Article Text

Ultrasound guidance for difficult peripheral venous access: systematic review and meta-analysis
  1. Grace Egan1,2,
  2. Donagh Healy1,
  3. Heidi O'Neill2,
  4. Mary Clarke-Moloney1,
  5. Pierce A Grace2,
  6. Stewart R Walsh1,2
  1. 1Vascular Research Unit, University Hospital Limerick, Limerick, Ireland
  2. 2Graduate Entry Medical School, University of Limerick, Limerick, Ireland
  1. Correspondence to Dr Mary Clarke-Moloney, Research Manager, Vascular Research Unit, University Hospital Limerick, Limerick, Ireland; mary.clarkemoloney{at}


Background Establishing intravenous access is often vital in an acute hospital setting but can be difficult. Ultrasound-guided cannulation increases success rates in prospective studies. However, these studies have often lacked a comparative group. This systematic review and meta-analysis aimed to determine the clinical effectiveness of Ultrasound-guided peripheral intravenous cannulation compared with the standard technique in patients known to have difficult access.

Methods Electronic abstract databases, trial registries, article reference lists and internet repositories were searched using the following search terms: ‘peripheral venous cannulation’, ‘peripheral venous access’. Studies meeting the following criteria were included: randomised controlled trial patients of all ages who required peripheral intravenous access; interventions were Ultrasound-guided versus standard cannulation technique; patients were identified as having difficult venous access; inclusion of at least one defined outcome (procedural success time to cannula placement; number of attempts).

Results 7 trials were identified (289 participants). Ultrasound guidance increases the likelihood of successful cannulation (pooled OR 2.42; 95% CI 1.26 to 4.68; p=0.008). There were no differences in time to successful cannulation, or number of percutaneous skin punctures.

Conclusion Ultrasound guidance increases the likelihood of successful peripheral cannulation in difficult access patients. We recommend its use in patients who have difficult venous access, and have failed venous cannulation by standard methods. Further randomised controlled trials (RCTs) with larger sample sizes would be of benefit to investigate if Ultrasound has any additional advantages in terms of reducing the procedure time and the number of skin punctures required for successful venous cannulation.

  • Ultrasound guidance
  • venous access
  • vascular-arterial
  • ultrasound
  • trauma
  • abdomen
  • abdomen- non-trauma
View Full Text

Statistics from


The traditional method of peripheral intravenous (PIV) cannulation requires knowledge of vascular anatomy to estimate target vessel location, and requires vessel visualisation or palpation for accurate puncture.1 PIV therapy emphasises utilisation of the upper extremities, typically via the basilic, cephalic and antecubital veins. Less common access sites include the dorsum of the foot, saphenous vein, external jugular vein, and the scalp veins in children and infants.1 Access can be difficult in certain patients due to factors such as obesity, oedema, chronic illness, hypovolaemia, intravenous drug use, vascular pathology2 and in the paediatric population.3 Patients with difficult access are oftentimes subjected to multiple insertion attempts by different operators utilising staff resources.4 In addition to increased discomfort and decreased patient satisfaction, such patients often have their blood draw and laboratory test results delayed.1 Importantly, patients who fail PIV access may have to undergo central venous access5 which is more invasive, time consuming and prone to serious complications—for example, pneumothorax.6

The use of ultrasonography to facilitate venous access in difficult access patients began in 1991 when Johns and colleagues investigated its use for venography in a patient with an oedematous leg.7 Its role in vascular access has expanded greatly since, largely for central venous access where it reduces complications and puncture attempts compared with the traditional landmark approach.8–10 This has led to increasing interest in Ultrasound to facilitate PIV access, particularly in patients with difficult access.5 ,11–13 Although the value of Ultrasound guidance for PIV access has been considered in a number of randomised controlled trials, to date there has been no pooled analysis. The primary objective of this systematic review and meta-analysis was to investigate the clinical effectiveness of Ultrasound guidance on PIV cannulation in difficult access patients. The primary outcome was successful PIV cannulation. Secondary outcomes included number of skin punctures and procedure time until successful cannulation.


This systematic review and meta-analysis were conducted in accordance with the Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) guidelines.14 The MEDLINE and EMBASE databases were searched from 1956 until January 2012 using the search terms ‘Peripheral venous cannulation’ and ‘peripheral venous access’. No limits were set on the electronic searches. No language restrictions were used. The clinical trials registry was searched in February 2012 to identify any unpublished trials. Article reference lists were scrutinised to identify any further eligible studies. Finally, a basic web search without restrictions was performed using the keywords ‘ultrasound’ and ‘peripheral venous access’ to identify any grey literature, including conference proceedings, meetings and government reports. All potentially eligible reports were reviewed independently by two authors (GE and SRW).

Studies were eligible for inclusion in the meta-analysis if they met the following criteria: (1) randomised controlled trial (2) patients of any age requiring PIV access, (3) patients randomised to Ultrasound versus standard technique for the placement of PIV cannulas; (4) patients were identified as having difficult venous access; (5) inclusion of at least one clinical outcome (procedural success, time to catheter placement, number of attempts). Eligibility criteria were defined in advance of the systematic review. A formal review protocol was not recorded.

The primary outcome was successful PIV cannulation. The secondary outcomes were number of skin punctures and procedure time until successful cannulation. We did not specify a definition for patients with difficult access. Patients were classified as difficult venous access using the criteria adopted by the original trial authors. We also did not specify the Ultrasound technique to be adopted, and accepted all studies, regardless of the Ultrasound technique, provided they met the eligibility criteria. Data were extracted from the trials for each of the outcomes and recorded in a Microsoft Excel (Microsoft, Redmond, Washington, USA) spreadsheet for analysis. Data were double checked by two authors (GE and SRW).

Trial quality was evaluated using the Jadad score. This validated scale consists of three items that assess methods used to generate random assignment, double-blinding and a description of dropouts and withdrawals by intervention group. Potential scores range from 0 to 5.15 A Jadad score of <2 indicates a lower-study quality, whereas a score of 2 or more indicates a higher-quality study. The weighted mean difference (WMD) was determined for continuous variables, and pooled ORs were calculated for categorical variables. Random effects models were used for all outcome measures.16 Heterogeneity was assessed using the Cochran Q-test. This is a null hypothesis test in which a result with a p value <0.05 indicates the presence of significant heterogeneity between the included RCTs. Bias was assessed qualitatively by visual inspection of funnel plots, and also quantitatively by calculation of the Egger test.17 Significance in all tests was set at 5%, and all p values reported are two-sided. The statistical analyses were performed using Statsdirect 2.5.7 (Statsdirect, Altrincham, UK).


Search and study selection

The search results are outlined in online figure 1 (Supplemental file). Fifteen thousand four hundred and five potentially relevant citations were identified by the initial searches. Title screening reduced this to 62 relevant publications for abstract screening. All non-comparative clinical studies and reviews were excluded, leaving 10 potentially relevant abstracts. This number was further reduced to five RCTs after closer inspection; two studies were found to be systematically allocated studies,5 ,11 and three RCTs did not directly compare Ultrasound with standard technique in patients with difficult access.18–20 A supplementary search for unpublished studies discovered an additional seven studies, but upon closer inspection, only two were RCTs21 ,22: one RCT was a published government report, while the other was available only as an abstract. Therefore, the overall search identified a total of seven RCTs for inclusion in this meta-analysis.

All seven studies included in the meta-analysis reported on one or more of several outcomes—success of procedure, time to catheter placement and number of attempts.3 ,21–26 The trials included a total of 289 patients scheduled for PIV cannulation. Six studies investigated real-time ultrasound compared with the standard technique of cannulation.3 ,21–23 ,25 ,26 One trial used an indirect method of ultrasound, whereby, there was no direct visualisation of the vein during the cannulation process.24 Six studies reported on cannulation success rates. One study only reported on cannulation success on the first attempt at cannulation.24 Details of the included trials are summarised in table 1.

Table 1

Characteristics of eligible trials

Primary outcome: successful cannulation

Six out of the seven studies3 ,22–26 reported on cannulation success using Ultrasound guidance versus the standard technique. Ultrasound-guidance was found to increase the likelihood of successful cannulation (107/136 Ultrasound group vs 84/136 ST group; pooled OR 2.42; 95% CI 1.26 to 4.68; p=0.008). There was no evidence of heterogeneity (Cochran's Q 1.79, 5 df, p=0.144) or bias (Egger test = −0.25 p=0.69). The funnel plot was symmetrical overall (figures 1–4).

Figure 1

Funnel plot of successful cannulation.

Figure 2

OR meta-analysis plot for successful cannulation.

Figure 3

Funnel plot of successful cannulation post-sensitivity analysis.

Figure 4

Successful cannulation post-sensitivity analysis.

Secondary outcome: procedure time

Five trials provided sufficient data regarding time to procedural success for analysis.3 ,22 ,23 ,25 ,26 Overall, there was no statistically significant difference in procedure time between Ultrasound and the ST group (WMD 1.18 min, 95% CI−3.55 to 5.90 min, p=0.63). There was evidence of heterogeneity (Cochran's Q 26.24, 4 df, p<0.0001), but no evidence of bias (Egger test=0.67, p=0.82).

Secondary outcome: number of attempts

Four trials provided sufficient data to evaluate the effect of Ultrasound on the number of cannulation attempts required3 ,22 ,23 ,26 Ultrasound guidance had no significant influence on the number of skin puncture attempts (WMD −0.64 attempts; 95% CI0.76 to −0.53; p<0.0001). There was evidence of heterogeneity (Cochran's Q 152.53; 3 df; p<0.0001) but no evidence of bias (Egger test=−2.92, p=0.72).

Sensitivity analysis

One study used an indirect method of Ultrasound, whereby Ultrasound was only used to identify the vein, and there was no real-time visualisation of the vein.24 This study also only recorded cannulation success rates on the first attempt. In view of these methodological differences, a sensitivity analysis was undertaken, excluding this trial's results. The apparent benefit of Ultrasound-guidance with respect to successful cannulation persisted (pooled OR 2.30; 95% CI 1.40 to 6.43; p=0.0075). Cochrane's Q test remained non-significant, indicating an absence of heterogeneity (Cochran's Q 0.65; 4 df; p=0.98). There was no evidence of bias (Egger test=−0.38; p=0.37).


Previous studies have supported the use of Ultrasound guidance in obtaining PIV access in difficult access patients,5 ,13 but to date there has been no systematic review or meta-analysis to determine the overall value of the technique. The results of our meta-analysis suggest that in patients with difficult venous access, the use of Ultrasound guidance significantly increases the likelihood of successful intravenous cannulation when compared with the standard technique. The increased overall success of Ultrasound-guided cannulation was not associated with any significant difference in the time taken to perform the procedure or in the number of skin punctures required to perform the procedure. This review will be of interest to many healthcare professionals, including emergency physicians, paediatricians, anaesthetists, nurses and emergency department technicians, because obtaining intravenous access is vital in many clinical settings, and patients with difficult venous access can represent a serious problem. Given that the number of skin puncture attempts and the time taken to perform the procedure were not significantly decreased by Ultrasound guidance in this meta-analysis, we cannot assume performing Ultrasound produces direct time savings or increased patient satisfaction. However, the overall increased success rate of Ultrasound guidance compared with the standard technique has the potential to decrease morbidity and mortality associated with attaining more invasive methods of access, particularly, central venous access with its associated risks and complications.6 By avoiding the need for central venous access, the use of Ultrasound-guided PIV access could likely increase patient satisfaction and reduce overall time and staff resources used.

The Jadad scores for the included trials ranged from one to three (see table 1). The studies suffered from the inability to double-blind the health professional performing venous cannulation. Two of the RCTs were abstracts, which may have affected their Jadad scores due to the limited space available to discuss the study characteristics. One trial excluded two patients before analysis of the results because they did not achieve successful venous cannulation.22 They did not state from which group the two failed access attempts came, which could have influenced the primary outcome of successful cannulation, although the effects would likely have been too small to affect significance. The characteristics of each trial also varied in terms of sample population, ultrasound technique adopted, level of experience of the operators and the level of training received (table 1). The criteria for defining difficult venous access constituted another source of clinical heterogeneity in this meta-analysis. These issues all warrant discussion as they may influence the results of study outcomes reported.

The indirect method of Ultrasound eschews direct visualisation of the vessel. Ultrasound is used only to identify the vessel. A temporary mark is placed on the skin corresponding to the vessel's subcutaneous position and the rest of the procedure is performed in the normal manner. This method was used in one of the included RCTs.24 The alternative is to perform needle placement under real-time (ie, direct) Ultrasound guidance which is visualised continuously; the operator holds the Ultrasound probe, visualises the target vessel and with the other hand, the needle punctures the skin and is advanced according to its progression on the monitor. In order to ensure that the one study that used an indirect approach did not confound the results, we performed a sensitivity analysis excluding this study. Cannulation success rates remained significantly higher when ultrasound was used in comparison with the standard technique. There was also no evidence of heterogeneity or bias.

Vessel visualisation can be performed in the long axis or the short axis. For the short-axis approach, the vessel is identified in the transverse plane and the needle is inserted at a 45° angle to the transducer. In contrast, the long-axis approach identifies the vessel in its long axis and the needle is inserted at a 30° angle to the skin surface.27 A previous study comparing the short- and long-axis approaches found that for novice users, the short-axis approach enabled quicker cannulation, however there were no statistically significant differences in terms of mean difficulty and number of skin breaks between groups.28 We did not investigate this technique in our meta-analysis as four of the RCTs did not specify method of axis, and one trial allowed sonographers to use whatever method they believed to be most effective.26 This may have influenced the time taken to perform the procedure and, thus, may have impacted on the secondary outcome of procedure time. Future studies should ensure that they specify the mode of axis used.

The number of operators involved in performing the procedure may impact on the success of PIV cannulation. Rose and colleagues compared the effects of the one-operator technique versus the two-operator technique in Ultrasound-guided cannulation of the basilic vein using a randomised controlled design.18 They found no statistically significant difference between the two. However, their study only contained 32 participants. Furthermore, although one versus two operators may not affect the overall success of the PIV placement, a single-operator approach could possibly prolong the time to achieve access and, therefore, since time is used as an outcome in most of the included studies, this could be an influential factor. Further studies are warranted to determine whether the number of operators involved in performing the procedure affects the outcomes of Ultrasound-guided PIV access.

A difficulty in undertaking this review was the varying criteria for classifying patients as ‘difficult peripheral intravenous access’. The number of attempts at intravenous access has been used in many studies. However, there is no clear definition of what exactly constitutes difficult venous access. Some studies use a cut-off of two failed attempts.3 ,21 ,26 Other studies use three failed attempts.5 ,29 Moreover, some studies take a ‘history of difficult access’ as an adequate definition. There are clearly inconsistencies with this approach. The skill of the healthcare professional attempting access by the standard technique will influence the perceived difficulty of attaining PIV access and the outcome of successful cannulation. People's skill and expertise can vary greatly and, thus, an attempt by a less experienced person may not mean that the patient has difficult access.30 As this is subjective, there may also be a bias effect because the operator knows that if attempts fail, an Ultrasound-guided technique will be used.30 Second, comparing this criterion with a ‘history of difficult venous access’ may not be acceptable, as this relies on a patient's memory which may be biased by the unpleasant nature of the procedure.

Given the clinical heterogeneity evident in the studies included in this meta-analysis, we used random-effects models (DerSimonian–Laird method). These models take into account additional sources of variation, such as random error and differences between the procedures and populations.16 We chose to use random-effects modelling as it incorporates statistical heterogeneity, which is reflected in the effect estimate.31 The selective publication of positive findings from randomised controlled trials is an important concern in meta-analytic reviews of the literature.17 Critical examination for the presence of publication and related biases must, therefore, become an essential part of meta-analytic studies.17 Funnel plots (plots of the trials' effect estimates against sample size), can be useful to assess the validity of meta-analyses.17 The results of this analysis demonstrate that for the primary outcome of successful cannulation, the funnel plot was relatively symmetrical. However, due to the fact that symmetry is generally defined informally through visual examination, unsurprisingly, this can result in differing interpretations by different observers.32 Egger's linear regression method is intended to quantify the bias captured by the funnel plot.17 The results of this meta-analysis demonstrated no evidence of publication bias for either the primary or secondary outcomes, as measured by the Egger test. However, since published trials are always more likely to show a positive treatment effect than unpublished trials, it is always possible that our search may have missed some unpublished studies.

When designing trials in the future, it would be important to assess and control for the many variables that may affect the outcomes. It may also be the case that this technique is more useful in certain situations (the procedure requires the patient to stay very still so it might not be ideal in a paediatric setting). Therefore, looking at the differing populations where Ultrasound-guided access has more success would be useful. Of note, it is also important that providers are still trained and experienced in the standard technique as there is a potential to deskill people, which becomes especially pertinent when using it to gain peripheral or central venous access in an emergency setting when Ultrasound is not available.8


The evidence presented in this review supports the use of Ultrasound guidance in obtaining difficult venous access in patients when standard attempts at obtaining access have failed. We found no evidence to support the use of Ultrasound in decreasing procedure time or the number of skin punctures required. However, we acknowledge the limitations of this review and recommend that further RCTs that directly compare Ultrasound-guided PIV access with standard approaches be performed to further investigate the effect Ultrasound may have on procedure time and number of punctures.

From the results of this meta-analysis, we recommend the use of ultrasound in patients who have difficult venous access and have failed venous cannulation by standard methods, particularly in an emergency setting when gaining access to a vein is vital.


View Abstract

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Files in this Data Supplement:


  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.