Randomized controlled trials (RCTs)

Seven RCTs met the selection criteria presented in Table 1. Three RCTs were designed as cluster randomized controlled trials.^10–12 This type of design simplified the randomized allocation of the device to operating rooms within the study center. In two of the clustered RCTs the presence of the device in the operating room also determined the device operator(s) for a defined period, either weekly,¹⁰ or daily.¹¹ Controls were patients where IV cannulation took place in an operating room where the device was not available. The other clustered RCT had one OR and the randomized weekly presence or absence of the device defined the intervention and control groups, respectively.¹² One RCT was a four-arm study,¹⁰ while the others only examined one imaging device and a control.^{3, 5, 6, 10, 12, 13} Two studies were published in 2011,^{6, 13} two in 2012,^{3, 5} and three were published in 2013.^10–12

Non-randomized studies (NRSs)

Three NRSs met the selection criteria presented in Table 1.^{1, 2, 15} One was a retrospective analysis of consecutive patients,¹ while the other two studies were conducted prospectively on consecutive patients.^{2, 15} In these studies patients were allocated to intervention or control based upon two timeframes, one that used the intervention and one that did not.^{1, 2, 15} All studies were two armed studies, one was published in 2015,¹ one in 2012,² and one in 2011.¹⁵

RCTs

Four included RCTs were conducted in the US,^{3, 5, 6, 13} while three were conducted in the Netherlands.^10–12 The studies from the Netherlands were conducted at, or in collaboration with the University Medical Center, Utrecht, the Netherlands, were all cluster RCTs, and all have at least one common author. The University Medical Centre in Utrecht has filed a patent for the VascuLuminator and the common author is listed as a co-inventor.^10–12 The remaining RCTs conducted in the US do not list common authors or institutions.^{3, 5, 6, 13} One RCT from the US also lists collaborators in Beirut and Australia.⁵

NRSs

Two NRSs were conducted in the Netherlands, published in 2011,¹⁵ and 2012,² and one was conducted in Germany, published in 2015.¹ The studies from the Netherlands were both conducted at the University Medical Centre in Utrecht which has filed a patent for the VascuLuminator and the first author of both studies is listed as a co-inventor of this device.^{2, 15}

No included studies were conducted in a Canadian healthcare setting.

RCTs

One RCT examined a population of 120 neonates, 43 of which were at least 1500g and 77 of which were less than 1500g. These enrolled preterm and term neonates were chosen for PICC placement in a level 3 neonatal intensive care facility.³ Four RCTs examined a mixed pediatric population that included neonates who underwent scheduled surgical interventions in an operating room at a children’s,^{5, 10, 11} or general hospital.¹² Two RCTs examined a mixed pediatric population that included neonates who required IV access in the emergency department (ED).^{6, 13} Inclusion criteria for two of these RCTs were ages 0 to 18 years,^{10, 11} one included patients aged 0 to 15 years,¹² one included patients aged 0 to 17 years,¹³ one included patients less than 20 years old,⁶ while one reported an enrolled patient age range of 0.18 to 17.1 years in the study.⁵ The largest RCT randomized 1913 patients,¹¹ followed by 770 randomized patients,¹⁰ 336 randomized patients,¹³ 148 randomized patients,⁵ 127 randomized patients,⁶ and 88 randomized patients,¹² for the smaller included RCTs that examined a mixed pediatric population. One of the studies aimed to evaluate device efficacy in a population with dark skin colour (Fitzpatrick skin colour scale types 5 and 6) and was conducted on the Dutch island of Curacao in the Caribbean.¹² Subgroup analyses were presented in five studies,^{5, 6, 10, 11, 13} and included patients under three years,^{10, 11} under two years,^{5, 13} patients at least two years old,⁵ patients less than 17 years,¹³ patients between 8 and 17 years,¹³ patients with a BMI over the 85th percentile,^{10, 11} difficult IV access score greater than three,¹¹ estimated as easy cannulation,⁵ estimated as difficult cannulation,⁵ dark skin colour (Fitzpatrick scale types 5 & 6),¹¹ medium or dark skin (undefined),⁵ light skin (undefined),⁵ profession of device operator,¹¹ and awake patients.¹¹ One study provided an extensive analysis of patient subgroups based on age, weight, ethnicity, chronic disease status, the reason for IV access requirement, the use or non-use of topical anesthetic, and the experience of the performer of the device.⁶ Exclusion criteria were already a cannula in situ,^10–12 cancelled surgery,^{10, 11} did not require cannula,¹² cannula gauge other than 22 required,⁵ non-English speaking guardians,¹³ need for emergent PIV,¹³ need for immediate resuscitation,⁶ and the presence of malformation or infection at potential insertion site.⁵

NRSs

The three included NRSs examined mixed pediatric populations.^{1, 2, 15} The most recent NRS examined patients 17 years old and younger requiring IV cannulation prior to surgical interventions.¹⁶ In 2012, Cuper et al., examined a mixed pediatric population of three years and younger that required arterial cannulation prior to cardiothoracic surgery,² and in 2011, Cuper et al., examined a population six years old and younger that required IV access for blood withdrawal.¹⁵ One study provided an exclusion criteria for patients that already had a cannula in situ, or for patients that had a cannula placed directly in the femoral artery for a clinical reason.² No population subgroup analyses were described in the NRSs included in this report.

RCTs

The RCTs included in this report evaluated three vascular transillumination devices; VeinViewer,^{3, 6, 11, 13} VascuLuminator,^10–12 and AccuVein AV300,^{5, 11} although it is not clear if similarly named devices were identical. One four-armed RCT examined all three of these devices compared to standard of care,¹¹ while the other six RCTs compared the use of one device to standard of care.^{3, 5, 6, 10, 12, 13} All RCTs provided a similar brief description of the investigated device and its use, as well as the training provided to the operators during the study. Three RCTs provided equipment supplier information on the VascuLuminator as De Koningh Medical Systems, Arnhem, NL.^10–12 One RCT refers to the VeinViewer as the VeinViewer Vision,¹¹ one as the Vein Viewer,³ and two as the VeinViewer.^{6, 13} Three RCTs listed this device as available from Christie Medical Corporation, Memphis, TN,^{3, 11, 13} and one lists the device supplier as Luminetx Corporation, Memphis TN.⁶ One RCT examined the AccuVein AV300 and listed the supplier as AccuVein LLC,⁵ and one as Avant Medical,¹¹ both were reported to be located in Cold Spring Harbor, NT, USA.

NRSs

The most recent NRS compared the efficacy of the AccuVein AV300 (AccuVein LLC, 40 Goose hill Rd, Cold Spring Harbor, NY) to standard of care.¹ The two NRSs, Cuper et al., 2011 and Cuper et al., 2012, examined an unbranded NIR vascular imaging system compared to standard of care.^{2, 15} However Cuper et al., 2012 stated that the University Medical Centre of Utrecht had filed a patent for the VascuLuminator and that two of the authors were listed as co-inventors.² In Cuper et al., 2011 the NIR imaging system was referred to as a prototype device.¹⁵

RCTs

One RCT examined the successful PICC placement and reported the radiographically confirmed success rate on the first session which included a maximum of four attempts. This RCT also reported the success rate within three sessions and the factors that were associated with successful PICC placement.³ Five RCTs reported the frequency of success on first attempt for venous access.^{5, 6, 10–12} Four RCTs reported the time required to achieve venous access.^{5, 10, 12, 13} Other reported outcomes included the number of attempts required,¹³ pain scores on a visual analogue scale (VAS),¹³ the perceived usefulness of the device,^{11, 12} whether a suitable vein was visible with the device,¹¹ and the success of the procedure over time to assess any effect of experience with the device.¹² None of the RCTs reported any data or methods for data collection on adverse event occurrence.^{3, 5, 6, 10–13}

NRSs

Three NRSs reported the frequency of success on the first venous access attempt,^{1, 2, 15} two also reported the time to success and the number of attempts before successful venous access.^{1, 2} Additionally one NRS reported the time to first flashback of blood,² and one reported the time of needle manipulation before venous access.¹⁵ None of the NRSs reported any data or methods for data collection on adverse event occurrence.^{1, 2, 15}

RCTs

One RCT was focused on placement of PICCs in a neonate-specific population and found a trend towards more successful PICC placement when the Vein Viewer was used, however this finding was not statistically significant.³ This RCT identified gestational age as a statistically significant factor in the overall success rate of PICC placement with or without use of the Vein Viewer.³ Regression analysis that accounted for the gestational age of all patients and all PICC placement attempts found that the Vein Viewer improved overall successful PICC placement. The authors of this study suggest that the Vein Viewer may provide additional benefits in more mature infants where visualization of vessels is often more challenging.³ The authors therefore also suggest that future studies should focus on a more mature infant population where the Vein Viewer may demonstrate a significant advantage.

The remaining six RCTs examined a mixed pediatric population perioperatively or in the ED.^{5, 6, 10–13} Five of these RCTs reported an outcome of venous access success on first attempt and none identified a statistically significant advantage of a vascular imaging device for the included patient population.^{5, 6, 10–12} No difference in the time required to achieve venous access was reported in the four RCTs that reported this outcome.^{5, 10, 12, 13} No advantage in the number of venous access attempts, or VAS pain scores was identified for the included patient population in the one RCT that reported these outcomes.¹³ Van der Woude et al. reported no change in first attempt success over time and also found that 32/43 operators found the VascuLuminator as at least partly helpful.¹² The one RCT that compared devices found the VeinViewer Vision and AccuVein AV300 devices visualized the vein of first choice more frequently and were rated as valuable by operators more frequently than the VascuLuminator. These findings did not translate into any significant clinical efficacy outcome improvements over standard of care for these three devices.¹¹ Subgroup analyses of included patient populations did not identify any statistically significant advantage of vascular imaging devices for patients under three years,^10–12 patient age categories,⁶ patient weight categories,⁶ patient ethnicity,⁶ chronic disease status,⁶ purpose of IV access,⁶ use of topical anesthetic,⁶ nurse experience,⁶ patients with a BMI over the 85th percentile,^{10, 11} Fitzpatrick skin grade 5 & 6,^{11, 12} awake patients,¹¹ when the operator was a trainee,¹¹ or for patients with a difficult IV access score of over 3,¹¹ for outcomes of venous access success on first attempt. One RCT, Van der Woude et al., found a statistically significant increase in successful first attempt for venous access in patients with an anticipated venous access difficulty of hard or very hard when the VascuLuminator was used.¹² Subgroup analyses of included patient populations did not identify a statistically significant decrease in the time required for vascular access for patients under three years,^{10, 12} patients with a BMI over the 85th percentile,¹⁰ or with an anticipated venous access difficulty of hard or very hard.¹² One RCT, identified a statistically significant decrease in the mean time to place PIV, and in nurse assessments of pain for patients under two years old when the VeinViewer was used.¹³ Pain assessment outcomes by parents and nurses were not different when the VeinViewer was used, as compared to standard of care, for other analyzed patient age subgroups including patients under two years as assessed by parents.¹³

NRSs

The NRSs included in this report did not identify any consensus of statistically significant advantages of vascular imaging devices for vascular access in the mixed pediatric populations examined.^{1, 2, 15} Standard of care demonstrated superior time to vascular access as compared to the AccuVein AV300 in all included patients, and patients less than 72 old in one study.¹ This study also demonstrated advantages of standard of care in outcomes of number of vascular access attempts, and the rate of success for first attempt for vascular access. These advantages were statistically significant for both all included patients and patients less than 72 months old.¹ In 2011, Cuper et al., found a greater success in vascular access upon first attempt using a NIR imaging device as compared to standard of care.¹⁵ In 2012, Cuper et al. did not find any statistically significant advantage to a NIR imaging device and suggested that vessel localization may not be a limiting factor in the time required and success rate of arterial cannulation in small children.²