Advice

Evidence review

Clinical and technical evidence

No publicly available evidence was found on the AccuVein AV400 model.

Two randomised controlled trials in children and 2 poster presentations on the AccuVein AV300 (provided by the device's UK distributor) were assessed.

Kaddoum et al. (2012) compared intravenous cannulation assisted by the AV300 device with the standard method of palpitation and visualisation alone (table 1).

Table 1 Summary of the Kaddoum et al. trial (2012)

Study component

Description

Objectives/hypotheses

To evaluate the efficacy of the AccuVein AV300 device in improving the first-time success rate of intravenous cannulation of anesthetised children. Use of the AV300 device was compared with the standard method of locating veins with palpation and visualisation alone.

Study design

Randomised controlled trial.

Setting

Conducted at a tertiary referral centre for children with cancer in the USA.

Inclusion/exclusion criteria

Inclusion: Patients had been referred for surgery or diagnostic imaging under anaesthesia, age <18 years, American Society of Anesthesiologists physical status I, II or III, no existing intravenous access, no need for interpreter.

Exclusion: Malformation or infection at potential insertion site, anticipated to need an intravenous cannula size other than 22 gauge.

Primary outcomes

First time success of intravenous cannulation as defined by an absence of signs of tissue infiltration after rapid administration of 5 ml of crystalloid solution via the cannula.

Statistical methods

An age stratified (<2 versus ≥2 years) block randomisation scheme was implemented to assign patients to a group.

A Fischer's exact test was used to calculate the sample size (146) and predicted a 90% power, a 5% probability of a Type I error and a CI of 95%.

Participants

146 children with cancer who had been referred for surgery or diagnostic imaging under anaesthesia.

Results

Visualisation and palpation alone

AccuVein AV300

Analysis

Randomised

n=74

Age (mean [range]) 5.4 [0.2–16.8]

BMI (mean [range])

17.9 [13.9–31.6]

Anticipated difficulty:

Easy 31, moderate 36, difficult 17.

n=72

Age (mean [range]) 5.7 [0.5–17.1]

BMI (mean [range])

17.3 [14–26.2]

Anticipated difficulty:

Easy 33, moderate 20, difficult 9.

Efficacy

n=54

n=54

Primary outcome: First attempt success rate

73%

(54/74)

75%

(54/72)

95% CI 62 to 82% for visualisation and palpation alone

95% CI 64 to 84% for AccuVein AV300

p=0.85

Selected secondary outcomes

Number of skin punctures

Mean=1.26

SD=0.75

Mean=1.33

SD=0.75

p=0.86

Time to successful cannulation (minutes)

Median=1.00

Range=0.38‑4.75

Median=1.18

Range=0.25–5.03

p=0.10

Conclusions

The trial found no statistically significant difference in first attempt success rates, number of skin punctures or the time to successful cannulation between the standard method of visualisation and palpation alone and the use of the AV300 device.

Abbreviations: CI, confidence interval; n, number of patients; SD, standard deviation.

Subgroups were identified from the trial population based on 3 factors generally associated with potential ease of cannulation. These were age (<2 versus ≥2 years), prior assessment of cannulation difficulty (easy versus difficult) and skin colour (light skin versus medium or dark skin).

No statistically significant differences in the first time success rates were found for these subgroup pairings. However, it should be noted that at least 1 subgroup in each of these pairings consisted of a small number of patients (11 or fewer) and therefore statistical power would have been low.

The authors reported that the AV300 device was easy to use and despite not being able to demonstrate any superiority to the standard method they did feel that visualisation of the veins was improved. They offered several possible reasons as to why this did not translate to a statistically significant increase in the cannulation success rate, including the lack of vein depth information and their perception that the device overestimated the size of superficial veins.

de Graaff et al. (2013) compared intravenous cannulation assisted by 3 devices that use infrared light. These were the VeinViewer Vision, the AccuVein AV300 and the Vasculuminator. The standard method of palpitation and visualisation alone was used as a control. This trial is summarised in table 2.

Table 2 Summary of the de Graaff et al. trial (2013)

Study component

Description

Objectives/hypotheses

To evaluate the effectiveness of 3 near-infrared devices in facilitating peripheral intravenous cannulation in children.

Study design

Cluster randomised clinical trial.

Setting

Conducted at a tertiary referral centre for children in the Netherlands.

Inclusion/exclusion criteria

Inclusion: Patients had been referred for non-cardiac surgery under anaesthesia, age <18 years.

Exclusion: Cannula already in situ.

Primary outcomes

Successful peripheral intravenous cannulation at the first attempt. An attempt was defined as a penetration of the skin with the needle.

Statistical methods

The Kruskal–Wallis test was used for comparison of continuous data. The chi-squared test was used for dichotomous data. Logistic regression analysis was used to calculate crude and adjusted odds ratios and correct for any bias on the profession of the operator.

Participants

1913 children who had been referred for non-cardiac surgery.

Results

Control Group (visualisation and palpation alone)

AccuVein AV300

VeinViewer

Vascu-luminator

Analysis

Randomised

n=444

n=292

n=357

n=290

Efficacy

n=328

Age (median [range]) 5 [0–18]

Dark skin/light skin:

20/436 (4.6%)

Anticipated difficulty:

(median [range])

2 [0–10]

n=218

Age (median [range]) 6 [0–17]

Dark skin/light skin:

15/289 (5.2%)

Anticipated difficulty:

(median [range])

2 [0–10]

n=267

Age (median [range]) 5 [0–18]

Dark skin/light skin:

23/344 (6.7%)

Anticipated difficulty:

(median [range])

2 [0–10]

n=211

Age (median [range]) 5 [0–18]

Dark skin/light skin:

10/287 (3.5%)

Anticipated difficulty:

(median [range])

2 [0–10]

Primary outcome: First attempt success rate

73.9%

(328/444)

74.7%

(218/292)

74.8%

(267/357)

72.8%

(211/290)

95% CI 69.8 to 78.0% for control

95% CI 69.7 to 79.6% for AV300

p=0.94

Selected secondary outcomes

Number of skin punctures (median,

[range])

1, [1–11]

1, [1–8]

1, [1–11]

1, [1–10]

p=0.95

Suitable vein not visible without device but visible with device

14.3%

(5/35)

23.7%

(9/38)

40.0%

(6/15)

p=0.14

Conclusions

The trial found no statistically significant difference in first attempt success rates or the median number of skin punctures between the standard method of palpation and visualisation alone and the use of the AV300 device.

Abbreviations: CI, confidence interval; n, number of patients.

de Graaff et al. concluded that although vein visibility was reported as enhanced, the near-infrared devices did not improve cannulation.

The investigators also divided their population into subgroups to see if improvements in cannulation rates were only improved if cannulation was deemed to be difficult. The subgroups used were: pre-assessed difficulty of cannulation (>3 on a 1–10 range [10 being extremely difficult] versus <3), age (<3 years versus ≥3 years), skin colour (Fitzpatrick types 5 and 6 [dark] versus other Fitzpatrick types) and BMI (>85th percentile versus <85th percentile). No statistically significant benefit in first time cannulation rates was found for any of these subgroups.

Frame et al. was a study presented at the Royal Medical Society's National Student Conference in Edinburgh 2012 in the form of a poster (table 3).

Table 3 Summary of the Frame et al. (2012) study

Study component

Description

Objectives/hypotheses

To evaluate the impact of a vein illumination device on the peripheral intravenous cannulation technique in the emergency department.

Study design

Prospective cohort study comparing the AccuVein AV300 device with a control of the standard technique of visualisation and palpation alone. This was actioned as a first audit cycle using the standard technique, and then a second audit cycle using the AV300.

Setting

Conducted in an emergency department. The patient age range was reported as 16–97 years with a mean age of 61 years.

Inclusion/exclusion criteria

None stated

Primary outcomes

Compliance to the University Hospitals of Leicester's

guidelines on peripheral intravenous cannulation.

In particular whether there was re-palpation of the

insertion site after the skin had been cleaned.

Statistical methods

Simple percentages and median values were compared. The difference in whether there was re-palpation of the insertion site after cleaning is reported as significant but the method used and confidence interval was not reported, so it is not clear if this is a statistical significance.

Participants

200 patients, split into 2 cycles of 100 patients each.

Results

1 st cycle (without the AV300)

2 nd cycle
(with the AV300)

Analysis

Grouping

n=100

n=100

There was no randomisation in this study.

Primary outcome: Re‑palpation of injection site after skin cleaning

41%

(41/100)

24%

(24/100)

Reported as significant, but no confidence intervals were given.

The target was 0%

Selected secondary outcomes

First time success rate

84%

76%

None reported.

Conclusions

The study reported a significant decrease in the re-palpation of the injection site after the skin clean when the AV300 device was used. This is an improved rate of compliance with the University Hospitals of Leicester's guidelines.

Abbreviations: n, number of patients.

Two more clinical trials were identified on ClinicalTrials.gov during the production of this briefing:

Costs and resource consequences

It is not possible to estimate the likely NHS usage because information on the number of patients for whom the technology would be used is not available or quantifiable.

No published evidence on resource consequences for either the AccuVein AV300 or the AccuVein AV400 devices was identified in literature searches. In the study by Kaddoum et al. (2012), there was no statistically significant difference in the time to successful cannulation with AccuVein AV300 compared with the standard methods by experienced paediatric anaesthetists without the device.

No evidence on the resource consequences of adopting the AccuVein AV400 was identified.

According to the manufacturer, both the AccuVein AV300 and the AccuVein AV400 can be used with minimal training, implying no substantial additional costs. There would be no need to change the way in which current services are organised. No additional facilities or equipment are needed to adopt the AccuVein AV300 or AV400 into practice.

Strengths and limitations of the evidence

Kaddoum et al. (2012)

Kaddoum et al. (2012) did not show a statistically significant improvement in efficacy with the device. The trial was a single centre study using a sample size of 146, and predicted a statistical power of 90%, a 5% probability of a type I error and a confidence interval of 95%. This would suggest that, if any improvement is offered by the device, this study had a 90% chance of observing it.

First attempt success rate is considered the most clinically relevant primary outcome. A recognised validation technique was used to verify the success of the cannulation. The secondary outcome of number of skin punctures is probably the most relevant to patient experience, and time to successful cannulation is useful for resource analysis.

Cannulation was performed by paediatric anaesthetists with a minimum of 8 years' intravenous cannula insertion experience and therefore all were considered experts. This may affect the generalisability of this study and may be a source of bias against the device, as any advantage over palpation may be minimised. The patients' arms and hands were inspected and a likely site was identified. Randomisation was by computer program and the assigned group revealed immediately before the first cannulation attempt. A separate observer recorded the results.

Only a 22 gauge cannula size was used in this study. This would have been a small size to use in older children and may have introduced a bias for this group as it would have been relatively easier to insert.

de Graaff et al. (2013)

de Graaff et al. (2013) was a larger, single centre trial involving a total sample size of 1913, with 292 allocated to the AV300 group and 444 allocated to the control group. The other patients were allocated to groups using other vein visualisation devices. The predicted statistical power of this study was 80% with a level of significance of 0.05 (assuming 2-sided testing).

The primary outcome was first attempt success rate, which would be considered the most clinically relevant. The use of a validation technique was not recorded in the study method, but it is standard practice to validate each cannulation. The secondary outcome of number of skin punctures is relevant for patient satisfaction.

Cannulation was performed by clinical staff with various professional roles, which may make this study more generalisable. Before the start of the study a 1‑month familiarisation period was used to acquaint all users with the devices even if they had previous experience. Logistic regression analysis was used to correct the outcomes for any potential bias by profession of the operator. The probability of a prolonged learning phase was assessed by checking for any trend in the time taken to cannulate the first 3 successive cohorts of patients. Randomisation was performed immediately before the first cannulation attempt and after the assignment of a team member to perform the cannulation. Recording of results was performed by a separate observer. Subgroups were assigned by each individual operator, which may have introduced some variability.

One of the authors of this study has declared that they have filed a patent for one of the other vein visualisation devices (Vasculuminator).

Frame et al. (2012) concludes that using the AV300 device improved compliance with a hospital's guidelines on intravenous cannulation by substantially reducing the rate at which the insertion site was re-palpated after the skin was cleaned. No confidence interval or details of the statistical method used to assess significance were provided for this finding, so it is not clear if this is a statistical significance. It was noted that the first time success rate of cannulation, when using the AccuVein, decreased by 8% compared with when AccuVein was not used, but no analysis was performed on this result to see if it was statistically significant.

Patient populations in both audit cycles were reported to be of a similar age, but additional factors that may influence the ease of cannulation, such as skin colour and BMI, were not considered. The study did not use randomisation and did not state whether these were consecutive patients. The study did not state whether cannulation was being attempted by the same group of people and does not correct for any operator bias. The study did not report who was responsible for recording the results.

The primary outcome in this study is relevant for compliance with good clinical practice, but its findings are limited in usefulness because the operators may have been aware of the auditing process. If so, or if additional interventions were applied between the 2 auditing cycles, then the improvement may not be entirely attributable to the use of the AV300.

Version of device

All of the evidence relates to the AV300 device and not the currently available AV400 device. Although there is uncertainty about whether the evidence for the AV300 is relevant to the AV400, it seems unlikely that the stated differences between the 2 devices would change the basic functionality of the unit. The modifications in the AV400 may improve usability and visualisation as veins with blood flow in all directions can be seen at once without adjusting the position of the device. No published evidence directly comparing the performance of the 2 models has been identified. There is no published economic evidence relating to either device.