a b s t r a c t

Background: When an intravenous (IV) catheter is needed and the common approach of inspection and palpation fails, an advanced access technique becomes necessary. Our objectives were to estimate pain scores, operator times, success rates, and complication rates when advanced techniques are used in a clinical setting.

Methods: We enrolled patients who had a need for advanced IV access and were able to give informed consent to participate in our study. We collected data on operator type, technique, initial success, number of attempts, skin punctures, operator time, pain scores, and complications. We estimated confidence intervals for proportions using normal binomial approximation or exact calculation.

Results: The registry documented 154 attempts in 116 patients. The median time from triage to establishment of an IV line was 203 minutes; multiple advanced attempts were required in 24% of cases. Most attempts (95%) used ei- ther ultrasound-guided cannulation of a peripheral vein (PUG) (108) or cannulated the external jugular vein (EJ)

(38). These 2 methods yielded similar pain scores (4.3-4.5), but PUG required more skin punctures (1.6 vs 1.2) and longer operator time (17.7 vs 11.9 minutes). The only complication was IV line failure, occurring in 6% (95% con- fidence interval, 0%-18%) of EJ approaches and 27% (95% confidence interval, 18%-38%) of the PUG scenarios. Conclusion: Most attempts to establish IV access used PUG or the EJ. External jugular vein cannulation was achieved more quickly, with fewer skin punctures and a lower rate of postinsertion failure, than PUG.

(C) 2015

Introduction

Intravenous (IV) catheter placement is vital for emergency care, allowing resuscitation, sedation, therapy, and pain relief. Inspection and palpation of peripheral veins, the traditional approach, is familiar to virtually all health care providers and is used in 97% of successful IV line insertions [1]. When inspection and palpation fail, one of several ad- vanced techniques can be used, but performance of these techniques is usually limited to a few skilled providers.

? Presented at the Annual Meeting of the Society for Academic Emergency Medi- cine, Atlanta, GA, 2013.

?? Funding for this project was provided by a grant from the Maryland Emergency

Medicine Network.

? Conflicts of interest: No author has a conflict of interest.

?? Authors’ contributions: MDW participated in study design, funding acquisition,

data collection, data analysis, manuscript preparation, and manuscript approval. SM par- ticipated in study design, funding acquisition, data collection, data analysis, manuscript preparation, and manuscript approval. YZ participated in study design, manuscript prepa- ration, and manuscript approval. CBM participated in data collection, manuscript prepara- tion, and manuscript approval.

* Corresponding author at: Department of Emergency Medicine, University of Maryland School of Medicine, 110 S Paca St, 6th Floor, Suite 200, Baltimore, MD 21201. Tel.: +1 410 328 8028.

E-mail address: [email protected] (M.D. Witting).

From the early days of venous cut-downs and central venous access, advanced techniques have evolved to include external jugular vein (EJ) access, ultrasound-guided access of a peripheral vein (PUG), and interosseous access. Studies in recent years have focused on PUG in awake patients, generally examining its use by trained study operators [1-8]. Most emergency departments (EDs) have operators with various levels of training in advanced techniques, and operators choose the technique they feel gives them the best chance for success. The purpose of this study was to quantify health care providers’ use of advanced ac- cess techniques, based on data from a local registry in which access at- tempts were documented. Our goal was to characterize success rates, complication rates, and pain associated with the techniques in a real- world setting.

Methods

Study design and setting

In this prospective cohort study, between June of 2011 and May of 2013, we enrolled patients who required advanced IV catheterization, defined as PUG, EJ access, or central venous access. The study was per- formed at 2 urban hospitals-a tertiary care center (53 000 ED visits per year) and a university-affiliated community hospital (58 000 ED

http://dx.doi.org/10.1016/j.ajem.2015.12.062

0735-6757/(C) 2015

Table 1

Characteristics of 116 patients

Characteristic n (%)

Diabetes mellitus 43 (37)

Hemodialysis 20 (17)

IV drug use 41 (35)

Obesity 36 (31)

Prior need for advanced access 100 (87)

Hypotension 13 (11)

visits per year). Both hospitals train residents from an emergency med- icine residency, which includes a 3-year categorical emergency medi- cine program and two 5-year programs combining emergency medicine with either pediatrics or internal medicine. Both EDs have ex- perienced nurses who perform most overall IV access attempts. Ultrasound-guided access of a peripheral vein is within the Scope of practice of nurses at both institutions, but most do not use it. Nurses in the tertiary care center ED, but not the community hospital, have the option to use the EJ for IV access. We excluded patients who were unable to provide written consent, such as those with altered conscious- ness or major instability. The institutional review board at each hospital approved the study protocol.

Data collection

Data were collected by 11 researchers; these were medical students, nurses, residents, and attending physicians. After obtaining written con- sent from each patient, we recorded the following patient factors: age, history of diabetes mellitus, history of renal dialysis, obesity (body mass index N 30), hypotension (systolic blood pressure b 90 mm Hg), prior need for advanced techniques for IV access, and intravenous drug use. Intravenous drug use was defined by self-report of a previous 6-month history of regular use or any use in the past 2 weeks.

For each attempt to insert an IV line, we recorded the operator type (registered nurse, resident, or attending physician), training level for residents, the technique selected (EJ access, PUG, or central access) and the anatomical location, and whether the attempt was initially suc- cessful, meaning that its status was confirmed by aspiration or injection without resistance or pain and the team felt the IV line was functional. We asked operators to record the amount of time they devoted to the insertion attempt, from the moment they gathered equipment to the time they deemed the line secure. We asked operators to indicate whether they administered a local anesthetic before the attempt. We defined an attempt as a single operator’s effort to access a vein in a given anatomical location; 1 attempt could have involved multiple skin punctures. If another operator attempted to access the same vein, we recorded this as an additional attempt. In addition, if the original op- erator tried another site, we recorded this as an additional attempt. We recorded the number of skin punctures associated with each attempt. At

the end of each attempt, we asked the patient to rate the pain associated with the entire attempt on a scale of 0 to 10. Data collectors were pres- ent in the ED with the operator and patient, frequently monitoring progress, but they were not continually present at the bedside.

We followed the patency of the IV lines during the patients’ ED stays. We defined failure as infiltration or other dysfunction requiring replace- ment, while in the ED, of a line that had been considered successfully inserted. We watched for major complications such as pneumothorax or Arterial puncture.

Data analysis

We analyzed data with attempt or patient as the unit of analysis, as appropriate. We used EpiCalc 2000 (Brixton Books, London, UK) for comparison of means and relative risk calculations, and we used SAS

9.3 (Cary, NC) for linear regression. The goal of linear regression was a precise and unbiased estimation of the effect of technique (PUG vs EJ ap- proach) on pain scores, analyzing only cases that involved 1 of these 2 techniques. We considered variables for the model based on an associ- ation with pain scores at a significance level less than 0.1. We eliminated variables from the model only if they affected bias of the technique es- timate by less than 0.1 and precision (SE) by less than 0.2. Our sample size was initially chosen to estimate complication rates for all tech- niques, including those chosen infrequently, but we did not prospec- tively know the distribution. Given this uncertainty, we received institutional review board permission to enroll more than 1000 pa- tients. Because of resource limitations and the relative rarity of central access, we chose to report our results based on available data.

Results

We collected information for 154 attempts in 116 patients (110 in the tertiary care hospital and 6 in the community hospital). The median age was 51 years (interquartile range, 43-58). Other population charac- teristics are shown in Table 1.

Technique selection

Operators for attempts were distributed as follows: registered nurse, 37 attempts; resident, 88 attempts; attending physician, 19 attempts; unspecified, 8 attempts; other (medical student, physician’s assistant), 2 attempts. The training level for residents was recorded in 84 attempts, including 16 by interns, 20 by second-year residents, and 48 by resi- dents with 2 or more prior years of experience. The data collector had a different training than the operator (eg, attending physician data col- lector, resident operator) in 82% of cases. The Figure shows the distribu- tion of technique choices by operators. Most attempts (95%) used PUG or the EJ, with physicians choosing PUG and nurses choosing the EJ. Cen- tral access was used in 8 (5%) of 154 of cases.

Figure. Technique choices by operator type in 154 attempts. “Other” category includes 1 physician’s assistant, 1 student, and 5 operators not specified.

Table 2

Results of initial IV line insertion attempts by various techniques

Technique	Attempts	Initial success (%)	Skin punctures, mean (SD)	Operator time in minutes, mean (SD)	pain level, mean (SD)
PUG	108	86 (80%)	1.6 (1.0)	17.7 (10.2)	4.3 (3.1)
EJ	38	33 (87%)	1.2 (0.5)	11.9 (7.1)	4.5 (3.3)
Central line	8	7 (88%)	1 (0)a	20.0 (11.7)	5.4 (3.0)a

^a Data were missing for 3 cases.

Results of cannulation attempts

Intravenous attempts were initially successful in 82% of cases. The median follow-up time was 178 minutes; 18% of IVs, initially thought to be successful, failed in the ED, with a median time to failure of 71 minutes. Table 2 summarizes the results for the techniques chosen. Success rates and pain scores were similar between the EJ and PUG groups, but the mean operator time was 5.8 minutes shorter for the EJ approach (95% confidence interval [CI], 2.8-9.3), which also had 0.5 fewer punctures per attempt (95% CI, 0.1-0.8). Intravenous patency data were available for 122 lines. Overall, 5 of the 6 Central lines (1 short line failed), 30 of the 32 (94%; 95% CI, 82%-100%) EJ lines, and 61 of the 84 (73%; 95% CI, 62-82%) PUG lines remained patent until the pa- tient left the ED. Thus, lines placed with PUG failed 4.3 times more often than those placed in the EJ (95% CI, 1.1-17.5). No cases of pneumothorax or major bleeding were reported in the registry.

Pain scores

The registry contained documentation of pain scores for 144 of 154 attempts. Anesthesia was administered for 4 (3%) of 145 procedures. The median pain score was 4, and the mean pain score was 4.3 (SD, 3.2). Pain scores were 8 or higher in 27 of the 144 cases (19%). Table 3 shows the adjusted association between pain level and technique. We found no association between pain scores and operator type (physician vs nurse), number of skin punctures preceding the attempt, patient age, obesity, diabetes, or hemodialysis. Pain scores were similar between the EJ approach and PUG, both in the crude scores shown in Table 2 and in the adjusted scores shown in Table 3. Pain scores were higher for those without a history of IV drug use or the need for advanced line insertion techniques in the past.

Results by patient

Table 4 shows the results analyzed by patient. Multiple advanced at- tempts, meaning multiple operators or body locations, were required in

advanced techniques, and each provider uses whatever technique he or she is familiar with and feels would work best in a given patient. Some providers are more experienced with these techniques than others, but those more experienced practitioners might not be available for a given attempt. Thus, a registry has the potential to describe the cur- rent state of advanced techniques for IV access, as it applies to a group of providers with various levels of expertise.

In our registry, most participating physicians used PUG, and most nurses used the EJ. These 2 techniques accounted for 95% of all attempts. Central line insertion was used in the remaining 5% of the attempts. Pa- tients frequently required multiple attempts and multiple skin punc- tures. Pain scores were moderate, averaging 4.3, and were unaffected by technique. The only complication we noted was postinsertion failure, which occurred mainly in lines inserted with PUG.

In 1999, Keyes et al [2] described the use of a 2-person ultrasound- guided technique in patients with difficult IV access. Since then, a wealth of studies have examined the use of PUG in EDs. In 2004, Costantino et al [3] confirmed the success of a 2-physician technique in patients who would otherwise have required central access. Several studies have shown that nurses and technicians can perform PUG suc- cessfully [4-8]. In a few settings, dedicated PUG protocols have de- creased the need for central venous access [9,10]. Our clinical settings do not require PUG by protocol, but training in this technique is avail- able to nurses and physicians. Ultrasound-guided access of a peripheral vein was used in 70%, confirming its popularity in our setting, particu- larly among physicians.

Studies have also revealed shortcomings of PUG. One showed that PUG was slower, by a median of 13 minutes, than inspection and palpa- tion in patients having 2 failed prior attempts [11]. Other studies have demonstrated limited survival of lines placed with PUG. Keyes et al noted a failure rate of 8% within the first hour after catheterization, and others noted Failure rates of 32% to 47% within 24 to 48 hours

Table 4

Summary of results from the patients’ perspective

24% of cases. Multiple skin punctures were required in 47% of patients,

including 13% requiring 4 or more.

Eventual IV line 112/116 (97%)

No. of attempts

1 88 (76%)

4. Discussion	2	22 (19%)
	3	4 (3%)
Our goal in creating the registry was to characterize modern ad- vanced techniques for establishing IV access in a real-world setting. The prior literature features results from research protocols involving 1, or perhaps 2, of these techniques, but most EDs are not doing ad-	4 or 5 Skin puncturesa 1 2 3	2 (2%) 62 (53%) 22 (19%) 13 (11%)
vanced IV access research [2-13]. Rather, most departments have a het-	4-6	15 (13%)
erogeneous group of providers with varying expertise in different	First operator

Table 3

Variables affecting pain scores

Variable	Effect on pain (95% CI)	P
No IV drug use	1.2 (0.2-2.3)	.02
No prior difficulty	1.5 (0-3.1)	.05
EJ (vs PUG)	0.2 (-1.0 to 1.3)	.7

MD 75 (66%)

RN 32 (28%)

Other/not noted 9 (8%)

IV timing (min)

Total operator time, median (IQR) 17 (12-30)

Triage to first IV line, median (IQR)^b 203 (133-378)

Abbreviation: IQR, interquartile range.

^a Data were missing for 4 patients.

^b Based on 93 patients who were enrolled before the first advanced IV line insertion attempt.

[2,12,13]. We found that 27% of lines placed via PUG failed while the pa- tient was in the ED and that operators took an average of 18 minutes on PUG attempts. We did not note any instances of arterial puncture; pre- vious studies reported rates of 2% to 3% [2,12].

Compared with the number of reports on PUG, relatively little has been published on the use of EJ catheterization in EDs. We found only 1 other study that compared EJ with PUG [13]. Those investigators found that physicians had a higher success rate with PUG than with an EJ ap- proach, in patients randomized to either group, as well as similar proce- dure times, not including the time needed to set up the ultrasound machine. In the subset with a visible EJ, success rates were similar for both techniques, and operator times were shorter for the EJ group. Our registry records showed that EJ cannulation was selected in 25% of cases. It was the most popular choice by nurses, although only 1 of the 2 facilities permits nurses to use it. We found that EJ attempts were associ- ated with shorter operator times, considering time to set up the ultra- sound machine. External jugular vein cannulation also had fewer skin punctures than PUG and a lower failure rate. We suspect that patients in our study were selected for the EJ approach based on the operator’s ability to locate the vein by inspection and palpation.

We found abundant literature describing the education of nurses and technicians in PUG but none describing their education in the use of EJ cannulation [4-8]. Interestingly, a policy statement issued by the American College of Emergency Physicians in 2011 promotes the train- ing of nonphysician practitioners in alternate access techniques, includ- ing EJ cannulation [14]. It is noteworthy that one of the hospitals in our study forbids the use of the EJ approach by nurses and technicians.

Among advanced techniques, central lines have a higher risk of major complications than PUG or EJ. One of our goals for the registry was to es- timate the real-world incidence of major complications. Prior trials, often including data from experienced operators, yielded low rates of major complications [15,16]. We expected that our setting, which includes phy- sicians in training, would have a higher incidence of complications, but our study did not enroll enough patients with central access to allow us to characterize complication rates. As in studies conducted in medical fa- cilities with dedicated PUG programs focusing on limiting central access, we found that central access was attempted infrequently in an environ- ment with operators trained in PUG or EJ cannulation [9,10].

Our patients experienced moderate pain in response to IV line inser- tion attempts, as reflected by their pain scores. The operators rarely pro- vided anesthesia. McNaughton et al [17] demonstrated that buffered lidocaine decreased the median pain score from 7 to 2 in a group of medical students and nurses. In our EDs, lidocaine administration re- quires a physician to write an order and a nurse to obtain the medication-steps that are prohibitive in patients who are often already facing delays. Our median pain score of 4 suggests a lesser potential ben- efit from injected anesthesia than in the McNaughton’s study group. We found no effect of prior skin punctures on pain scores, in contrast to the results of another recent study [18]. Importantly, we found no differ- ence in pain scores in patients undergoing EJ cannulation and those in whom PUG was used. Anecdotally, we have found that some patients are reluctant to have an IV line placed in their neck without receiving anesthesia. Our results suggest that this procedure is no more painful than PUG when provided without anesthesia.

Limitations

Patients were enrolled for this study by a limited number of research assistants, generally while they were working clinical shifts. In some cases, these assistants recorded the results of their own IV line insertion attempts. On one hand, this tendency toward convenience sampling might have led to overrepresentation and clustering by certain opera- tors, particularly those with an interest in advanced IV access. On the other hand, most attempts were recorded at the tertiary care hospital, which has many residents and students seeking training in advanced techniques. These 2 factors likely biased in opposite directions with

regard to performance characteristics. Because this was not a clinical trial, patient selection for technique was not randomized. Patients may have been chosen for EJ catheterization based on the operator’s ability to inspect and palpate the vein. These factors may have biased to- ward improved performance by operators choosing EJ catheterization. We relied on the operators to measure the time they spent on the procedures. Their estimates could have been inaccurate; a more accu-

rate method would have been observation by a third party.

Some variables might have been affected by unmeasured influences. For example, time to IV access could have been affected by ED crowding, which we did not quantify.

Conclusion

Most health care providers in our study group chose EJ cannulation or PUG when an advanced technique for establishing IV access was needed. Physician operators tended to select PUG, whereas nurses fa- vored EJ. Postinsertion failure occurred in 27% of PUG, vs 6% with EJ. The 2 approaches had similar initial success rates and produced similar pain scores, but the EJ route had fewer complications, shorter operator time, and fewer skin punctures.

Acknowledgments

The authors thank the following research assistants for enrolling pa- tients in the study: Sandra Beverly, Bryan Stover, Matthew Zeitler, Steven Biederman, Jennifer Scott, Alexander Cruise, Colleen Holley, Kenneth Nugent, and Ryan Spangler. The authors also thank Linda J. Kesselring, MS, ELS, for copyediting this manuscript.

References

1. Witting MD, Moayedi S, Beverly SK, Stover BJ, Miller AC. Incidence of advanced IV ac- cess in 2 urban EDs. Am J Emerg Med 2015;33:705-7.
2. Keyes LE, Frazee BW, Snoey ER, Simon BC, Christy D. Ultrasound-guided brachial and basilic vein cannulation in emergency department patients with difficult intrave- nous access. Ann Emerg Med 1999;34:711-4.
3. Costantino AK, Parikh WA, Satz WA, Fojtik JP. Ultrasonography-guided peripheral in- travenous access versus traditional approaches in patients with difficult intravenous access. Ann Emerg Med 2005;46:456-61.
4. Brannan L, Blaivas M, Lyon M. emergency nurses‘ utilization of ultrasound guidance for placement of peripheral intravenous lines in difficult-access patients. Acad Emerg Med 2004;11:1361-3.
5. Bauman M, Braude D, Crandall C. Ultrasound-guidance versus standard technique in patients with difficult IV access by ED technicians. Am J Emerg Med 2009;27: 135-40.
6. Miles G, Salcedo A, Spear D. Implementation of a successful registered nurse periph- eral ultrasound-guided intravenous catheter program in an emergency department. J Emerg Nurs 2012;38:353-6.
7. Schoenfeld E, Boniface K, Shokoohi H. ED technicians can successfully place ultrasound-guided intravenous catheters in patients with poor vascular access. Am J Emerg Med 2001;29:496-501.
8. Weiner SG, Sarff AR, Esener DE, Shroff SD, Budhram GR, Switkowski KM, et al. Single-operator ultrasound-guided Intravenous line placement by emergency nurses reduces the need for physician intervention in patients with difficult-to- establish intravenous access. J Emerg Med 2013;44:653-60.
9. Au AK, Rotte MJ, Grzybowski RJ, Ku BS, Fields JM. Decrease in central venous catheter placement due to use of ultrasound guidance for peripheral intravenous catheters. Am J Emerg Med 2012;30:1950-4.
10. Shokoohi H, Boniface K, McCarthy M, Al-tiae TK, Sattarian M, Ding R, et al. Ultrasound-guided peripheral intravenous access program is associated with a marked reduction in central venous catheter use in noncritically ill emergency de- partment patients. Ann Emerg Med 2013;61:198-203.
11. Stein J, George B, River G, Hebig A, McDermott D. Ultrasonographically guided pe- ripheral intravenous cannulation in emergency department patients with difficult intravenous access: a randomized trial. Ann Emerg Med 2009;54:33-40.
12. Dargin JM, Rebholz CM, Lowenstein RA, Mitchell PM, Feldman JA. Ultrasonographically- guided peripheral intravenous catheter survival in ED patients with difficult access. Am J Emerg Med 2008;28:1-7.
13. Costantino TG, Kirtz JF, Satz WA. Ultrasound-guided Peripheral venous access versus the external jugular vein as the initial approach to the patient with difficult vascular access. J Emerg Med 2010;39:462-7.
14. American College of Emergency Physicians. Policy statement: alternative methods to vascular access in the emergency department. Ann Emerg Med 2011;58:402.
15. Tseng M, Sadler D, Wong J, et al. Radiologic placement of central venous catheters: rates of success and immediate complications in 3412 cases. Can Assoc Radiol J 2001;52:379-84.
16. Reusch S, Walder B, Tramer M. Complications of central venous catheters: internal jug- ular versus subclavian access-a systematic review. Crit Care Med 2002;30:454-60.
17. McNaughton C, Zhou C, Robert L, Storrow A, Kennedy R. A randomized crossover comparison of injected buffered lidocaine, lidocaine cream, and no analgesia for pe- ripheral intravenous cannula insertion. Ann Emerg Med 2009;54:214-20.
18. Fields JM, Piela NE, Ku BS. Association between multiple IV attempts and perceived pain levels in the emergency department. J Vasc Access 2014;15:514-8.