We tested the hypothesis that ECG guidance for central venous catheter (CVC) placement in a large paediatric population of neurosurgical patients in routine daily practice is not inferior to chest radiograph confirmation. However, the primary clinical goal in routine practice was to decrease the incidence of misplacement. Owing to the retrospective, noninterventional nature of this study and an anonymised review process of the radiographs, there was a waiver for further review by the Institutional Review Board of the University of Würzburg, Würzburg, Germany (correspondence on 21 December 2009) and records were reviewed in compliance with German federal data protection law. One hundred and fifty-five records from procedures between 2007 and 2009 were included in the study.
Standard operating procedures during this time required maximal barrier precautions, real-time 2-D ultrasound guidance of the vessel puncture and confirmation of the catheter tip position by ECG with documentation in the chart. Radiopaque double lumen paediatric central lines Arrow Blueflex (Teflex Medical GmbH, Kernen, Germany) were in use and the Seldinger wire was connected via an alligator clamp to a Certodyn adapter (B. Braun, Melsungen, Germany) for the intravascular ECG recording. Catheters were considered to be placed correctly if a clear rise in p-wave amplitude was observed (‘p-atriale’).
Digital radiographs (taken on average 60 min after the end of the operation on arrival at the paediatric ICU) were analysed using the Patient Archiving and Communication System (Image Devices, Idstein, Germany).
The peak of the concave shadow, which is visible through superposition of the distal superior vena cava (SVC) and the right atrium, was defined as radiographic transition of the SVC into the right atrium. The carina was defined as landmark for an extracardiac position of the central line tip because the pericardium is not visible on conventional chest radiographs. A distance between the catheter tip and the carina of ΔC = ±0.5 cm was defined as carina alignment and extracardiac catheter position and a catheter placed >2 cm above the carina was considered too high and a catheter with ΔC > 1 cm and placement >1 cm below the transition of the SVC into the right atrium too low.
In addition, the angle between the lateral wall of the vena cava and the catheter tip (α ct) was measured and an angle of α ct more than 40° was considered to be an increased risk of vessel perforation and was registered as such.
Normal distribution was tested using the Kolgomorov–Smirnov test. Age, sex and cannulated vessel were tested with regard to ΔC using multivariate analysis and post-hoc Duncan test (P = 0.05) (SPSS 6.0; IBM Deutschland, Ehningen, Germany).
Of the 155 identified central lines, 137 were placed via the right internal jugular vein (RIJ), nine via the left internal jugular vein (LIJ), two via the right subclavian vein (RSV) and seven via the left subclavian vein (LSV). The sex distribution was 87 boys and 68 girls, the age range from 0 to 14 years with 44 infants between 0 and12 months, 47 toddlers between 13 and 72 months and 64 children between 3 and 14 years.
One hundred and thirty-eight catheter tips (89%) were positioned correctly according to the definitions used in our protocol. In 17 cases (11%), a malposition was diagnosed by chest radiograph despite ECG confirmation during placement (Table 1). Of those, 11 were placed via the RIJ, three via the LIJ, one via the RSV and two via the LSV. Eight catheters (all placed via the RIJ) were too high. However, the catheters were not relocated and the clinical relevance of this malposition is negligible. If those are excluded and only clinically relevant malpositions are counted, the clinical success rate increased to 94%.
Although nine of 17 malpositions were recorded in the group of toddlers between 12 and 72 months of age, neither age nor sex or height were a significant factor in the multivariate analysis. Complications such as pneumothorax, vessel perforation or thrombosis were not observed.
The success rate of 89% per protocol and 94% for clinical relevance in our population is comparable to the 92% described in adults by Salmela and Aromaa1 and David et al.2 and the 90.5% observed by Neubauer3 in a paediatric population. Weber et al.4 were successful in 44 of 50 patients between 0 and17 years. In studies without ECG guidance, success rates of only 70% have been reported.2,5,6 The observed increased success rates with ECG guidance are of independent clinical importance whether chest radiograph confirmation is routinely performed or not. A missing p-atriale can be a sign of a serious complication (e.g. extravascular placement, aberrant intravascular placement). Without ECG guidance, there are no such warning signs even if real-time ultrasound is used.
Five out of 16 catheters placed via the left side were repositioned and all cases with an angle of α ct more than 40° occurred in the left-sided group in which malpositions are generally more common.7 This might be because of the fact that catheters were pulled back to the point where an increase of the p-wave started. Although this is a reasonable manoeuvre for right-sided CVC placement, left-sided catheters should not be pulled back and the position double-checked by radiograph.
This study is limited by its retrospective design and the relatively low number of left-sided catheters. Nevertheless, it reflects daily practice in a large paediatric cohort and is therefore relevant for clinical decision pathways regarding CVC placement and position confirmation, reducing the amount of unnecessary chest radiographs.
Acknowledgements relating to this article
Assistance with the study: the authors would like to thank W Schummer for expert help and valuable discussions.
Financial support and sponsorship: funding for this study was provided by departmental funds from the Klinik und Poliklinik für Anästhesiologie, Universitätsklinikum Würzburg, Julius-Maximilians-University of Würzburg, Würzburg, Germany.
Conflicts of interest: SSS is shareholder and employee of Deutsches Beratungszentrum für Hygiene (BZH GmbH).
1. Salmela L, Aromaa U. Verification of the position of a central venous catheter by intra-atrial ECG. When does this method fail? Acta Anaesthesiol Scand
2. David JS, Tazarourte K, Perfus JP, Savary D. Is ECG-guidance a helpful method to correctly position a central venous catheter during prehospital emergency care? Acta Anaesthesiol Scand
3. Neubauer AP. Percutaneous central i.v. access in the neonate: experience with 535 silastic catheters. Acta Paediatr
4. Weber F, Buitenhuis M, Lequin MH. Determination of the optimal length of insertion of central venous catheters in pediatric patients by endovascular ECG. Minerva Anestesiol
5. Conces DJ Jr, Holden RW. Aberrant locations and complications in initial placement of subclavian vein catheters. Arch Surg
6. Peres PW. Positioning central venous catheters: a prospective survey. Anaesth Intensive Care
7. Kremser J, Kleemann F, Reinhart K, Schummer W. Optimized method for correct left-sided central venous catheter placement under electrocardiographic guidance. Br J Anaesth