The use of central lines in the intensive care setting is critical to the clinical management of patients of all ages and complex conditions. One very common procedure used in the neonatal intensive care unit (NICU) is the placement of an umbilical venous catheter (UVC). The UVC is used to provide short-term central intravenous access for parenteral nutrition, venous laboratory access, administration of caustic medications, and fluid resuscitation in premature and critically ill infants. This group of infants includes those with extremely low birth weight, respiratory distress, hypoglycemia, hypoxic ischemic encephalopathy, pulmonary hypertension, and various neonatal conditions and congenital anomalies.1
The utilization of central lines in the NICU depends upon successful UVC placement. The ideal position of an UVC is within the right atrial/inferior vena cava (RA/IVC) junction or thoracic IVC (Figure 1).2,3 A malpositioned UVC may be placed too deep in the correct vascular track (Figure 2) or placed in a different vascular track. Common malpositioned UVCs are located within the right portal vein vasculature. Complications associated with the use of UVCs include, but are not limited to, arrhythmias, cardiac tamponade, thrombus formation, hepatic injury, bleeding, peritoneal cavity perforation, abscess formation, pulmonary infarction, sepsis, endocarditis, portal hypertension, and fluid collection within the hepatic parenchyma, cardiac, and pleural spaces.4,5 Ensuring accurate UVC position decreases the risks of many complications, such as arrhythmias, cardiac tamponade, and cast formation. Current NICU practice is to verify UVC position with an x-ray (radiograph).4 When the UVC is positioned too deep, it is adjusted by pulling back on the catheter followed by evaluating for an appropriate placement verified by a repeat x-ray. Exposing neonates to multiple x-rays for one procedure increases their risk of future malignancies.6
Three articles published as far back as 1982 discussed the use of ultrasonography for evaluating UVC placement.7–9 Since then, multiple studies have recommended the use of ultrasound for UVC placement confirmation for a more reliable interpretation compared with radiograph.2–6,10,11 Currently, “point-of-care ultrasound (POCUS)” is used in emergency departments and some adult intensive care unit settings.12 Although it is widely accepted that the use of radiographs can lead to an imprecise and less accurate interpretation of UVC position, the use of ultrasound has still not been adopted as the new standard of care,4 likely due to a lack of education and training programs being offered. More recently, the use of ultrasonography and ultrasound training programs has increased; however, these programs primarily target neonatologists even though neonatal nurse practitioners (NNP) frequently insert the UVC.13–16 While other specialized nurse practitioners are trained to use ultrasonography in their critical care practice,12 to the author's knowledge only one program focused on neonatal education invited nurse practitioners and other advanced care providers to attend.17 NNPs trained to use ultrasonography would promote the adoption of this technique for UVC placement, evaluation, and confirmation.
The overall goal of this project was to evaluate the feasibility of an NNP performing ultrasounds and the accuracy of their interpretation of the placement of the UVC within neonatal vasculature and cardiac anatomy. The reliability between the NNP's interpretation of UVC placement with that of a pediatric cardiologist was determined. In addition, ultrasonography was compared to radiographic images to evaluate the reliability of these 2 methods in determining proper UVC placement.
This evidence-based practice project evaluated the use of ultrasound by an NNP to ascertain proper UVC placement in infants in a level III NICU of a perinatal regional center with a 46-bed capacity. The NICU healthcare team included registered nurses, pediatric residents, neonatal fellows, neonatologists, and NNPs. The UVCs utilized in the NICU were 3.5F or 5F Argyle polyurethane catheters. The project design began with a retrospective chart review from January 2018 to June 2018 to evaluate the number of x-rays and line adjustments involved in the use of radiograph to determine proper UVC placement. In the second portion of the project, ultrasound was used in addition to the current unit practice (using radiographs) to assess interpretation of UVC placement. Before starting the data collection, the Stony Brook University institutional review board (IRB) deemed the project a “quality improvement/quality advancement (QI/QA) project”; thus, IRB review was not required (exempt). Administrative approval from the hospital executive leadership team was also obtained.
Prior to the beginning of the ultrasound evaluation, the first author was trained by a pediatric cardiologist, a pediatric nurse practitioner with subspecialty training in pediatric cardiology, and licensed sonographers. Training consisted of reviewing approximately 15 hours of online ultrasound videos, as well as approximately 4 to 5 months of intermittent hands-on experience whenever a pediatric cardiology consult was conducted in the NICU (∼15 hours of hands-on and observational experience of sonographers/cardiologists conducting echocardiograms). When a NICU provider consulted the pediatric cardiology team, the first author attempted to locate anatomical landmarks via ultrasound within the parasternal short-axis view. This view was selected because previous research identified the parasternal short-axis view as a more reliable view and could easily be learned by novice ultrasound users.3
Three measures were evaluated from the retrospective chart review to describe the process of radiographic UVC placement confirmation, including (1) radiologist interpretation of the radiographic UVC placement on initial and final x-rays (throughout life of the line), (2) the number of readjustments after the UVC was placed (how many times the UVC was pulled back after placement due to malpositioning), and (3) the number of radiographs required to confirm acceptable placement of the UVC (including the initial x-ray when the UVC was placed). A secondary data source provided a list of patients who had central lines attempted and/or successfully placed, which served as the source of infants with UVCs. Data were collected on successful UVC placements defined as a UVC that was inserted and located at or near the RA/IVC junction. Six UVCs with insufficient data available were excluded (ie, interfacility transport, inadequate documentation). Data were categorized based on anatomical locations. The range of acceptability of UVC placement within the right atrium is variable within the literature, with one older study accepting of UVC placement in the right atrium, but addressing the pathologic consequences that are associated with right atrial placement of UVCs.2 For the purpose of this project, the right atrium was considered a malpositioned UVC since it is not ideal for placement.2 UVC adjustments were counted mostly via adjustment notes that were found in the documentation section of the electronic medical record (EMR). Within the radiology report there was a section that provided a reason why the x-ray was ordered. If the reason for the x-ray stated, for example, “line adjustment” “or “UVC adjustment,” and did not correlate with an adjustment noted in the record, it was counted as an adjustment. The number of x-rays counted for the life of the UVC line was determined via the radiology section of the EMR. The number of x-rays included both anterior-posterior and lateral views. The reason for x-ray had to state “UVC,” “line,” or “adjustment” to be counted as evaluation of proper UVC placement. This method was included since many x-rays are conducted to evaluate endotracheal tube placement, and evaluate lung pathology or lung expansion, rather than evaluating UVC placement. Ultrasound data collection included 7 measures: (1) the day of life (DOL) the UVC was placed; (2) the UVC length at the point of insertion into the umbilicus (ie, suturing 8 cm at insertion to umbilicus means that there is 8 cm of UVC catheter indwelling in the patient) at the time of the initial and final x-ray (final x-ray that confirmed successful placement of the UVC and allowed staff to start using the central line); (3) the number of adjustments required for proper placement; (4) the number of x-rays required for proper placement; (5) the UVC length up to the point of insertion to the umbilicus at time of the ultrasound; (6) the time between UVC placement and the ultrasound; and (7) the time between the most recent x-ray available and the time of the ultrasound (eg, an x-ray 30 minutes after ultrasound is more recent than 4 hours prior to ultrasound). The UVC insertion length that was sutured at the entry to the umbilicus was measured at 3 different points in time: (1) from when the UVC was initially placed (at time of initial x-ray), (2) when it was deemed acceptable to use after potential adjustments (at time of final x-ray), and (3) at time of ultrasound. These lengths were recorded and later compared to when the ultrasound was being completed by the first author. Infant-specific data, including gestational age and birth weight, were also collected to describe the sample.
The primary outcome measure for the project was the reliability between the NNP and pediatric cardiologist's assessment of UVC placement based on the image obtained by the NNP. Differences in placement interpretations were compared between the cardiologist's interpretation of the ultrasound placement and the radiologist's interpretation on the x-ray. The clinical radiologist interpretations of UVC placement were based on the radiograph that was completed closest to the time of the ultrasound (most recent x-ray, before or after ultrasound). The radiologist and the cardiologist were blinded to each other's interpretations. The cardiologist was also blinded to the NNP's ultrasound interpretation except in 15 cases where the cardiologist was informed of the NNP's interpretation due to the critical finding of a malpositioned UVC. Multiple radiologist interpretations were used in this project. No adjustments of the UVC were made based on ultrasound findings, only with radiographic interpretation and measurement.
Ultrasound Evaluation Process
The ultrasound was performed by the first author after any patient had a successfully placed UVC (confirmed via radiograph prior to ultrasound). Prior to completing the ultrasound, parents were informed that an ultrasound would also be completed to assist in verifying UVC placement. The Zonare ZS3 Ultrasound System was used to collect all cardiac images. The C10-3 transducer was used for all encounters, and one patient encounter included use of both the C10-3 and a C9-3 transducer because the infant was large for gestational age. The C9-3 transducer is larger and rounded and in this patient encounter, more effective in visualizing UVC position. For the C10-3 transducer, 33 Hz was used, and for the C9-3 transducer, 25 Hz was used.
Descriptive statistics were used to describe the sample and UVC placement. Interrater reliability data were calculated using κ tests. Interpretations of UVC placements were categorized into binary categories for data analysis as acceptable placement or malpositioned. To evaluate interrater agreements between raters, κ tests were used. Power analysis supported enrolling 20 or more patients to evaluate interrater reliability.
Retrospective Chart Review Results
Fifty-four infants had successful UVC placement over the 6 months of retrospective chart review. Of these UVC placements, only 48 were available for formal review, and 6 had insufficient data available to analyze (ie, interfacility transport, inadequate documentation). Of the 48 initial radiologist interpretations of UVC placement, 35% (n = 17) were considered acceptable, 60% (n = 29) were unacceptable, and 4% (n = 2) could not be categorized. Even when an initial x-ray showed an appropriately placed UVC, the medical team performed subsequent x-rays to monitor the placement of the UVC over the course of its use. Thirty-nine final interpretations were collected since 9 did not require a subsequent x-ray to evaluate UVC placement. Successful UVC placement occurred in 46% (n = 18), 51% (n = 20) were unacceptable, and 2% (n = 1) were uncategorizable. There were a total of 64 adjustments (range 0-4) and 180 x-rays (range 1-11). The large number of total x-rays was mostly due to subsequent x-rays to monitor UVC placement or evaluate the UVC after an adjustment.
Over the course of 2 months, 20 infants with a total of 21 ultrasound encounters were evaluated. One patient had 2 separate ultrasound encounters after their UVC had become malpositioned and was then replaced. Infant birth weights ranged from 600 to 3920 g (average 1798.1 g, ± 880.9 g) and gestational ages ranged from 24 to 39 weeks (average 321/5 weeks, ± 3.69 weeks). The DOL when the UVC was placed occurred primarily on DOL 1 (76%, range 1-4 DOL).
Of the 21 encounters, 7 had different indwelling UVC catheter lengths between the time of final x-ray to confirm placement and the time of ultrasound. The discrepancies between the lengths of UVC ranged from −0.5 (0.5 cm removed from patient) to +0.5 cm (0.5 cm inserted into patient) (average 0.01-cm difference). The range of time between UVC placement and the ultrasound being completed was 2 to 86 hours, with an average time difference of 23.6 hours. Most infants (71.4%) had ultrasounds completed before 24 hours from when the UVC was placed. The assessments greater than 24 hours were delayed due to the patient's being clinically unstable. Discrepancy between UVC lengths at the time of final x-ray and time of ultrasound occurred in 3 of the 5 encounters when an ultrasound was more than 24 hours from UVC placement (range −0.25 cm to +0.5 cm). There were a total of 10 adjustments (range 0-2) and 33 x-rays (range 1-3) required to obtain adequate UVC placement using radiograph. The range of time between the most recent x-ray that evaluated UVC placement and the time of ultrasound was 0.5 to 77 hours, with an average time difference of 12.7 hours.
Interrater reliability was performed comparing the ratings of 1 NNP and 1 pediatric cardiologist. The NNP had no previous ultrasound experience, while the pediatric cardiologist had 15 years of experience interpreting ultrasounds, specific to the pediatric/neonatal population. There was 86% agreement on ultrasound interpretations with a κ score of 0.667 (good interrater reliability). There was also 86% agreement between the pediatric cardiologist's ultrasound interpretation and the radiologist's radiograph interpretation (of most recent x-ray) with a κ score of 0.690 (good interrater reliability). Sixty-six percent (n = 14) of the pediatric cardiologist ultrasound assessments were considered malpositioned after the NICU team determined the UVC to be placed acceptably using radiographs. Malpositioned lines included UVC placement in the right atrium or touching the atrial septum (n = 3) or through the foramen ovale in the left atrium (n = 11) (Figure 2). The pediatric cardiologists also made formal recommendations of how much the UVC should be adjusted to achieve proper placement via ultrasound. Of these interpretations, one measurement was as long as 1.7 cm.
This project demonstrated that NNPs can competently use ultrasound for evaluation of UVC placement and quickly identify the true anatomical position of the UVC. In addition, 66% of UVCs that were deemed acceptable on radiographs by NICU healthcare providers were actually found to be malpositioned on ultrasound. The use of ultrasound can significantly decrease radiation exposure, allow for the UVC to be adjusted in real time, and therefore prevent morbidity and mortality secondary to malpositioned UVCs.2 Furthermore, other benefits include better efficiency than radiography, leading to improved time management and workflow for providers and nurses due to availability for immediate line adjustment.
In 1 retrospective review, 1 patient was exposed to 11 x-rays to monitor or adjust the UVC. Multiple UVC line adjustments increase the risk of transection of the UVC, accidental malpositioning with pulling on the line, and infection; thus can significantly impact the neonatal clinician's workflow and patient safety. Ultrasound enables UVC adjustments in real time, and therefore minimizes exposure to radiation in this group of infants who often require x-rays for other reasons and are already at increased risk of future malignancy.6
This retrospective analysis also found that UVCs are malpositioned when evaluated by radiographs. Unlike prior research,4 this project revealed good interrater reliability for UVC placement between the radiologist's interpretation of the x-ray and the pediatric cardiologist's interpretation via ultrasound. This result was most likely due to the small sample of 21, the use of binary coding to compare interpretations (acceptable vs unacceptable placement) instead of coding for more specific anatomical position (ie, right atrium, left atrium, etc), and the use of multiple radiologist interpretations using radiographs. Even if radiography was as reliable as ultrasound, the additional radiation exposure is still a concern for the patient's health and outcomes.
Documenting the length of insertion of the UVC at multiple points in time assisted with identifying the potential limitation of displacing the UVC from the time between completion of the final x-ray to verify UVC placement and collection of the ultrasound imaging by the first author. With 5 ultrasound encounters occurring more than 24 hours after the UVC was placed, only 3 of these cases showed a discrepancy in UVC length of insertion between the time of final x-ray that confirmed placement and the ultrasound. Initial and frequent assessment of UVC length is an important component of UVC use. UVC position can fluctuate due to patient handling by NICU staff, handling by parents, infant activity level, drying of the cord, and changes within the extravascular and intravascular fluid compartments. Therefore, using ultrasound for subsequent UVC evaluations can help to quickly restore proper UVC placement.
This study has several limitations. First, only 21 patient encounters were evaluated; therefore, more evidence-based practice studies with larger samples will improve the generalization of the findings. Further research that focuses on reducing the number of UVC adjustments required and therefore limits radiation exposure is warranted. Second, the pediatric cardiologist was not blinded to some NNP interpretations prior to providing their own interpretation of potentially malpositioned lines. This occurred in most patient encounters, with 15 out of 21 interpretations determined to be malpositioned by the NNP, and this introduced a potential limitation of the pediatric cardiologist's interpretation. When lines were found to be officially malpositioned by the pediatric cardiologist, the NICU medical team was immediately made aware. One radiological interpretation was seen by the NNP prior to conducting the ultrasound, risking skewing the NNP interpretation. Only having 2 raters to conduct an interrater reliability between the pediatric cardiologist and NNP analyses may also be a limitation. Furthermore, interpretations were conducted by different radiologists, and this added to the potential for skewed data when comparing interrater scores.
Further evaluation of NNP use of ultrasound for identification of proper UVC placement should be conducted. Further studies should include multiple NNPs to identify the factors that may hinder a rater's reliability, such as years of experience with the UVC procedure. The time required to obtain acceptable UVC placement in real time using ultrasound should be compared with the time required to make multiple line adjustments using radiographs to assist in nursing workflow (ie, start intravascular fluids and administer medications, golden hour care).
NNP education and training conducted by experts in the field, such as physicians, and specifically radiologists and cardiologists, will be critical to any successful implementation of ultrasound monitoring of UVC placement by NNPs. Learning opportunities, such as courses or conferences that focus on didactic and hands-on experiences, should be created, evaluated, and more accessible to NNPs. Hospitals need to ensure routine quality assurance programs that test and evaluate the accuracy of noncardiology professionals using ultrasound to evaluate UVC placement.
Education and training are considered significant barriers to the NNP adopting ultrasound into their practice. This technology is readily available, but barriers to its implementation must be challenged. Future studies should focus on interviewing key stakeholders about the barriers to implementation of this technique in clinical practice. Other potential barriers include funding, certification, maintaining competency, and provider billing.
The use of ultrasonography by neonatal healthcare providers has been discussed for decades, but its adoption has been limited. This study provides initial data to support NNPs performing ultrasonography to evaluate UVC placement. Collaboration between NNPs and interdisciplinary teams, such as pediatric cardiology, is key, and high-risk infants can only benefit from improvements in care.
What we know:
This project provides initial data to support ultrasonography use by NNPs to evaluate UVC position.
NNP education and training will be central to any successful implementation of ultrasound.
Interdisciplinary collaboration can assist with implementation of ultrasound use.
High-risk infants can benefit from improvements in care and the future workforce of NNPs can continue to challenge the status quo.
What needs to be studied:
This technology is readily available, but barriers to its implementation must be challenged.
For future projects and progress, interviewing key stakeholders as to what barriers are in place for implementing this technique into practice needs to be evaluated.
Potential barriers besides education include funding, certification, maintaining competency, and provider billing.
Research with larger samples and studying other factors to assess NNP reliability to utilize ultrasound can assist with translation into practice.
What we can do today:
Learning opportunities such as courses or conferences that focus on didactic and hands-on experiences should be created, evaluated, and accessible to APRNs.
Hospitals and clinical administration need to create quality assurance programs that annually test and evaluate the accuracy of noncardiology professionals that are using ultrasound to evaluate UVC position.
- Stony Brook Children's Department of Pediatrics, Division of Neonatology
- Stony Brook Children's Department of Pediatrics, Division of Pediatric Cardiology
- Stony Brook Children's NICU RNs
- Stony Brook Pediatric Cardiology Sonographers
- Damaris Salguedo, RDCS
- Barbara Blizzard, RDCS, RDMS
- Ryan Cherry, RDCS
- Heather Ferguson, MS, RN
- Daniela Titchiner, MD
- Julie Thompson, PhD
- Debra Sansoucie, EdD, NNP-BC
- Patricia Mele, DNP, NNP-BC
1. Rorke JM, Ramasethu J, Chahine AA. In: Macdonald MG Jr, Rais-Bahrami K, eds. Atlas of Procedures in Neonatology. 5th ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2013.
2. Ades A, Sable C, Cummings S, Cross R, Markle B, Martin G. Echocardiographic evaluation of umbilical venous catheter
placement. J Perinatol. 2003;23(1):24–28.
3. Karber BC, Nielsen JC, Balsam D, Messina C, Davidson D. Optimal radiologic position of an umbilical venous catheter
tip as determined by echocardiography in very low birth weight newborns. J Neonatal Perinatal Med. 2017;10(1):55–61.
4. Sharma D, Farahbakhsh N, Tabatabaii SA. Role of ultrasound for central catheter tip localization in neonates: a review of the current evidence. J Matern Fetal Neonatal Med. 2018;32(14):2429–2437.
5. Simanovsky N, Ofek-Shlomai N, Rozovsky K, Ergaz-Shaltiel Z, Hiller N, Bar-Oz B. Umbilical venous catheter
position: evaluation by ultrasound. Eur Radiol. 2011;21(9):1882–1886.
6. Michel F, Brevaut-Malaty V, Pasquali R, et al. Comparison of ultrasound and x-ray in determining the position of umbilical venous catheters. Resuscitation. 2012;83(6):705–709.
7. George L, Waldman JD, Cohen ML, et al. Umbilical vascular catheters: localization by two-dimensional echocardio/aortography. Pediatr Cardiol. 1982;2(3):237–243.
8. Oppenheimer DA, Carroll BA, Garth KE, Parker BR. Sonographic localization of neonatal umbilical catheters. AJR Am J Roentgenol. 1982;138(6):1025–1032.
9. Oppenheimer DA, Carroll BA. Ultrasonic localization of neonatal umbilical catheters. Radiology. 1982;142(3):781–782.
10. Franta J, Harabor A, Soraisham AS. Ultrasound assessment of umbilical venous catheter
migration in preterm infants: a prospective study. Arch Dis Child Fetal Neonatal Ed. 2017;102(3):F251–F255.
11. Pulickal AS, Charlagorla PK, Tume SC, Chhabra M, Narula P, Nadroo AM. Superiority of targeted neonatal echocardiography for umbilical venous catheter
tip localization: accuracy of a clinician performance model. J Perinatol. 2013;33(12):950–953.
12. Brunhoeber LA, King J, Davis S, Witherspoon B. Nurse practitioner use of point-of-care ultrasound in critical care. J Nurse Pract. 2018;14(5):383–388.
13. El-Khuffash AF, McNamara PJ. Neonatologist-performed functional echocardiography in the neonatal intensive care unit. Semin Fetal Neonatal Med. 2011;16(1):50–60.
14. Poon WB, Wong KY. Neonatologist-performed point-of-care functional echocardiography in the neonatal intensive care unit. Singapore Med J. 2017;58(5):230–233.
15. Neonatal Research Institute. Bedside Ultrasound Training Program for Neonatologist. http://newbornresearch.org/usprogram/
. Accessed June 25, 2019.
17. CHOP. Neonatology and Critical Care Point-of-care Ultrasound Course. https://chop.cloud-cme.com/default.aspx?P=5&EID=660
. Accessed June 25, 2019.