- Email: [email protected]
Ultrasonography Versus Landmark for Peripheral Intravenous Cannulation: A Randomized Controlled Trial
GENERAL MEDICINE/ORIGINAL RESEARCH
Ultrasonography Versus Landmark for Peripheral Intravenous Cannulation: A Randomized Controlled Trial Melissa L. McCarthy, ScD*; Hamid Shokoohi, MD, MPH; Keith S. Boniface, MD, RDMS; Russell Eggelton, EMT-B, EDT; Andrew Lowey, BS, EMT-P; Kelvin Lim, BS; Robert Shesser, MD; Ximin Li, MS; Scott L. Zeger, PhD *Corresponding Author. E-mail: [email protected].
Study objective: Randomized controlled trials report inconsistent findings when comparing the initial success rate of peripheral intravenous cannulation using landmark versus ultrasonography for patients with difficult venous access. We sought to determine which method is superior for patients with varying levels of intravenous access difficulty. Methods: We conducted a 2-group, parallel, randomized, controlled trial and randomly allocated 1,189 adult emergency department (ED) patients to landmark or ultrasonography, stratified by difficulty of access and operator. ED technicians performed the peripheral intravenous cannulations. Before randomization, technicians classified subjects as difficult, moderately difficult, or easy access according to visible or palpable veins and perception of difficulty with a landmark approach. If the first attempt failed, we randomized subjects a second time. We compared the initial and second-attempt success rates by procedural approach and difficulty of intravenous access, using a generalized linear mixed regression model, adjusted for operator. Results: The 33 participating technicians enrolled a median of 26 subjects (interquartile range 9 to 55). The initial success rate was 81% but varied significantly by technique and difficulty of access. The initial success rate by ultrasonography was higher than landmark for patients with difficult access (48.0 more successes per 100 tries; 95% confidence interval [CI] 35.6 to 60.3) or moderately difficult access (10. 2 more successes per 100 tries; 95% CI 1.7 to 18.7). Among patients with easy access, landmark yielded a higher success rate (10.6 more successes per 100 tries; 95% CI 5.8 to 15.4). The pattern of second-attempt success rates was similar. Conclusion: Ultrasonographic peripheral intravenous cannulation is advantageous among patients with difficult or moderately difficult intravenous access but is disadvantageous among patients anticipated to have easy access. [Ann Emerg Med. 2015;-:1-9.] Please see page XX for the Editor’s Capsule Summary of this article. 0196-0644/$-see front matter Copyright © 2015 by the American College of Emergency Physicians. http://dx.doi.org/10.1016/j.annemergmed.2015.09.009
INTRODUCTION Intravenous cannulation of a peripheral vein is the most common invasive procedure performed in hospitals worldwide.1 The traditional landmark approach involves visual inspection and palpation of the extremity to locate a vein, followed by needle puncture and catheter insertion. The landmark failure rate on first attempt ranges from 12% to 26% among adults and is twice as high in difficultaccess patients.2-4 When multiple cannulation attempts are necessary, patients experience more pain and treatment delays. Ultrasonography increases the success rate of central venous line placement, especially at the internal jugular site, but studies evaluating ultrasonographic peripheral intravenous line placement have reported inconsistent results.5,6 Eight randomized controlled trials have Volume
-,
no.
-
:
-
2015
compared the initial or final peripheral intravenous line placement success rate between landmark and ultrasonography in difficult-access patients.7-14 The majority of the trials9,11-14 favored ultrasonography and reported clinically meaningful differences of 10% or greater in initial13 or overall success rate.9,11,12,14 However, each study’s results were based on a small sample size (range of 35 to 60 total subjects). Meta-analyses of the trials’ results have reported conflicting findings because of heterogeneous patient populations, varying definitions of difficult access, and operator experience.15-17 This randomized controlled trial of 1,189 adult emergency department (ED) patients seeks to determine whether the initial peripheral intravenous line placement success rate is higher with landmark or ultrasonography among patients with difficult, moderately difficult, or easy venous Annals of Emergency Medicine 1
McCarthy et al
Ultrasonography Versus Landmark for Peripheral Intravenous Cannulation
Editor’s Capsule Summary
What is already known on this topic Intravenous access is important but without technical aids can sometimes be difficult. Ultrasonography offers an alternative to standard insertion methods. What question this study addressed Does ultrasonographically guided intravenous insertion aid in all patients or just those deemed in advance to present more challenging access? What this study adds to our knowledge Using a randomized, controlled assignment in 1,189 patients, trained technicians had higher success with landmark approaches in those deemed to have easier access, but ultrasonographic intravenous insertion attempts produced higher success in the moderateand highest-difficulty groups. How this is relevant to clinical practice Selective use of ultrasonographic intravenous placement approaches targeting more difficult patients is pragmatic and evidence based.
access by randomly assigning the procedural method within each difficulty of intravenous access group and operator. We originally hypothesized that the initial success rate would be at least 5% higher when ultrasonography was used among patients with difficult or moderately difficult peripheral intravenous access but no different (ie, <5%) among patients with easy access. MATERIALS AND METHODS Study Design and Setting We conducted a randomized controlled trial with a 2-group parallel design at an academic ED in Washington, DC. The university’s institutional review board approved the study protocol and we registered the trial at clinicaltrials.gov. At this ED, peripheral intravenous cannulation is the primary responsibility of technicians. The ones who work in our ED have previous training as emergency medicine technicians, certified medical assistants, certified nursing assistants, or are nursing students. Since 2008, our ED has trained the technicians working in the ED to perform ultrasonographic peripheral intravenous catheterization. During the first 3 months of employment, each technician completes instruction in ultrasonographic peripheral intravenous cannulation that includes a 2-hour lecture and hands-on demonstration of 2 Annals of Emergency Medicine
the skill. On class completion, under preceptor supervision the technician must perform 10 successful ultrasonographic peripheral intravenous procedures.18 ED nurses at our institution are not trained in ultrasonographic peripheral intravenous cannulation. Their role in peripheral intravenous cannulation is limited to cases in which the treating technician is busy and the nurse believes he or she can successfully place the peripheral intravenous catheter with landmark. Selection of Participants We enrolled adult ED patients deemed capable of providing informed consent who required a peripheral intravenous line in their upper extremity, but not hands. We excluded high-acuity patients (all triage level 1 patients plus triage level 2 for whom the care team deemed it unsafe to delay peripheral intravenous placement).19 If a physician ordered a diagnostic test or treatment that required peripheral intravenous access, the research assistant introduced himself or herself to the technician and patient, explained the study, and asked the patient to participate. Patients who agreed signed a written consent form. Before beginning study enrollment, we also consented the technicians so that we could document the outcomes by technician. Enrollment occurred from August 2013 to August 2014 between 9 AM and 9 PM. We anticipated there would be more eligible patients with easy access than moderately difficult or difficult-access patients, so during the first 4 months of enrollment, we instructed the research assistants to enroll a patient with easy access only after 2 patients with more difficult access were enrolled to ensure that target enrollment in the easy-access group was not achieved much earlier than that of the other 2 access groups. We used SAS software (version 9.3; SAS Institute, Inc, Cary, NC) to generate a randomization strategy that randomly allocated the procedure technique to patients with equal probability to landmark or ultrasonography, stratified by technician and difficulty of intravenous access. We blocked randomization into groups of 6 to distribute the treatment assignment equally by stratum. We uploaded the randomization strategy into REDCap (version 6.5.12; Nashville, TN), a flexible, Web-based, secure data collection system; the research assistants used REDCap for randomization and real-time data capture.20 If the first attempt failed, we randomized the subject a second time, using the same randomization strategy. This gave us the opportunity to compare the 2 treatment methods among a subgroup of the subjects who were similar in all baseline characteristics (because of randomization) and had a failed first attempt. If the second attempt failed, the technique used to secure intravenous access was at the technician’s discretion. Volume
-,
no.
-
:
-
2015
McCarthy et al
Ultrasonography Versus Landmark for Peripheral Intravenous Cannulation
For subjects randomly assigned to landmark, the technician secured a tourniquet around the chosen arm and palpated to identify an acceptable vein. The technician prepped the site with a chlorhexidine swab, inserted the needle, and advanced the catheter into the vein. The technician attached a saline solution lock adapter and used 10 mL of normal saline solution to flush the lock. While flushing the saline solution lock, the technician visualized and palpated the site to confirm placement. For subjects randomly assigned to ultrasonography, the technician used an ultrasonography machine (Sonosite M-Turbo or Zonare ultra) to visualize an acceptable vein. The technician prepped the site with a chlorhexidine swab and sterile lubrication gel. Next, the technician used a dynamic, single-operator technique that involved holding the probe in the nondominant hand and concurrently inserting the needle and advancing the catheter into the vein with the other hand, using the ultrasonography machine to visualize and guide the needle. After insertion, the technician tested the catheter in the same manner as landmark. Before randomization, the technician placed a tourniquet around the brachial region of each subject’s upper extremities and visually inspected and palpated the arms to evaluate suitability of the vessels for intravenous access. If the technician could not see or palpate a vein in either extremity, we classified the subject as having difficult access. If the technician could see or palpate at least 1 vein but anticipated difficulty using landmark, we classified the subject as having moderately difficult access. If the technician could see or palpate at least 1 vein and thought intravenous access would be easy with landmark, we classified the subject as having easy access. The research assistants briefly interviewed subjects to obtain their weight and height and ascertain whether they had any health conditions or were receiving treatments previously associated with difficult intravenous access.2,4 The research assistants observed the peripheral intravenous procedure and recorded which arm and vein was used, the catheter gauge and length, the technique used, whether the attempt was successful, and whether any complications occurred. If more than 2 attempts were required, the research assistant documented the total number of attempts, the final method used to gain intravenous access, and who obtained it. Technicians who agreed to participate in the study completed a brief survey before their first subject enrollment. We asked them to report the number of years of experience they had placing peripheral intravenous lines and years placing ultrasonographic peripheral intravenous lines. We also asked each technician to estimate how many peripheral intravenous lines and ultrasonographic Volume
-,
no.
-
:
-
2015
peripheral intravenous lines they placed on average during a typical 12-hour shift. Outcome Measures The main outcomes were success or failure on the initial and second attempts. Secondary outcomes included the occurrence of a complication, patient-reported pain associated with the procedure (0¼no pain at all to 10¼extremely painful), and duration of the first attempt, defined as the time from onset of tourniquet placement until the saline solution had been flushed through the peripheral intravenous line. We defined an intravenous attempt as 1 percutaneous needle puncture, regardless of the amount of subcutaneous exploration from the single puncture site. We considered a cannulation successful if the technician was able to infuse fluid without infiltration. If it was unsuccessful, the research assistant also recorded whether infiltration and arterial or nerve puncture occurred. Primary Data Analysis We conducted separate analyses for the first and second attempts. When calculating the target sample size, we assumed 80% power and initial success rates according to a previous study of 65% for difficult-access patients, 74% for moderately difficult ones, and 85% for those with easy access, using a landmark approach.3 To detect a 5% difference in the initial success rate for each intravenous access group, we estimated that the sample size required was 6,314 subjects (1,377 subjects with difficult access, 1,094 with moderately difficult access, and 686 with easy access for each treatment group). This proved too large, given available resources. We revised our target sample size after 4 months of study enrollment. We assumed the same initial success rates with landmark as we did when we estimated the original sample size estimates3 for each treatment group and the same level of power (ie, 80%). The only change we made when revising the sample size estimates was to increase the difference we were able to detect from 5% to 10% or greater in the first attempt between the 2 techniques for the groups with easy access and moderately difficult access and to 20% or greater for the group with difficult access. The revised target sample size was 948 (73, 260, and 141 subjects, respectively, in each treatment group for difficult, moderately difficult, and easy access). We analyzed all data according to intention to treat; the statisticians were blinded to procedural assignment (X.L. and S.L.Z.). A priori, our objective was to evaluate the influence of procedural technique, difficulty of intravenous access, and operator on the initial success rate. We Annals of Emergency Medicine 3
McCarthy et al
Ultrasonography Versus Landmark for Peripheral Intravenous Cannulation
examined the influence of other factors on the initial success rate, such as body mass index, arm circumference, age, and the presence of different comorbidities; however, the inclusion of these other factors made no clinically or statistically meaningful difference on the influence of the a priori variables of interest. Thus, we present the results based only on the 3 a priori variables of interest: procedural technique, the difficulty of intravenous access, and operator skill. In a prespecified subgroup analysis, we tested for a significant interaction effect between technique and difficulty of access group on the first- and second-attempt success rates with the Breslow-Day test. Within each intravenous access group, we calculated the risk ratio and 95% confidence interval (CI) for the categorical outcomes by study group. We report the medians of the pain rating and procedure duration because both were positively skewed. Within each intravenous access group, we estimated the median difference and 95% CI of the pain rating and procedure duration by treatment group. With the revised sample size, we did not have the power to test for a significant interaction effect between treatment method and operator. Instead, we used the Breslow-Day test to conduct a post hoc subgroup analysis between procedural method and technician skill on the first- and second-attempt success rates. To account for possible correlation of subjects treated by the same technician, we fit a generalized linear mixed
logistic regression model that included fixed effects for treatment group, difficult intravenous access group, and an interaction between the 2 factors, as well as a random effect for technician.21 To examine the influence of technician skill on success, we fit a logistic regression model that included the average number of ultrasonographic peripheral intravenous lines placed during a typical shift (<5 versus 5), treatment method, an interaction of the two, and the fixed effects included in the generalized linear mixed model. RESULTS During the 13-month study period, we screened 1,662 and identified 1,617 eligible patients. The main reasons eligible subjects were not enrolled were they refused (N¼217) or they were classified as having easy access (N¼181) and were not selected during the beginning of the study when we were enrolling a subject with easy access only after 2 subjects with more difficult access (Figure 1). Among the 36 technicians who worked during study enrollment hours, 33 (92%) participated. The participating technicians enrolled a median of 26 subjects (interquartile range 9 to 55). More than three quarters (82%) of the 33 participating technicians reported 1 or more years of experience placing ultrasonographic peripheral intravenous lines. During a typical 12-hour shift, 39% of the technicians reported placing 5 or more ultrasonographic peripheral intravenous lines.
Figure 1. Derivation of the study sample. US, Ultrasound; LM, landmark; tx, treatment.
4 Annals of Emergency Medicine
Volume
-,
no.
-
:
-
2015
McCarthy et al
Ultrasonography Versus Landmark for Peripheral Intravenous Cannulation
Table 1 displays the distribution of patient and clinical characteristics of the 2 treatment groups by difficulty of intravenous access. The mean age of the study population was 45.4 years, 63% were women, and 37% were obese. Approximately three quarters (76%) of the peripheral intravenous lines were placed in the antecubital fossa; an additional 18% were placed in the forearm. The initial success rate was 81% but varied substantially by technique and access difficulty (Table 2, Figure 2). First-attempt success rate for ultrasonographic peripheral intravenous line placement ranged between 82% and 86%, regardless of difficulty of intravenous access. In contrast, the initial landmark success rate varied from 35% to 97% by difficulty of intravenous access. The initial success rate was 48.0 per 100 attempts (95% CI 35.6 to 60.3) higher for patients with difficult access and 10.2 higher per 100 attempts (95% CI 1.7 to 18.7) for those with moderately difficult access by ultrasonography versus landmark. Among patients with easy intravenous access, landmark yielded a 10.6 per 100 attempts higher success rate (95% CI 5.8 to 15.4) compared with ultrasonography. When more skilled versus less skilled
providers used ultrasonography, the initial success rate was higher: 9.4 per 100 attempts with difficult intravenous access (95% CI 2.5 to 16.3), 9.7 per 100 attempts with moderately difficult access (95% CI 2.8 to 16.6), and 7.9 per 100 attempts with easy access (95% CI 2.1 to 13.7) (Figure 2). For the 227 subjects whose catheter failed on the first attempt, we successfully randomized 197 (87%) a second time. Overall, 69% of the 197 randomized catheter insertions were successful on the second attempt. Similar to that of the first attempt, the success rate was 59.9 per 100 attempts higher for patients with difficult access (95% CI 40.7 to 79.2) and 8.8 per 100 attempts higher for those with moderately difficult access (95% CI –11.7 to 0.3); landmark was superior with patients with easy access (31.8 per 100 attempts; 95% CI 5.4 to 58.2) (Table 2, Figure 2). The second-attempt success rates were also higher with ultrasonography among technicians with more ultrasonographic peripheral intravenous skill compared with those with less skill (Figure 2). There was not a significant difference in the complication rate by procedural method (Table 2). For all 3 intravenous
Table 1. Percentage distribution of patient and clinical characteristics by difficulty of intravenous access and treatment assignment of first attempt.* Difficult Access, % Patient and Clinical Characteristics Age, y 18–34 35–54 55 Male sex Line placement attempted before enrollment in study Yes No Body mass index, kg/m2 <18.5 (underweight) 18.5–24.9 (normal weight) 25–29.9 (overweight) 30 (obese) Presence of comorbidities Cancer (past or present) Diabetes End-stage renal disease Intravenous drug use (past or present) Sickle cell disease Present steroid use Frequent blood draws or admissions, or living in a nursing home
Moderately Difficult Access, %
Easy Access, %
Overall, N[1,189, %
Ultrasonography, N[98
LM, N[94
Ultrasonography, N[202
LM, N[199
Ultrasonography, N[305
LM, N[291
33.1 35.5 31.4 37.0
24.5 35.7 39.8 36.7
20.2 39.4 40.4 25.5
29.7 34.7 35.6 33.7
31.7 34.2 34.1 33.2
39.0 36.1 24.9 39.3
37.5 35.0 27.5 43.3
6.8 93.2
20.4 79.6
12.8 87.2
6.4 93.6
9.5 90.5
3.0 97.0
2.7 97.3
2.3 32.6 28.2 36.9
1.0 35.7 18.4 44.9
2.1 28.7 31.9 37.3
1.5 26.2 26.2 46.1
0.6 27.1 27.6 44.7
5.2 35.1 28.5 31.2
1.4 38.5 31.6 28.5
8.2 16.1 4.0 4.0
10.2 25.5 12.2 15.3
11.7 29.8 8.5 11.7
9.9 22.3 5.9 3.5
9.5 16.6 3.0 4.5
4.9 10.5 1.3 1.0
7.9 10.0 1.7 1.0
2.1 4.4 14.3
4.1 5.1 25.5
7.4 9.6 29.8
1.5 4.5 14.4
3.0 4.5 16.6
0.7 1.3 7.5
1.0 5.5 11.0
LM, Landmark. *Treatment assignment according to intention to treat.
Volume
-,
no.
-
:
-
2015
Annals of Emergency Medicine 5
Difficult Access Outcomes First attempt Initial success rate, % Complications, % None Infiltration Arterial puncture Nerve puncture Median duration, s‡ Second attempt Success rate, % Complications, % None Infiltration Arterial puncture Nerve puncture Overall§ Median pain rating
Overall
Ultrasonography
LM
N¼1,189 80.9
N¼98 81.6
N¼94 35.1
95.7 3.6 0.3 0.4 125 N¼197 69.0
95.9 2.1 1.0 1.0 211 N¼45 84.4
94.7 3.2 1.1 1.0 123 N¼28 25.0
94.9 4.6 0.5 0 N¼1,185 2
97.8 0 2.2 0 N¼98 2
96.4 3.4 0 0 N¼94 1
Moderately Difficult
Risk Ratio (95% CI) 2.32 (1.74–3.11) 1.01 0.66 0.91 1.00 88
(0.95–1.08) (0.11–3.84) (0.06–14.33) (0.06–15.76) (52–120)
3.38 (1.76–6.49) 1.01 (0.93–1.10) † † †
1 (0.5–3.0)
Easy Access
Ultrasonography
LM
Risk Ratio (95% CI)
Ultrasonography
LM
Risk Ratio (95% CI)
N¼202 81.2
N¼199 71.4
1.14 (1.02–1.27)
N¼305 85.9
N¼291 96.6
0.89 (0.85–0.94)
94.5 4.0 0.5 1.0 199 N¼36 72.2
93.5 6.5 0 1.0 103 N¼46 63.0
95.8 3.9 0 0.3 159 N¼14 64.3
98.3 1.7 0 0 80 N¼28 96.4
94.4 4.6 0 0 N¼201 3
91.3 8.7 0 0 N¼199 2
85.7 14.3 0 0 N¼302 3
100.0 0 0 0 N¼291 2
1.01 (0.96–1.06) 0.62 (0.26–1.45) †
1.00 (0.14–7.03) 97 (77–116) 1.15 (0.85–1.55) 1.04 (0.92–1.17) 0.53 (0.10–2.73) † †
1 (1.0–3.0)
0.97 (0.95–1.00) 2.29 (0.82–6.43) † †
79 (63–92) 0.67 (0.45–0.99) 0.86 (0.69–1.06) †
Ultrasonography Versus Landmark for Peripheral Intravenous Cannulation
6 Annals of Emergency Medicine Table 2. Primary and secondary outcomes by difficulty of intravenous access group and treatment assignment.*
† †
1 (0–2.0)
*Treatment assignment according to intention to treat. † Relative risk ratio not calculated because of zero cell. ‡ Missing procedure time on 7 subjects. § Missing pain rating on 4 subjects.
Volume -, -
: -
2015
McCarthy et al
no.
McCarthy et al
Ultrasonography Versus Landmark for Peripheral Intravenous Cannulation
The initial technician (63%) or a different technician (27%) ultimately obtained access in all but 6 cases. LIMITATIONS The results of this study must be considered in light of the following limitations. The main study limitation is generalizability. The research assistants were not able to enroll every consecutively eligible patient. The research assistants were enrolling and collecting data for 9 other clinical research studies being conducted in our ED at the same time as this one, dealing simultaneously with screening, enrollment, and data collection across the 10 studies. Because we successfully enrolled patients with various levels of intravenous access and we involved 33 technicians, the study sample should be representative of our ED population. Furthermore, randomization ensured that the subjects enrolled in the 2 treatment groups were similar in regard to all baseline characteristics. We also believe that our results should generalize to other experienced operators, to children,8,13 and to other settings (ie, other EDs and ICUs) that treat acutely ill patients. Other study limitations are that we evaluated only 1 type of ultrasonographic method, we did not time the entire procedure if the first attempt failed, and only the statisticians were blinded to the treatment method.
Figure 2. The odds ratio and 95% CI of first- and secondattempt success for ultrasonography versus landmark for each intravenous access group, estimated from the generalized linear mixed logistic regression model. Also shown are estimates of odds ratio and 95% CI of first- and second-attempt success for ultrasonography versus landmark by intravenous access group and operator skill, using logistic regression.
access groups, the procedure took approximately twice as long with ultrasonography compared with landmark (range of median difference 79 to 97 seconds longer with ultrasonography). The median pain rating associated with the peripheral intravenous procedure was 1 point higher on the 11-point scale when performed with ultrasonography compared with landmark, regardless of difficulty of intravenous access (Table 2). Of the 1,189 enrolled subjects, 64 (5.4%) had 3 or more peripheral intravenous cannulation attempts (maximum of 7). In the majority of these cases (64%), ultrasonography was the final method used. One subject had a central line placed, another had an external jugular line, 2 had no lines placed, and 1 received an arterial or venous stick instead. Volume
-,
no.
-
:
-
2015
DISCUSSION This randomized controlled trial demonstrated that the success rate of the procedural technique used to obtain peripheral intravenous access in the upper extremity of adult ED patients depends on the difficulty of intravenous access. When a technician can easily see or palpate a vein, landmark is superior in terms of initial success rate, pain mitigation, and procedural duration. Among patients with easy access, landmark resulted in an estimated 11 additional first-attempt successes per 100 patients treated. In contrast, ultrasonography is more successful when a vein is not clearly visible or palpable. We estimate 10 and 48 additional initial successes per 100 attempts for patients with moderately difficult and difficult access, respectively, who are treated with ultrasonography. There were no clinically meaningful differences in the complication rate between the 2 methods. Our results suggest that the benefit of success is modest for either approach among patients with easy or moderately difficult intravenous access but large when ultrasonography is used with patients with difficult access. There is no widely accepted definition of difficult access. Past studies have used 1 or more failed attempts, a history of failed intravenous access, clinician suspicion of difficult Annals of Emergency Medicine 7
McCarthy et al
Ultrasonography Versus Landmark for Peripheral Intravenous Cannulation
access, or absence of a visible or palpable vein.7-14,22 We prospectively characterized intravenous access by using a combination of objective and subjective criteria. We found clinically meaningful differences in the initial success rate among our 3 intravenous access groups. Even among the subgroup of patients for whom the first attempt resulted in failure, the pattern of second-attempt success was similar to that of the first attempt: landmark was superior for patients with easy access; ultrasonography, for those with difficult and moderately difficult access. Previous studies have compared landmark and ultrasonographic peripheral intravenous cannulation among patients with difficult intravenous access.7-14,22 It is difficult to compare the results of this study to those of previous studies because of numerous methodological differences. Four of the 9 previous studies were performed in more controlled settings (ie, the operating room or the ICU).7,12,13,22 Of the 5 studies conducted in the ED, 2 comprised pediatric patients younger than 10 years.8,9 Only 2 studies used the same definition of difficult intravenous access as this study.12,13 In 7 of the 9 studies, the operators received limited training in ultrasonographic peripheral intravenous cannulation before enrollment started,7-9,14 or no details were provided on the training received.12,13,22 Of the 9 studies conducted to date, only 3 reported the initial success rate by procedural approach.7,8,13 If initial success rate, overall success rate, average number of attempts, or frequency of physician rescue attempt is considered, 7 of the 9 studies favored ultrasonography over landmark for patients with difficult access.8,9,11-14,22 To our knowledge, this is the first study to demonstrate that ultrasonography is superior among patients with moderately difficult intravenous access, those for whom the operator can see or palpate a vein but anticipates some difficulty using a landmark approach. The patients classified as having moderately difficult access enjoyed a 10% higher initial and second-attempt success rate with ultrasonography compared with landmark. Typically, peripheral intravenous cannulation is the primary responsibility of nursing.2,14,23 In this study, ED technicians performed all peripheral intravenous catheterizations. However, the technicians achieved success rates similar to those of nurse or physician operators.15,16 Regardless of peripheral intravenous access difficulty, the ultrasonographic success rate was approximately 80%. In contrast, the technicians attained a much higher landmark success rate (ie, >95%) when peripheral intravenous access was easy. Since 2008, the study ED has relied on a program that trains technicians to perform ultrasonographic peripheral intravenous line placement. Consequently, 8 Annals of Emergency Medicine
there is always 1 or more providers working in the ED at all times who can place a peripheral intravenous line with either method. As a result of this program, our ED has substantially reduced its use of central venous line placement.24 This study used a large group of experienced technicians, so the results are not confounded by a learning curve for most of the operators. More than 80% of the technicians had 1 or more years of experience with ultrasonographic peripheral intravenous line placement at study onset. We believe these results reflect operator skill and are not biased by differential expertise because operators prefer to use landmark with patients who have easy access.25 We randomized the treatment approach by operator and difficulty of intravenous access, so each technician placed an approximately equal number of catheters with or without ultrasonography in each intravenous access group. Technicians who reported placing more ultrasonographic peripheral intravenous lines during a typical shift had higher success rates with ultrasonography compared with those who reported placing fewer lines. When more skilled providers used ultrasonography, our results estimated an additional 8 to 10 initial successes per 100 patients, depending on intravenous access difficulty. The strengths of our study include a design that randomized patients within strata defined by operator and difficulty of intravenous access to control for these important sources of variation, a second randomized treatment assignment if the first attempt failed, a large sample of patients with various degrees of difficulty of peripheral intravenous access, and a sizeable cohort of technicians experienced in peripheral intravenous catheterization with both approaches. Within this context, our results suggest that landmark is the superior method for patients with easy intravenous access, whereas ultrasonography is superior for those with moderately difficult or difficult intravenous access. The authors acknowledge the valuable contributions of the 33 technicians who participated in the study. Supervising editor: Donald M. Yealy, MD Author affiliations: From the Department of Health Policy and Management (McCarthy) and Department of Emergency Medicine (McCarthy, Shokoohi, Boniface, Shesser), George Washington University Hospital, Washington, DC (Lim, Eggelton, Lowey); and the Department of Biostatistics, Johns Hopkins University Bloomberg School of Public Health, Baltimore, MD (Li, Zeger). Author contributions: MLM, HS, KSB, RE, and SLZ conceived the study and designed the trial. MLM and RS obtained research Volume
-,
no.
-
:
-
2015
McCarthy et al
Ultrasonography Versus Landmark for Peripheral Intravenous Cannulation
funding. MLM, HS, RE, AL, and KL supervised the conduct of the trial and data collection and undertook recruitment of patients. XL and SLZ provided statistical advice on study design and analyzed the data. MLM managed the data, including quality control, and drafted the article, and all authors contributed substantially to its revision. MLM takes responsibility for the paper as a whole. Funding and support: By Annals policy, all authors are required to disclose any and all commercial, financial, and other relationships in any way related to the subject of this article as per ICMJE conflict of interest guidelines (see www.icmje.org). The authors have stated that no such relationships exist and provided the following details: This study was supported in part by award K01HS017957 from the Agency for Healthcare Research and Quality. For every 3 subjects enrolled, the authors gave the technician a $5 gift card as a token of appreciation for assisting with study enrollment and data collection in addition to his or her clinical responsibilities. Publication dates: Received for publication June 25, 2015. Revisions received August 12, 2015, and August 24, 2015. Accepted for publication September 2, 2015. Clinical trial registration number: NCT01859559 The sponsor was not involved in the design and conduct of the study; collection, management, analysis, and interpretation of the data; or preparation, review, or approval of the article. The content is solely the responsibility of the authors and does not represent the official views of the Agency for Healthcare Research and Quality.
REFERENCES 1. Alexandrou E. The One Million Global Catheters PIVC Worldwide Prevalence study. Br J Nurs. 2014;23:S16-S17. 2. Sabri A, Szalas J, Holmes KS, et al. Failed attempts and improvement strategies in peripheral intravenous catheterization. Biomed Mater Eng. 2013;23:93-108. 3. Sebbane M, Claret PG, Lefevbre S, et al. Predicting peripheral venous access difficulty in the emergency department using body mass index and a clinical evaluation of venous accessibility. J Emerg Med. 2013;44:299-305. 4. Crowley M, Brim C, Proehl J, et al. Emergency nursing resource: difficult intravenous access. 2011. Emergency Nurses Association. 1-21-2012. Available at: https://www.ena.org/practice-research/research/CPG/ Documents/DifficultIVAccessCPG.pdf. Accessed September 2015. 5. Hind D, Calvert N, McWilliams R, et al. Ultrasonic locating devices for central venous cannulation: meta-analysis. BMJ. 2003;327:1-7. 6. Calvert N, Hind D, McWilliams R, et al. Ultrasound for central venous cannulation: economic evaluation of cost-effectiveness. Anaesthesia. 2004;59:1116-1120. 7. Aponte H, Acosta S, Rigamonti D, et al. The use of ultasound for placement of intravenous catheters. AANA J. 2007;75:212-216. 8. Bair AE, Rose JS, Vance CW, et al. Ultrasound-assisted peripheral venous access in young children: a randomized controlled trial and pilot feasibility study. West J Emerg Med. 2008;9:219-224.
Volume
-,
no.
-
:
-
2015
9. Doniger SJ, Ishimine P, Fox JC, et al. Randomized controlled trial of ultrasound-guided peripheral intravenous catheter placement versus traditional techniques in difficult-access pediatric patients. Pediatr Emerg Care. 2009;25:154-159. 10. Stein J, George B, River G, et al. Ultrasonographically guided peripheral intravenous cannulation in emergency department patients with difficult intravenous access: a randomized trial. Ann Emerg Med. 2009;54:33-40. 11. River G, Hebig M, McAlpine I, et al. Nurse-operated ultrasound for difficult intravenous access: a randomized trial. Ann Emerg Med. 2009;54:S87. 12. Kerfone T, Petitpas F, Frasca D, et al. Ultrasound-guided peripheral venous access in severely ill patients with suspected difficult vascular puncture. Chest. 2012;141:279-280. 13. Benkhadra M, Collignon M, Fournel I, et al. Ultrasound guidance allows faster peripheral IV cannulation in children under 3 years of age with difficult venous access: a prospective randomized study. Pediatr Anesth. 2012;22:449-454. 14. Weiner SG, Sarff AR, Esener DE, et al. Single-operator ultrasoundguided intravenous line placement by emergency nurses reduces the need for physician intervention in patients with difficult-to-establish intravenous access. J Emerg Med. 2013;44:653-660. 15. Heinrichs J, Fritze Z, Vandermeer B, et al. Ultrasonographically guided peripheral intravenous cannulation of children and adults: a systematic review and meta-analysis. Ann Intern Med. 2013;61: 444-454. 16. Egan G, Healy D, O’Neill H, et al. Ultrasound guidance for difficult peripheral venous access: systematic review and meta-analysis. Emerg Med J. 2013;30:521-526. 17. Liu YT, Alsaawi A, Bjornsson HM. Ultrasound-guided peripheral venous access: a systematic review of randomized-controlled trials. Eur J Emerg Med. 2014;21:18-23. 18. Schoenfeld E, Shokoohi H, Boniface K. ED technicians can succesfully place ultrasound-guided intravenous catheters in patients with poor vascular access. Am J Emerg Med. 2011;29:496-501. 19. Wuerz RC, Milne LW, Eitel DR, et al. Reliability and validity of a new five-level triage instrument. Acad Emerg Med. 2000;7:236-242. 20. Harris PA, Taylor R, Thielke R, et al. Research Electronic Data Capture (REDCap)—a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009;42:377-381. 21. McCulloch CE, Searle SR, Neuhaus JM. Generalized, Linear and Mixed Models. 2nd ed. Hoboken, NJ: John Wiley & Sons; 2012. 22. Pappas NL, Michaud TE, Wolbers RM, et al. Ultrasound guidance as a rescue technique for peripheral intravenous cannulation. C104-1878, 1–16. 2006. Bethesda, Maryland, Uniformed Services University of the Health Sciences. 9-11-2012. Available at: https://www.ena.org/ practice-research/research/CPG/Documents/DifficultIVAccessCPG. pdf. Accessed September 2015. 23. Jacobson AF, Winslow EH. Variables influencing intravenous catheter insertion difficulty and failure: an analysis of 339 intravenous catheter insertions. Heart Lung. 2005;34:345-359. 24. Shokoohi H, Boniface K, McCarthy M, et al. Ultrasound-guided peripheral intravenous access program is associated with a marked reduction in central venous catheter use in noncritically ill emergency department patients. Ann Emerg Med. 2013;61:198-203. 25. Devereaux PJ, Bhandri M, Clarke M, et al. Need for expertise based randomized controlled trials. BMJ. 2005;330:1-5.
Annals of Emergency Medicine 9
Ultrasonography Versus Landmark for Peripheral Intravenous Cannulation: A Randomized Controlled Trial Melissa L. McCarthy, ScD*; Hamid Shokoohi, MD, MPH; Keith S. Boniface, MD, RDMS; Russell Eggelton, EMT-B, EDT; Andrew Lowey, BS, EMT-P; Kelvin Lim, BS; Robert Shesser, MD; Ximin Li, MS; Scott L. Zeger, PhD *Corresponding Author. E-mail: [email protected].
Study objective: Randomized controlled trials report inconsistent findings when comparing the initial success rate of peripheral intravenous cannulation using landmark versus ultrasonography for patients with difficult venous access. We sought to determine which method is superior for patients with varying levels of intravenous access difficulty. Methods: We conducted a 2-group, parallel, randomized, controlled trial and randomly allocated 1,189 adult emergency department (ED) patients to landmark or ultrasonography, stratified by difficulty of access and operator. ED technicians performed the peripheral intravenous cannulations. Before randomization, technicians classified subjects as difficult, moderately difficult, or easy access according to visible or palpable veins and perception of difficulty with a landmark approach. If the first attempt failed, we randomized subjects a second time. We compared the initial and second-attempt success rates by procedural approach and difficulty of intravenous access, using a generalized linear mixed regression model, adjusted for operator. Results: The 33 participating technicians enrolled a median of 26 subjects (interquartile range 9 to 55). The initial success rate was 81% but varied significantly by technique and difficulty of access. The initial success rate by ultrasonography was higher than landmark for patients with difficult access (48.0 more successes per 100 tries; 95% confidence interval [CI] 35.6 to 60.3) or moderately difficult access (10. 2 more successes per 100 tries; 95% CI 1.7 to 18.7). Among patients with easy access, landmark yielded a higher success rate (10.6 more successes per 100 tries; 95% CI 5.8 to 15.4). The pattern of second-attempt success rates was similar. Conclusion: Ultrasonographic peripheral intravenous cannulation is advantageous among patients with difficult or moderately difficult intravenous access but is disadvantageous among patients anticipated to have easy access. [Ann Emerg Med. 2015;-:1-9.] Please see page XX for the Editor’s Capsule Summary of this article. 0196-0644/$-see front matter Copyright © 2015 by the American College of Emergency Physicians. http://dx.doi.org/10.1016/j.annemergmed.2015.09.009
INTRODUCTION Intravenous cannulation of a peripheral vein is the most common invasive procedure performed in hospitals worldwide.1 The traditional landmark approach involves visual inspection and palpation of the extremity to locate a vein, followed by needle puncture and catheter insertion. The landmark failure rate on first attempt ranges from 12% to 26% among adults and is twice as high in difficultaccess patients.2-4 When multiple cannulation attempts are necessary, patients experience more pain and treatment delays. Ultrasonography increases the success rate of central venous line placement, especially at the internal jugular site, but studies evaluating ultrasonographic peripheral intravenous line placement have reported inconsistent results.5,6 Eight randomized controlled trials have Volume
-,
no.
-
:
-
2015
compared the initial or final peripheral intravenous line placement success rate between landmark and ultrasonography in difficult-access patients.7-14 The majority of the trials9,11-14 favored ultrasonography and reported clinically meaningful differences of 10% or greater in initial13 or overall success rate.9,11,12,14 However, each study’s results were based on a small sample size (range of 35 to 60 total subjects). Meta-analyses of the trials’ results have reported conflicting findings because of heterogeneous patient populations, varying definitions of difficult access, and operator experience.15-17 This randomized controlled trial of 1,189 adult emergency department (ED) patients seeks to determine whether the initial peripheral intravenous line placement success rate is higher with landmark or ultrasonography among patients with difficult, moderately difficult, or easy venous Annals of Emergency Medicine 1
McCarthy et al
Ultrasonography Versus Landmark for Peripheral Intravenous Cannulation
Editor’s Capsule Summary
What is already known on this topic Intravenous access is important but without technical aids can sometimes be difficult. Ultrasonography offers an alternative to standard insertion methods. What question this study addressed Does ultrasonographically guided intravenous insertion aid in all patients or just those deemed in advance to present more challenging access? What this study adds to our knowledge Using a randomized, controlled assignment in 1,189 patients, trained technicians had higher success with landmark approaches in those deemed to have easier access, but ultrasonographic intravenous insertion attempts produced higher success in the moderateand highest-difficulty groups. How this is relevant to clinical practice Selective use of ultrasonographic intravenous placement approaches targeting more difficult patients is pragmatic and evidence based.
access by randomly assigning the procedural method within each difficulty of intravenous access group and operator. We originally hypothesized that the initial success rate would be at least 5% higher when ultrasonography was used among patients with difficult or moderately difficult peripheral intravenous access but no different (ie, <5%) among patients with easy access. MATERIALS AND METHODS Study Design and Setting We conducted a randomized controlled trial with a 2-group parallel design at an academic ED in Washington, DC. The university’s institutional review board approved the study protocol and we registered the trial at clinicaltrials.gov. At this ED, peripheral intravenous cannulation is the primary responsibility of technicians. The ones who work in our ED have previous training as emergency medicine technicians, certified medical assistants, certified nursing assistants, or are nursing students. Since 2008, our ED has trained the technicians working in the ED to perform ultrasonographic peripheral intravenous catheterization. During the first 3 months of employment, each technician completes instruction in ultrasonographic peripheral intravenous cannulation that includes a 2-hour lecture and hands-on demonstration of 2 Annals of Emergency Medicine
the skill. On class completion, under preceptor supervision the technician must perform 10 successful ultrasonographic peripheral intravenous procedures.18 ED nurses at our institution are not trained in ultrasonographic peripheral intravenous cannulation. Their role in peripheral intravenous cannulation is limited to cases in which the treating technician is busy and the nurse believes he or she can successfully place the peripheral intravenous catheter with landmark. Selection of Participants We enrolled adult ED patients deemed capable of providing informed consent who required a peripheral intravenous line in their upper extremity, but not hands. We excluded high-acuity patients (all triage level 1 patients plus triage level 2 for whom the care team deemed it unsafe to delay peripheral intravenous placement).19 If a physician ordered a diagnostic test or treatment that required peripheral intravenous access, the research assistant introduced himself or herself to the technician and patient, explained the study, and asked the patient to participate. Patients who agreed signed a written consent form. Before beginning study enrollment, we also consented the technicians so that we could document the outcomes by technician. Enrollment occurred from August 2013 to August 2014 between 9 AM and 9 PM. We anticipated there would be more eligible patients with easy access than moderately difficult or difficult-access patients, so during the first 4 months of enrollment, we instructed the research assistants to enroll a patient with easy access only after 2 patients with more difficult access were enrolled to ensure that target enrollment in the easy-access group was not achieved much earlier than that of the other 2 access groups. We used SAS software (version 9.3; SAS Institute, Inc, Cary, NC) to generate a randomization strategy that randomly allocated the procedure technique to patients with equal probability to landmark or ultrasonography, stratified by technician and difficulty of intravenous access. We blocked randomization into groups of 6 to distribute the treatment assignment equally by stratum. We uploaded the randomization strategy into REDCap (version 6.5.12; Nashville, TN), a flexible, Web-based, secure data collection system; the research assistants used REDCap for randomization and real-time data capture.20 If the first attempt failed, we randomized the subject a second time, using the same randomization strategy. This gave us the opportunity to compare the 2 treatment methods among a subgroup of the subjects who were similar in all baseline characteristics (because of randomization) and had a failed first attempt. If the second attempt failed, the technique used to secure intravenous access was at the technician’s discretion. Volume
-,
no.
-
:
-
2015
McCarthy et al
Ultrasonography Versus Landmark for Peripheral Intravenous Cannulation
For subjects randomly assigned to landmark, the technician secured a tourniquet around the chosen arm and palpated to identify an acceptable vein. The technician prepped the site with a chlorhexidine swab, inserted the needle, and advanced the catheter into the vein. The technician attached a saline solution lock adapter and used 10 mL of normal saline solution to flush the lock. While flushing the saline solution lock, the technician visualized and palpated the site to confirm placement. For subjects randomly assigned to ultrasonography, the technician used an ultrasonography machine (Sonosite M-Turbo or Zonare ultra) to visualize an acceptable vein. The technician prepped the site with a chlorhexidine swab and sterile lubrication gel. Next, the technician used a dynamic, single-operator technique that involved holding the probe in the nondominant hand and concurrently inserting the needle and advancing the catheter into the vein with the other hand, using the ultrasonography machine to visualize and guide the needle. After insertion, the technician tested the catheter in the same manner as landmark. Before randomization, the technician placed a tourniquet around the brachial region of each subject’s upper extremities and visually inspected and palpated the arms to evaluate suitability of the vessels for intravenous access. If the technician could not see or palpate a vein in either extremity, we classified the subject as having difficult access. If the technician could see or palpate at least 1 vein but anticipated difficulty using landmark, we classified the subject as having moderately difficult access. If the technician could see or palpate at least 1 vein and thought intravenous access would be easy with landmark, we classified the subject as having easy access. The research assistants briefly interviewed subjects to obtain their weight and height and ascertain whether they had any health conditions or were receiving treatments previously associated with difficult intravenous access.2,4 The research assistants observed the peripheral intravenous procedure and recorded which arm and vein was used, the catheter gauge and length, the technique used, whether the attempt was successful, and whether any complications occurred. If more than 2 attempts were required, the research assistant documented the total number of attempts, the final method used to gain intravenous access, and who obtained it. Technicians who agreed to participate in the study completed a brief survey before their first subject enrollment. We asked them to report the number of years of experience they had placing peripheral intravenous lines and years placing ultrasonographic peripheral intravenous lines. We also asked each technician to estimate how many peripheral intravenous lines and ultrasonographic Volume
-,
no.
-
:
-
2015
peripheral intravenous lines they placed on average during a typical 12-hour shift. Outcome Measures The main outcomes were success or failure on the initial and second attempts. Secondary outcomes included the occurrence of a complication, patient-reported pain associated with the procedure (0¼no pain at all to 10¼extremely painful), and duration of the first attempt, defined as the time from onset of tourniquet placement until the saline solution had been flushed through the peripheral intravenous line. We defined an intravenous attempt as 1 percutaneous needle puncture, regardless of the amount of subcutaneous exploration from the single puncture site. We considered a cannulation successful if the technician was able to infuse fluid without infiltration. If it was unsuccessful, the research assistant also recorded whether infiltration and arterial or nerve puncture occurred. Primary Data Analysis We conducted separate analyses for the first and second attempts. When calculating the target sample size, we assumed 80% power and initial success rates according to a previous study of 65% for difficult-access patients, 74% for moderately difficult ones, and 85% for those with easy access, using a landmark approach.3 To detect a 5% difference in the initial success rate for each intravenous access group, we estimated that the sample size required was 6,314 subjects (1,377 subjects with difficult access, 1,094 with moderately difficult access, and 686 with easy access for each treatment group). This proved too large, given available resources. We revised our target sample size after 4 months of study enrollment. We assumed the same initial success rates with landmark as we did when we estimated the original sample size estimates3 for each treatment group and the same level of power (ie, 80%). The only change we made when revising the sample size estimates was to increase the difference we were able to detect from 5% to 10% or greater in the first attempt between the 2 techniques for the groups with easy access and moderately difficult access and to 20% or greater for the group with difficult access. The revised target sample size was 948 (73, 260, and 141 subjects, respectively, in each treatment group for difficult, moderately difficult, and easy access). We analyzed all data according to intention to treat; the statisticians were blinded to procedural assignment (X.L. and S.L.Z.). A priori, our objective was to evaluate the influence of procedural technique, difficulty of intravenous access, and operator on the initial success rate. We Annals of Emergency Medicine 3
McCarthy et al
Ultrasonography Versus Landmark for Peripheral Intravenous Cannulation
examined the influence of other factors on the initial success rate, such as body mass index, arm circumference, age, and the presence of different comorbidities; however, the inclusion of these other factors made no clinically or statistically meaningful difference on the influence of the a priori variables of interest. Thus, we present the results based only on the 3 a priori variables of interest: procedural technique, the difficulty of intravenous access, and operator skill. In a prespecified subgroup analysis, we tested for a significant interaction effect between technique and difficulty of access group on the first- and second-attempt success rates with the Breslow-Day test. Within each intravenous access group, we calculated the risk ratio and 95% confidence interval (CI) for the categorical outcomes by study group. We report the medians of the pain rating and procedure duration because both were positively skewed. Within each intravenous access group, we estimated the median difference and 95% CI of the pain rating and procedure duration by treatment group. With the revised sample size, we did not have the power to test for a significant interaction effect between treatment method and operator. Instead, we used the Breslow-Day test to conduct a post hoc subgroup analysis between procedural method and technician skill on the first- and second-attempt success rates. To account for possible correlation of subjects treated by the same technician, we fit a generalized linear mixed
logistic regression model that included fixed effects for treatment group, difficult intravenous access group, and an interaction between the 2 factors, as well as a random effect for technician.21 To examine the influence of technician skill on success, we fit a logistic regression model that included the average number of ultrasonographic peripheral intravenous lines placed during a typical shift (<5 versus 5), treatment method, an interaction of the two, and the fixed effects included in the generalized linear mixed model. RESULTS During the 13-month study period, we screened 1,662 and identified 1,617 eligible patients. The main reasons eligible subjects were not enrolled were they refused (N¼217) or they were classified as having easy access (N¼181) and were not selected during the beginning of the study when we were enrolling a subject with easy access only after 2 subjects with more difficult access (Figure 1). Among the 36 technicians who worked during study enrollment hours, 33 (92%) participated. The participating technicians enrolled a median of 26 subjects (interquartile range 9 to 55). More than three quarters (82%) of the 33 participating technicians reported 1 or more years of experience placing ultrasonographic peripheral intravenous lines. During a typical 12-hour shift, 39% of the technicians reported placing 5 or more ultrasonographic peripheral intravenous lines.
Figure 1. Derivation of the study sample. US, Ultrasound; LM, landmark; tx, treatment.
4 Annals of Emergency Medicine
Volume
-,
no.
-
:
-
2015
McCarthy et al
Ultrasonography Versus Landmark for Peripheral Intravenous Cannulation
Table 1 displays the distribution of patient and clinical characteristics of the 2 treatment groups by difficulty of intravenous access. The mean age of the study population was 45.4 years, 63% were women, and 37% were obese. Approximately three quarters (76%) of the peripheral intravenous lines were placed in the antecubital fossa; an additional 18% were placed in the forearm. The initial success rate was 81% but varied substantially by technique and access difficulty (Table 2, Figure 2). First-attempt success rate for ultrasonographic peripheral intravenous line placement ranged between 82% and 86%, regardless of difficulty of intravenous access. In contrast, the initial landmark success rate varied from 35% to 97% by difficulty of intravenous access. The initial success rate was 48.0 per 100 attempts (95% CI 35.6 to 60.3) higher for patients with difficult access and 10.2 higher per 100 attempts (95% CI 1.7 to 18.7) for those with moderately difficult access by ultrasonography versus landmark. Among patients with easy intravenous access, landmark yielded a 10.6 per 100 attempts higher success rate (95% CI 5.8 to 15.4) compared with ultrasonography. When more skilled versus less skilled
providers used ultrasonography, the initial success rate was higher: 9.4 per 100 attempts with difficult intravenous access (95% CI 2.5 to 16.3), 9.7 per 100 attempts with moderately difficult access (95% CI 2.8 to 16.6), and 7.9 per 100 attempts with easy access (95% CI 2.1 to 13.7) (Figure 2). For the 227 subjects whose catheter failed on the first attempt, we successfully randomized 197 (87%) a second time. Overall, 69% of the 197 randomized catheter insertions were successful on the second attempt. Similar to that of the first attempt, the success rate was 59.9 per 100 attempts higher for patients with difficult access (95% CI 40.7 to 79.2) and 8.8 per 100 attempts higher for those with moderately difficult access (95% CI –11.7 to 0.3); landmark was superior with patients with easy access (31.8 per 100 attempts; 95% CI 5.4 to 58.2) (Table 2, Figure 2). The second-attempt success rates were also higher with ultrasonography among technicians with more ultrasonographic peripheral intravenous skill compared with those with less skill (Figure 2). There was not a significant difference in the complication rate by procedural method (Table 2). For all 3 intravenous
Table 1. Percentage distribution of patient and clinical characteristics by difficulty of intravenous access and treatment assignment of first attempt.* Difficult Access, % Patient and Clinical Characteristics Age, y 18–34 35–54 55 Male sex Line placement attempted before enrollment in study Yes No Body mass index, kg/m2 <18.5 (underweight) 18.5–24.9 (normal weight) 25–29.9 (overweight) 30 (obese) Presence of comorbidities Cancer (past or present) Diabetes End-stage renal disease Intravenous drug use (past or present) Sickle cell disease Present steroid use Frequent blood draws or admissions, or living in a nursing home
Moderately Difficult Access, %
Easy Access, %
Overall, N[1,189, %
Ultrasonography, N[98
LM, N[94
Ultrasonography, N[202
LM, N[199
Ultrasonography, N[305
LM, N[291
33.1 35.5 31.4 37.0
24.5 35.7 39.8 36.7
20.2 39.4 40.4 25.5
29.7 34.7 35.6 33.7
31.7 34.2 34.1 33.2
39.0 36.1 24.9 39.3
37.5 35.0 27.5 43.3
6.8 93.2
20.4 79.6
12.8 87.2
6.4 93.6
9.5 90.5
3.0 97.0
2.7 97.3
2.3 32.6 28.2 36.9
1.0 35.7 18.4 44.9
2.1 28.7 31.9 37.3
1.5 26.2 26.2 46.1
0.6 27.1 27.6 44.7
5.2 35.1 28.5 31.2
1.4 38.5 31.6 28.5
8.2 16.1 4.0 4.0
10.2 25.5 12.2 15.3
11.7 29.8 8.5 11.7
9.9 22.3 5.9 3.5
9.5 16.6 3.0 4.5
4.9 10.5 1.3 1.0
7.9 10.0 1.7 1.0
2.1 4.4 14.3
4.1 5.1 25.5
7.4 9.6 29.8
1.5 4.5 14.4
3.0 4.5 16.6
0.7 1.3 7.5
1.0 5.5 11.0
LM, Landmark. *Treatment assignment according to intention to treat.
Volume
-,
no.
-
:
-
2015
Annals of Emergency Medicine 5
Difficult Access Outcomes First attempt Initial success rate, % Complications, % None Infiltration Arterial puncture Nerve puncture Median duration, s‡ Second attempt Success rate, % Complications, % None Infiltration Arterial puncture Nerve puncture Overall§ Median pain rating
Overall
Ultrasonography
LM
N¼1,189 80.9
N¼98 81.6
N¼94 35.1
95.7 3.6 0.3 0.4 125 N¼197 69.0
95.9 2.1 1.0 1.0 211 N¼45 84.4
94.7 3.2 1.1 1.0 123 N¼28 25.0
94.9 4.6 0.5 0 N¼1,185 2
97.8 0 2.2 0 N¼98 2
96.4 3.4 0 0 N¼94 1
Moderately Difficult
Risk Ratio (95% CI) 2.32 (1.74–3.11) 1.01 0.66 0.91 1.00 88
(0.95–1.08) (0.11–3.84) (0.06–14.33) (0.06–15.76) (52–120)
3.38 (1.76–6.49) 1.01 (0.93–1.10) † † †
1 (0.5–3.0)
Easy Access
Ultrasonography
LM
Risk Ratio (95% CI)
Ultrasonography
LM
Risk Ratio (95% CI)
N¼202 81.2
N¼199 71.4
1.14 (1.02–1.27)
N¼305 85.9
N¼291 96.6
0.89 (0.85–0.94)
94.5 4.0 0.5 1.0 199 N¼36 72.2
93.5 6.5 0 1.0 103 N¼46 63.0
95.8 3.9 0 0.3 159 N¼14 64.3
98.3 1.7 0 0 80 N¼28 96.4
94.4 4.6 0 0 N¼201 3
91.3 8.7 0 0 N¼199 2
85.7 14.3 0 0 N¼302 3
100.0 0 0 0 N¼291 2
1.01 (0.96–1.06) 0.62 (0.26–1.45) †
1.00 (0.14–7.03) 97 (77–116) 1.15 (0.85–1.55) 1.04 (0.92–1.17) 0.53 (0.10–2.73) † †
1 (1.0–3.0)
0.97 (0.95–1.00) 2.29 (0.82–6.43) † †
79 (63–92) 0.67 (0.45–0.99) 0.86 (0.69–1.06) †
Ultrasonography Versus Landmark for Peripheral Intravenous Cannulation
6 Annals of Emergency Medicine Table 2. Primary and secondary outcomes by difficulty of intravenous access group and treatment assignment.*
† †
1 (0–2.0)
*Treatment assignment according to intention to treat. † Relative risk ratio not calculated because of zero cell. ‡ Missing procedure time on 7 subjects. § Missing pain rating on 4 subjects.
Volume -, -
: -
2015
McCarthy et al
no.
McCarthy et al
Ultrasonography Versus Landmark for Peripheral Intravenous Cannulation
The initial technician (63%) or a different technician (27%) ultimately obtained access in all but 6 cases. LIMITATIONS The results of this study must be considered in light of the following limitations. The main study limitation is generalizability. The research assistants were not able to enroll every consecutively eligible patient. The research assistants were enrolling and collecting data for 9 other clinical research studies being conducted in our ED at the same time as this one, dealing simultaneously with screening, enrollment, and data collection across the 10 studies. Because we successfully enrolled patients with various levels of intravenous access and we involved 33 technicians, the study sample should be representative of our ED population. Furthermore, randomization ensured that the subjects enrolled in the 2 treatment groups were similar in regard to all baseline characteristics. We also believe that our results should generalize to other experienced operators, to children,8,13 and to other settings (ie, other EDs and ICUs) that treat acutely ill patients. Other study limitations are that we evaluated only 1 type of ultrasonographic method, we did not time the entire procedure if the first attempt failed, and only the statisticians were blinded to the treatment method.
Figure 2. The odds ratio and 95% CI of first- and secondattempt success for ultrasonography versus landmark for each intravenous access group, estimated from the generalized linear mixed logistic regression model. Also shown are estimates of odds ratio and 95% CI of first- and second-attempt success for ultrasonography versus landmark by intravenous access group and operator skill, using logistic regression.
access groups, the procedure took approximately twice as long with ultrasonography compared with landmark (range of median difference 79 to 97 seconds longer with ultrasonography). The median pain rating associated with the peripheral intravenous procedure was 1 point higher on the 11-point scale when performed with ultrasonography compared with landmark, regardless of difficulty of intravenous access (Table 2). Of the 1,189 enrolled subjects, 64 (5.4%) had 3 or more peripheral intravenous cannulation attempts (maximum of 7). In the majority of these cases (64%), ultrasonography was the final method used. One subject had a central line placed, another had an external jugular line, 2 had no lines placed, and 1 received an arterial or venous stick instead. Volume
-,
no.
-
:
-
2015
DISCUSSION This randomized controlled trial demonstrated that the success rate of the procedural technique used to obtain peripheral intravenous access in the upper extremity of adult ED patients depends on the difficulty of intravenous access. When a technician can easily see or palpate a vein, landmark is superior in terms of initial success rate, pain mitigation, and procedural duration. Among patients with easy access, landmark resulted in an estimated 11 additional first-attempt successes per 100 patients treated. In contrast, ultrasonography is more successful when a vein is not clearly visible or palpable. We estimate 10 and 48 additional initial successes per 100 attempts for patients with moderately difficult and difficult access, respectively, who are treated with ultrasonography. There were no clinically meaningful differences in the complication rate between the 2 methods. Our results suggest that the benefit of success is modest for either approach among patients with easy or moderately difficult intravenous access but large when ultrasonography is used with patients with difficult access. There is no widely accepted definition of difficult access. Past studies have used 1 or more failed attempts, a history of failed intravenous access, clinician suspicion of difficult Annals of Emergency Medicine 7
McCarthy et al
Ultrasonography Versus Landmark for Peripheral Intravenous Cannulation
access, or absence of a visible or palpable vein.7-14,22 We prospectively characterized intravenous access by using a combination of objective and subjective criteria. We found clinically meaningful differences in the initial success rate among our 3 intravenous access groups. Even among the subgroup of patients for whom the first attempt resulted in failure, the pattern of second-attempt success was similar to that of the first attempt: landmark was superior for patients with easy access; ultrasonography, for those with difficult and moderately difficult access. Previous studies have compared landmark and ultrasonographic peripheral intravenous cannulation among patients with difficult intravenous access.7-14,22 It is difficult to compare the results of this study to those of previous studies because of numerous methodological differences. Four of the 9 previous studies were performed in more controlled settings (ie, the operating room or the ICU).7,12,13,22 Of the 5 studies conducted in the ED, 2 comprised pediatric patients younger than 10 years.8,9 Only 2 studies used the same definition of difficult intravenous access as this study.12,13 In 7 of the 9 studies, the operators received limited training in ultrasonographic peripheral intravenous cannulation before enrollment started,7-9,14 or no details were provided on the training received.12,13,22 Of the 9 studies conducted to date, only 3 reported the initial success rate by procedural approach.7,8,13 If initial success rate, overall success rate, average number of attempts, or frequency of physician rescue attempt is considered, 7 of the 9 studies favored ultrasonography over landmark for patients with difficult access.8,9,11-14,22 To our knowledge, this is the first study to demonstrate that ultrasonography is superior among patients with moderately difficult intravenous access, those for whom the operator can see or palpate a vein but anticipates some difficulty using a landmark approach. The patients classified as having moderately difficult access enjoyed a 10% higher initial and second-attempt success rate with ultrasonography compared with landmark. Typically, peripheral intravenous cannulation is the primary responsibility of nursing.2,14,23 In this study, ED technicians performed all peripheral intravenous catheterizations. However, the technicians achieved success rates similar to those of nurse or physician operators.15,16 Regardless of peripheral intravenous access difficulty, the ultrasonographic success rate was approximately 80%. In contrast, the technicians attained a much higher landmark success rate (ie, >95%) when peripheral intravenous access was easy. Since 2008, the study ED has relied on a program that trains technicians to perform ultrasonographic peripheral intravenous line placement. Consequently, 8 Annals of Emergency Medicine
there is always 1 or more providers working in the ED at all times who can place a peripheral intravenous line with either method. As a result of this program, our ED has substantially reduced its use of central venous line placement.24 This study used a large group of experienced technicians, so the results are not confounded by a learning curve for most of the operators. More than 80% of the technicians had 1 or more years of experience with ultrasonographic peripheral intravenous line placement at study onset. We believe these results reflect operator skill and are not biased by differential expertise because operators prefer to use landmark with patients who have easy access.25 We randomized the treatment approach by operator and difficulty of intravenous access, so each technician placed an approximately equal number of catheters with or without ultrasonography in each intravenous access group. Technicians who reported placing more ultrasonographic peripheral intravenous lines during a typical shift had higher success rates with ultrasonography compared with those who reported placing fewer lines. When more skilled providers used ultrasonography, our results estimated an additional 8 to 10 initial successes per 100 patients, depending on intravenous access difficulty. The strengths of our study include a design that randomized patients within strata defined by operator and difficulty of intravenous access to control for these important sources of variation, a second randomized treatment assignment if the first attempt failed, a large sample of patients with various degrees of difficulty of peripheral intravenous access, and a sizeable cohort of technicians experienced in peripheral intravenous catheterization with both approaches. Within this context, our results suggest that landmark is the superior method for patients with easy intravenous access, whereas ultrasonography is superior for those with moderately difficult or difficult intravenous access. The authors acknowledge the valuable contributions of the 33 technicians who participated in the study. Supervising editor: Donald M. Yealy, MD Author affiliations: From the Department of Health Policy and Management (McCarthy) and Department of Emergency Medicine (McCarthy, Shokoohi, Boniface, Shesser), George Washington University Hospital, Washington, DC (Lim, Eggelton, Lowey); and the Department of Biostatistics, Johns Hopkins University Bloomberg School of Public Health, Baltimore, MD (Li, Zeger). Author contributions: MLM, HS, KSB, RE, and SLZ conceived the study and designed the trial. MLM and RS obtained research Volume
-,
no.
-
:
-
2015
McCarthy et al
Ultrasonography Versus Landmark for Peripheral Intravenous Cannulation
funding. MLM, HS, RE, AL, and KL supervised the conduct of the trial and data collection and undertook recruitment of patients. XL and SLZ provided statistical advice on study design and analyzed the data. MLM managed the data, including quality control, and drafted the article, and all authors contributed substantially to its revision. MLM takes responsibility for the paper as a whole. Funding and support: By Annals policy, all authors are required to disclose any and all commercial, financial, and other relationships in any way related to the subject of this article as per ICMJE conflict of interest guidelines (see www.icmje.org). The authors have stated that no such relationships exist and provided the following details: This study was supported in part by award K01HS017957 from the Agency for Healthcare Research and Quality. For every 3 subjects enrolled, the authors gave the technician a $5 gift card as a token of appreciation for assisting with study enrollment and data collection in addition to his or her clinical responsibilities. Publication dates: Received for publication June 25, 2015. Revisions received August 12, 2015, and August 24, 2015. Accepted for publication September 2, 2015. Clinical trial registration number: NCT01859559 The sponsor was not involved in the design and conduct of the study; collection, management, analysis, and interpretation of the data; or preparation, review, or approval of the article. The content is solely the responsibility of the authors and does not represent the official views of the Agency for Healthcare Research and Quality.
REFERENCES 1. Alexandrou E. The One Million Global Catheters PIVC Worldwide Prevalence study. Br J Nurs. 2014;23:S16-S17. 2. Sabri A, Szalas J, Holmes KS, et al. Failed attempts and improvement strategies in peripheral intravenous catheterization. Biomed Mater Eng. 2013;23:93-108. 3. Sebbane M, Claret PG, Lefevbre S, et al. Predicting peripheral venous access difficulty in the emergency department using body mass index and a clinical evaluation of venous accessibility. J Emerg Med. 2013;44:299-305. 4. Crowley M, Brim C, Proehl J, et al. Emergency nursing resource: difficult intravenous access. 2011. Emergency Nurses Association. 1-21-2012. Available at: https://www.ena.org/practice-research/research/CPG/ Documents/DifficultIVAccessCPG.pdf. Accessed September 2015. 5. Hind D, Calvert N, McWilliams R, et al. Ultrasonic locating devices for central venous cannulation: meta-analysis. BMJ. 2003;327:1-7. 6. Calvert N, Hind D, McWilliams R, et al. Ultrasound for central venous cannulation: economic evaluation of cost-effectiveness. Anaesthesia. 2004;59:1116-1120. 7. Aponte H, Acosta S, Rigamonti D, et al. The use of ultasound for placement of intravenous catheters. AANA J. 2007;75:212-216. 8. Bair AE, Rose JS, Vance CW, et al. Ultrasound-assisted peripheral venous access in young children: a randomized controlled trial and pilot feasibility study. West J Emerg Med. 2008;9:219-224.
Volume
-,
no.
-
:
-
2015
9. Doniger SJ, Ishimine P, Fox JC, et al. Randomized controlled trial of ultrasound-guided peripheral intravenous catheter placement versus traditional techniques in difficult-access pediatric patients. Pediatr Emerg Care. 2009;25:154-159. 10. Stein J, George B, River G, et al. Ultrasonographically guided peripheral intravenous cannulation in emergency department patients with difficult intravenous access: a randomized trial. Ann Emerg Med. 2009;54:33-40. 11. River G, Hebig M, McAlpine I, et al. Nurse-operated ultrasound for difficult intravenous access: a randomized trial. Ann Emerg Med. 2009;54:S87. 12. Kerfone T, Petitpas F, Frasca D, et al. Ultrasound-guided peripheral venous access in severely ill patients with suspected difficult vascular puncture. Chest. 2012;141:279-280. 13. Benkhadra M, Collignon M, Fournel I, et al. Ultrasound guidance allows faster peripheral IV cannulation in children under 3 years of age with difficult venous access: a prospective randomized study. Pediatr Anesth. 2012;22:449-454. 14. Weiner SG, Sarff AR, Esener DE, et al. Single-operator ultrasoundguided intravenous line placement by emergency nurses reduces the need for physician intervention in patients with difficult-to-establish intravenous access. J Emerg Med. 2013;44:653-660. 15. Heinrichs J, Fritze Z, Vandermeer B, et al. Ultrasonographically guided peripheral intravenous cannulation of children and adults: a systematic review and meta-analysis. Ann Intern Med. 2013;61: 444-454. 16. Egan G, Healy D, O’Neill H, et al. Ultrasound guidance for difficult peripheral venous access: systematic review and meta-analysis. Emerg Med J. 2013;30:521-526. 17. Liu YT, Alsaawi A, Bjornsson HM. Ultrasound-guided peripheral venous access: a systematic review of randomized-controlled trials. Eur J Emerg Med. 2014;21:18-23. 18. Schoenfeld E, Shokoohi H, Boniface K. ED technicians can succesfully place ultrasound-guided intravenous catheters in patients with poor vascular access. Am J Emerg Med. 2011;29:496-501. 19. Wuerz RC, Milne LW, Eitel DR, et al. Reliability and validity of a new five-level triage instrument. Acad Emerg Med. 2000;7:236-242. 20. Harris PA, Taylor R, Thielke R, et al. Research Electronic Data Capture (REDCap)—a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009;42:377-381. 21. McCulloch CE, Searle SR, Neuhaus JM. Generalized, Linear and Mixed Models. 2nd ed. Hoboken, NJ: John Wiley & Sons; 2012. 22. Pappas NL, Michaud TE, Wolbers RM, et al. Ultrasound guidance as a rescue technique for peripheral intravenous cannulation. C104-1878, 1–16. 2006. Bethesda, Maryland, Uniformed Services University of the Health Sciences. 9-11-2012. Available at: https://www.ena.org/ practice-research/research/CPG/Documents/DifficultIVAccessCPG. pdf. Accessed September 2015. 23. Jacobson AF, Winslow EH. Variables influencing intravenous catheter insertion difficulty and failure: an analysis of 339 intravenous catheter insertions. Heart Lung. 2005;34:345-359. 24. Shokoohi H, Boniface K, McCarthy M, et al. Ultrasound-guided peripheral intravenous access program is associated with a marked reduction in central venous catheter use in noncritically ill emergency department patients. Ann Emerg Med. 2013;61:198-203. 25. Devereaux PJ, Bhandri M, Clarke M, et al. Need for expertise based randomized controlled trials. BMJ. 2005;330:1-5.
Annals of Emergency Medicine 9