Placement of a peripheral intravenous (PIV) catheter is among the most frequently performed procedures in hospitals. Approximately one-quarter of patients treated and discharged from the emergency department (ED) and almost all admitted ED patients will have a PIV line placed during their stay, often for critical interventions 1. In most cases, this is a relatively straightforward procedure based on the traditional landmark method that involves visualization and palpation of a vein. However, it is estimated that PIV access is difficult to achieve in ∼10% of ED patients 2 because of obesity, intravenous drug abuse, dehydration, chronic illness, or vasculopathy 3. For patients with difficult access, repeated attempts are often necessary before achieving successful catheter placement. Each additional pass of the needle increases patient discomfort and anxiety, delays the initiation of treatment, increases the risk of complications, and utilizes more clinician time; thus, success on the first attempt is ideal. If a PIV catheter cannot be placed, the other access options include the placement of a temporary intraosseous line, a peripherally inserted central catheter (PICC), or a central venous catheterization. These alternatives to PIV are associated with an increased risk of serious complications and higher cost 4.
The use of ultrasound (US) guidance for PIV placement has the potential to improve the initial and overall success rate among patients with difficult peripheral venous access 5. US allows real-time visualization of target veins that are otherwise invisible to naked inspection or palpation, it provides precise information on the size and location (including depth from skin surface) of the target as well as the relative distance from neighboring arteries or nerves that need to be avoided. Dynamic US guidance enables redirection of the needle without additional transdermal needle sticks. For central venous catheters (CVCs), a meta-analysis of randomized-controlled trials (RCTs) found that US guidance resulted in higher initial and overall success rates, fewer complications, and faster cannulation compared with the traditional landmark method 6. In addition, US guidance of CVCs has been shown to be cost-effective because it requires less clinician time and causes fewer complications 7.
The objective of this study was to systematically review RCTs that compared US guidance with the traditional landmark method for PIV placement among patients with difficult peripheral access to determine whether one approach is superior to the other.
Materials and methods
We developed a systematic review protocol that included a search strategy, inclusion/exclusion criteria, and methods of analysis a priori. We followed the PRISMA statement for the purpose of reporting this systematic review 8. The primary outcome in our review was the success rate of PIV placement defined as initial and/or the overall success rate. Secondary outcomes included the number of attempts and time to successful PIV placement. There were two protocol changes during the review. First, during the electronic database search phase, we agreed to exclude EMBASE because EMBASE (1974 to current year) is embedded in the SCOPUS database. Second, during the title review phase, we decided to exclude studies that focused on the placement of PICCs.
We developed a comprehensive search strategy overseen by a health science librarian. Our six-step search strategy involved looking for eligible studies in biomedical research journal databases, more general journal databases, clinical trial registries, conference proceedings, bibliographies of studies we reviewed, and by contacting authors. First, we developed a search strategy in Ovid MEDLINE (1966 – present) with broad keywords to ensure a high capture rate (see Fig. 1, http://links.lww.com/EJEM/A39). The keywords that we used in our search strategy were ultrasonography, peripheral vein, and catheterization. We imposed no language, publication date, or publication status restrictions. Second, we tailored our search strategy and executed it in the indexing languages of CINAHL (1996 to present), SCOPUS, and Google Scholar. Third, we used the search strategy to search four clinical trial registries: the Cochrane Central Register of Controlled Trials (CENTRAL), the National Institute of Health Research at ClinicalTrials.gov, the National Research Register Archive from the National Research Register Archive of the United Kingdom, and the World Health Organization International Clinical Trials Registry Platform.
Fourth, we searched the following society conference proceedings that were available online using the same keyword search applied for Ovid MEDLINE: the American College of Emergency Physicians (2003–2011), the Society of Academic Emergency Medicine (1997–2011), the American Society of Anesthesiologists (2000–2011), the American Institute of Ultrasound in Medicine (2000–2012), the European Society of Emergency Medicine (2007–2010), the Australasian Society of Emergency Medicine (2003–2012), and the World Interactive Network Focused on Critical Ultrasound (2009–2010). Conference proceedings of the European Society of Intensive Care Medicine (2009–2011) were searched manually. Fifth, we contacted experts in the field and queried them about any relevant unpublished or ongoing work. Finally, we searched the reference lists of studies on this topic for other possible relevant studies.
Selection of studies
Although we did not limit our search to RCTs, we focused the systematic review on RCTs to ensure that the studies we included fulfilled a relatively high level of methodological rigor on the basis of the Australian Government National Health and Medical Research Council (NHRMC) evidence hierarchy (level II, RCTs, or above). To be included in our systematic review, the study had to be an RCT that compared the success rate of dynamic US-guided PIV placement with the traditional landmark approach among patients with difficult PIV access (Table 1). Because there is no accepted definition of difficult PIV access, we included all definitions of difficult PIV access used by the original trial investigators.
We excluded studies for the following reasons: (a) allocation was systematic rather than random; (b) US guidance was static rather than dynamic; (c) US was compared with other access options (e.g. CVC, external jugular vein catheter or PICC) rather than the traditional landmark method; (d) data were not reported on the primary outcome (i.e. the success rates of the two methods); and (e) the trial had not undergone any peer review (i.e. available as a report but never published in a peer-review journal).
Two of the authors performed the search (A.A., H.M.B.). Two authors (A.A., H.M.B.) independently scanned the titles and abstracts identified by the electronic searches to identify studies that fulfilled the inclusion criteria. The two reviewers classified the articles as ‘relevant’, ‘not relevant’, or ‘potentially relevant’. We retrieved the full text of all ‘relevant’ or ‘potentially relevant’ articles identified by at least one abstractor as well as all articles without abstracts that could not be evaluated by the title alone.
We developed a data abstraction form (Fig. 2, http://links.lww.com/EJEM/A40) that the reviewers used to document key RCT principles, important study design details, and all primary and secondary outcomes. The reviewers recorded whether each trial reported the randomization method, use of concealment, intention to treat, and exclusion of participants after randomization. The reviewers also recorded the study setting, patient population (number of patients, age, definition of difficult peripheral access), operator type and experience, and whether the US guidance involved single or dual operator. Finally, the reviewers documented all primary (initial success rate and/or overall success rate) and secondary outcomes (number of attempts, time to successful cannulation) reported by the trial investigators as well as details that affected the measurement of the outcomes such as the definition of successful placement.
Two reviewers (A.A., H.M.B.) independently abstracted the data and a third author (Y.T.L.) arbitrated two papers where the reviewers disagreed (both were excluded). Duplicate publications of the same study material were excluded. Data check and analysis were carried out by one author (Y.T.L.).
We identified 1778 titles and/or abstracts from the electronic database searches and three additional articles from the gray literature search (see Fig. 3, http://links.lww.com/EJEM/A41). The two reviewers examined the titles and abstracts and identified 50 ‘potentially relevant’ or ‘relevant’ studies. After a full-text review of the 50 articles, six articles fulfilled the inclusion and exclusion criteria. We excluded 29 of the studies because they were not an RCT; 13 studies because they did not compare dynamic US guidance with the traditional landmark method; and two studies because they did not report the initial or overall success rate and all attempts to contact study authors for the information were unsuccessful.
Table 2 shows the characteristics of the included studies with respect to the study setting, patient population, and operator characteristics. Three of the studies were carried out in the ED setting and two out of the six studies included focused only on pediatric patients. The largest study included 60 patients 12. The definitions of difficult PIV access used by the different studies included failed attempts by a provider, a history of difficulty as reported by the patient or recorded in the chart, inability to visualize or palpate veins, or provider judgment. The studies varied in terms of whether the operators employed single or dual operators and the extent of experience they had with US guidance for PIV.
All the studies included reported randomly allocating the study interventions; however, Aponte et al. 9 and River et al. 13 did not report whether concealment was achieved (Table 3). Stein et al. 14 randomly allocated all patients to one study group or the other but the investigators were not able to administer the intervention to all patients. Intention-to-treat analysis was not specified in the Aponte, River, or Doniger studies but all studies reported that all patients received the allocated interventions.
A total of 316 patients were enrolled in the six studies; 153 were randomized to the control group and 163 were randomized to the US-guided PIV group 9–14. We were able to determine both first-attempt success rate and overall success rate in two out of the six studies. The studies varied in terms of when the procedure was considered a failure. For example, Benkhadra et al. 10 considered the procedure a failure if cannulation was not achieved within 15 min. In contrast, other studies defined failure if successful cannulation was not achieved within two 12 to four 11 attempts. Two studies detected a significantly higher success rate for US compared with the traditional method on the basis of initial success 10,12 or overall success 12.
The two pediatric studies found that time to successful cannulation was shorter and fewer attempts were required to achieve success for patients randomized to US compared with the traditional method 10,11. In the study by Benkhadra et al. 10, in accordance with the protocol, the study was terminated after interim analysis because time to successful cannulation was found to be significantly shorter in the US guidance group. None of the studies based on adult patients reported any statistically significant differences in the secondary outcomes between the two study groups (Tables 4 and 5).
The most important limitation of this systematic review is the methodological heterogeneity of the studies included and the small number of patients included for final review. The studies varied meaningfully in terms of the patient populations, definition of difficult venous access, operator experience and technique, and reported outcome variables. Because of the substantial methodological differences and the small number of patients, we did not carry out a meta-analysis of the results of the studies.
Second, we limited the systematic review to RCTs even though numerous observational studies have been carried out on this topic. Although the observational studies would have increased the sample size, they would have introduced more heterogeneity in study design characteristics and quality of the evidence. Further, many of the observational studies did not include a control group.
Publication bias can bias systematic reviews, but given that the majority of study results concluded no difference between the US-guidance group and the traditional technique group, we assume that publication bias does not influence this review.
Finally, quality of the trials was not evaluated formally with a score, but important design features were reviewed and discussed.
On the basis of this review, routine use of US-guidance for difficult venous access patients is not strongly supported by the literature. We found that there is a paucity of quality research (NHRMC evidence hierarchy level II or higher) on this topic. Closer scrutiny shows recurring study issues that merit exposition.
Among the six studies that were included in the final review, the largest study enrolled 60 patients. Having a relatively small number of participants may underpower the study to detect a difference between the research and the control arms. Looking beyond the six studies included in this review, two other studies achieved the next highest level of evidence and were not included because their participants were systematically allocated and not randomized. In Bauman’s research of 75 patients, the success rate was not different but US guidance was superior to the traditional method in time to success 16. In a study of 60 patients by Costantino et al. 15, the US group was more successful (97%) compared with the traditional group (33%), and the US group also took less time (4±5.6 vs. 15±11.8 min).
The target patient population, defined as those with difficult intravenous access, varied between studies and affected the results of the studies. The two studies that reported an improved success rate with US defined difficult venous access as those veins that are not visible or palpable. The four studies that did not find a difference in the success rate defined difficult venous access either with failed attempts or a history of difficult venous access. In addition, the review found that US conferred pediatric patients the greatest benefit with respect to decreasing number of attempts and time to cannulation.
Appropriate eligibility for peripheral intravenous access
Inappropriate use of US-guided PIV access may lead to artificially poor performance. In some instances of difficult access, it is more appropriate to go directly to an alternative approach such as placing a central venous or intraosseus access rather than attempting an US-guided PIV. One study has shown that when no vein was visible by US, PIV cannulation was never successful 5. In a study by Panebianco et al. 17, it was found that no attempt at US-guided PIV placement was successful when the vein was at over 16 mm depth and that success was more likely when the vessel diameter was 3.0 mm or more. Another study found the ideal depth of vessel to be 3–15 mm and that a diameter of at least 4 mm significantly improved the success rate of US-guided PIV placement 18. In the study by Kerforne et al.12, patients in the US group required to have a vein transverse diameter of at least 2.5 mm to be included but vein depth was not specified. None of the other studies included in our review addressed this issue.
Most people agree that diagnostic US is operator dependent and can lead to varying results on the basis of skill level 19. The operators in this review were various level of medical providers and their US experience ranged from 1 h of training to fellowship trained or American College of Emergency Physicians credentialed physician sonographer. It should also be pointed out that in the Doniger and Stein studies, those performing the traditional method for the control group may not have been the same as those for the US guidance experimental group 11,14. As with all procedures in medicine, there is a learning curve to achieving US-guided PIV access. A study by Witting et al. 18 found that those who had previously placed 20 or more US-guided catheters had almost twice the success rate compared with those who had done fewer than 20. It must be considered possible that a study where all the operators had a significant previous experience in using EUS guidance for intravenous placement would lead to a result different from our findings.
Definition of outcome variables
The nonstandardized approach to outcome definition led to difficulty in comparing results, but also hampered the ability to show a clinically meaningful effect. For this reason, the authors decided against carrying out a meta-analysis for this review. ‘Success rate’ is interpreted by different authors as the overall success rate or the first-attempt success rate. The definition of overall success rate is not agreed upon. For example, Benkhadra and colleagues 10–12 defined overall success as PIVs achieved within 15 min, some use different definitions such as success within two attempts or four attempts whereas others did not apply any formal parameter for this measurement.
‘Time to success’ is also not quantified uniformly. Start time may begin upon patient enrollment, as in the Stein study, or when the tourniquet is placed and secured, as in the Benkhadra trial 10,14. This may account for the wide range of procedure time, as short as 1 min in the Benkhadra study, to as long as 60 min in the Stein study. Even ‘number of attempts’ may be qualified as number of skin punctures without restriction on the number of subcutaneous redirections, or number of needle redirections, irrespective of the number of cutaneous punctures.
Implications for clinical practice and future research
One question we must ask for future research is what ideal measures are we looking for? There is no consensus but we stipulate that from a patient’s perspective, first-attempt success should be the preferred measure. In addition, it may be better to detect the superiority of one method to another, as the overall failure rate is likely to be low. Eventually, most will be able to achieve PIV if allowed enough attempts or time. We advocate defining time 0 as the time when a tourniquet is placed and secured as this isolates the data to procedural time. Admittedly, this does not take into account the time spent locating the US machine. We also recommend counting the number of cutaneous sticks as number of attempts. In both traditional and US guidance techniques, dynamic subcutaneous adjustments can be subtle and numerous and these minor adjustments are difficult to quantify.
In summary, the existing literature offers limited support for US guidance in the placement of PIV for difficult access patients. The greatest benefit in success rate is found in patients whose veins are neither visible nor palpable. The greatest benefits in time to success and number of attempts are found in the pediatric population. Specific study-related issues have been highlighted that outline how future studies can improve upon the existing literature and better show the utility of US guidance in PIV placement.
The authors thank Melissa McCarthy and Elaine Sullo for their assistance on this project.
Conflicts of interest
There are no conflicts of interest.
1. Pitts SR, Niska RW, Xu J, Burt CW.National hospital ambulatory medical care survey. Natl Health Stat Report
2. Jones SE, Nesper TP, Alcouloumre E.Prehospital intravenous line placement: a prospective study.Ann Emerg Med1989;18:244–246.
3. Dychter SS, Gold DA, Carson D, Haller M.Intravenous therapy.J Infus Nurs2012;35:84–91.
4. McGee DC, Gould MK.Preventing complications of central venous catheterization.N Engl J Med2003;12:1123–1133.
5. Schnadower D, Lin S, Perera P, Smerling A, Dayan P.A pilot study of ultrasound
analysis before pediatric peripheral vein cannulation
attempt.Acad Emerg Med2007;14:483–485.
6. Hind D, Calvert N, McWilliams R, Davidson A, Paisley S, Beverley C, et al..Ultrasonic locating devices for central venous cannulation
7. Calvert N, Hind D, McWilliams R, Davidson A, Beverley CA, Thomas SM.Ultrasound
for central venous cannulation
: economic evaluation of cost-effectiveness.Anaesthesia2004;59:1116–1120.
8. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gøtzsche PC, Ioannidis JPA, et al..The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration.PLoS Med2009;6:e1000100.
9. Aponte H, Acosta S, Rigamonti D, Sylvia B, Austin P, Samolitis T.The use of ultrasound for placement of intravenous catheter
10. Benkhadra M, Collignon M, Fournel I, Oeuvrad C, Rollin P, Perrin M, et al..Ultrasound
guidance allows faster peripheral IV cannulation
in children under 3 years of age with difficult venous access: a prospective randomized study.Paediatr Anaesth2012;22:449–454.
11. Doniger SJ, Ishimine P, Fox JC, Kanegaye JT.Randomized controlled trial of ultrasound
-guided peripheral intravenous
catheter placement versus traditional techniques in difficult-access pediatric patients.Pediatr Emerg Care2009;25:154–159.
12. Kerforne T, Petitpas F, Frasca D, Goudet V, Robert R, Mimoz O.Ultrasound
-guided peripheral venous access in severely ill patients with suspected difficult vascular puncture.Chest2012;141:279–280.
13. River G, Hebig A, McAlpine I, Stein J.Nurse-operated ultrasound
for difficult intravenous access: a randomized trial.Ann Emerg Med2009;54:33–40.
14. Stein J, George B, River G, Hebig A, McDermott D.Ultrasonographically guided peripheral intravenous cannulation
in emergency department patients with difficult intravenous access: a randomized trial.Ann Emerg Med2009;54:33–40.
15. Costantino T, Parikh A, Satz W, Fojtik J.Ultrasonography-guided peripheral intravenous
access versus traditional approaches in patients with difficult intravenous access.Ann Emerg Med2005;46:456–461.
16. Bauman M, Braude D, Crandall C.Ultrasound
-guidance vs. standard technique in difficult vascular access patients by ED technicians.Am J Emerg Med2009;27:135–140.
17. Panebianco NL, Fredette JM, Szyld D, Sagalyn EB, Pines JM, Dean AJ.What you see (sonographically) is what you get: vein and patient characteristics associated with successful ultrasound
-guided peripheral intravenous
placement in patients with difficult access.Acad Emerg Med2009;16:1298–1303.
18. Witting MD, Schenkel SM, Lawner BJ, Euerle BD.Effects of vein width and depth on ultrasound
-guided peripheral intravenous
success rates.J Emerg Med2010;39:70–75.
19. Moore CL, Copel JA.Point-of-care ultrasonography.N Engl J Med2011;14:749–757.
cannulation; peripheral intravenous; ultrasound
Supplemental Digital Content
© 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins