 5.22; 100 participants; Analysis 7.4), downgraded due to imprecision (small sample size (early study termination) and few events) and inconsistency (wide CIs).

7.4. Analysis.

Comparison 7: Closed‐ versus open‐end tip, Outcome 4: All‐cause mortality

Secondary outcomes

Catheter failure

Three trials examined this outcome (Hoffer 2001; Miyagaki 2012; Ong 2010). Open‐ended PICCs may slightly reduce risk of catheter failure (RR 0.54, 95% CI 0.42 to 0.70; I² = 0%; 517 participants; low certainty evidence; Analysis 7.5). We downgraded the certainty of evidence due to risk of bias (unclear allocation concealment) and imprecision (all studies were terminated early). This possible slight benefit was lost after applying sensitivity analysis by removing two studies that performed randomisation at the PICC level (Hoffer 2001; Ong 2010) (RR 0.39, 95% CI 0.04 to 3.79; 1 study; 25 participants).

7.5. Analysis.

Comparison 7: Closed‐ versus open‐end tip, Outcome 5: Catheter failure

PICC‐related BSI

No data were reported for this intervention.

Catheter breakage

Three trials examined this outcome (Hoffer 2001; Miyagaki 2012; Ong 2010). Miyagaki 2012 (25 participants) reported no events, thereby precluding sensitivity analysis. Pooled findings showed that use of open‐ended PICCs may slightly reduce risk of catheter breakage (RR 0.21, 95% CI 0.05 to 0.84; I² = 0%; 517 participants; low certainty evidence; Analysis 7.6). We downgraded the certainty of evidence for risk of bias (unclear allocation concealment) and imprecision (all studies were terminated early).

7.6. Analysis.

Comparison 7: Closed‐ versus open‐end tip, Outcome 6: Catheter breakage

PICC dwell time

Four trials reported this outcome. Three trials reported dwell time in median days (Hoffer 2001; Miyagaki 2012; Ong 2010), and one as mean days (Johnston 2012), precluding the pooling of data. Ong 2010 reported a median dwell of 23.3 days (IQR 1 to 168) and 27.8 days (IQR 2 to 245) in distal side slit and open‐ended PICCs, respectively. Miyagaki 2012 reported a median dwell of 16 days (IQR 9 to 39) and 16 days (IQR 5 to 52) in distal side slit and open‐ended PICCs, respectively. Hoffer 2001 reported a median dwell of 35.1 days (IQR 1 to 211) and 37.5 days (IQR 1 to 306) in distal side slit and open‐ended PICCs, respectively.

We included data from one study in the meta‐analysis. We are uncertain whether use of PICCs with open‐ended tips improves PICC dwell as the certainty of evidence is very low (MD −2.74, 95% CI −6.74 to 1.26; 1 study; 101 participants; Analysis 7.7) (Johnston 2012), downgraded due to risk of bias (uncertain allocation concealment and sequence generation), imprecision (small sample size (early study termination)), and inconsistency (wide CIs).

7.7. Analysis.

Comparison 7: Closed‐ versus open‐end tip, Outcome 7: PICC dwell time

Other safety endpoints ‐ skin reaction

No data were reported for this intervention.

Subgroup analysis: age (paediatric versus adult)

No data were reported for this intervention.

Comparison 8: Catheters without anti‐thrombogenic surface modification versus anti‐thrombogenic surface‐modified catheters

Two studies compared anti‐thrombogenic surface‐modified catheters with catheters without surface modification (Gavin 2020; Kleidon 2018). See Table 2 Anti‐thrombogenic surface modification compared to no anti‐thrombogenic surface modification for PICC design.

Primary outcomes

VTE

Two trials examined this outcome (Gavin 2020; Kleidon 2018). We are uncertain whether use of anti‐thrombogenic surface‐modified catheters reduces VTE risk as the certainty of evidence is very low (RR 0.67, 95% CI 0.13 to 3.54; I² = 15%; 257 participants; Analysis 8.1), downgraded due to imprecision (few events, small sample size, and wide CIs).

8.1. Analysis.

Comparison 8: Anti‐thrombogenic surface modification versus non‐anti‐thrombogenic surface modification, Outcome 1: Venous thromboembolism

PICC‐associated BSI

Two trials reported this outcome (Gavin 2020; Kleidon 2018). We are uncertain whether use of anti‐thrombogenic surface‐modified catheters reduces PICC‐associated BSI risk as the certainty of evidence is very low (RR 0.20, 95% CI 0.01 to 4.00; 257 participants; Analysis 8.3), downgraded due to imprecision (few events, small sample size, and wide CIs).

8.3. Analysis.

Comparison 8: Anti‐thrombogenic surface modification versus non‐anti‐thrombogenic surface modification, Outcome 3: PICC‐associated BSI

Occlusion

Two trials reported this outcome (Gavin 2020; Kleidon 2018). We are uncertain whether use of anti‐thrombogenic surface‐modified catheters reduces occlusion risk as the certainty of evidence is very low (RR 0.69, 95% CI 0.04 to 11.22; I² = 70%; 257 participants; Analysis 8.4), downgraded due to imprecision (few events, small sample size, and wide CIs).

8.4. Analysis.

Comparison 8: Anti‐thrombogenic surface modification versus non‐anti‐thrombogenic surface modification, Outcome 4: Occlusion

All‐cause mortality

One trial reported this outcome (Gavin 2020). We are uncertain whether use of anti‐thrombogenic surface‐modified catheters reduces all‐cause mortality risk as the certainty of evidence is very low (RR 0.49, 95% CI 0.05 to 5.26; 111 participants), downgraded due to imprecision (few events, small sample size, and wide CIs).

Secondary outcomes

Catheter failure

Two trials reported this outcome (Gavin 2020; Kleidon 2018). Use of anti‐thrombogenic surface‐modified catheters may make little or no difference to risk of catheter failure (RR 0.76, 95% CI 0.37 to 1.54; I² = 46%; 257 participants; low certainty evidence; Analysis 8.6). We downgraded the certainty of evidence for imprecision (few events and small sample size). Subgroup analysis showed no clear difference in paediatric patients (RR 0.51, 95% CI 0.23 to 1.13; 1 study; 146 participants) and adult patients (RR 1.06, 95% CI 0.53 to 2.12; 1 study; 111 participants; Analysis 8.7).

8.6. Analysis.

Comparison 8: Anti‐thrombogenic surface modification versus non‐anti‐thrombogenic surface modification, Outcome 6: Catheter failure

8.7. Analysis.

Comparison 8: Anti‐thrombogenic surface modification versus non‐anti‐thrombogenic surface modification, Outcome 7: Catheter failure (subgroup analysis ‐ age)

PICC‐related BSI

One trial examined this outcome (Gavin 2020; 111 participants) and reported no events. We are uncertain whether use of anti‐thrombogenic surface‐modified catheters reduces PICC‐related BSI risk as the certainty of evidence is very low, downgraded for imprecision (no events, small sample size, and wide CIs).

Catheter breakage

Two trials examined this outcome (Gavin 2020; Kleidon 2018), Gavin 2020 reported no events, thereby precluding subgroup analysis. We are uncertain whether use of anti‐thrombogenic surface‐modified catheters reduces risk of catheter breakage as the certainty of evidence is very low (RR 0.15, 95% CI 0.01 to 2.79; 257 participants; Analysis 8.8), downgraded for imprecision (few events, small sample size, and wide CIs).

8.8. Analysis.

Comparison 8: Anti‐thrombogenic surface modification versus non‐anti‐thrombogenic surface modification, Outcome 8: Catheter breakage

PICC dwell time

Two trials (Gavin 2020; Kleidon 2018; 257 participants) compared anti‐thrombogenic surface‐modified catheters with catheters without surface modification and reported dwell time in median catheter days. Kleidon 2018 reported PICC dwell of 12.9 days (IQR 9.0 to 14.1) and 13.8 days (IQR 10.0 to 17.3) for non‐modified and surface‐modified catheters, respectively. Gavin 2020 reported PICC dwell of 8 days (IQR 5 to 15) and 12 days (IQR 5 to 21) for non‐modified and surface‐modified catheters, respectively. We are uncertain whether use of anti‐thrombogenic surface‐modified catheters improves PICC dwell time as the certainty of evidence is very low, downgraded for imprecision (few events, small sample size, and wide CIs).

Other safety endpoints ‐ skin reaction

One trial (111 participants) reported skin reaction (local allergic reaction) (Gavin 2020). We are uncertain whether use of anti‐thrombogenic surface‐modified catheters reduces risk of skin reaction as the certainty of evidence is very low (RR 2.95, 95% CI 0.12 to 70.82; 111 participants; Analysis 8.9), downgraded for imprecision (few events, small sample size, and wide CIs).

8.9. Analysis.

Comparison 8: Anti‐thrombogenic surface modification versus non‐anti‐thrombogenic surface modification, Outcome 9: Skin reaction

Subgroup analysis: age (paediatric versus adult)

One study included paediatric patients only (Kleidon 2018). We used these data in meta‐analyses.

VTE

One study included only children (Kleidon 2018). This study showed no clear impact of the intervention on VTE risk in children (RR 0.41, 95% CI 0.08 to 2.05; 146 participants). One study included only adult patients (Gavin 2020). Meta‐analysis of these data showed no clear impact of intervention on VTE risk (RR 2.95, 95% CI 0.12 to 70.82; 1 study; 111 participants; Analysis 8.2).

8.2. Analysis.

Comparison 8: Anti‐thrombogenic surface modification versus non‐anti‐thrombogenic surface modification, Outcome 2: Venous thromboembolism (subgroup analysis ‐ age)

PICC‐associated BSI

Kleidon 2018 reported no events, thereby precluding subgroup analyses.

Occlusion

One study included only children (Kleidon 2018). This study showed no clear impact of the intervention on occlusion risk in children (RR 0.21, 95% CI 0.02 to 1.72; 146 participants). One study included only adult patients (Gavin 2020). Meta‐analysis of these data showed no clear impact of intervention on occlusion risk (RR 2.95, 95% CI 0.32 to 27.47; 1 study; 111 participants; Analysis 8.5).

8.5. Analysis.

Comparison 8: Anti‐thrombogenic surface modification versus non‐anti‐thrombogenic surface modification, Outcome 5: Occlusion (subgroup analysis ‐ age)

All‐cause mortality

No study reported all‐cause mortality in paediatric patients.

Comparison 9: No antimicrobial‐impregnated catheters versus antimicrobial‐impregnated catheters

Three studies compared antimicrobial‐impregnated catheters with no antimicrobial‐impregnated catheters (Gilbert 2019; Sheretz 1997; Storey 2016). No study reported other safety endpoints (skin reaction). See Table 3 Antimicrobial impregnation compared to no antimicrobial impregnation for PICC design.

Primary outcomes

VTE

One trial reported this outcome (Storey 2016). We are uncertain whether use of antimicrobial‐impregnated catheters reduces VTE risk as the certainty of evidence is very low (RR 0.54, 95% CI 0.05 to 5.88; 167 participants; Analysis 9.1), downgraded due to imprecision (few events, small sample size (early study termination), and wide CIs).

9.1. Analysis.

Comparison 9: Non‐antimicrobial‐impregnated catheters versus antimicrobial‐impregnated catheters, Outcome 1: Venous thromboembolism

PICC‐associated BSI

Two trials reported this outcome (Gilbert 2019; Storey 2016); however, incomplete information precluded the pooling of data.

Storey 2016 (167 participants) reported three events. We are uncertain whether use of antimicrobial‐impregnated catheters reduces PICC‐associated BSI risk as the certainty of evidence is very low (RR 2.17, 95% CI 0.20 to 23.53; 167 participants; Analysis 9.2), downgraded for imprecision (few events, small sample size (early study termination), and wide CIs).

9.2. Analysis.

Comparison 9: Non‐antimicrobial‐impregnated catheters versus antimicrobial‐impregnated catheters, Outcome 2: PICC‐associated BSI

Gilbert 2019 (861 participants) examined catheter‐related BSI per 1000 PICC days. The study reported no clear difference in PICC‐associated BSI incidence between groups (RR 0.78, 95% CI 0.27 to 2.25).

Occlusion

Two trials reported this outcome (Gilbert 2019; Sheretz 1997). Antimicrobial‐impregnated catheters probably make little or no difference to occlusion risk (RR 1.00, 95% CI 0.57 to 1.74; I² = 0%; 1025 participants; moderate certainty evidence; Analysis 9.3). We downgraded the certainty of evidence to moderate due to risk of bias (uncertain sequence generation).

9.3. Analysis.

Comparison 9: Non‐antimicrobial‐impregnated catheters versus antimicrobial‐impregnated catheters, Outcome 3: Occlusion

All‐cause mortality

Two trials reported this outcome (Gilbert 2019; Sheretz 1997). Antimicrobial‐impregnated catheters probably make little or no difference to risk of all‐cause mortality (RR 1.12, 95% CI 0.71 to 1.75; I² = 0%; 1082 participants; moderate certainty evidence; Analysis 9.4). We downgraded the certainty of evidence to moderate due to risk of bias (uncertain sequence generation).

9.4. Analysis.

Comparison 9: Non‐antimicrobial‐impregnated catheters versus antimicrobial‐impregnated catheters, Outcome 4: All‐cause mortality

Secondary outcomes

Catheter failure

One trial reported this outcome (Sheretz 1997). Antimicrobial‐impregnated catheters may make little or no difference to risk of catheter failure (RR 1.04, 95% CI 0.82 to 1.30; 221 participants; low certainty evidence; Analysis 9.5). We downgraded the certainty of evidence to low due to risk of bias (uncertain sequence generation) and imprecision (small sample size).

9.5. Analysis.

Comparison 9: Non‐antimicrobial‐impregnated catheters versus antimicrobial‐impregnated catheters, Outcome 5: Catheter failure

PICC‐related BSI

Two trials examined rate of PICC‐related BSI during their study period (Gilbert 2019; Sheretz 1997). Sheretz 1997 reported no events. Antimicrobial‐impregnated catheters probably make little or no difference to PICC‐related BSI risk (RR 1.05, 95% CI 0.71 to 1.55; 1082 participants; moderate certainty evidence; Analysis 9.6). We downgraded the certainty of evidence to moderate due to risk of bias (uncertain sequence generation).

9.6. Analysis.

Comparison 9: Non‐antimicrobial‐impregnated catheters versus antimicrobial‐impregnated catheters, Outcome 6: PICC‐related BSI

Catheter breakage

One study evaluated this outcome and reported seven events (Gilbert 2019). Antimicrobial‐impregnated catheters may make little or no difference to risk of catheter breakage (RR 0.86, 95% CI 0.19 to 3.83; 804 participants; low certainty evidence; Analysis 9.7). We downgraded the certainty of evidence to low for imprecision (few events and wide CIs).

9.7. Analysis.

Comparison 9: Non‐antimicrobial‐impregnated catheters versus antimicrobial‐impregnated catheters, Outcome 7: Catheter breakage

PICC dwell time

Three studies reported (1192 participants) catheter dwell, two as mean days (Sheretz 1997; Storey 2016), and one as median days (Gilbert 2019; 804 participants), precluding the pooling of data. Gilbert 2019 reported a median dwell time of 8.20 days (IQR 4.77 to 12.13) for antimicrobial‐impregnated catheters and 7.86 days (IQR 5 to 12.53) for non‐antimicrobial‐impregnated catheters.

We included data from two studies in the meta‐analysis (Sheretz 1997; Storey 2016). Antimicrobial‐impregnated catheters may make little or no difference to PICC dwell time (MD 0.13, 95% CI −2.17 to 2.44; I² = 0%; 388 participants; low certainty evidence; Analysis 9.8). We downgraded the certainty of evidence for risk of bias (uncertain sequence generation) and imprecision (small sample size).

9.8. Analysis.

Comparison 9: Non‐antimicrobial‐impregnated catheters versus antimicrobial‐impregnated catheters, Outcome 8: PICC dwell time

Other safety endpoints ‐ skin reaction

No data were reported for this intervention.

Subgroup analysis: age (paediatric versus adult)

No data were reported for this intervention.

Discussion

The primary goal of this review was to compare the effects of different PICC material and designs on catheter complications. PICCs are frequently complicated by mechanical and infectious complications. These complications increase medical costs, prolong hospital stay, and contribute to treatment delays (Al‐Asadi 2019; Ullman 2022b). We included 12 RCTs of catheter design or material interventions for hospitalised patients in this review. The evidence suggests that uncertainty remains around the benefits versus harms of PICC material and design options. The findings of this review demonstrate that there is limited evidence to support the use of specific PICC material or design innovations in practice.

Summary of main results

We included 12 studies with approximately 2913 participants in the review: five studies compared no valve versus integrated valve technology; three studies compared distal valve technology versus proximal valve technology; two studies compared distal valve technology versus proximal valve technology with PASV; one study compared proximal valve technology versus proximal valve technology with PASV; three studies compared tapered catheter versus non‐tapered catheter; three studies compared silicone catheter versus polyurethane catheter; four studies compared closed‐ (distal side splits) versus open‐end tip catheters; two studies compared catheters without surface modification versus anti‐thrombogenic surface‐modified catheters; and three studies compared non‐antimicrobial‐impregnated catheters versus antimicrobial‐impregnated catheters.

There were substantial variations in intervention and control groups and outcome measures in the trials, which limited our ability to combine findings. The included trials were mainly undertaken in middle‐ or high‐income countries. Overall, we judged the certainty of the evidence as very low, with the exception of evidence from some comparisons that was assessed as low or moderate certainty.

Primary outcomes

There may be little difference in the incidence of VTE in PICCs with integrated valve technology (data from 437 participants) when compared with no valve. We are uncertain whether use of anti‐thrombogenic surface‐modified catheters (257 participants) or antimicrobial‐impregnated catheters (data from 167 participants) reduces VTE risk in hospitalised patients. We are uncertain whether integrated valve technology (data from 257 participants), anti‐thrombogenic surface modification (data from 257 participants), or antimicrobial impregnation (data from 167 participants) reduces PICC‐associated BSI risk when compared with catheters without these innovations. Integrated valve technology may make little or no difference to occlusion risk (data from 900 participants) when compared to catheters with no valve technology. We are uncertain whether use of anti‐thrombogenic surface‐modified catheters reduces occlusion risk (data from 257 participants), while antimicrobial‐impregnated catheters probably make little or no difference to occlusion risk (data from 1025 participants) when compared to catheters without modification or impregnation. We are uncertain whether use of integrated valve technology (data from 473 participants) or anti‐thrombogenic surface‐modified catheters (data from 111 participants) reduces all‐cause mortality risk. Use of antimicrobial‐impregnated catheters probably makes little or no difference to all‐cause mortality (data from 1082 participants).

Secondary outcomes

Use of integrated valve technology (data from 720 participants), anti‐thrombogenic surface‐modified catheters (data from 257 participants), or antimicrobial‐impregnated catheters (data from 221 participants) may make little or no difference to risk of catheter failure. We are uncertain whether use of integrated valve technology (data from 542 participants) or anti‐thrombogenic surface‐modified catheters (data from 111 participants) reduces PICC‐related BSI risk. Use of antimicrobial‐impregnated catheters probably makes little or no difference to PICC‐related BSI risk (data from 1082 participants). We are uncertain whether use of integrated valve technology (data from 799 participants) or anti‐thrombogenic surface‐modified catheters reduces catheter breakage risk (data from 257 participants). Antimicrobial‐impregnated catheters may make little or no difference to risk of catheter breakage (data from 804 participants).

Contextual factors such as the setting and focus on the intervention may play a role in defining the applicability of the evidence. The methodological quality of the included studies was variable, and we performed sensitivity analysis to assess the effects of study design and quality on our primary outcomes. When studies were excluded due to i) early termination or ii) randomising PICCs, no change in intervention effect was seen across review outcomes. Only one study included children (Kleidon 2018). Not all studies provided sufficient data for inclusion in the meta‐analysis. Subgroup analysis was useful for investigating participant age. However, data were limited for most outcomes across the review, thus limiting conclusions from these subgroup analyses.

In this healthcare era, technology is being rapidly designed. New catheter innovations are emerging into the market often before the health service and researcher have the capacity to evaluate the benefit or harm of the product (Ullman 2022a). As a result, studies are often of poor quality or include small sample sizes, making the drawing of definitive conclusions challenging. In this review we found low certainty evidence that catheter failure may be reduced through the use of open‐ended tips or proximal valve technology. Further, catheter breakage may be reduced through the use of proximal valve technology PICCs; however, the certainty of evidence was assessed as low. Important outcomes that require high event rates are poorly studied and impacted by inconsistent outcome definitions, which could be minimised by using international recommendations for reporting catheter care quality (Schults 2021).

The results of our review are limited due to the small number of RCTs meeting our inclusion criteria and differences in treatments tested and the control arms. The certainty of the evidence presented is in general low, and we cannot draw any strong conclusions on the benefits and harms of PICC material or design innovations for hospitalised patients.

Overall completeness and applicability of evidence

The included trials involved small numbers of participants, with variation in intervention and control groups and outcome measures, which limited our ability to combine trial findings. Given the generally small sample size of the included trials, caution is advised in interpreting the evidence presented in this review. Further studies with larger sample sizes are needed to permit definitive conclusions. Future trials need to consider the role of catheter materials in ensuring the longevity of the catheter while reducing harmful adverse events such as BSI or VTE. Moreover, specific studies comparing antimicrobial and antithrombogenic PICC material in high‐risk patient populations (e.g. patients with chronic disease) are urgently needed to clarify how reduced risk of catheter failure can be achieved to optimise the course of intravenous therapy and patient health outcomes. These data are needed to inform policy and consumer (and caregiver) decision‐making in the context of device selection and appropriate catheter use for individual patient circumstance. Finally, the number of catheter lumens and the impact on complication occurrence may be explored in future review updates.

Our review has limitations. Firstly, we did not identify reports of rare adverse events in the included studies. This may inform catheter selection in clinical practice, and therefore such decisions should be guided by the treating medical team. Secondly, attempts to obtain unpublished data via contact with study authors were largely unsuccessful and may introduce publication bias.

Quality of the evidence

Overall, we judged the certainty of the evidence as very low, with the exception of evidence from some comparisons that was assessed as low or moderate certainty (see Table 1; Table 2; Table 3). The major reasons for downgrading the certainty of the evidence were related to risk of bias and imprecision. Methodological limitations included unclear randomisation methods and allocation concealment, as well as lack of outcome assessor blinding and, as anticipated, a lack of participant blinding. Based on the available data, four studies did not adequately report random sequence generation (Hoffer 1999; Johnston 2012; Pittiruti 2014; Sheretz 1997), or allocation concealment (Hoffer 1999; Hoffer 2001; Johnston 2012; Pittiruti 2014), putting them at risk of selection bias. We assessed no studies as at overall low risk of bias. We also downgraded the certainty of the evidence for imprecision of effect estimates due to the small numbers of studies with limited participants contributing to some outcomes (these studies had a total number of participants less than 400), and for inconsistency of the results (due to wide CIs across analyses). Statistical heterogeneity impacted subgroup analyses and our ability to answer subgroup review questions. There is a need for high‐quality studies with larger sample sizes to assess the effects of PICC material and design for hospitalised patients in clinical settings including outpatient clinics. The available evidence is not of sufficient certainty to draw robust conclusions for the outcomes evaluated in this review regarding the use of PICC material and design to prevent PICC complications and failure.

Potential biases in the review process

An extensive literature search was performed by Cochrane Vascular, and later repeated by the University of Queensland librarians. Two review authors independently determined study eligibility, and two review authors independently extracted data and performed quality assessment in order to reduce bias and subjectivity. There were no significant disagreements during the review process. We included only randomised clinical trials in our review. We are confident that all potential sources of data included in this review were carefully cross‐checked and vetted.

Agreements and disagreements with other studies or reviews

To date, limited published RCTs have directly evaluated the effectiveness of different PICC materials and designs and compared these catheter innovations with conventional PICC design and materials to prevent complication and PICC failure. Despite this, consumers report this as an important quality and safety issue (Schults 2022; Ullman 2023).

Previous meta‐analyses have shown similar ineffectiveness, or no evidence of effectiveness, when including all central venous catheter types, not just PICCs (Slaughter 2020), with meta‐analysis demonstrating that neither catheter material nor design alone or in combination had a significant impact on thrombosis risk. Results from recent observational studies in the USA also suggest that antimicrobial and antithrombogenic PICCs use is not associated with a reduction in major catheter complications (Ullman 2022a). However, when considering BSI, a 2015 systematic review found that antimicrobial PICCs were in fact associated with a significant reduction in central‐line‐associated BSI (Kramer 2015). This finding was supported by a pre‐/post‐implementation study from Northwest Indiana (DeVries 2021), which found that implementation of an antimicrobial PICC quality improvement project reduced baseline BSI rate of 1.83/1000 PICC days to 0.162/1000 PICC days (91.15% reduction, P < 0.001).

Authors' conclusions

Implications for practice.

Twelve studies involving 2913 participants compared peripherally inserted central catheter (PICC) material and design interventions in hospitalised patients. We found there may be little or no difference between groups for our primary and secondary outcomes. There is some low certainty evidence that use of proximal compared to distal valved PICCs may reduce catheter failure and breakage, and that use of open‐ versus closed‐ended PICCs may reduce catheter failure and breakage. Clinical application of PICC material and designs should be determined in each case after considering the advantages and disadvantages of each product and indication for vascular access device. The certainty of evidence varied from moderate to very low across studies, with almost all evidence subject to some limitations.

Implications for research.

An important finding of this review is in regard to the quality of included trials. It is important that future trials present more detailed descriptions of the randomisation process and provide information about blinded outcome assessors and sample size calculations. Given that small sample size and few events was an important limitation across studies included in the review, larger sample sizes are needed in future clinical trials. In addition, standardised outcome measurement (e.g. adverse events and pain) and time points should be used to enable data pooling alongside complete reporting of outcomes with appropriate measures of variance. There remains a need for studies exploring comparisons of products now being widely consumed in practice, such as antithrombogenic and antimicrobial PICCs. Such studies must have high‐quality methods and large sample sizes.

History

Protocol first published: Issue 7, 2019

Notes

Parts of the methods section of this protocol are based on a standard template established by Cochrane Vascular.

Appendices

Appendix 1. Sources searched and search strategies

Source	Search strategy	Hits retrieved
1. VASCULAR REGISTER IN CRSW (Date of most recent search: 26 October 2021)*	#1 catheter* AND INREGISTER AND 01/01/2019_TO_26/01/2021:CRSCREATED #2 Peripheral* AND INREGISTER AND 01/01/2019_TO_26/01/2021:CRSCREATED #3 #1 AND #2	July 2019: 124 Jan 2021: 169 Oct 2021: 22
2. CENTRAL via CRSO (Date of most recent search: 26 October 2021)	#1 MESH DESCRIPTOR Catheterization, Central Venous EXPLODE ALL TREES 762 #2 MESH DESCRIPTOR Catheterization, Peripheral EXPLODE ALL TREES 867 #3 MESH DESCRIPTOR Central Venous Catheters EXPLODE ALL TREES 103 #4 PICC*:TI,AB,KY 441 #5 (Peripherally inserted central venous catheter*):TI,AB,KY 132 #6 (PIC line):TI,AB,KY 2 #7 (percutaneous indwelling central catheter*):TI,AB,KY 0 #8 (Peripherally inserted central catheter*):TI,AB,KY 214 #9 #1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7 OR #8 2020 #10 MESH DESCRIPTOR Equipment Design EXPLODE ALL TREES 8042 #11 MESH DESCRIPTOR POLYURETHANES EXPLODE ALL TREES 420 #12 MESH DESCRIPTOR SILICONES EXPLODE ALL TREES 974 #13 antimicrobial‐impregnated:TI,AB,KY 27 #14 (Antiseptic coated):TI,AB,KY 17 #15 bonding:TI,AB,KY 2823 #16 (catheter wall):TI,AB,KY 0 #17 (Chemical bonded):TI,AB,KY 0 #18 Clamp:TI,AB,KY 4351 #19 coating:TI,AB,KY 1542 #20 design*:TI,AB,KY 231970 #21 flexibil*:TI,AB,KY 3869 #22 impregnat*:TI,AB,KY 1032 #23 (injectable pressure):TI,AB,KY 0 #24 (injection pressure*):TI,AB,KY 42 #25 material*:TI,AB,KY 69981 #26 mechanical:TI,AB,KY 18544 #27 nonstick:TI,AB,KY 2 #28 non‐stick:TI,AB,KY 6 #29 non‐valved:TI,AB,KY 16 #30 polymer:TI,AB,KY 1901 #31 polyurethane:TI,AB,KY 505 #32 (Power injectable):TI,AB,KY 9 #33 rigid*:TI,AB,KY 4377 #34 silicone:TI,AB,KY 2317 #35 stretch*:TI,AB,KY 5243 #36 (surface modif*):TI,AB,KY 132 #37 valve:TI,AB,KY 7285 #38 valved:TI,AB,KY 155 #39 (compare adj2 complications):TI,AB,KY 269 #40 #10 OR #11 OR #12 OR #13 OR #14 OR #15 OR #16 OR #17 OR #18 OR #19 OR #20 OR #21 OR #22 OR #23 OR #24 OR #25 OR #26 OR #27 OR #28 OR #29 OR #30 OR #31 OR #32 OR #33 OR #34 OR #35 OR #36 OR #37 OR #38 OR #39 322408 #41 #9 AND #40 854	July 2019: 854 Jan 2021: 141 Oct 2021: 35 May 2023: 78
3. Clinicaltrials.gov (Date of most recent search: 16 May 2023)	[INTERVENTION]Peripherally inserted central catheter OR Peripheral Catheterization OR Peripherally inserted central venous catheter OR Central Venous Catheter OR PICC | [OTHER TERMS] Equipment Design OR POLYURETHANE OR SILICONES OR antimicrobial impregnated OR Antiseptic coated OR polymer OR rigid OR stretch OR valve OR design OR bonding OR coating OR Clamp OR flexible OR material OR non stick OR mechanical	July 2019: 139 Jan 2021: 32 Oct 2021: 15 May 2023: 32
4. ICTRP Search Portal (Date of most recent search: 16 May 2023)	[INTERVENTION]Peripherally inserted central catheter OR Peripheral Catheterization OR Peripherally inserted central venous catheter OR Central Venous Catheter OR PICC | [OTHER TERMS] Equipment Design OR POLYURETHANE OR SILICONES OR antimicrobial impregnated OR Antiseptic coated OR polymer OR rigid OR stretch OR valve OR design OR bonding OR coating OR Clamp OR flexible OR material OR non stick OR mechanical	July 2019: 11 Jan 2021: Oct 2021: 3 16 May 2023: 4
5. Medline (Ovid MEDLINE® Epub Ahead of Print, In‐Process & Other Non‐Indexed Citations, Ovid MEDLINE® Daily and Ovid MEDLINE®) 1946 to present (Date of most recent search: 16 May 2023)	1 exp Catheterization, Central Venous/ 2 exp Catheterization, Peripheral/ 3 exp Central Venous Catheters/ 4 PICC*.ti,ab. 5 "Peripherally inserted central venous catheter*".ti,ab. 6 "PIC line".ti,ab. 7 "percutaneous indwelling central catheter*".ti,ab. 8 "Peripherally inserted central catheter*".ti,ab. 9 or/1‐8 10 exp Equipment Design/ 11 exp POLYURETHANES/ 12 exp SILICONES/ 13 "antimicrobial‐impregnated ".ti,ab. 14 "Antiseptic coated".ti,ab. 15 bonding.ti,ab. 16 "catheter wall".ti,ab. 17 "Chemical bonded".ti,ab. 18 Clamp.ti,ab. 19 coating.ti,ab. 20 design*.ti,ab. 21 flexibil*.ti,ab. 22 impregnat*.ti,ab. 23 "injectable pressure*".ti,ab. 24 "injection pressure*".ti,ab. 25 material*.ti,ab. 26 mechanical.ti,ab. 27 micropattern.ti,ab. 28 nonstick.ti,ab. 29 non‐stick.ti,ab. 30 non‐valved.ti,ab. 31 polymer.ti,ab. 32 polyurethane.ti,ab. 33 "Power injectable".ti,ab. 34 rigid*.ti,ab. 35 silicone.ti,ab. 36 stretch*.ti,ab. 37 "surface modif*".ti,ab. 38 valve.ti,ab. 39 valved.ti,ab. 40 (compare adj2 complications).ti,ab. 41 or/10‐40 42 9 and 41 43 randomized controlled trial.pt. 44 controlled clinical trial.pt. 45 randomized.ab. 46 placebo.ab. 47 drug therapy.fs. 48 randomly.ab. 49 trial.ab. 50 groups.ab. 51 or/43‐50 52 exp animals/ not humans.sh. 53 51 not 52 54 42 and 53	July 2019: 1910 Jan 2021: 280 Oct 2021: 105 May 2023: 88
6. EMBASE via OVID (Date of most recent search: 16 May 2023)	1 exp central venous catheterization/ 2 exp central venous catheter/ 3 PICC*.ti,ab. 4 "Peripherally inserted central venous catheter*".ti,ab. 5 "PIC line".ti,ab. 6 "percutaneous indwelling central catheter*".ti,ab. 7 "Peripherally inserted central catheter*".ti,ab. 8 or/1‐7 9 exp equipment design/ 10 exp polyurethan/ 11 exp silicone derivative/ 12 "antimicrobial‐impregnated ".ti,ab. 13 "Antiseptic coated".ti,ab. 14 bonding.ti,ab. 15 "catheter wall".ti,ab. 16 "Chemical bonded".ti,ab. 17 Clamp.ti,ab. 18 coating.ti,ab. 19 design*.ti,ab. 20 flexibil*.ti,ab. 21 impregnat*.ti,ab. 22 "injectable pressure*".ti,ab. 23 "injection pressure*".ti,ab. 24 material*.ti,ab. 25 mechanical.ti,ab. 26 micropattern.ti,ab. 27 nonstick.ti,ab. 28 non‐stick.ti,ab. 29 non‐valved.ti,ab. 30 polymer.ti,ab. 31 polyurethane.ti,ab. 32 "Power injectable".ti,ab. 33 rigid*.ti,ab. 34 silicone.ti,ab. 35 stretch*.ti,ab. 36 "surface modif*".ti,ab. 37 valve.ti,ab. 38 valved.ti,ab. 39 (compare adj2 complications).ti,ab. 40 or/9‐39 41 8 and 40 42 randomized controlled trial/ 43 controlled clinical trial/ 44 random$.ti,ab. 45 randomization/ 46 intermethod comparison/ 47 placebo.ti,ab. 48 (compare or compared or comparison).ti. 49 ((evaluated or evaluate or evaluating or assessed or assess) and (compare or compared or comparing or comparison)).ab. 50 (open adj label).ti,ab. 51 ((double or single or doubly or singly) adj (blind or blinded or blindly)).ti,ab. 52 double blind procedure/ 53 parallel group$1.ti,ab. 54 (crossover or cross over).ti,ab. 55 ((assign$ or match or matched or allocation) adj5 (alternate or group$1 or intervention$1 or patient$1 or subject$1 or participant$1)).ti,ab. 56 (assigned or allocated).ti,ab. 57 (controlled adj7 (study or design or trial)).ti,ab. 58 (volunteer or volunteers).ti,ab. 59 trial.ti. 60 or/42‐59 61 41 and 60	July 2019: 2114 Jan 2021: 474 Oct 2021: 228 May 2023: 465
7. CINAHL via EBSCO (Date of most recent search: 16 May 2023)	S58 S42 AND S57 S57 S43 OR S44 OR S45 OR S46 OR S47 OR S48 OR S49 OR S50 OR S51 OR S52 OR S53 OR S54 OR S55 OR S56 S56 MH "Random Assignment" S55 MH "Triple‐Blind Studies" S54 MH "Double‐Blind Studies" S53 MH "Single‐Blind Studies" S52 MH "Crossover Design" S51 MH "Factorial Design" S50 MH "Placebos" S49 MH "Clinical Trials" S48 TX "multi‐centre study" OR "multi‐center study" OR "multicentre study" OR "multicenter study" OR "multi‐site study" S47 TX crossover OR "cross‐over" S46 AB placebo* S45 TX random* S44 TX trial* S43 TX "latin square" S42 S9 AND S41 S41 S10 OR S11 OR S12 OR S13 OR S14 OR S15 OR S16 OR S17 OR S18 OR S19 OR S20 OR S21 OR S22 OR S23 OR S24 OR S25 OR S26 OR S27 OR S28 OR S29 OR S30 OR S31 OR S32 OR S33 OR S34 OR S35 OR S36 OR S37 OR S38 OR S39 OR S40 S40 TX compare N2 complications S39 TX valved S38 TX valve S37 TX surface modif* S36 TX stretch* S35 TX silicone S34 TX rigid* S33 TX Power injectable S32 TX polyurethane S31 TX polymer S30 TX non‐valved S29 TX non‐stick S28 TX nonstick S27 TX micropattern S26 TX mechanical S25 TX material* S24 TX injection pressure* S23 TX injectable pressure* S22 TX impregnat* S21 TX flexibil* S20 TX design* S19 TX coating S18 TX Clamp S17 TX Chemical bonded S16 TX catheter wall S15 TX bonding S14 TX Antiseptic coated S13 TX antimicrobial‐impregnated S12 (MH "Silicones+") S11 (MH "Polyurethanes") S10 (MH "Equipment Design+") S9 S1 OR S2 OR S3 OR S4 OR S5 OR S6 OR S7 OR S8 S8 TX Peripherally inserted central catheter* S7 TX percutaneous indwelling central catheter* S6 TX PIC line S5 TX Peripherally inserted central venous catheter* S4 TX PICC* S3 (MH "Central Venous Catheters+") S2 (MH "Catheterization, Peripheral+") S1 (MH "Catheterization, Central Venous+")	July 2019: 834 Jan 2021: 149 Oct 2021: 47 May 2023: 788
TOTAL before de‐duplication		July 2019: 5986 Jan 2021: 1245 Oct 2021: 455 May 2023: 1451
TOTAL after de‐duplication		July 2019: 4215 Jan 2021: 1039 Oct 2021: 391 May 2023: 1366
*Due to the Cochrane Vascular Group no longer being active (https://www.cochrane.org/about‐us/our‐global‐community/review‐groups) searches could not be updated/rerun by the Cochrane Vascular Information Specialist and the Specialised Register is no longer maintained

Data and analyses

Comparison 1. No valve versus integrated valve technology.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1.1 Venous thromboembolism	3	437	Risk Ratio (IV, Random, 95% CI)	0.71 [0.19, 2.63]
1.2 Venous thromboembolism (subgroup analysis ‐ age)	3		Risk Ratio (IV, Random, 95% CI)	Subtotals only
1.2.1 Paediatric patients	1	146	Risk Ratio (IV, Random, 95% CI)	0.41 [0.08, 2.05]
1.2.2 Adult patients	2	291	Risk Ratio (IV, Random, 95% CI)	2.09 [0.22, 19.81]
1.3 Venous thromboembolism (sensitivity analysis)	2	257	Risk Ratio (IV, Random, 95% CI)	0.67 [0.13, 3.54]
1.4 PICC‐associated BSI	2	257	Risk Ratio (IV, Random, 95% CI)	0.20 [0.01, 4.00]
1.5 Occlusion	5	900	Risk Ratio (IV, Random, 95% CI)	0.86 [0.53, 1.38]
1.6 Occlusion (subgroup analyses ‐ age)	5		Risk Ratio (IV, Random, 95% CI)	Subtotals only
1.6.1 Paediatric patients	1	146	Risk Ratio (IV, Random, 95% CI)	0.21 [0.02, 1.72]
1.6.2 Adult patients	4	754	Risk Ratio (IV, Random, 95% CI)	0.92 [0.57, 1.50]
1.7 Occlusion (sensitivity analyses)	5		Risk Ratio (IV, Random, 95% CI)	Subtotals only
1.7.1 Patient unit of randomisation	4	538	Risk Ratio (IV, Random, 95% CI)	0.87 [0.51, 1.49]
1.7.2 Not terminated early	3	619	Risk Ratio (IV, Random, 95% CI)	0.71 [0.18, 2.91]
1.8 All‐cause mortality	2	473	Risk Ratio (IV, Random, 95% CI)	0.85 [0.44, 1.64]
1.9 Catheter failure	4	720	Risk Ratio (IV, Random, 95% CI)	0.80 [0.62, 1.03]
1.10 Catheter failure (sensitivity analyses)	3	716	Risk Ratio (IV, Random, 95% CI)	0.76 [0.53, 1.09]
1.10.1 Patient unit of randomisation	3	358	Risk Ratio (IV, Random, 95% CI)	0.76 [0.46, 1.27]
1.10.2 Not terminated early	3	358	Risk Ratio (IV, Random, 95% CI)	0.76 [0.46, 1.27]
1.11 Incidence of PICC‐related BSI	2	542	Risk Ratio (IV, Random, 95% CI)	0.51 [0.19, 1.32]
1.12 Catheter breakage	4	799	Risk Ratio (IV, Random, 95% CI)	1.05 [0.22, 5.06]
1.13 Catheter breakage (sensitivity analyses)	4		Risk Ratio (IV, Random, 95% CI)	Subtotals only
1.13.1 Patient unit of randomisation	3	437	Risk Ratio (IV, Random, 95% CI)	0.71 [0.03, 15.65]
1.13.2 Not terminated early	3	619	Risk Ratio (IV, Random, 95% CI)	0.64 [0.07, 5.84]
1.14 PICC dwell time	2	463	Mean Difference (IV, Fixed, 95% CI)	‐0.90 [‐4.39, 2.59]
1.15 Skin reaction	1	111	Risk Ratio (IV, Random, 95% CI)	2.95 [0.12, 70.82]

Comparison 2. Distal valve technology versus proximal valve technology.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
2.1 PICC‐associated BSI	1	393	Risk Ratio (IV, Random, 95% CI)	0.39 [0.08, 1.99]
2.2 Occlusion	3	560	Risk Ratio (IV, Random, 95% CI)	0.79 [0.51, 1.22]
2.3 All‐cause mortality	1	100	Risk Ratio (IV, Random, 95% CI)	1.23 [0.29, 5.22]
2.4 Catheter failure	2	492	Risk Ratio (IV, Random, 95% CI)	0.54 [0.42, 0.70]
2.5 PICC‐related BSI	1	100	Risk Ratio (IV, Random, 95% CI)	0.31 [0.01, 7.39]
2.6 Catheter breakage	2	492	Risk Ratio (IV, Random, 95% CI)	0.21 [0.05, 0.84]
2.7 PICC dwell time	1	67	Mean Difference (IV, Fixed, 95% CI)	‐4.00 [‐8.20, 0.20]

Comparison 3. Distal valve technology versus proximal valve technology with PASV.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
3.1 PICC‐associated BSI	1	392	Risk Ratio (IV, Random, 95% CI)	0.39 [0.08, 2.00]
3.2 Occlusion	2	459	Risk Ratio (IV, Random, 95% CI)	0.77 [0.49, 1.20]
3.3 Catheter failure	1	392	Risk Ratio (IV, Random, 95% CI)	0.56 [0.42, 0.73]
3.4 Catheter breakage	1	392	Risk Ratio (IV, Random, 95% CI)	0.28 [0.06, 1.33]
3.5 PICC dwell time	1	67	Mean Difference (IV, Fixed, 95% CI)	‐4.00 [‐8.20, 0.20]

Comparison 4. Proximal valve technology versus proximal valve technology with PASV.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
4.1 Venous thromboembolism	1	121	Risk Ratio (IV, Random, 95% CI)	3.05 [0.13, 73.40]
4.2 Occlusion	1	121	Risk Ratio (IV, Random, 95% CI)	0.25 [0.03, 2.21]
4.3 Catheter breakage	1	121	Risk Ratio (IV, Random, 95% CI)	0.15 [0.01, 2.75]

Comparison 5. Non‐tapered catheter versus tapered catheter.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
5.1 Venous thromboembolism	2	511	Risk Ratio (IV, Random, 95% CI)	0.82 [0.30, 2.26]
5.2 Occlusion	3	613	Risk Ratio (IV, Random, 95% CI)	1.08 [0.64, 1.83]
5.3 All‐cause mortality	1	332	Risk Ratio (IV, Random, 95% CI)	1.30 [0.46, 3.67]
5.4 Catheter breakage	1	180	Risk Ratio (IV, Random, 95% CI)	13.55 [0.71, 258.15]
5.5 PICC dwell time	1	101	Mean Difference (IV, Fixed, 95% CI)	‐0.48 [‐4.58, 3.62]

Comparison 6. Silicone versus polyurethane catheter.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
6.1 PICC‐associated BSI	1	392	Risk Ratio (IV, Random, 95% CI)	0.39 [0.08, 2.00]
6.2 Occlusion	3	518	Risk Ratio (IV, Random, 95% CI)	0.86 [0.58, 1.27]
6.3 Occlusion ‐ sensitivity analysis	2	126	Risk Ratio (IV, Random, 95% CI)	0.92 [0.50, 1.67]
6.4 Catheter failure	3	518	Risk Ratio (IV, Random, 95% CI)	0.56 [0.42, 0.73]
6.5 Catheter failure ‐ sensitivity analysis	2	126	Risk Ratio (IV, Random, 95% CI)	0.44 [0.08, 2.50]
6.6 Catheter breakage	2	417	Risk Ratio (IV, Random, 95% CI)	0.28 [0.06, 1.33]
6.7 PICC dwell time	1	101	Mean Difference (IV, Fixed, 95% CI)	‐2.74 [‐6.74, 1.26]

Comparison 7. Closed‐ versus open‐end tip.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
7.1 PICC‐associated BSI	1	392	Risk Ratio (IV, Random, 95% CI)	0.28 [0.06, 1.33]
7.2 Occlusion	4	618	Risk Ratio (IV, Random, 95% CI)	0.88 [0.59, 1.29]
7.3 Occlusion ‐ sensitivity analysis	2	126	Risk Ratio (IV, Random, 95% CI)	0.92 [0.50, 1.67]
7.4 All‐cause mortality	1	100	Risk Ratio (IV, Random, 95% CI)	1.23 [0.29, 5.22]
7.5 Catheter failure	3	517	Risk Ratio (IV, Random, 95% CI)	0.54 [0.42, 0.70]
7.6 Catheter breakage	3	517	Risk Ratio (IV, Random, 95% CI)	0.21 [0.05, 0.84]
7.7 PICC dwell time	1	101	Mean Difference (IV, Fixed, 95% CI)	‐2.74 [‐6.74, 1.26]

Comparison 8. Anti‐thrombogenic surface modification versus non‐anti‐thrombogenic surface modification.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
8.1 Venous thromboembolism	2	257	Risk Ratio (IV, Random, 95% CI)	0.67 [0.13, 3.54]
8.2 Venous thromboembolism (subgroup analysis ‐ age)	2		Risk Ratio (IV, Random, 95% CI)	Subtotals only
8.2.1 Paediatric patients	1	146	Risk Ratio (IV, Random, 95% CI)	0.41 [0.08, 2.05]
8.2.2 Adult patients	1	111	Risk Ratio (IV, Random, 95% CI)	2.95 [0.12, 70.82]
8.3 PICC‐associated BSI	2	257	Risk Ratio (IV, Random, 95% CI)	0.20 [0.01, 4.00]
8.4 Occlusion	2	257	Risk Ratio (IV, Random, 95% CI)	0.69 [0.04, 11.22]
8.5 Occlusion (subgroup analysis ‐ age)	2		Risk Ratio (IV, Random, 95% CI)	Subtotals only
8.5.1 Paediatric patients	1	146	Risk Ratio (IV, Random, 95% CI)	0.21 [0.02, 1.72]
8.5.2 Adult patients	1	111	Risk Ratio (IV, Random, 95% CI)	2.95 [0.32, 27.47]
8.6 Catheter failure	2	257	Risk Ratio (IV, Random, 95% CI)	0.76 [0.37, 1.54]
8.7 Catheter failure (subgroup analysis ‐ age)	2		Risk Ratio (IV, Random, 95% CI)	Subtotals only
8.7.1 Paediatric patients	1	146	Risk Ratio (IV, Random, 95% CI)	0.51 [0.23, 1.13]
8.7.2 Adult patients	1	111	Risk Ratio (IV, Random, 95% CI)	1.06 [0.53, 2.12]
8.8 Catheter breakage	2	257	Risk Ratio (IV, Random, 95% CI)	0.15 [0.01, 2.79]
8.9 Skin reaction	1	111	Risk Ratio (IV, Random, 95% CI)	2.95 [0.12, 70.82]

Comparison 9. Non‐antimicrobial‐impregnated catheters versus antimicrobial‐impregnated catheters.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
9.1 Venous thromboembolism	1	167	Risk Ratio (IV, Random, 95% CI)	0.54 [0.05, 5.88]
9.2 PICC‐associated BSI	1	167	Risk Ratio (IV, Random, 95% CI)	2.18 [0.20, 23.53]
9.3 Occlusion	2	1025	Risk Ratio (IV, Random, 95% CI)	1.00 [0.57, 1.74]
9.4 All‐cause mortality	2	1082	Risk Ratio (IV, Random, 95% CI)	1.12 [0.71, 1.75]
9.5 Catheter failure	1	221	Risk Ratio (IV, Random, 95% CI)	1.04 [0.82, 1.30]
9.6 PICC‐related BSI	2	1082	Risk Ratio (IV, Random, 95% CI)	1.05 [0.71, 1.55]
9.7 Catheter breakage	1	804	Risk Ratio (IV, Random, 95% CI)	0.86 [0.19, 3.83]
9.8 PICC dwell time	2	388	Mean Difference (IV, Fixed, 95% CI)	0.13 [‐2.17, 2.44]

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Gavin 2020.

Study characteristics
Methods	Study design: Single‐centre RCT Method of randomisation: Computer‐generated Concealment of allocation: Web‐based service
Participants	Country: Australia Number: 111 participants in multiple wards or HITH (hospital in the home). Group 1: 55 PICCs. Group 2: 56 PICCs. Age: Group 1: 62 years (±16) Group 2: 59 years (±15) Sex (M/F): 54/57 Inclusion criteria: PICC inserted in department of medical imaging, length of stay > 24 h, ≥ 18 years, provided informed consent, transferred to HITH service within 24 hours of PICC insertion Exclusion criteria: Current bloodstream infection, allergy to the study product, PICC was to be inserted through diseased, burned, or scarred skin, PICC was inserted in other department/hospital, they could not provide consent without an interpreter, previous enrolment in study, had an existing central venous device including pulmonary artery catheters
Interventions	Group 1: Arrow International ‐ power‐injectable with external clamp ‐ polyurethane ‐ 4/5 Fr ‐ single/double lumens Group 2: Navilyst Medical BioFlo PICC ‐ Endexo power‐injectable with PASV proximal valve technology ‐ polyurethane ‐ 4/5 Fr ‐ single/double lumens
Outcomes	Primary: Feasibility was assessed as a "composite analysis of elements" including: eligibility, recruitment, retention and attrition, protocol adherence, missing data, patient and healthcare professional satisfaction. PICC failure was defined as "the following complications associated with PICC removal": i. CLABSI defined as "A laboratory­ confirmed BSI that is not secondary to an infection at another body site (excludes Mucosal Barrier Injury LCBSI), with PICC in place for > 2 calendar days on the day of the BSI (day of PICC placement being with PICC in place for > 2 calendar days on the day of the BSI (day of PICC placement being Day 1) and the PICC was in place on the date of the event or the day before, when all elements of LCBSI, were first present together (see CDC NHSN for full criteria) confirmed by a blinded infectious disease specialist using de‐­identified clinical and microbiological data". ii. Local infection defined as "purulent phlebitis confirmed with a positive skin swab (> 15 colony forming units [cfu]) or PICC tip culture, but with negative or no blood culture, confirmed by blinded infectious disease specialist". iii. Device occlusion defined as "≥ 1 lumen cannot be flushed, aspirated, or resolved post ­thrombolytic dwell". iv. Venous thrombosis defined as " Suspected: Removed as too painful for patient to tolerate, or Confirmed: Ultrasound­ or venographically ­confirmed thrombosed deep vessel (basilica, brachial, axillary or subclavian) at the PICC site in a patient with symptoms of thrombosis such as arm pain, swelling, redness, and tenderness at PICC site10,30, or a symptomatic patient with a thrombus or fibrin sheath occluding ≥ 1 lumen at PICC removal". v. PICC fracture or dislodgement defined as "visible split in PICC material with leakage or radiographic evidence of extravasation or infiltration into tissue, causing removal". Secondary: Phlebitis defined as "any sign of chemical, mechanical, or infective phlebitis, determined by patient complaint of pain and nurse examining PICC site". Safety endpoints defined as "local or systemic allergic reactions". PICC dwell defined as "time in days".
Notes	"Published protocol by Kleidon and colleagues defines PICC failure in further detail" Trial registration: The study was prospectively registered on the Australia and New Zealand Clinical Trials Registry: ACTRN12616001578493. The trial protocol was published a priori (Kleidon et al., Vasc Access 3:15–21, 2017). Funding source: "AngioDynamics (the BioFlo® PICC manufacturer) provided partial funds to undertake this research with an unrestricted donation to Griffith University (but not to the study authors). Queensland Health provided in kind support to fund the remainder of the trial. The funders had no role in the study design, collection, analysis, or interpretation of the data, writing of the report, or decision to submit the article for publication."
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Evidence: “Patients who agreed to participate were randomised via computer‐generated randomisation” p. 789
Allocation concealment (selection bias)	Low risk	Evidence: “Patients…were randomised [sic] immediately before PICC insertion via a web‐based service to ensure allocation concealment” p. 789
Blinding (performance bias and detection bias) All outcomes	Unclear risk	Comment: not reported
Blinding of participants and personnel (performance bias) All outcomes	High risk	Evidence: "It was not possible to blind the PICC inserters, patients, or healthcare professionals caring for the patients enrolled in the study" p. 795
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Evidence: “The infection control physician…radiologist…and data analyst were blinded to the study allocation” p. 789
Incomplete outcome data (attrition bias) All outcomes	Low risk	Evidence: “No patients withdrew from the study and there were no missing data” p. 787
Selective reporting (reporting bias)	Low risk	Comment: Stated outcomes were reported according to protocol
Other bias	Low risk	Evidence: “AngioDynamics (the BioFlo® PICC manufacturer) provided partial funds to undertake this research with an unrestricted donation to Griffith University (but not to the study authors). Queensland Health provided in kind support to fund the remainder of the trial. The funders had no role in the study design, collection, analysis, or interpretation of the data, writing of the report, or decision to submit the article for publication” p. 796 "We followed our patients for up to of 4 weeks rather than until PICC removal. This might explain why only four incidences of occlusion resulting in PICC failure occurred" p. 795 "Another limitation was that only patients showing clinical symptoms of thrombus were referred for an ultrasound for confirmation. This may explain the low rate of PICC‐associated thrombus in this cohort. Future studies should consider routine ultrasounds as approximately two‐thirds of PICC‐associated thrombi are asymptomatic" p. 795

Gilbert 2019.

Study characteristics
Methods	Study design: Multicentre RCT Method of randomisation: Computer‐generated Concealment of allocation: Centrally controlled by external body
Participants	Country: England, UK Number: 861 participants in 18 neonatal units. Group 1: 431 PICCs. Group 2: 430 PICCs. Age: Group 1: 3.90 days (IQR 1.90 to 6.12) Group 2: 4.12 days (IQR 2.04 to 5.93) Sex (M/F): 439/422 Inclusion criteria: All babies requiring a narrow‐gauge (French gauge 1) PICC Exclusion criteria: Previous enrolment in study, allergy to the study product
Interventions	Group 1: Vygon Premicath PICC; standard ‐ polyurethane ‐ 1 Fr ‐ single lumen Group 2: Vygon Premistar PICC; miconazole‐ and rifampicin‐impregnated PICC ‐ polyurethane ‐ 1 Fr ‐ single lumen
Outcomes	Primary:Time to first bloodstream infection defined as a "positive blood culture (including fungal BSI) taken between 24 hours after randomisation until 48 hours after removal". Secondary: Secondary outcomes related to infection: "the type of organism isolated from bloodstream infection..., the rate of bloodstream infection... per 1000 days with PICC, occurrence of one or more bloodstream infections, rate of catheter‐related bloodstream infection... per 1000 days with PICC, rifampicin resistance in any isolate from blood or CSF culture, rifampicin resistance in any isolate from PICC tips, and rifampicin resistance in any isolate from blood or CSF culture or from the PICC tip..." Outcomes measured to detect potential biases in sampling or treatment on the basis of knowledge of PICC allocation were as follows: rate of blood or CSF culture sampling per 1000 days with PICC, duration of antimicrobial exposure from randomisation up to 48 h after line removal, and time to PICC removal. Clinical secondary outcomes included: chronic lung disease (defined as requiring respiratory support or supplemental oxygen at 36 weeks’ postmenstrual age), necrotising enterocolitis, treatment for retinopathy of prematurity, abnormalities on cranial ultrasound, time from random allocation to full milk feeds, total duration of parenteral nutrition from random allocation until discharge from neonatal care, and death before discharge to go home from neonatal care. Death within 6 months of randomisation and time to death were recorded from linked death registration data. Adverse events, both expected and unexpected, for all babies who had a PICC successfully inserted until 48 h after PICC removal.
Notes	Trial registration: This trial is registered with ISRCTN, number 81931394. Funding source: UK National Institute for Health Research Health Technology Assessment programme funded the study. "The funder appointed independent members to the Trial Steering Committee and Data Monitoring Committee, approved all protocol amendments, and monitored study progress against agreed milestones. The funder had no involvement in data interpretation or writing of the report."
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Evidence: “Participants were randomly assigned (1:1)… by use of a secure web‐based randomisation programme by the principal investigator or a delegated other individual at the site” p. 383
Allocation concealment (selection bias)	Low risk	Evidence: “The randomisation programme was centrally controlled by the Clinical Trials Research Centre (University of Liverpool, Liverpool, UK) to ensure allocation concealment” p.383
Blinding (performance bias and detection bias) All outcomes	Unclear risk	Comment: not reported
Blinding of participants and personnel (performance bias) All outcomes	High risk	Evidence: “Masking of clinicians to PICC allocation was impractical because rifampicin caused brown staining of the antimicrobial‐impregnated PICC” p. 383 Comment: The trial co‐ordinators found "a slightly increased rate of blood culture sampling in the antimicrobial‐impregnated PICC group" p. 388
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Evidence: “participant inclusion in analyses and occurrence of outcome events were determined by following an analysis plan that was specified before individuals saw unblinded data” p.383
Incomplete outcome data (attrition bias) All outcomes	Low risk	Evidence: “achieved almost complete follow‐up and assessment of the primary outcome” p. 388
Selective reporting (reporting bias)	Low risk	Comment: Clinical trial registration (ISRCTN81931394). All outcomes reported as described in publication.
Other bias	Low risk	Evidence: “The trial was funded by the UK National Institute for Health Research Technology Assessment programme" p. 390. "Intention to treat analysis" p. 387.

Hoffer 2001.

Study characteristics
Methods	Study design: Single‐centre RCT Method of randomisation: Computer‐generated Concealment of allocation: Not stated
Participants	Country: USA Number: 98 participants received 100 PICCs. Group 1: 48 PICCs. Group 2: 52 PICCs Age: 46 years Sex (M/F): 66/34 Inclusion criteria: ≥ 18 years, requires a single‐lumen upper extremity PICC Exclusion criteria: Nil additional reported.
Interventions	Group 1: Bard Groshong PICC ‐ distal‐valved ‐ silicone ‐ 4 Fr ‐ single lumen Group 2: Catheter Innovations ‐ proximal valve with PASV technology ‐ silicone ‐ 4 Fr ‐ single lumen
Outcomes	Primary:Catheter‐related infection was defined as "fever or elevated white blood cell count (or both) and positive blood culture from the PICC, positive PICC tip culture, or positive peripheral blood culture with no other source and clinical improvement after catheter removal". Entry site infection defined by "purulence at the site". Phlebitis was diagnosed by "local pain and a palpable cord along the vein or by positive sonographic evaluation in conjunction with erythema and oedema of the extremity". Occlusion defined as "the inability to use the catheter for the assigned therapy, administration of antibiotics in most cases. Inability to aspirate blood did not warrant thrombolysis unless the purpose of the catheter was access for blood draws". Catheter fractures "included cracks in the external shaft or hub of the catheter that resulted in leakage of infused materials. Fractures were repaired with the manufacturer’s recommended repair kits when available".
Notes	Trial registration: Not stated Funding source: Not stated
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Evidence: “patients were randomized from a computer‐generated list” p. 1174
Allocation concealment (selection bias)	Unclear risk	Comment: Computer generated list
Blinding (performance bias and detection bias) All outcomes	Unclear risk	Comment: not reported
Blinding of participants and personnel (performance bias) All outcomes	High risk	Evidence: “A deficiency of this study was the inability to blind the participants to the type of catheter used. This was not feasible as a result of the different physical appearance of the catheters” p. 1177
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Comment: not reported
Incomplete outcome data (attrition bias) All outcomes	High risk	Comment: Study wasterminated early due to catheter fractures.
Selective reporting (reporting bias)	Unclear risk	Comment: Published protocol not available
Other bias	Unclear risk	Evidence: "The high fracture incidence was an unexpected finding and led to the early termination of the study" p. 1175

Hoffer 1999.

Study characteristics
Methods	Study design: Single‐centre RCT Method of randomisation: Not reported Concealment of allocation: Not reported
Participants	Country: USA Number: 333 participants received 365 PICCs. Group 1: 182 PICCs. Group 2: 180 PICCs Age: 46 (±16) Sex (M/F): 233/129 Inclusion criteria: ≥ 18 years, requiring a single‐lumen PICC placed in an upper extremity (placed by the hospital interventional radiology service) Exclusion criteria: Nil additional reported.
Interventions	Group 1: Cook ‐ clamped non‐valved ‐ 5 Fr ‐ single lumen Group 2: Catheter Innovations Clampless valved PU PICC (proximal PASV) ‐ 5 Fr ‐ single lumen
Outcomes	Primary:Catheter‐related infection was defined as "fever or elevated white blood cell count (or both) and positive blood culture from the PICC, positive PICC tip culture, or positive peripheral blood culture with no other source and clinical improvement after catheter removal". Entry site infection defined by "purulence at the site". Phlebitis was diagnosed by "local pain and a palpable cord along the vein or by positive sonographic evaluation in conjunction with erythema and oedema of the extremity". Occlusion defined as "the inability to use the catheter for the assigned therapy, administration of antibiotics in most cases. Inability to aspirate blood did not warrant thrombolysis unless the purpose of the catheter was access for blood draws". Catheter fractures "included cracks in the external shaft or hub of the catheter that resulted in leakage of infused materials. Fractures were repaired with the manufacturer’s recommended repair kits when available".
Notes	Trial registration: Not stated Funding source: Not disclosed
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Comment: not reported
Allocation concealment (selection bias)	Unclear risk	Comment: not reported
Blinding (performance bias and detection bias) All outcomes	Unclear risk	Comment: not reported
Blinding of participants and personnel (performance bias) All outcomes	High risk	Evidence: “A deficiency of this study was the inability to blind the participants to the type of catheter used. This was not feasible as a result of the different physical appearance of the catheters and different flushes administered” p. 1397
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Comment: not reported
Incomplete outcome data (attrition bias) All outcomes	Low risk	Evidence: “Five patients were lost to follow up: two after 90 days (considered discontinued at the last date of follow up), and one each at 1, 3, and 15 days after placement (these were excluded from the study)” p.1395
Selective reporting (reporting bias)	Unclear risk	Comment: Published protocol not available
Other bias	Unclear risk	Comment: not reported

Itkin 2014.

Study characteristics
Methods	Study design: Single‐centre RCT Method of randomisation: Computer‐generated Concealment of allocation: Randomised code associated with the participant's randomised allocation was placed in participant's chart.
Participants	Country: USA Number: 339 participants in multiple wards. Group 1: 139 PICCs, Group 2: 135 PICCs Age: Group 1: 53.9 years (±15.1) Group 2: 53.9 years (±14.3) Sex (M/F): 153/120 Inclusion criteria: Requiring PICC > 2 weeks, ≥ 18 years Exclusion criteria: Coagulopathy or thrombocytopenia, renal insufficiency, skin‐related problems around insertion site, history of mastectomy or axillary dissection at insertion site, current upper extremity or central venous thrombosis, known hypercoagulable state, prior enrolment in the study, therapy required within < 1 hour
Interventions	Group 1: Teleflex ‐ non‐tapered ‐ 5 Fr ‐ double lumen Group 2: Bard ‐ reverse‐tapered ‐ 5 Fr ‐ double lumen
Outcomes	Primary:Symptomatic thrombosis defined as "the presence of clinical symptoms (pain, arm swelling) and confirmation of venous thrombosis by US". Asymptomatic thrombosis defined as "the presence of clot in the accessed vein (as seen by US) without clinical signs or symptoms". Partial thrombosis defined as "the presence of clot in the vessel lumen with part of the lumen remaining patent and the presence of Doppler detectable flow". Complete thrombosis defined as "complete obliteration of the vessel lumen by clot". Ease of insertion: not reported Bleeding rate: not reported Complications: not reported
Notes	Trial registration: Not stated Funding source: The research was funded by Teleflex Medical. One of the authors was a paid consultant for Teleflex Medical at the time of the study. Teleflex Medical clinical trials team participated in study design and some aspects of data collection. What role the funder had in data interpretation or writing of the report is not disclosed.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Evidence: “Randomization within each of the above‐listed strata was computer generated in balanced blocks of four such that exactly two were randomly assigned to receive a non tapered PICC and two were randomly assigned to receive a tapered PICC in a random order that varied from block to block” p. 91e1
Allocation concealment (selection bias)	Low risk	Evidence: “When anticoagulation therapy status and bacteremia status were determined, the next available randomized allocation was obtained from the randomization list for those strata. A randomization code associated with that randomized allocation was recorded into the patient’s record” p. 91e1
Blinding (performance bias and detection bias) All outcomes	Unclear risk	Comment: not reported
Blinding of participants and personnel (performance bias) All outcomes	High risk	Comment: neither participants nor clinical personnel were blinded to allocation
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Evidence: "Two board‐certified radiologists interpreted the US scans independently. Both readers were blinded to the type of PICC. In case of any discrepancy, the final interpretation was achieved by consensus" p. 88 Comment: Analysis of outcomes: ease of insertion, bleeding rate and complications were unblinded
Incomplete outcome data (attrition bias) All outcomes	Low risk	Evidence: “not all of the patients had US performed; this relates to hospital discharge and the difficulty in getting patients back for US scanning. The large number of patients enrolled mitigates this limitation” p.90 Comment: Each arm of the study had an attrition rate of 29 participants resulting in a 17% loss to follow up.
Selective reporting (reporting bias)	Unclear risk	Comment: Clinical trials registration (NCT00728819), there is no definition for outcomes ease of insertion, complications and bleeding rate. Additional outcomes reported in publication not present in protocol.
Other bias	High risk	Comment: The paper reports the "research was funded by Teleflex Medical. S.O.T. is a paid consultant for Teleflex Medical" p.85 the manufacturer of one of the study products. Evidence: "Teleflex Medical clinical trials team participated in study design and some aspects of data collection. The statistical analysis was performed independently by Biomedical Statistical Consulting." p. 85

Johnston 2012.

Study characteristics
Methods	Study design: Single‐centre RCT Method of randomisation: Not reported Concealment of allocation: Not reported
Participants	Country: UK Number: 102 participants in an ICU ward. Group 1: 34 PICCs, Group 2: 33 PICCs, Group 3: 34 PICCs Age: Group 1: 53.7 years (±18.7), Group 2: 58.5 years (±16.1), Group 3: 60.7 years (±14.9) Sex (M/F): 54/47 Inclusion criteria: ICU patients requiring PICC Exclusion criteria: Failed to consent, < 18 years, contraindication to PICC insertion
Interventions	Group 1: Cook Turbo‐Flo PICC ‐ non‐valved ‐ polyurethane ‐ 5 Fr ‐ double lumen Group 2: Navilyst Medical Vaxcel PICC ‐ proximal valve with PASV technology ‐ polyurethane ‐ 5 Fr ‐ double lumen Group 3: Bard Groshong PICC ‐ distal valve ‐ silicone ‐ 5 Fr ‐ double lumen
Outcomes	Primary:Occlusions were "recorded and dealt with according to the following protocol. Both withdrawal occlusions and total occlusions were dealt with in the same way, and we did not distinguish between the two or document which sort had occurred".
Notes	Trial registration: Not stated Funding source: The authors received no financial support for the study.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Comment: not reported
Allocation concealment (selection bias)	Unclear risk	Comment: not reported
Blinding (performance bias and detection bias) All outcomes	Unclear risk	Comment: not reported
Blinding of participants and personnel (performance bias) All outcomes	High risk	Comment: not reported
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Comment: not reported
Incomplete outcome data (attrition bias) All outcomes	High risk	Evidence: “The study was discontinued early after 102 patients were recruited because of four episodes of haemolysis in blood samples taken from the Navilyst PASV PICC” p.422 Comment: The study did not reach sample size calculation therefore the results are not sufficiently powered and should be interpreted with caution.
Selective reporting (reporting bias)	Unclear risk	Comment: Published protocol not available
Other bias	Unclear risk	Comment: Table 1 indicates that the groups were not balanced at baseline in regard to gender p. 423

Kleidon 2018.

Study characteristics
Methods	Study design: Single‐centre RCT Method of randomisation: Computer‐generated Concealment of allocation: Web‐based via third party to ensure allocation concealment until study entry
Participants	Country: Australia Number: 150 participants in medical/surgical wards. Group 1: 75 PICCs; Group 2: 75 PICCs Age: Group 1: 7.5 years (±4.9); Group 2: 7.1 years (±5.1) Sex (M/F): 86/64 Inclusion criteria: PICC insertion, age < 18 years, predicted hospital stay > 24 hours, single‐lumen PICC, and written informed consent by an English‐speaking, legal parent or guardian. Exclusion criteria: Patients were excluded if they had a current (< 48 hours) BSI, vessel size < 2 mm, could not speak English without an interpreter, required a multilumen PICC, or were previously enrolled in the study.
Interventions	Group 1: Cook Turbo‐Ject PICC ‐ power‐injectable, external clamp ‐ polyurethane ‐ 3/4 Fr ‐ single lumen Group 2: Navilyst MedicalBioFlo PICC ‐ Endexo and PASV proximal valve technology ‐ polyurethane ‐ 3/4 Fr ‐ single lumen
Outcomes	Primary:Feasibility was assessed as a "composite analysis of elements" including: eligibility, recruitment, retention and attrition, protocol adherence, missing data, patient and healthcare professional satisfaction. PICC failure was defined as "the following complications associated with PICC removal": i. CLABSI defined as "A laboratory­ confirmed BSI that is not secondary to an infection at another body site (excludes Mucosal Barrier Injury LCBSI), with PICC in place for > 2 calendar days on the day of the BSI (day of PICC placement being with PICC in place for > 2 calendar days on the day of the BSI (day of PICC placement being Day 1) and the PICC was in place on the date of the event or the day before, when all elements of LCBSI, were first present together (see CDC NHSN for full criteria) confirmed by a blinded infectious disease specialist using de‐­identified clinical and microbiological data". ii. Local infection defined as "purulent phlebitis confirmed with a positive skin swab (> 15 colony forming units [cfu]) or PICC tip culture, but with negative or no blood culture, confirmed by blinded infectious disease specialist". iii. Device occlusion defined as "≥ 1 lumen cannot be flushed, aspirated, or resolved post ­thrombolytic dwell". iv. Venous thrombosis defined as " Suspected: Removed as too painful for patient to tolerate, or Confirmed: Ultrasound­ or venographically ­confirmed thrombosed deep vessel (basilica, brachial, axillary or subclavian) at the PICC site in a patient with symptoms of thrombosis such as arm pain, swelling, redness, and tenderness at PICC site10,30, or a symptomatic patient with a thrombus or fibrin sheath occluding ≥ 1 lumen at PICC removal". v. PICC fracture or dislodgement defined as "visible split in PICC material with leakage or radiographic evidence of extravasation or infiltration into tissue, causing removal". Secondary:Adverse events: not defined Pain: not defined Redness at the insertion site: not defined PICC dwell defined as "time in days".
Notes	Trial registration: The study was prospectively registered with the Australian Clinical Trials Registry (ACTRN12615001290583), and the research protocol was published. Funding source: "AngioDynamics (the BioFlo® PICC manufacturer) provided partial funds to undertake this research via an unrestricted donation to Griffith University (but not the study authors). Queensland Health provided in‐kind support to fund the remainder of the trial. The funders had no role in the study design, collection, analysis, or interpretation of the data, writing of the report, or decision to submit the article for publication."
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Evidence: “Patients were randomly assigned in a 1:1 ratio with computer‐generated and randomly varied block sizes of 2 and 4.” p. 518
Allocation concealment (selection bias)	Low risk	Evidence: “Randomization was computer generated, and web based via Griffith University… to ensure allocation concealment until study entry” p.518
Blinding (performance bias and detection bias) All outcomes	Unclear risk	Comment: not reported
Blinding of participants and personnel (performance bias) All outcomes	High risk	Evidence: “The study could not be blinded because study products had to be visible to the clinical and research staff” p. 524 Comment: classified as high risk because the investigator was involved in all stages of the study
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Evidence: “A blinded radiologist and infectious disease specialist reviewed and diagnosed thrombosis of deep veins and catheter‐associated BSI outcomes, respectively” p. 519
Incomplete outcome data (attrition bias) All outcomes	Low risk	Evidence: “No participants were lost to follow‐up, and no primary outcome data were missing” p.519
Selective reporting (reporting bias)	Low risk	Comment: Clinical trial registration (ACTRN12615001290583) and published protocol. All outcomes reported as described in publication.
Other bias	Low risk	Comment: the paper reports "unrestricted, investigator‐initiated research or educational grants to support the research from product manufacturers 3M, Adhezion Inc, AngioDynamics, Bard Medical, Baxter, B. Braun Medical Inc, Becton Dickinson, CareFusion, Centurion Medical Products, Cook Medical, Entrotech, FloMedical, ICU Medical Inc, Medical Australia Limited, Medtronic, Smiths Medical, and Teleflex. Griffith University has received consultancy payments on behalf of C.R.M., A.J.U., and T.K. from manufacturers 3M, AngioDynamics, Bard Medical, B. Braun Medical Inc, Becton Dickinson, CareFusion, Mayo Healthcare Inc, ResQDevices, and Smiths Medical. AngioDynamics (the BioFlo® PICC manufacturer) provided partial funds to undertake this research via an unrestricted donation to Griffith University (but not the study authors). Queensland Health provided in‐kind support to fund the remainder of the trial. The funders had no role in the study design, collection, analysis, or interpretation of the data, writing of the report, or decision to submit the article for publication." p. 524

Miyagaki 2012.

Study characteristics
Methods	Study design: Single‐centre RCT Method of randomisation: Computer‐generated Concealment of allocation: Computer‐generated and concealed
Participants	Country: Japan Number: 26 participants with oncological diagnosis. Group 1: 11 PICCs; Group 2: 14 PICCs Age: Group 1: 67 years (IQR 50 to 80); Group 2: 64 years (IQR 56 to 80) Sex (M/F): 24/1 Inclusion criteria: PICC is needed clinically for chemotherapy or intravenous nutrition, between 20 and 80 years. Exclusion criteria: Taking warfarin or an antiplatelet medicine within 2 weeks of an insertion day; past history of pulmonary embolism, deep vein thrombosis, and endocarditis; patients judged to be inappropriate as a participant
Interventions	Group 1: Bard Groshong ‐ distal valve side slits ‐ silicone ‐ 4 Fr ‐ single lumen Group 2: PI Catheter, Covidien ‐ open‐end tip ‐ polyurethane ‐ 4 Fr ‐ single lumen
Outcomes	Primary: The primary endpoint was the completion rate of PICC indication. Secondary:Suspected BSI "The catheter was removed prematurely when any catheter related blood stream infection (CR‐BSI) was suspected according to Centers for Disease Control and Prevention guidelines." Haemorrhage at the insertion site defined as any haemorrhage at the insertion site, i.e. bleeding within 24 h after insertion or after removal of PICC catheter, was also registered with a 3‐grade scale: fair, little, and none. Phlebitis‐related complications: not clearly defined; citation provided, but primary source did not define phlebitis‐related complications. Vein thrombosis: not defined Catheter malfunction defined as "partial/complete occlusion with drip disturbance". Fracture: not defined
Notes	Trial registration: The study was registered in the UMIN Clinical Trial Registry (UMIN000005451). Funding source: The study was funded by Nippon Sherwood Medical Industries Ltd. (Tokyo, Japan). The authors did not disclose whether the funders had any role in the study design, collection, analysis, or interpretation of the data, writing of the report, or decision to submit the article for publication.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Evidence: “A coordinating centre in our institution was responsible for creating the treatment allocation code using a computer‐generated randomization table” p. 49
Allocation concealment (selection bias)	Low risk	Evidence: ‘Treatment allocation was arranged prior to catheter placement” p. 49
Blinding (performance bias and detection bias) All outcomes	Unclear risk	Comment: not reported
Blinding of participants and personnel (performance bias) All outcomes	High risk	Comment: not reported
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Comment: not reported
Incomplete outcome data (attrition bias) All outcomes	High risk	Evidence: “As Groshong arm became unable to be continued because of a recall due to insufficient product registration document, we were obliged to terminate this study at the time that 26 patients were enrolled in total” p.49 Comment: The study did not reach sample size calculation therefore the results are not sufficiently powered and should be interpreted with caution.
Selective reporting (reporting bias)	Low risk	Comment: Clinical trial registration (UMIN000005451). All outcomes reported as described in trial registry
Other bias	Low risk	Comment: This study was supported financially by Nippon Sherwood Medical Industries Ltd. (Tokyo, Japan).

Ong 2010.

Study characteristics
Methods	Study design: Single‐centre RCT Method of randomisation: Randomised by statistician Concealment of allocation: Sequentially numbered, sealed envelopes
Participants	Country: Singapore Number: 392 PICCs inserted in 326 participants Age: Mean age 50.4 years Sex (M/F): 247/145 Inclusion criteria: Written informed consent Exclusion criteria: Not reported
Interventions	Group 1: Bard Groshong PICC ‐ distal valve ‐ silicone ‐ 4 Fr Group 2: Navilyst Medical Vaxcel PICC ‐ proximal valve with PASV valve technology ‐ polyurethane ‐ 4 Fr
Outcomes	Primary:Phlebitis "diagnosed if erythema or induration, warmth, pain, or tenderness existed around the catheter exit site". Exit‐site infection "diagnosed when erythema, induration, or tenderness was noted around the catheter exit site, associated with other signs and symptoms of infection, such as fever or pus emerging from the exit site, with or without concomitant bloodstream infection". Definite CRBSI defined as "isolation of the same organism (identical species and antibiogram) from the catheter segment and the peripheral blood culture in a patient with clinical symptoms of bloodstream infection and no other apparent source of infection". Probable CRBSI defined as "positive culture either from catheter segment or peripheral blood in a patient with clinical symptoms of bloodstream infection and no other apparent source of infection, defervescence within 48 hours of catheter removal, and initiation of appropriate antibiotic therapy". Occluded "when resistance, which significantly compromised the use of the catheter for infusion of its assigned therapy, was encountered. Inability to draw blood through the catheter by itself was not considered as occlusion".
Notes	Comment: "393 patients received a total of 459 PICC insertions" page 1192 Trial registration: Not stated Funding source: Nil reported.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Evidence: “The 500 PICC insertions that were to be performed (250 for each of the two catheter groups) were randomized with the help of a statistician” p. 1192
Allocation concealment (selection bias)	Low risk	Evidence: “Five hundred sealed envelopes labeled with consecutive numbers were prepared, each containing a name of the PICC to be used per the randomized list. An envelope was opened by an intervention nurse in consecutive order just before each procedure, and the corresponding PICC was inserted for that particular patient” p. 1192
Blinding (performance bias and detection bias) All outcomes	Unclear risk	Comment: not reported
Blinding of participants and personnel (performance bias) All outcomes	High risk	Evidence: “The radiologists who performed the PICC insertions could not be blinded to the type of PICC used because of the different appearances of the PICCs” p. 1195
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Comment: not reported
Incomplete outcome data (attrition bias) All outcomes	High risk	Evidence: “Sixty‐seven patients, each receiving one PICC, were excluded from the study because of (i) death from underlying morbidity while the catheters were in situ (n = 38), (ii) no follow‐up data due to transfer to other institutions (n = 21), and (iii) self‐request of removal of catheters and discharge against medical advice (n = 8)” p. 1192 Comment: The study did not reach set recruitment numbers set at 500. The final study cohort comprised of 326 participants.
Selective reporting (reporting bias)	Unclear risk	Comment: Pubilshed protocol not available, unclear whether the trial was prosectively registered.
Other bias	Unclear risk	Comment: Intention to treat and research funding support not specified.

Pittiruti 2014.

Study characteristics
Methods	Study design: Single‐centre RCT Method of randomisation: Not reported Concealment of allocation: Not reported
Participants	Country: Italy Number: 180 participants in an oncology unit. Group 1: 61 PICCs. Group 2: 60 PICCs. Group 3: 59 PICCs Age: Group 1: 64 years (±12.1). Group 2: 61 years (±10.1). Group 3: 62 years (±14.5) Sex (M/F): 64/116 Inclusion criteria: Patients requiring single‐lumen 4‐French PICC Exclusion criteria: Patients with local contraindications to PICC insertion or who refused to participate in the study were excluded.
Interventions	Group 1: Bard PowerPICC Solo 2‐ open‐ended, reverse‐tapered, power‐injectable, proximal valves ‐ polyurethane ‐ 4 Fr ‐ single lumen Group 2: Navilyst MedicalXcela PICC ‐ open‐ended, power‐injectable, proximal valve with PASV valve technology ‐ polyurethane ‐ 4 Fr ‐ single lumen Group 3: MedcompProPICC ‐ open‐ended, power‐injectable with no valve ‐ polyurethane ‐ 4 Fr ‐ single lumen
Outcomes	Primary:Occlusion defined as "complete occlusion (no infusion, no withdrawal); partial occlusion (difficult infusion or withdrawal); PWO—partial withdrawal occlusion (infusion ok but no withdrawal); occlusions solved by flushing; occlusions solved by pharmacological action (urokinase); PICCs removed because of irreversible occlusion". Secondary:Catheter‐related bloodstream infection as defined by the current international guidelines* Asymptomatic and symptomatic venous thrombosis detected by ultrasound Mechanical complications defined as "dislocation, ‘minor’, <4 cm, or ‘major’, >4 cm; catheter rupture".
Notes	*O’Grady NP, Alexander M, Burns LA, et al.; Healthcare Infection Control Practices Advisory Committee. Guidelines for the prevention of intravascular catheter‐related infections. Am J Infect Control. 2011;39(4)(Suppl 1):S1‐34. Trial registration: Not stated Funding source: The authors have no financial disclosures.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Comment: not reported
Allocation concealment (selection bias)	Unclear risk	Comment: not reported
Blinding (performance bias and detection bias) All outcomes	Unclear risk	Comment: not reported
Blinding of participants and personnel (performance bias) All outcomes	High risk	Comment: not reported
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Comment: not reported
Incomplete outcome data (attrition bias) All outcomes	High risk	Evidence: "The study was interrupted at 180 patients (approximately 60 for each group) because of three episodes of rupture of Power Solo PICCs." p. 521 Comment: The study did not reach sample size calculation therefore the results are not sufficiently powered and should be interpreted with caution.
Selective reporting (reporting bias)	Low risk	Comment: not reported
Other bias	High risk	Comment: No sample size calculation

Sheretz 1997.

Study characteristics
Methods	Study design: Single‐centre RCT Method of randomisation: Not reported Concealment of allocation: Each package was marked with a randomly assigned study number.
Participants	Country: USA Number: 226 participants. Group 1: 116 PICCs; Group 2: 105 PICCs Age: Not reported Sex (M/F): 91/130 Inclusion criteria: > 18 years, expected duration of catheterisation > 24 hours Exclusion criteria: Forearm dermatitis, inability to keep the forearm dry, were on an anti‐inflammatory drug, had an allergy to a skin disinfectant, pregnant
Interventions	Group 1: Vialon‐ standard ‐ polyurethane ‐ 16/20 gauge Group 2: Vialon ‐ chlorhexidine coated ‐ polyurethane ‐ 16/20 gauge
Outcomes	Primary:Phlebitis was defined as "erythema, swelling, tenderness, induration, a palpable cord (2.5cm), and purulence; phlebitis was diagnosed if two or more conditions were present described by Maki and Ringer". Secondary:Catheter‐related bacteraemia defined as "catheter colonisation plus a blood culture growing the same organism". Local catheter infection defined as "catheter colonisation plus fever not attributable to another cause or catheter site purulence plus a negative blood culture".
Notes	Trial registration: Not stated Funding source: Nil disclosed.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Comment: not reported
Allocation concealment (selection bias)	Low risk	Evidence: “Each package was marked with a randomly assigned study number that was assigned at the time of insertion” p. 231
Blinding (performance bias and detection bias) All outcomes	Unclear risk	Comment: not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Evidence: “All patients and study personnel were blinded as to whether or not the catheter was coated with chlorhexidine” p. 231
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Evidence: “All patients and study personnel were blinded as to whether or not the catheter was coated with chlorhexidine” p. 231
Incomplete outcome data (attrition bias) All outcomes	Low risk	Evidence: “Catheter cultures were not done on 47 patients; 28 catheters (60%) were removed by nurses or other non‐study personnel, 17 (36%) either were pulled out by the patients or came out accidentally, and 2 (4%) came out for unknown reasons” p. 233
Selective reporting (reporting bias)	Unclear risk	Comment: not reported
Other bias	Unclear risk	Comment: not reported

Storey 2016.

Study characteristics
Methods	Study design: Single‐centre RCT Method of randomisation: Conducted by a third party Concealment of allocation: Sealed envelopes
Participants	Country: USA Number: 167 participants from cardiovascular thoracic, medical intensive care unit, and oncology. Group 1: 87 PICCs; Group 2: 80 PICCs Age: Group 1: 64 years (21); Group 2: 62 years (22) Inclusion criteria: Requires PICC, ≥ 18 years, no allergy to CHG, length of stay ≥ 48 hours, required insertion of a single‐ or double‐lumen PICC line Exclusion criteria: Pregnant, difficult PICC insertion requiring placement in vascular lab
Interventions	Group 1: Power‐injected PICC ‐ 5 Fr/other ‐ single/double lumen Group 2: Chlorhexidine‐impregnated, power‐injected PICC ‐ 5 Fr/other ‐ single/double lumen
Outcomes	Primary:CLABSI was defined as "laboratory‐confirmed CLABSIs not secondary to an infection at another body site, were reviewed and verified by one certified infection prevention specialists based on the criteria established in the Centre for Disease Control (CDC) guidelines (CDC, 2015)". VTE "identified through clinical assessment of symptoms and diagnostic tests as ordered per standard practice for suspected occurrence". Post‐insertion bleeding was defined as "either moderate or severe dependent on the type of dressing required to control the bleeding".
Notes	Trial registration: Not stated Funding source: Not disclosed
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Evidence: “To reduce bias, randomization was conducted by a third party who randomly mixed and selected envelopes containing study assignment group for each unit” p. 5
Allocation concealment (selection bias)	Low risk	Evidence: “Sixty envelopes per unit were divided evenly (30 in each group) and randomly assigned to either group A (CHG PICC) or B (non‐CHG). The randomized envelope(s) were selected and placed in the enrolment folder” p.5
Blinding (performance bias and detection bias) All outcomes	Unclear risk	Comment: not reported
Blinding of participants and personnel (performance bias) All outcomes	High risk	Evidence: “Blinding of the study was not possible as the catheters differ in appearance and may have introduced bias into the study” p. 12
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Comment: not reported
Incomplete outcome data (attrition bias) All outcomes	High risk	Comment: attrition data reported in consort diagram (Appendix Figure 1) No further patients lost to follow up or discontinued intervention. The consort diagram is incorrect, they have included withdrawn patients and inability to place CHQ PICC before randomisation – these would have occurred after randomisation. Comment: Sample size calculation was set for 180 patients but the study was discontinued early due to slow recruitment from the cardiovascular thoracic unit resulting in only 167 patients enrolled in the study. Therefore, the results are not sufficiently powered and should be interpreted with caution.
Selective reporting (reporting bias)	Unclear risk	Comment: Pubilshed protocol not available, unclear whether the trial was prosectively registered.
Other bias	Unclear risk	Comment: Unequal baseline characteristics

BSI: bloodstream infection
CDC: Centers for Disease Control and Prevention
CHG: chlorhexidine gluconate
CLABSI: central line bloodstream infection
CSF: cerebrospinal fluid
ICU: intensive care unit
IQR: interquartile range
PICC: peripherally inserted central catheter
RCT: randomised controlled trial
US: ultrasound

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Alport 2012	Not an RCT
Bach 1998	Not an RCT
Bennegard 1982	Not an RCT
Chu 2007	Not an RCT, and study did not compare PICC material or design
Chu 2010	Not an RCT
Di Giacomo 2009	Not an RCT
Garland 2008	Did not compare PICC material or design
Kagan 2019	Not an RCT
Kang 2016	Not an RCT, and study did not compare PICC material or design
Khaldi 2009	Did not compare PICC material or design
Leowenthal 2014	Not an RCT
Lozano 2012	Study was not an RCT; did not compare PICC material or design; and did not report any of the primary or secondary outcomes of interest for this review.
Mermel 2007	Wrong study design
NCT00621712	Clinical trial: did not examine participants requiring a PICC
Parker 1995	Not an RCT, and study did not compare PICC material or design
Poli 2016	Did not examine participants requiring a PICC
Ridyard 2017	Did not examine participants requiring a PICC
Rupp 2005	Did not examine participants requiring a PICC
Toh 2013	Not an RCT
Treotola 2010	Not an RCT, and study did not compare PICC material or design
van Vliet 2001	Did not examine participants requiring a PICC
Wheeler 1992	Did not examine participants requiring a PICC
Wu 2019	Did not examine participants requiring a PICC
Yang 2012	Study was not an RCT; did not compare PICC material or design; and did not report any of the primary or secondary outcomes of interest for this review.
Zampieri 2012	Study was not an RCT; did not compare PICC material or design; and did not report any of the primary or secondary outcomes of interest for this review.

PICC: peripherally inserted central catheter
RCT: randomised controlled trial

Characteristics of studies awaiting classification [ordered by study ID]

Cassim 2019.

Methods	Randomised controlled trial
Participants	127
Interventions	Chlorag+ard PICC versus Bard PICC
Outcomes	Central line bloodstream infection, catheter‐related upper extremity deep vein thrombosis
Notes	Conference abstract, additional unpublished data sought

Yoon 2016.

Methods	Randomised controlled trial
Participants	Uncertain
Interventions	Endexo PICC technology
Outcomes	Unclear
Notes	Conference abstract (SIR 2016 Annual Scientific Meeting Program). Insufficent detail, principal invesigator did not respond to email contact

PICC: peripherally inserted central catheter

Characteristics of ongoing studies [ordered by study ID]

ACTRN12616001354471.

Study name	Comparing the effectiveness of PowerPICC with BioFlo on peripherally‐inserted central catheter (PICC)‐related occlusion and infection rates in the oncology/haematology setting: a randomised controlled trial
Methods	RCT
Participants	Country: Australia Target sample size: 240 Age: > 18 years Inclusion criteria: adults, patients referred to Queensland X‐Ray Mater Private Hospital Brisbane who require a PICC line, able to provide written consent or have access to a ‘person responsible’ and can provide the consent on behalf of the patient, English speaking or have access to an interpreter Exclusion criteria: patients younger than 18 years, patients who are from a non‐English speaking background and do not have access to an interpreter, patients who are unable to consent and do not have access to a ‘person responsible’
Interventions	Group 1: BioFlo PICC (AngioDynamics) Group 2: unknown
Outcomes	Primary outcomes: CRBSI: defined as a laboratory‐confirmed bloodstream infection where the PICC line was in situ for 2 days or more and must be in place when the infection occurred Occlusion: will be judged based on the nurses’ documentation of the incidence of blockage (e.g. resistance, sluggish flow) in which interventions were employed to resolve it Thrombosis: will be determined based on the results of the “PICC‐o‐gram” or ultrasound to monitor for occlusion from time of PICC line insertion until time of central line removal Secondary outcome: Adverse events: e.g. catheter rupture, sensitivity or allergy, severe pain
Starting date	29 September 2016
Contact information	carajoyce.cabilan@mater.org.au
Notes

ACTRN12619000022167.

Study name	Comparing peripherally inserted central catheter (PICC) materials to prevent infections and blood clots: a randomized controlled trial
Methods	RCT
Participants	Country: Australia Target sample size: 1150 Age: < 5 years old Inclusion: require PICC for fluid or medication, able to supply informed consent, vascular size able to support 4‐French PICC or larger Exclusion: previous enrolment in the study, current BSI, thrombosis in the vein where PICC is to be placed, non‐English speaking
Interventions	Group 1: BioFlo PICC with PASV (by AngioDynamics) ‐ 4/5 Fr Group 2: Arrowg+ard Blue Advance PICC (by Teleflex) ‐ 4/5 Fr
Outcomes	Primary outcome: PICC failure: composite primary outcome of thrombotic (venous thrombosis, breakage, occlusion) and infective complications (PICC‐associated bloodstream infection, local infection) (all defined below) severe enough to cause cessation of PICC function prior to therapy completion. Each of these complications will be assessed via a review of medical records (including pathology results) and patient assessment. Secondary outcomes: PICC‐associated BSI: a laboratory‐confirmed BSI (LCBSI) that is not secondary to an infection at another body site (excludes mucosal barrier injury LCBSI), with PICC in place for > 2 calendar days on the day of the BSI (day of PICC placement being Day 1) and the PICC in place on the date of the event or the day before, when all elements of LCBI, were first present together (see CDC Device‐associated Module BSI for full criteria) confirmed by a blinded infectious disease specialist, using de‐identified data Local infection: purulent phlebitis confirmed with a positive (> 15 cfu) swab, but with negative or no blood culture, confirmed by a blinded infectious disease specialist Occlusion: complete: at least 1 lumen cannot be flushed or aspirated, or resolved post‐thrombolytic dwell; partial: decreased ability of at least 1 lumen to either infuse blood or fluid and/or withdraw blood or fluid from at least 1 lumen despite the use of thrombolytic Breakage: visible split in PICC material with leakage or radiographic evidence of extravasation/infiltration into tissue, in a PICC formerly flushed to clear occlusion Venous thrombosis: ultrasound/venographic/image confirmed thromboses occurring within the same vessel as the PICC location within 1 week of PICC removal, in a symptomatic patient (pain, swelling), confirmed by blinded radiologist All‐cause PICC complication during treatment: (as above, but may/may not require PICC removal). A composite of thrombotic and infective PICC‐associated complications (thrombotic complication, infective complication, individual complications, and adverse events). Collected via patient assessment and medical record (including pathology result) review Thrombotic complication: composite of venous thrombosis, occlusion and breakage at any stage of PICC dwell. Collected via patient assessment and medical record (including pathology result) review Infective complication: composite of PICC‐associated BSI and local infection at any stage during the PICC dwell. Collected via patient assessment and medical record (including pathology result) review Adverse events: any local or systemic allergic reaction (e.g. pruritus), pain, mortality. Collected via patient assessment and medical record (including pathology result) review PICC dwell time: hours from insertion until removal. Collected via patient assessment and medical record (including pathology result) review Staff satisfaction: using 0‐to‐10 numeric rating scales. Collected at PICC insertion and removal through discussion Healthcare costs: estimate of direct product costs, healthcare resource utilisation (including additional equipment, staff time) and failure‐associated resource usage. These data will be collected using a combination of individual patient assessment, Medicare data, and study‐specific questionnaires. These data will be collected using a combination of individual patient assessment, Medicare data, and study‐specific questionnaires. [At PICC failure, completion of treatment, or 8 weeks after PICC insertion (which ever is soonest).]
Starting date	9 January 2019
Contact information	a.ullman@griffith.edu.au
Notes

ChiCTR2000030971.

Study name	A comparative study of the incidence of occlusion in two different types of PICC catheters in outpatients
Methods	Trial
Participants	Patients who insist on maintenance in our outpatient department once a week Aged 18 years or over Patients who can move freely Patients who can sign informed consent
Interventions	Experimental group: inserting a front‐end, high‐pressure resistant PowerPICC catheter Control group: inserting a 3‐directional valve PICC catheter
Outcomes	Occlusion Bleeding Phlebitis Displacement Thrombus Catheter damage
Starting date	20 March 2020
Contact information	Sun Wenyan; +86 13910108189; email: s13910108189@126.com Beijing Union Medical College Hospital
Notes	trialsearch.who.int/Trial2.aspx?TrialID=ChiCTR2000030971

NCT05278507.

Study name	Comparing Arrow PICC catheters w/ Arrowga+rd Blue Advanced protection performance and safety to unprotected PICC's
Methods	Prospective, randomised, multicentre study
Participants	444 participants will be enrolled in the study and randomised to receive 1 of the 2 PICC devices.
Interventions	The Arrow PICC with Arrowga+rd Blue Advanced protection (Teleflex Medical Incorporated, Morrisville, NC, USA; hereafter referred to as "AGBA") is an FDA‐cleared pressure injectable device that offers both antimicrobial and anti‐thrombogenic protection for at least 30 days. The application of Arrowga+rd Blue Advanced protection uses a proprietary process whereby chlorhexidine is chemically bonded to the intra‐ luminal catheter surfaces from tip to hub, and extra‐luminal catheter body. The device is cleared for marketing in the United States by the Food and Drug Administration, and has obtained the CE mark for marketing in the European Union. The French size and length selected for use will be documented.
Outcomes	First‐attempt PICC insertion success PICC placement success Confirmation of tip patency by aspirating each lumen for brisk blood return and flushing each lumen with at least 10 mL of 0.9% sodium chloride injection USP Incidence of CRBSI assessed by laboratory‐confirmed bloodstream infection
Starting date	16 January 2020
Contact information	Linda Wu; 15010697631; email: lindi.wu@teleflex.com Jovan Xu; 86106411; email: Jovan.Xu@teleflex.com
Notes

RBR‐48342g.

Study name	Peripheral insertion central catheter with Bioflo technology versus Power Picc: incidence and predictors of thrombosis
Methods	RCT
Participants
Interventions	Device E07.132 Intervention group: insertion of PICC with endexo polymer in 517 hospitalised adult patients with a medical order of PICC Control group: insertion of PICC without endexo polymer in 517 hospitalised adult patients with a medical order of PICC Steps to be followed for both groups: Step 1: Request for assessment for insertion of PICC by the titular physician in the patient's electronic health record. Step 2: The nurse at the target unit of the patient will contact the infusion therapy team (ITT) by telephone. The ITT nurse assigned to the procedure should: Step 3: Randomise the patient. Step 4: Evaluate the venous network in the upper limbs with the Site Rite BARD Ultrasound, following the institutional instrument: initial assessment for PICC catheter insertion; evaluation of the criteria related to the risks of deep venous thrombosis according to the instruments of venous thromboembolism and clinical algorithm. Step 5: Request signature of the informed consent form by the patient or guardian, and free and informed consent form for invasive procedures and surgeries. Step 6: Insert the catheter, according to the institutional instrument PICC insertion and request an antero‐posterior chest X‐ray in the bed, with a report signed by a radiologist. Step 7: Evaluate the patient daily, according to the institutional instrument daily assessment of patients with PICC.
Outcomes	Primary outcome: a statistically significant decrease in the incidence of deep vein thrombosis diagnosed through Doppler ultrasound in the presence of 1 of the signs and symptoms: pain, hyperaemia, and oedema
Starting date	18 April 2019
Contact information	eduarda.santos@einstein.br
Notes

BSI: bloodstream infection
CDC: Centers for Disease Control and Prevention
cfu: colony‐forming units
CHG: chlorhexidine gluconate
CRBSI: catheter‐related bloodstream infection
PICC: peripherally inserted central catheter
RCT: randomised controlled trial

Differences between protocol and review

We performed post hoc sensitivity analyses excluding studies that were terminated early.

Contributions of authors

JAS: conceived the review question, developed the protocol, and co‐ordinated protocol development. Completed the first draft, made an intellectual contribution, and approved the final version prior to submission. Guarantor of the review.

TK: conceived the review question, edited the protocol, made an intellectual contribution, and approved the final version prior to submission.

KC: supported data review and extraction, made an intellectual contribution, and approved the final version prior to submission.

EY: supported data analysis, made an intellectual contribution, and approved the final version prior to submission.

AU: conceived the review question, edited the protocol, made an intellectual contribution, and approved the final version prior to submission.

Sources of support

Internal sources

• The University of Queensland, School of Nursing and Midwifery, Other

  Health librarian and information specialist support

External sources

• Chief Scientist Office, Scottish Government Health Directorates, The Scottish Government, UK

  The Cochrane Vascular editorial base is supported by the Chief Scientist Office.

Declarations of interest

JAS: none known.

TK: The University of Queensland has received investigator‐initiated research grants unrelated to the current study from BD‐Bard and 3M. Over five years ago, AngioDynamics provided funding to TK's former employer (Griffith University) for part of a randomised controlled trial that was eligible for inclusion in this Cochrane review. This grant was investigator‐initiated, and AngioDynamics did not play any part in study development, data collection, analysis, or decision to publish. TK did not review this study. The review of this study was undertaken by other authors of the review. All potential conflicts of interest have been declared in the statements above. TK declares that she does not have any additional conflicts of interest. TK declares that all payments have been made to her institution, and she has not received any personal gain from these bodies and none of the declared funding will bias the review.

KC: none known.

EY: none known.

AU: The University of Queensland has received investigator‐initiated research grants unrelated to the current study from BD‐Bard and 3M. Over five years ago, AngioDynamics provided funding to AU's former employer (Griffith University) for part of a randomised controlled trial that was eligible for inclusion in this Cochrane review. This grant was investigator‐initiated, and AngioDynamics did not play any part in study development, data collection, analysis, or decision to publish. AU did not review this study. The review of this study was undertaken by other authors of the review. All potential conflicts of interest have been declared in the statements above. AU declares that she does not have any additional conflicts of interest. AU declares that all payments have been made to her institution, and she has not received any personal gain from these bodies and none of the declared funding will bias the review.

New