Rates of Intracardiac Umbilical Venous Catheter Placement in Neonates

Transcription

1 ORIGINAL RESEARCH Rates of Intracardiac Umbilical Venous Catheter Placement in Neonates Andrei Harabor, MD, Amuchou Soraisham, MD, DM, MSc Received September 27, 2013, from the Section of Neonatology, Department of Pediatrics, University of Calgary, Calgary, Alberta, Canada. Revision requested November 18, Revised manuscript accepted for publication December 16, Address correspondence to Andrei Harabor, MD, Section of Neonatology, Department of Pediatrics, Foothills Medical Center, th St NW, Room 780, Calgary, AB T2N 2T9, Canada. - services.ca Abbreviations CI, confidence interval; ELBW, extremely low birth weight; IVC, inferior vena cava; LA, left atrium; NICU, neonatal intensive care unit; OR, odds ratio; QA, quality assurance; QI, quality improvement; RA, right atrium; TNE, targeted neonatal echocardiography; UVC, umbilical venous catheter doi: /ultra Objectives To review umbilical venous catheter (UVC) placement in neonates who underwent targeted neonatal echocardiography (TNE) and to correlate catheter tip placement on TNE and anteroposterior thoracoabdominal radiography. Methods We conducted a retrospective analysis of 51 neonates who had UVC positions assessed by TNE and radiography in a neonatal intensive care unit (NICU). A single operator performed all TNE examinations. The final radiographic catheter placement was taken from the image closest to the time of echocardiography. Fisher exact, χ 2, and t tests were used as appropriate. Results Among the 51 neonates who had catheters placed for 24 hours or more, TNE was performed on 48 in the first 48 hours, 2 at day 6, and 1 at day 9. Thirty-six neonates were extremely low birth weight (ELBW; <1000 g). Twenty-nine had good catheter tip positions, and 22 had catheters inside the heart (10 in the right atrium [RA], 3 at the foramen ovale, and 9 in the left atrium [LA]). Twenty neonates with catheter tips in the heart were ELBW, including 8 with catheters in the LA. The ELBW neonates were more likely to have catheters in the heart than non-elbw neonates (20 of 36 versus 2 of 15; P =.01; odds ratio [OR], 8.1; confidence interval [CI], ). Good placement on TNE varied widely in relation to thoracic vertebral landmarks on radiography: from the T7 8 interspace to T11. When radiography showed a catheter tip at T9 T10, there was no difference in the proportion of neonates with a good catheter position versus malposition (8 of 22 versus 8 of 29; P =.55; OR, 0.67; CI, ). Conclusions A high proportion of ELBW neonates in a busy NICU had UVCs placed with the tips in the RA or LA despite common placement practices. We recommend adding TNE to radiography to position UVCs, especially in ELBW neonates. Key Words echocardiography; neonate; point-of-care ultrasound; radiography; sonography; umbilical venous catheter Umbilical venous catheters (UVCs) have been used in neonates for more than 60 years. 1 Initially they were viewed as emergency vascular access devices only and were used for exchange transfusion or blood administration. With time and improvement in materials, prolonged use as central catheters became widespread, and data pointed to value in preserving them in place for as long as 28 days in very low-birth-weight neonates. 2 As practices evolved, most units including ours use UVCs for the first 7 to 10 days of life by the American Institute of Ultrasound in Medicine J Ultrasound Med 2014; 33:

2 Complications of UVC use include thrombosis, pericardial or peritoneal effusion, liver tissue necrosis, arrhythmia, and death. The first reports were published soon after UVCs were used on a larger scale and continued to appear over the ensuing decades, mostly associated with malposition of the catheter tip There is still some debate about ideal placement, and although by far the most often advocated position is at the junction of the inferior vena cava (IVC) and right atrium (RA), there are still references accepting a UVC tip position in the RA 13 or recommending placement at a certain distance from the cardiac shadow to avoid the pericardial space. 12 A UVC is usually inserted blindly at a distance computed using either the shoulder-to-umbilicus length graph of Dunn 14 or the formula of Shukla and Ferrara based on birth weight. 15 In a 2008 online tutorial on UVC insertion, the New England Journal of Medicine suggested using the formula of Shukla and Ferrara 15 [(weight 3 + 9)/2 + 1, in centimeters] for intended placement in the IVC just outside the RA. 16 After initial placement, the catheter position is adjusted on the basis of chest and abdominal radiography. 17 In the 1960s and 1970s, several articles reviewed the assessment of UVC position with chest and abdominal radiography, and there have been more recent reviews as well Anteroposterior thoracoabdominal radiography supplemented with a lateral radiographic view of the chest and abdomen was considered standard initially. In all those articles, there is precious little direct information about bony landmarks and the UVC tip position, and mostly indirect references are made to the diaphragm and heart shadow when cases are discussed. Generally, the emphasis is on avoiding intrahepatic placement. In 1982, George et al 24 were the first to publish on localizing the catheter tip by sonography. Five more articles on this topic were published since then, 3 of them in 2011, in keeping with the recent spread of bedside sonographic capabilities in neonatal intensive care units (NICUs) across the world There are some conflicting findings but 1 commonality is the excellent ability of determining intracardiac placement, as expected by the principle of the method. In many units, as in ours, thoracoabdominal radiography is still the main imaging modality used at the time of UVC placement. This quality improvement/quality assurance (QI/QA) study was prompted by the clinical observation during targeted neonatal echocardiography (TNE) that a substantial number of the UVCs were found in the heart. We thus proceeded to a limited internal audit of the placement of UVCs. We reviewed the radiographs of a number of neonates who had TNE studies done within the first few days after birth to document the position of the UVC tip on both TNE and thoracoabdominal radiography and to establish any correlation and practical pitfalls in placing the UVC tip in the IVC between the hepatic vein confluence and the RA opening. Materials and Methods We conducted a retrospective analysis of neonates who had UVC positions assessed both by thoracoabdominal radiography and TNE in a tertiary care NICU between November 2009 and April As a QI/QA project, we reviewed the radiographs of neonates who had both an UVC in place for more than 24 hours as well as 1 or more TNE studies. The position of the UVC tip with regard to thoracic vertebrae on radiography closest to the time of the TNE study was noted. All radiographic examinations were done less than 24 hours from the TNE study. Serial radiographs were reviewed, and if placement was not consistent, the charts were checked for evidence of repositioning by the medical team unless the TNE report mentioned repositioning of the catheter as an outcome of the TNE study. Umbilical venous catheter placement is attempted in our NICU for neonates weighing less than 1250 g for secure access in the first week of life, for neonates of any size with high illness acuity on the first day of life, including resuscitation at birth, and in neonates with very difficult peripheral venous access. The insertion length is calculated beforehand using either the Dunn graph 14 or the formula of Shukla and Ferrara 15 according to operator preference. After placement, a thoracoabdominal radiograph is taken, and the tip position is adjusted if needed. The aim is to have the catheter tip no higher than the T8 9 vertebral interspace and no lower than the T10 vertebral body, ideally on the T9 vertebral body. If the diaphragm position is very high or very low, the catheter tip might be aimed slightly more cephalad or caudal than usual. The choice of repeating radiography immediately after catheter repositioning is left to the operator. In our unit, it is not routine to perform lateral chest and abdominal radiography for UVC placement. As the diaphragm has a central concavity with regard to the heart and large vessels, some approximation is used for tip location on a lateral radiograph as well. 26 That factor, combined with the issue of manipulating a fragile baby receiving continuous positive airway pressure or intubation under a sterile field, weighed into our decision to use only anteroposterior thoracoabdominal radiography routinely for umbilical line placement J Ultrasound Med 2014; 33:

3 Targeted neonatal echocardiography is performed for accepted indications and according to the current protocols. 30,31 It is ordered at the discretion of the attending neonatologist. We use a Vivid I ultrasound machine (GE Healthcare, Milwaukee, WI) with a 10S-RS probe usually set at 10 MHz. Results are reported right away both verbally and on a standardized form placed in the patient s chart. During TNE, in all neonates who have a UVC in situ, it is standard to visualize the suprahepatic IVC and the position of the catheter tip in the IVC or within the heart. If the catheter is not visualized in the heart or the proximal IVC, we do not routinely look for an intrahepatic or a more distal location of the UVC tip. All studies were performed by a single operator (A.H.). Our local and provincial guidelines allow for publication of QI/QA project data without Institutional Ethics Board review. The projects are screened for fulfilling the QI/QA criteria with a tool based on the publications of Lynn et al 32 and Baily et al. 33 In addition, the chair of our Institutional Ethics Board was consulted and confirmed the QI/QA nature of the project. In this study, we considered the UVC position to be good if the tip of the UVC was located in the IVC between the hepatic vein confluence and the RA opening. We examined the position of the UVC tip on TNE and with regard to thoracic vertebrae on thoracoabdominal radiography. We used a Student t test for continuous variables and a χ 2 or Fisher exact test for categorical variables when appropriate. P <.05 was considered significant. Results Between November 2009 and April 2011, 51 neonates who had 1 or more TNE studies also had UVCs placed for 24 hours or more. Targeted neonatal echocardiography was performed on 48 of those neonates within the first 48 hours after birth, 2 at day 6, and 1 at day 9. Of the 51 neonates, 36 were extremely low birth weight (<1000 g). Figure 1 shows a flow diagram of the study participants. Of the 51 neonates, 29 (57%) had good UVC positions in the suprahepatic IVC up to the RA entry. The remaining 22 neonates had UVCs inside the heart: 10 in the RA, 3 with the tip at the foramen ovale opening, and 9 in the left atrium (LA). Of the 51 neonates, 9 had the UVC repositioned after the initial radiographic examination. Seven of these neonates still had UVC tips in the heart on TNE. Of the 22 neonates with UVCs in the heart, 20 were ELBW, including 8 neonates with UVCs in the LA. Compared to non-elbw neonates, ELBW neonates were more likely to have UVCs in the heart (20 of 36 [55.5%] versus 2 of 15 [13.3%]; P =.01; odds ratio [OR], 8.1; confidence interval [CI], ). Good UVC placement as assessed by TNE varied widely in relation to thoracic vertebral landmarks on radiog - raphy, ranging from the T7 8 interspace to T11 in our study. The UVC tip was at T9 or lower in 21 of 29 neonates with good UVC positions compared to 8 of 22 with intra - cardiac placement (P =.015; Figure 2). The catheter tip was at T9 T10 on radiography in 8 neonates with both good positions and intracardiac placement (P = 0.55; OR, 0.67; CI, ; Figure 3). Figure 4 shows the positions of the UVC tips on TNE compared to thoracic vertebral levels on radiography. Between T7 T8 and T10, there were as many catheters inside the heart as outside for any given level. Four catheters were adjusted with TNE and had radiography afterward; tips were placed at T10 in 2 neonates at T9 and T11 in 1 each. During the same period, there were 58 ELBW neonates who had UVCs placed that were considered in good positions and who did not undergo TNE. The vertebral level distribution of the catheter tips in this group was no different than that in the group who had TNE (30 of 58 versus 16 of 36 had the tip at T9 T10; P =.53; OR, 0.75; CI, ). Discussion Umbilical venous catheterization is a frequently performed procedure in the NICU. The umbilical vein offers quick central access in a small or sick neonate and in cases of difficult peripheral vein access. It provides secure access with a relatively low infection risk for at least 1 week. It is also an essential neonatal skill taught to pediatric residents and neonatology fellows and as such will continue to be largely used in academic centers. The currently most accepted ideal location of the catheter tip is the short portion of the IVC between hepatic vein drainage and entry into the RA. Figure 1. Flow diagram of the study participants. J Ultrasound Med 2014; 33:

4 The usual formulas used for initial placement have been criticized for low performance. 34 Some investigators have tried to circumvent formulas and ensure good placement by inserting the UVC under electrocardiographic guidance by using it essentially as an electrocardiographic electrode. This approach showed good results in 2 small studies, 35,36 but the final position confirmation was done by radiography alone. We are not aware of routine use of this technique at this time. Our study further shows considerable difficulty in locating an ideal catheter tip position by using thoracoabdominal radiography, especially in ELBW neonates. Interpretation of our data should take into account specific limitations. First, as this study was an audit prompted by a clinical observation, it was retrospective in nature, and a selection bias due to unknown factors is possible. However, as the other ELBW neonates who did not undergo TNE had UVCs placed at a similar level, the selection bias was unlikely to substantially influence the conclusions. Figure 2. Patients with UVCs placed in the heart at T8 or higher versus T9 or lower. The sample size, although of the same order of magnitude as in previous studies, was small overall, and statistical calculations have obvious limitations in this context. Second, variable time had passed between radiography and TNE, although all of the TNE studies were performed less than 24 hours after radiography showing UVC placement. It is theoretically possible that some migration of the catheter had occurred between the time of the radiography and the TNE. In our unit, the catheter position at the umbilical stump is checked and noted several times a day by nursing staff, and if any appreciable displacement is noted, another radiographic examination is done. There is anecdotal mention of umbilical lines migrating proximally within the first 24 to 48 hours, especially in smaller neonates, which is attributed to drying of the Wharton jelly and shortening of the cord. Although neonatologists and NICU staff mention this factor conversationally, it is not clearly documented in any scientific literature or textbook. We could find only 1 article that mentioned UVC migration, and in that case, it was linked to abdominal distention. 37 Our personal experience speaks to the fact that, in general, the UVC tip is frequently placed in the heart in smaller neonates, even if they are scanned immediately after radiography. Our data show that aiming at particular thoracic vertebrae with the catheter tip does not guarantee appropriate placement. Between T7 8 and T10, there were as many catheters inside the heart as outside for any given level (Figure 4). Of course, the lower one aims, the lesser the risk that the catheter will be in the heart, but many of them will be left in or below the ductus venosus. Indeed, we have seen several cases of catheters, considered in good positions at T9 10 or T10 initially on thoracoabdominal radiography but not visualized on TNE, that were clearly Figure 3. Patients with UVCs placed in the heart when the tips were at T9 T10 versus other positions. Figure 4. Percentages of UVC tips inside the LA or RA or outside the heart by thoracic vertebral level. Decimal numbers indicate vertebral interspaces J Ultrasound Med 2014; 33:

5 malpositioned in one of the portal branches on subsequent radiographs and were removed at that point. A study by Greenberg et al 25 used sonography rather than radiography for initial placement of UVCs. It still showed that an appropriately positioned tip could be anywhere from T7 to T10. It seems from the data provided that the imaging was done by the neonatal team as a specially acquired general sonographic skill, specifically to check the UVC location. Our study was close methodologically to that of Ades et al, 26 and our results were similar overall to theirs and the results of Michel et al. 28 Our population, though, had more ELBW neonates than either study. To our knowledge, this study is the first to document UVC tip position in the context of standardized TNE, performed by a neonatal team rather than radiologists or cardiologists. On the basis of the results of this QA project, we are now using TNE assessment more often for UVC insertion to localize the tip position, especially for neonates born before 32 weeks. Overall, we interpret our data as further evidence that thoracoabdominal radiography alone is not the most adequate imaging modality for positioning the UVC tip in an ideal location. This finding is especially true for ELBW neonates, in whom the margin for ideal placement is only a few millimeters, so it is not very surprising that more than half end up with the tip in the heart. The appropriate advanced statistical method for a study such as this is very difficult to determine. Calculation of specificity and sensitivity requires that one of the methods be considered a reference standard. That is not the case here. The methods are mostly complementary. Radiography is much better at localizing the UVC tip when it is in the liver, and that factor was the focus of the original articles talking about assessing placement. Targeted neo - natal echocardiography is clearly better at localizing the UVC tip in the heart. They are both expected to do well when the UVC location is grossly too far. Radiography is the standard of care due to availability and historical development, but it is not the diagnostic reference standard in a true sense, which led Michel et al 28 to choose a subjective reference standard for their study. Statistical techniques that compare agreement of two methods with categorical results vary from more traditional, simpler, and widely used to new and complex The appropriateness and interpretation of the most commonly used parameter, the κ statistic (with its variants), has been deemed problematic. 38,41 In this context, using a simple description of the results is likely to be the most honest and clear option. The clinical question that prompted the QI/QA project Is the performance of thoracoabdominal radiography alone adequate in daily clinical practice? was answered fairly when half of the ELBW neonates still had the UVC tip in the heart. We suggest that sonography is a necessary tool for achieving correct UVC placement as an add-on to thoraco - abdominal radiography, especially in ELBW neonates. This approach should be straightforward in the context of TNE, although it could also be a general point-of-care sonographic skill. We believe that this factor should be a further driving reason for acquisition of TNE capabilities in level 3 NICUs. The availability of affordable and portable ultrasound equipment will be of great assistance in this regard. References 1. Diamond LK, Allen FH Jr, Thomas WO Jr. Erythroblastosis fetalis, VII: treatment with exchange transfusion. N Engl J Med 1951; 244: Butler-O Hara M, Buzzard CJ, Reubens L, McDermott MP, DiGrazio W, D Angio CT. A randomized trial comparing long-term and short-term use of umbilical venous catheters in premature infants with birth weights of less than 1251 grams. Pediatrics 2006; 118:e25 e Butler-O Hara M, D'Angio CT, Hoey H, Stevens TP. An evidence-based catheter bundle alters central venous catheter strategy in newborn infants. J Pediatr 2012; 160: e2 4. Oski FA, Allen DM, Diamond LK. Portal hypertension a complication of umbilical vein catheterization. Pediatrics 1963; 31: Walker D, Pellett JR. Pericardial tamponade secondary to umbilical vein catheters. J Pediatr Surg 1972; 7: Emmrich P, Baumann W, Stechele U. The possibilities and dangers of longterm infusion therapy in severely ill newborn and premature infants. Eur J Intensive Care Med 1975; 1: Kulkarni PB, Dorand RD. Hydrothorax: a complication of intracardiac placement of umbilical venous catheter. J Pediatr 1979; 94: Onal EE, Saygili A, Koç E, Türkyilmaz C, Okumus N, Atalay Y. Cardiac tamponade in a newborn because of umbilical venous catheterization: is correct position safe? Paediatr Anaesth 2004; 14: Hermansen MC, Hermansen MG. Intravascular catheter complications in the neonatal intensive care unit. Clin Perinatol 2005; 32: Sehgal A, Cook V, Dunn M. Pericardial effusion associated with an appropriately placed umbilical venous catheter. J Perinatol 2007; 27: Haase R, Hein M, Thäle V, Vilser C, Merkel N. Umbilical venous catheters: analysis of malpositioning over a 10-year period [in German]. Z Geburtshilfe Neonatol 2011; 215: Nowlen TT, Rosenthal GL, Johnson GL, Tom DJ, Vargo TA. Pericardial effusion and tamponade in infants with central catheters. Pediatrics 2002; 110: Wortham BM, Rais-Bahrami K. Umbilical vein catheterization. In: MacDonald MG, Ramasethu J (eds). Atlas of Procedures in Neonatology. Philadelphia, PA: Lippincott Williams & Wilkins; 2007: Dunn PM. Localization of the umbilical catheter by post-mortem measurement. Arch Dis Child 1966; 41: J Ultrasound Med 2014; 33:

6 15. Shukla H, Ferrara A. Rapid estimation of insertional length of umbilical catheters in newborns. Am J Dis Child 1986; 140: Anderson J, Leonard D, Braner DA, Lai S, Tegtmeyer K. Videos in clinical medicine: umbilical vascular catheterization. N Engl J Med 2008; 359:e Ringer SA, Gray JE. Common neonatal procedures. In: Cloherty JP, Eichenwald EC, Hansen AR, Stark AR (eds). Manual of Neonatal Care. Philadelphia, PA: Lippincott Williams & Wilkins; 2012: Baker DH, Berdon WE, James LS. Proper localization of umbilical arterial and venous catheters by lateral roentgenograms. Pediatrics 1969; 43: Campbell RE. Roentgenologic features of umbilical vascular catheterization in the newborn. Am J Roentgenol Radium Ther Nucl Med 1971; 112: Weber AL, DeLuca S, Shannon DC. Normal and abnormal position of the umbilical artery and venous catheter on the roentgenogram and review of complications. Am J Roentgenol Radium Ther Nucl Med 1974; 120: Sanders CF. The placement of the umbilical venous catheter in the newborn and its relationship to the anatomy of the umbilical vein, ductus venosus and portal venous system. Clin Radiol 1978; 29: Narla LD, Hom M, Lofland GK, Moskowitz WB. Evaluation of umbilical catheter and tube placement in premature infants. Radiographics1991; 11: Oestreich AE. Umbilical vein catheterization: appropriate and inappropriate placement. Pediatr Radiol 2010; 40: George L, Waldman JD, Cohen ML, et al. Umbilical vascular catheters: localization by two-dimensional echocardio/aortography. Pediatr Cardiol 1982; 2: Greenberg M, Movahed H, Peterson B, Bejar R. Placement of umbilical venous catheters with use of bedside real-time ultrasonography. J Pediatr 1995; 126: Ades A, Sable C, Cummings S, Cross R, Markle B, Martin G. Echocardiographic evaluation of umbilical venous catheter placement. J Perinatol 2003; 23: Fleming SE, Kim JH. Ultrasound-guided umbilical catheter insertion in neonates. J Perinatol 2011; 31: 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: 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: Kluckow M, Seri I, Evans N. Functional echocardiography: an emerging clinical tool for the neonatologist. J Pediatr 2007; 150: Mertens L, Seri I, Marek J, et al. Targeted neonatal echocardiography in the neonatal intensive care unit: practice guidelines and recommendations for training. Writing group of the American Society of Echocardiography (ASE) in collaboration with the European Association of Echocardiography (EAE) and the Association for European Pediatric Cardiologists (AEPC). J Am Soc Echocardiogr 2011; 24: Lynn J, Baily MA, Bottrell M, et al. The ethics of using quality improvement methods in health care. Ann Intern Med 2007; 146: Baily MA, Bottrell M, Lynn J, Jennings B; Hastings Center. The ethics of using QI methods to improve health care quality and safety. Hastings Cent Rep 2006; 36:S1 S Verheij GH, Te Pas AB, Witlox RS, Smits-Wintjens VE, Walther FJ, Lopriore E. Poor accuracy of methods currently used to determine umbilical catheter insertion length. Int J Pediatr 2010; 2010: Tsui BC, Richards GJ, Van Aerde J. Umbilical vein catheterization under electrocardiogram guidance. Paediatr Anaesth 2005; 15: Nodari S, Agostino R, Bucci G. Proper location of umbilical venous catheters by internal electrocardioscopy. Acta Paediatr Scand 1973; 62: Salvadori S, Piva D, Filippone M. Umbilical venous line displacement as a consequence of abdominal girth variation. J Pediatr 2002; 141: Grant JM. The fetal heart rate trace is normal, isn t it? Observer agreement of categorical assessments. Lancet 1991; 337: Lin L, Hedayat AS, Wu W. A unified approach for assessing agreement for continuous and categorical data. J Biopharm Stat 2007; 17: Henriques T, Antunes L, Bernardes J, Matias M, Sato D, Costa-Santos C. Information-based measure of disagreement for more than two observers: a useful tool to compare the degree of observer disagreement. BMC Med Res Methodol 2013; 13: Nelson JC, Pepe MS. Statistical description of interrater variability in ordinal ratings. Stat Methods Med Res 2000; 9: J Ultrasound Med 2014; 33: