Worldwide, over a billion peripheral intravenous (IV) catheters are placed each year.1 Establishing IV access in pediatric patients can be challenging due to age, prematurity, body habitus, or other underlying medical conditions. Often, multiple attempts are required before successful IV placement, leading to patient/caregiver anxiety, pain, and dissatisfaction.2 The difficult intravenous access (DIVA) score is a 4-variable proportionally weighted rule in which subjects with a score of 4 or greater are more likely to have a failed IV placement on the first attempt.3Table 1 outlines the weight of factors included in the DIVA score: prematurity, age, palpability, and visibility of a vein.
| Vein visible after tourniquet | Visible, 0 points | Not visible, 2 points | |
| Vein palpable after tourniquet | Palpable, 0 points | Not palpable, 2 points | |
| Age | 3+ years old, 0 points | 1–2 years old, 1 point | <1 year old, 3 points |
| History of prematurity | Full term, 0 points | Premature, 3 points |
In the initial validation of the DIVA score with nurse-performed IV placement, a DIVA score of 4 or greater was associated with a 50% or greater likelihood of a failed first attempt.3 External validation found a DIVA score of 4 or greater predicted a more than 60% likelihood of failed first attempt IV catheterization by nurses and nearly 40% by physicians.4,5
Emergency department technicians (EDTs) have assisted nurses and physicians in a variety of procedural tasks, including peripheral IV catheterization. There are few studies that evaluate EDTs’ skills with IV access and, to our knowledge, no studies evaluating the DIVA score with EDT-performed IV placement. The primary outcome of this study was to assess EDTs’ skill with peripheral IV placement in patients with DIVA scores of 4 or greater. The secondary outcome was to evaluate the association of experience of EDTs with first-attempt IV success rates. We also assessed the impact of experience on EDTs’ ability to place peripheral IVs in patients with high DIVA scores.
Methods
Study design
This was a prospective, cohort study conducted between May 2017 and December 2018. The Children's National Health System Institutional Review Board granted study approval (Pro00008461).
A convenience sample of pediatric patients, 18 years of age or younger, were approached if they presented to 1 of 2 pediatric emergency departments (EDs), a freestanding tertiary care children's hospital with an annual ED census of approximately 90 000 and an ED satellite with a census of approximately 37 000. Patients were subsequently enrolled if it was determined by the primary physician caring for the patient that they needed an IV. Patients were excluded from the study if they were deemed critically ill or unstable by the examining physician, if the initial IV placement was started by someone other than an EDT, or if the EDT or family declined to participate in the study. Table 2 shows the DIVA score components of enrolled patients. A total of 29 EDTs, with experience ranging from less than 6 months to greater than 5 years, placed IVs in enrolled patients as illustrated in Table 3.
| Patients with DIVA <4, n (%) | Patients with DIVA ≥4, n (%) | Total patients, n (%) | |
|---|---|---|---|
| Age | |||
| <1 year | 4 (12) | 75 (51) | 79 (44) |
| 1–2 years | 4 (12) | 25 (17) | 29 (16) |
| >2 years | 25 (76) | 48 (32) | 73 (40) |
| Prematurity status | |||
| Premature | 0 (0) | 66 (45) | 66 (36) |
| Full term | 33 (100) | 82 (55) | 115 (64) |
| Veins palpability | |||
| Veins are palpable | 28 (85) | 64 (43) | 92 (51) |
| Veins are not palpable | 5 (15) | 84 (57) | 89 (49) |
| Vein visibility | |||
| Veins are visible | 31 (94) | 60 (41) | 91 (50) |
| Veins are not visible | 2 (6) | 88 (59) | 90 (50) |
| Years of experience placing IVs | EDTs, n (%) |
|---|---|
| <6 months | 1 (4) |
| 6 months–1 year | 2 (7) |
| 1–2 years | 2 (7) |
| 2–5 years | 3 (10) |
| >5 years | 21 (72) |
EDT=emergency department technician
EDTs were randomized to attempting the IV for each patient. The EDT determined the most preferable anatomic site for IV insertion. Prematurity status (gestational age less than 38 weeks) was established with the attending physician and/or family-reported status. After tourniquet application, the EDT performing the procedure assessed and scored vein visibility and palpability. The enrolling research staff member calculated the DIVA score and the duration of the procedure from the time of contact between the first needle piercing the skin to the placement of the first adhesive to secure the successful IV. Duration of the procedure (time to IV placement) included time for multiple attempts and changes in staff members attempting IV placement.
In order to address our primary outcome, we used the same definition of a successful peripheral IV catheterization on first attempt as the original derivation study, defined as an IV attempt in which a saline flush could be injected without compromising the vein.3 At the conclusion of all attempts, despite ultimate success or failure, research staff members conducted surveys with the initial EDT as well as the enrolled patient or caregiver. Variables of interest included the number of IV attempts before a successful placement, the total time to a successful IV placement, the number of years of experience placing IVs, and the number of years that the EDT has been employed in this capacity.
Data collection and analysis
The primary outcome variable was success or failure of first attempted peripheral IV placement by an EDT. The total number of attempts before placement or failure was documented, as was the median time to IV placement. First-time failure rates for IV access were calculated for all patients with DIVA scores of 4 or greater and for patients with DIVA scores less than 4. We calculated test characteristics and the area under the receiver operator curve (ROC) using the DIVA score as a continuous variable. A ROC curve is a graphic plot that illustrates the diagnostic ability of specific classifiers or predictors. In this case those predictors were first-time IV success based on each DIVA score.
We performed a post hoc analysis of the impact of EDT experience on successful first-attempt IV placement. Years of EDT experience were treated as a continuous variable. Years of IV placement experience were reported as an ordinal variable, but was evaluated as a dichotomous variable based on the reasoning that skill level plateaus after 5 years of experience. The association between years of IV placement experience and first-attempt success was analyzed using χ2 analysis, and the relationship between years of EDT experience and median time to IV placement was assessed with linear regression. We also examined this relationship using multiple linear regression adjusted for DIVA score.
Results
One hundred eighty-one children and 29 EDTs were recruited for this study. Technicians employed in their role of EDT had variable experience, ranging from 6 months to greater than 35 years in both patient groups. Twenty-one (72%) of the EDTs had greater than 5 years of IV placement experience and 8 (28%) had less than 5 years of IV placement experience. Forty-four percent of enrolled patients were younger than 1 year of age (Table 2). Among the patients with DIVA scores of 4 or more, the distribution of the 4 DIVA score components was relatively balanced.
DIVA scores in our sample ranged from 0 to 10. For patients with DIVA scores of less than 4, only 3 children required a second attempt to obtain successful intravenous catheterization, resulting in a first-attempt IV placement failure rate of 9% (95% CI, 3–24). In contrast, for patients with DIVA scores of 4 or greater, 61 children required multiple IV attempts, resulting in a first IV placement attempt failure rate of 41% (95% CI, 33–49). The average number of attempts for this group was 1.6 (95% CI, 1.41–1.79) and ranged from 1 to 11 (Table 4). Fifteen of the 29 EDTs initially attempted placement of IV catheters in children with DIVA scores of less than 4, and all 29 participating EDTs initially attempted placement in children with DIVA scores of 4 or greater. Of the 15 EDTs that attempted IV placement in children with DIVA scores of less than 4, 5 had less than 5 years of IV experience, whereas 10 had greater than 5 years of IV experience. Overall, EDTs with greater than 5 years of experience had a first-time IV success rate of 69% (95% CI, 60–76) compared to 55% (95% CI, 41–68) for those with less than 5 years of experience. For patients with DIVA scores of 4 or greater, first-time success rates and median time to IV placement for each EDT are illustrated in Table 5. In these patients with higher DIVA scores, an attempt by an EDT with more than 5 years of experience was associated with nearly 3-fold greater odds of first-attempt success (OR 2.8, 95% CI, 1.03–7.63).
| DIVA score | Patients, n (%) | First-attempt IV placement failure, n (%) |
|---|---|---|
| 0 | 18 (10) | 0 (0) |
| 1 | 4 (2) | 1 (25) |
| 2 | 7 (4) | 0 (0) |
| 3 | 4 (2) | 2 (50) |
| 4 | 40 (22) | 11 (28) |
| 5 | 53 (29) | 23 (43) |
| 6 | 16 (9) | 8 (50) |
| 7 | 17 (9) | 10 (59) |
| 8 | 15 (8) | 4 (27) |
| 9 | 0 (0) | 0 (0) |
| 10 | 7 (4) | 5 (71) |
| EDT | Years of experience | First-time success rate, % | Median time to IV, mm:ss | IQR to IV, mm:ss | Patients, n |
|---|---|---|---|---|---|
| 1 | 0 | 0 | 07:09 | N/A | 1 |
| 2 | 0 | 0 | 1:37:35 | N/A | 1 |
| 3 | 0 | 100 | 01:30 | N/A | 1 |
| 4 | 1 | 0 | 16:37 | 14:41–18:33 | 2 |
| 5 | 1 | 100 | 01:18 | N/A | 1 |
| 6 | 2 | 40 | 33:46 | 04:09–64:57 | 4 |
| 7 | 2 | 67 | 03:18 | 02:21–12:09 | 3 |
| 8 | 2 | 0 | 1:23:04 | N/A | 1 |
| 9 | 4 | 25 | 25:19 | 06:28–47:59 | 4 |
| 10 | 5 | 100 | 01:11 | 01:02–01:21 | 2 |
| 11 | 5 | 63 | 03:56 | 02:22–14:28 | 8 |
| 12 | 5 | 0 | 08:26 | 07:18–16:35 | 3 |
| 13 | 6 | 67 | 03:21 | 01:01–05:35 | 9 |
| 14 | 7 | 58 | 05:47 | 01:31–08:39 | 12 |
| 15 | 7 | 100 | 01:09 | 00:53–03:30 | 3 |
| 16 | 8 | 71 | 01:35 | 00:42–19:08 | 7 |
| 17 | 10 | 0 | 17:54 | N/A | 1 |
| 18 | 13 | 56 | 04:52 | 03:01–13:41 | 9 |
| 19 | 14 | 50 | 06:56 | 03:22–46:07 | 10 |
| 20 | 15 | 100 | 03:45 | 03:15–04:16 | 2 |
| 21 | 15 | 50 | 10:50 | 05:48–15:53 | 2 |
| 22 | 20 | 100 | 00:49 | 00:42–00:56 | 2 |
| 23 | 20 | 27 | 04:26 | 02:22–17:52 | 12 |
| 24 | 25 | 100 | 06:30 | N/A | 1 |
| 25 | 28 | 100 | 00:46 | N/A | 1 |
| 26 | 30 | 100 | 03:30 | N/A | 1 |
| 27 | 30 | 60 | 00:42 | 00:35–01:00 | 5 |
| 28 | 35 | 68 | 03:27 | 01:59–06:16 | 38 |
| 29 | 35 | 100 | 00:50 | 00:35–01:05 | 2 |
EDT=emergency department technician; IQR=interquartile range; IV=intravenous; mm:ss=minutes:seconds; N/A=not applicable
The median time to IV placement for patients with DIVA scores of less than 4 was 75 seconds (interquartile range [IQR] 90–816 seconds) compared to 255 seconds (IQR 42–163 seconds) for patients with DIVA scores of 4 or greater (P < 0.01). For patients with a DIVA score of 4 or greater, time to IV placement was 23 minutes less with each year of EDT experience (P = 0.026, R2 = 0.033). Similarly, in patients with DIVA scores of 4 or greater, a multiple linear regression analysis adjusting for the DIVA score revealed that with every year of EDT experience, the time to IV placement was 25 minutes less (P = 0.02, R2 = 0.05). The linear relationship between EDT experience and time to IV placement explained 3% of the variation in our data, while adding a DIVA score in the linear model accounted for an additional 2% of the variation.
Figure 1 shows the ROC curve for the DIVA scores. The areas under the curve are similar to the original derivation study, 0.67 for their model compared to 0.65 for our study (95% CI, 0.57–0.73). In addition, consistent with the original derivation study, a DIVA score with the highest discriminative ability was 4 (Table 6).
| DIVA score | Sensitivity | Specificity | Positive LR | Negative LR |
|---|---|---|---|---|
| 0 | 1 | 0 | 1 | N/A |
| 1 | 1 | 0.1538 | 1.181754 | 0 |
| 2 | 0.984375 | 0.179487 | 1.199707 | 0.087054 |
| 3 | 0.984375 | 0.239316 | 1.294066 | 0.06529 |
| 4 | 0.953125 | 0.25641 | 1.281788 | 0.182813 |
| 5 | 0.78125 | 0.4701 | 1.474335 | 0.465327 |
| 6 | 0.40625 | 0.735043 | 1.533268 | 0.807776 |
| 7 | 0.296875 | 0.803419 | 1.510192 | 0.875166 |
| 8 | 0.140625 | 0.880342 | 1.175224 | 0.976183 |
| 9 | N/A | N/A | N/A | N/A |
| 10 | 0.78125 | 0.982906 | 45.70317 | 0.222554 |
LR=likelihood ratio
Discussion
Research evaluated the ability of EDTs to place IVs on the first attempt and calculated the median time it took to place IVs in patients with varying DIVA scores. In post hoc analysis, it was found that EDTs with greater than 5 years of IV placement experience were more likely to successfully place IVs on the first attempt, specifically in patients with DIVA scores greater than or equal to 4. In addition, the more experience an EDT had, the less time it took to obtain IV access.
Prior DIVA validation studies evaluated the applicability of the score when IVs were placed by nurses and physicians.4,5 This is, to our knowledge, the first study to evaluate the ability of EDTs to place IVs in patients with varying DIVA scores. In this study, EDTs had similar outcomes to those of physicians, with approximately 40% overall first-time failed IV placement in children with DIVA scores of 4 or greater. In addition, this is, we believe, the first study to evaluate the effect of years of experience on the ability of EDTs to place IVs in pediatric patients with difficult IV access.
The DIVA score allows clinicians to prospectively identify which patients may experience multiple IV attempts. Not only is this information is useful to set expectations for patients and caregivers, but also it allows an opportunity to seek out the most experienced provider to attempt the IV first. Identifying a patient with potentially difficult access a priori will also make way for the use of alternative technology, such as infrared vein-finder lights or point-of-care ultrasound.6–8 The data showed that the DIVA score was helpful but limited to first-attempt IV failure. A significant variation in time to IV access was noted. While the DIVA score has limitations in predicting likelihood of first-time IV placement failure, the need for higher discriminative ability is not critical as misclassification of a patient has low risk and the current standard of care does not discriminate between difficult and easy access empirically. In order to effectively decrease the number of attempts and consistently improve overall success, future studies should be directed toward adapting resources and strategies among pediatric patients with DIVA scores of 4 or greater presenting with a need for IV cannulation. Resources and strategies can be, but not limited to, ultrasound guidance, IV placement nursing specialists, and/or highly experienced ED personnel.8
Conclusion
The DIVA score may be generalizable to IVs placed by experienced EDTs. For patients with high DIVA scores, greater than or equal to 5 years of technician experience was associated with nearly 3-fold higher odds of successful first-time IV placement. Greater than 10 years of experience did not account for any additional benefit. Several minutes of time to IV cannulation were saved by experienced technicians compared to less experienced technicians. Evaluations of DIVA score and other predictors of IV access difficulty should incorporate provider experience.
Limitations
The foremost limitation of this study is that patients were enrolled using convenience sampling, which likely resulted in selection bias. Furthermore, patients were enrolled at 2 urban pediatric EDs of varying censuses and acuity without overlap in technicians at each site. This study enrolled patients at a limited number of sites with well-incorporated roles for EDTs and may not be generalizable. As with previous studies attempting to validate the DIVA score, acutely ill children were excluded. Finally, our study was not originally designed to assess the link between time to IV placement and years of experience, and therefore, the number of IV encounters was not evenly balanced across years of EDT experience.
Recommendations for practice
Apply the DIVA score to each pediatric patient requiring IV access to prospectively identify those who will be difficult IV-access patients and adjust technique, tools, or personnel appropriately.