Perinatal/Neonatal Case Presentation - Nature

Perinatal/Neonatal Case Presentation &

&

&

&

&

&

&

&

&

&

&

&

&

&

Central Venous Catheter Tip in the Right Atrium: A Risk Factor for Neonatal Cardiac Tamponade Jonathan C. Darling, MD Simon J. Newell, MD Omar Mohamdee, MB, ChB Orhan Uzun, MD Catherine J. Cullinane, MB, BAO, BCh Peter R. F. Dear, MD Fatal cardiac tamponade is a well recognised complication of the use of central venous catheters in neonatal patients. There is controversy over optimum catheter tip position to balance catheter performance against risk of adverse events. We report a series of five cases of tamponade occurring in one neonatal unit over a 4 - year period, related to catheter tip placement in the right atrium. Right atrial catheter angulation, curvature or looping ( CA ) was present in all five cases on plain radiograph. It was infrequently seen in other patients over the same period. Review of the literature indicates that CA was present in 6 of the 11 previous cases where the presence or absence of CA can be determined. Where right atrial catheter tip placement is accepted, clinicians should be aware of this characteristic catheter configuration, which is a major risk factor for cardiac tamponade. We recommend that catheter tips should not be placed in the right atrium to avoid risk of tamponade. Journal of Perinatology 2001; 21:461 – 464.

INTRODUCTION Pericardial tamponade is a well recognized complication of the use of central venous catheters in neonatal patients.1 – 10 The mortality is about 65%.2,11 Right atrial catheter tip placement has been considered acceptable practice,5,12,13 and is still widely used in the UK. However, there is increasing consensus, particularly in North America, that catheter tips should not be placed in the right atrium, but ideally in the superior vena cava.2,6,11 Although there are many case reports, only one has

Department of Paediatrics and Child Health ( J.C.D., S.J.N., P.R.F.D. ), University of Leeds and United Leeds Teaching Hospitals Trust, Leeds, UK; Department of Histopathology ( O.M., C.J.C. ), University of Leeds and United Leeds Teaching Hospitals Trust, Leeds, UK; Department of Paediatric Cardiology ( O.U. ), University of Leeds and United Leeds Teaching Hospitals Trust, Leeds, UK. Address correspondence and reprint requests to Jonathan C. Darling, MD, Academic Unit of Paediatrics and Child Health, St. James’ University Hospital, Leeds LS9 7TF, UK.

suggested that distal angulation of the catheter may be a risk factor.14 We report a case series that indicates that distal catheter angulation, curvature or looping (CA) is an important risk factor for tamponade, in addition to the primary risk factor of right atrial tip placement. CASE DESCRIPTIONS In a 4-year period (1993 to 1997) in a regional neonatal unit, five cases of tamponade occurred, three of whom died (Table 1). All five were preterm, required ventilation for hyaline membrane disease, and received artificial surfactant and broad-spectrum antibiotics according to standard protocol. Each case was associated with CA within the right atrium (Figure 1). We defined CA as any angulation or curvature in the distal 5 cm of the catheter not explained by the anatomy of the great veins, including looping where the catheter crosses itself. All except Case 5 presented with acute, unexplained collapse, preceded by tachycardia for 1 to 2 hours (an elevation in rate of about 20 beats per minute above baseline), and followed by severe bradycardia leading to arrest. Of these, only Case 4 was resuscitated successfully, following aspiration of 37 ml pericardial fluid by emergency pericardiocentesis. Tense pericardial effusions containing 8 to 15 ml of TPN fluid were demonstrated at post mortem in Cases 1 to 3. In each case, there was evidence of perforation of the right atrium, with histologic appearances similar to previous descriptions.10,15 Case 5 presented with a subacute deterioration and enlarged cardiac shadow, and tamponade was diagnosed on urgent echocardiography and treated successfully by pericardiocentesis, with aspiration of 10 ml of fluid. In both survivors, there was a dramatic improvement following pericardiocentesis. CA was present in all five cases of tamponade that occurred over a 4-year period in our unit (Figure 1). The unit receives between 500 and 600 admissions annually, and about 125 percutaneous catheters are inserted each year. Therefore over the 4-year period, tamponade occurred in 1% of catheters inserted, and death due to tamponade in 0.6%. To determine the background rate of CA, we reviewed radiographs for a sample of 43 catheters inserted in the Unit during the same period. We found three cases (7%) of CA, all with the tips in the right atrium. These catheter positions were maintained for between 2 and 8 days, with no apparent ill effects. Because 7% of catheters in this sample had CA, we can estimate that over the 4-year period this risk factor would have been present in 35 of 500 catheters placed. Five of these suffered tamponade. The probability of this difference in

Journal of Perinatology 2001; 21:461 – 464 # 2001 Nature Publishing Group All rights reserved. 0743-8346/01 $17

www.nature.com/jp

461

Darling et al.

Central Venous Catheter Tip in the Right Atrium: A Risk Factor for Neonatal Cardiac Tamponade

Figure 1. Radiographic catheter positions in five cases of tamponade. To emphasise catheter configurations and for greater clarity, line drawings have been used rather than radiograph reproductions. The radiographs depicted in (a) to (e) correspond to Cases 1 to 5 in Table 1, respectively, and show catheter positions at the time of tamponade.

frequency of CA in the tamponade and nontamponade groups occurring by chance is less than 0.0003 (Fisher’s exact test). Five catheters were noted to migrate (i.e, their tips moved further in by at least 2 cm) after initially being secured. DISCUSSION We have described five cases of neonatal pericardial tamponade that occurred over a 4-year period in our unit. CA was present in every 462

case. The background rate of CA was low at around 7%. CA is therefore strongly associated with tamponade. A review of the English language literature identified 16 cases of neonatal tamponade due to central venous catheters. In 11 of these, the presence or absence of CA could be ascertained or inferred (Table 2), although it had not been highlighted as a risk factor. Of these, CA was present in 6 (55%). The overall mortality was 7/16 (44%). Thus, the literature indicates that CA is present in many cases of tamponade. Journal of Perinatology 2001; 21:461 – 464


Darling et al.

Table 1 Five Cases of Tamponade Occurring 1993 to 1997 Case no.

1 2 3 4 5

Gestation (wk)

Birth weight (g)

Problems

Catheter type*

Age at catheter insertion (d)

Age at tamponade (d)

Presentation of tamponade

Outcome

24 29 32 25 28

600 1550 1740 890 1130

BPD, PDA HMD HMD BPD, PDA BPD, ileal perforation

SIL SIL SIL PU SIL

2 1 3 2 24

18 3 5 28 46

Acute collapse Acute collapse Acute collapse Acute collapse Subacute deterioration

Died Died Died Survived Survived

BPD, bronchopulmonary dysplasia; PDA, patent ductus arteriosus; HMD, hyaline membrane disease. *SIL (Silastic), Vygon Epicutaneo-Cava-Catheter Art.-No. 2184; PU (polyurethane), Vygon Nutriline Pic Line Code 1352-301.

Table 2 Literature Review of Catheter Angulation and Tamponade Year

Reference

Time to tamponade*

Outcome

Catheter tip positiony

1981 1984 1992 1993 1995 1995 1995 1995 1995 1995 1997

Kulkani et al.20 Agarwal et al.5 Aiken et al.21 Beattie et al.22 Keeney and Richardson2 Keeney and Richardson2 Keeney and Richardson2 Keeney and Richardson2 Keeney and Richardson2 Keeney and Richardson2 Cade and Puntis14

37 d 24 hr 17 hr 4d 22 hr 9 hr 18 hr 4d 3d 5 hr 12 d

Survived Survived Survived Survived Died Died Died Survived Survived Died Survived

In superior vena cava (T) Distal curvature in right atrium (R) Partly looped in right atrium (T) & (R) Coiled in right atrium (T) & (R) Right atrium (T) Superior vena cava (T) Junction right atrium and SVC (T) Junction right atrium and SVC (T) Right atrium (T) Left subclavian vein (T) In right atrium (R)

Catheter angulation (CA)

No No No No

No Yes Yes Yes Yes (inferred) (inferred) (inferred) Yes (inferred) Yes

Table compiled from all neonatal cases reported in the English literature where it was possible to determine whether distal catheter angulation, looping or curvature was present or not either from the text, an included radiograph or diagram, or by inference (because tip position not in the atrium). Gestation range from 25 to 42 weeks. Weight range from 750 to 6260 g. Line types: four Silastic or silicon; four polyurethane; three polyethylene. *Time in hours (hr) or days (d). yPosition determined from report text (T) or radiograph included in report (R).

The presence of CA means firstly that the tip must be touching the atrial wall, and secondly that there is a pressure applied on the wall by the tip. These factors combine with the hyperosmolality of the TPN, the thinness of the wall, and its repeated pulsing against the catheter tip, to lead to inflammation, necrosis, and finally perforation and tamponade.5 – 10 Greater angles of incidence to the atrial wall and catheter stiffness are also likely to be relevant.16 Catheter tip placement in the right atrium is not considered acceptable in adults because of the recognized risk of tamponade.17,18 On the basis of our case series and the previous literature,2,6,11 we recommend that central venous catheter tips should not be placed in the right atrium in neonatal patients. Even placement at the junction of the right atrium and the superior vena cava is hazardous because catheters can migrate several centimeters into the atrium, as seen here and previously reported.19 Journal of Perinatology 2001; 21:461 – 464

Catheters placed in the superior or inferior vena cava may be complicated by thrombosis or hydrothrorax, but these complications are usually more benign than pericardial tamponade.1,11,12 We suggest the following strategy will prevent deaths from tamponade: 1. Use central venous catheters only when there is a clear indication. 2. Do not place catheter tips in the right atrium but in either the superior or inferior vena cava such that their tips are at least 0.5 cm outside the cardiac outline on chest radiograph in small infants, or 1 cm outside in larger infants. 3. Ensure catheter visibility, either through use of radio -opaque catheters or contrast injection when checking position. 4. Perform regular radiographs to identify migration or CA rapidly. 5. Pull back immediately any catheter that has migrated into the right atrium or which shows CA. 6. Train staff to consider emergency pericardiocentesis for acute collapse 463

Darling et al.


when initial resuscitation fails in an infant with a central catheter in place, and to avoid using central line for fluid boluses which may exacerbate tamponade.

We have described five cases of neonatal tamponade over a 4-year period, three of which were fatal. All had CA on plain radiograph. CA was rarely present in other infants who did not have tamponade. CA was present in 6 of 11 cases described in the literature. Thus, we conclude that right atrial catheter tip placement should be avoided. Where right atrial catheter tip placement is accepted, clinicians should be aware that the characteristic catheter configuration we have described is a major warning sign indicating that cardiac tamponade may be imminent, and requires immediate catheter withdrawal. Implementation of the strategies we have described will prevent unnecessary neonatal deaths. Acknowledgments We thank J. Gibbs for his expert cardiological assistance, and the staff in the Neonatal Unit. We also thank A. Craig for comments on the manuscript.

References 1. Goutail - Flaud MF, Sfez M, Berg A, et al. Central venous catheter - related complications in newborns and infants — a 587 case survey. J Pediatr Surg 1991;26:645 – 50. 2. Keeney SE, Richardson CJ. Extravascular extravasation of fluid as a complication of central venous lines in the neonate. J Perinatol 1995;15:284 – 8. 3. Opitz JC, Toyama W. Cardiac tamponade from central venous catheterization: two cases in premature infants with survival. Pediatrics 1982;70:139 – 40. 4. Franciosi RA, Ellefson RD, Uden D, Drake RM. Sudden unexpected death during central hyperalimentation. Pediatrics 1982;69:305 – 7. 5. Agarwal KC, Ali K, Falla A, Amato JJ. Cardiac perforation from central venous catheters: survival after cardiac tamponade in an infant. Pediatrics 1984;73:333 – 8. 6. Sasidharan P, Billman D, Heimler R, Nelin L. Cardiac arrest in an extremely low birth weight infant: complication of percutaneous central venous catheter hyperalimentation. J Perinatol 1996;16:123 – 6.

464

7. Wirrell EC, Pelausa EO, Allen AC, Stinson DA, Hanna BD. Massive pericardial effusion as a cause for sudden deterioration of a very low birthweight infant. Am J Perinatol 1993;10:419 – 23. 8. Garg M, Chia - Chi C, Merritt RJ. An unusual case presentation: pericardial tamponade complicating central venous catheter. J Perinatol 1989;9:456 – 7. 9. Cherng YG, Cheng YJ, Chen TG, Wang CM, Liu CC. Cardiac tamponade in an infant. Anaesthesia 1994;49:1052 – 4. 10. Byard RW, Bourne AJ, Moore L, Little KE. Sudden death in early infancy due to delayed cardiac tamponade complicating central venous line insertion and cardiac catheterization. Arch Pathol Lab Med 1992;116:654 – 6. 11. Giacoia GP. Cardiac tamponade and hydrothorax as complications of central venous parenteral nutrition in infants. J Parenter Enteral Nutr 1991;15:110 – 3. 12. Chathas MK, Paton JB, Fisher DE. Percutaneous central venous catheterization. Three years’ experience in a neonatal intensive care unit. Am J Dis Child 1990;144:1246 – 50. 13. Anonymous. Central venous access in children [ editorial ] . Lancet 1991;338:1301 – 2. 14. Cade A, Puntis JW. Pericardial parenteral nutrition infusion following acute angulation of a neonatal polyurethane central venous catheter tip. Clin Nutr 1997;16:263 – 4. 15. Rogers BB, Berns SD, Maynard EC, Hansen TW. Pericardial tamponade secondary to central venous catheterization and hyperalimentation in a very low birthweight infant. Pediatr Pathol 1990;10:819 – 23. 16. Gravenstein N, Blackshear RH. In vitro evaluation of relative perforation potential of central venous catheters: comparsion of materials, selected models, number of lumens, and angles of incidence to simulated membrane. J Clin Monit 1991;7:1 – 6. 17. Collier PE, Goodman GB. Cardiac tamponade caused by central venous catheter perforation of the heart: a preventable complication. J Am Coll Surg 1995;181:459 – 63. 18. Harford FJ, Kleinsasser J. Fatal cardiac tamponade in a patient receiving total parenteral nutrition via a Silastic central venous catheter. J Parenter Enteral Nutr 1984;8:443 – 6. 19. Fischer GW, Scherz RG. Neck vein catheters and pericardial tamponade. Pediatrics 1973;52:868 – 71. 20. Kulkani PB, Dorand RD, Simmons EJ. Pericardial tamponade: complication of total parenteral nutrition. J Pediatr Surg 1981;16:735 – 6. 21. Aiken G, Porteous L, Tracey M, Richardson V. Cardiac tamponade from a fine silastic central venous catheter in a premature infant. J Paediatr Child Health 1992;28:325 – 7. 22. Beattie PG, Kuschel CA, Harding JE. Pericardial effusion complicating a percutaneous central venous line in a neonate. Acta Paediatr 1993;82:105 – 7.

Journal of Perinatology 2001; 21:461 – 464