Difficulty in using ultrasonography for central venous cannulation in children: ‘a case of old dogs and new tricks’?

* E-mail: sverghese{at}cnmc.org

Editor—We would like to comment on the conclusions reached by Grebenik and colleagues in a recent paper titled ‘NICE guidelines for central venous catheterization in children. Is the evidence base sufficient?’1 In this paper, the authors questioned the recommendations on the use of the ultrasound from recent guidelines from the National Institute for Clinical Excellence (NICE).2 Based on their own study, Grebenik and colleagues concluded that there is currently insufficient evidence to support the use of ultrasound guidance as the preferred method for elective central venous catheterization in children.1 While not wishing to elaborate on whether or not the available evidence justifies the NICE guidelines, we would like to make some comments regarding Grebenik and colleagues' interpretation of the methodology and conclusions of two studies from our institution that were used to support the NICE guidelines.3 4

In both our studies, the procedures were performed by anaesthesia fellows ‘who were closely guided and supervised’ by experienced practitioners. The apparent prolongation of catheterization time is attributable more to definition than ‘operator inexperience’. Grebenik and colleagues defined catheterization time as time ‘from the moment of needle insertion through the skin to the successful placement of the guidewire’. We defined catheterization as the time from skin puncture until the aspiration of blood from both ports of the double lumen catheter. This included passing of guide wire, then a dilator, replacing the dilator with the double lumen catheter, removal of the guide wire and ultimately aspirating blood from both the catheter ports. Clearly our definition includes more steps of the process.

In our first study, there were 95 patients, all of whom were <12 months of age and <10 kg.3 Fifty-two infants underwent catheterization using the landmarks method, while 43 infants had ultrasound-assisted catheterization. We attribute the complete absence of carotid puncture in the ultrasound group to the ability to differentiate, visually, the jugular from the carotid vessel. Incidentally, our second study published later but completed first, involved an entirely different set of patients from the first.4

Grebenik's comment regarding the relative success rate and carotid puncture rates among the different studies must be viewed in the context of patient ages and sizes. They also noted that in their study ‘all failures except one occurred in patients who are <10 kg in weight and under one year of age’. However, they do not provide a denominator for patients in this age and size category in their study. Thus it is impossible to compare results with our studies in which all the subjects were <10 kg and <1 yr old.

It has been clear to us that experience with the ultrasound device is critical to a consistent high success and low carotid puncture rates, especially for those of us who did not have the opportunity to develop excellent hand–eye coordination on computers and arcade games in our youth. Learning to watch the screen while manipulating the needle was painfully slow. Before starting the study, we initially used the method described by Alderson, slowly developing increased ability to perform ‘real-time’ punctures.5 With our direct hands-on supervision and direction, our fellows were able to develop these capabilities much faster.

There is no question that the Site-rite probe is cumbersome. We have abandoned attempting to pass the needle through the needle guide and merely use the needle guide bracket as an indicator of the center of the probe. Hopefully newer probes can be smaller and easier to use. Most of our staff initially opposed the use of ultrasound guided catheter placement, but have been generally won over. This very similar to the experience of D. H. Scott, who reports on the inherent fear of new technologies that at best can de-skill practitioners in what used to be an everyday task or, at worst, prove to be dangerous.6 We never claimed that not using ultrasound guidance constitutes substandard practice. However, in newborns and infants who have a higher degree of anatomical variation in the position of their neck vessels, we find the use of imaging invaluable.7 When skilled practitioners are able to supplement their manual skill with technology that improves success and virtually eliminates carotid puncture, it is hard to look back. It is time to accept that the complexity of modern medicine exceeds the capacity of unaided human skills. May be our hope will lie with the younger generation of practitioners who are not afraid of relying on technology to improve patient safety.

S. T. Verghese* and W. A. McGill

Washington, DC, USA


 
* E-mail: thomas.grau{at}med.uni-heidelberg.de

Editor—Central venous cannulation in infants can be difficult and challenging. This procedure is characterized by a learning curve and regular practice is needed to achieve skill and competence.

In Grebenik's study,1 there were 65 patients in the control group and 69 patients in the ultrasound group. Ten patients (14%) dropped out in the ultrasound group owing to unavailability of the ultrasound machine. The text provides limited information on how the puncture was performed, and what ultrasound landmarks were used to detect the jugular vein. Were both the artery and the vein detected? How were the landmarks defined? Did they use ultrasound in two perpendicular planes and what type of probe was used?

Grebenik and colleagues had success rates of ~89% in the landmark group, which is 10% higher than that found by Verghese3 4 and Alderson,5 and their results in the ultrasound group were ~20% lower than other studies.

In their analysis, the authors made a mistake when comparing the success rates. Proper application of the {chi}2-test results in a non-significant difference between both procedures. The success rates of the landmark group (58/65) vs the ultrasound group (46/59) results in a {chi}2 with Yates correction of 2.1 with one degree of freedom (P=0.14). Fisher's exact test showed a P-value of 0.142 and for arterial punctures ({chi}2=0.6, P=0.423; Fisher's P=0.348). This contrasts with the results of Grebenik and colleagues (P<0.002 and P<0.03). Statistical analysis was performed with Primer Biostatistics 4.04 from Stanton A. Glantz McGraw Hill (NY, USA) and with StatXact-5 Release 5.0.3 from Cytel-Software Cooperation (Cambridge, MA, USA).

In our opinion, the authors have drawn inappropriate conclusions because of the high drop out rate and their statistical analysis.

T. Grau*, J. Kessler and U. Mansmann

Heidelberg, Germany


 
E-mail: o.dearlove{at}man.ac.uk

Editor—I congratulate Grebenik and colleagues on their paper,1 which questions whether NICE guidelines that apply to adults automatically apply to children. I had also noticed that the guidelines might not be evidence-based concerning children. Also there seems to be a trend towards issuing the guidelines and then thereafter seeking the evidence base for the course recommended.

I was interested in the results of the group which had CVP lines inserted using only landmarks, and the complications that resulted. Although there were five authors, the number of children they described was quite small, 124 cases. It is not clear if every child who had a line insertion during the study period was randomized. The great advantage is that Grebenik's paper is double-blinded, and is therefore placed high in the hierarchy of information. Another question for the reader is whether the research applies in one's own area of practice. One answer is to record the results in a long series and see if the results are similar to the treatment group. I have retained records since 2000 of every CVP line I have placed in patients undergoing paediatric surgery. None of these was a cardiac by-pass patient. This is a completely different population from Grebenik's, but more likely to be met outside a specialist hospital. Only 10% are <1 yr and 50% are <5 yr of age.

I include my own personal results for landmark-based right internal jugular vein cannulation, and where I directly supervised a trainee. I recorded the insertion as difficult as anything that was not a straight stab. This is stricter than the criterion given in Grebenik's paper.

In the 4 yr 2000–2003, in 74 patients, 70 right internal jugular lines were correctly placed, 4 failed although one was subsequently placed by a colleague. Success rate was therefore 94%. Carotid artery puncture occurred in five patients (4.5%). This is not significantly different from the published landmark rates in the paper.

The legal nature of guidelines means that one may be criticized for not following guidelines. The difficulty about randomized control trials is that they may not show harm as an outcome as informed consent issues means that it would be unethical to perform such a study. Harm may fairly be judged from a long series of the activity and auditing it as I have done above. The advantage of reporting a personal consecutive series is that every insertion is recorded within the time limits given.

In order for NICE guidelines involving ultrasound to succeed in their purpose to prevent harm, it must be shown that the insertion success rate using ultrasound is >94% and the arterial stab rate is <4.5%. The number needed to detect a halving of difference in a setting such as this is ~1200.

O. R. Dearlove

Manchester, UK


 
E-mail: kategreb{at}btinternet.com

Editor—I am grateful for the opportunity to respond to the comments on our study from Drs Verghese and McGill, Grau and colleagues, and Dearlove.

As a cardiac anaesthetist, I feel that I belong to a group of clinicians that is not generally seen as ‘Luddite’ in adopting new technology, and has proved itself to be quite adept in learning ‘new tricks’; for example even quite ‘old dogs’ have succeeded in embracing intraoperative transoesophageal echocardiography, ventricular assist devices, continuous cardiac output monitoring and thrombelastography, to name but a few. As a group, we were disappointed that the use of ultrasound did not help us to avoid the scenario of difficult central venous access in small infants.

With regard to methodology, I appreciate that we were timing a smaller part of the process of inserting a central line than were Verghese and colleagues. Nonetheless, once a Seldinger wire has been placed in a central vein, the remainder of the procedure is straightforward and usually completed within 1 min.

Our problems with ultrasound-guided cannulation may, as stated in the paper, have been related to inexperience with the use of ultrasound, but as Drs Verghese and Gill state, the Site-rite probe is cumbersome and bulky in relation to a small infant. Whilst developing the necessary hand–eye coordination may take a number of attempts, I would agree with Scott that only fairly minimal training is required to perform real-time ultrasound-guided central venous cannulation in adults.6 The situation may not be the same in infants. Since the acquisition of ultrasound imaging, and the study that we described, my practice in small infants has changed. Currently, I use ultrasound to determine whether the internal jugular vein is normally sized and situated, and then proceed with cannulation without real-time ultrasonic guidance—the technique used by Alderson and colleagues.5

In response to Dr Grau and colleagues, both artery and vein were detected, but not in two perpendicular planes. A 9.0 MHz probe was used with a focal length of 0.5 cm. There was an excessive number of drop-outs in the ultrasound group, because the ultrasound device we were using at that time was battery-operated and shared between a number of different sites—we frequently found both batteries to be flat when we came to use it, or alternatively the machine was in use elsewhere. We now have a dedicated machine for the operating theatre that is mains-powered. There will be times when ultrasound is not available, and it is important to maintain the ability to cannulate central vessels without ultrasound.

I am grateful to Dr Grau and colleagues for statistical correction. Nevertheless, we were unable to replicate the success of Verghese and colleagues3 with ultrasound-guided cannulation in children. The point of our article was to demonstrate how flimsy the evidence is for the use of ultrasound guidance in central venous catheterization in children, and in particular the very small number of patients in the few published trials. For infants and children, I think that there is insufficient evidence to support a meta-analysis.

I appreciate Dr Dearlove's power calculation for demonstrating a real improvement in success rate with ultrasound; it seems unlikely that a prospective trial of around 1200 patients will be performed.

In conclusion, we did not suggest that clinicians should not use ultrasound guidance for central venous cannulation in children. Indeed, I believe it to be a useful tool both to delineate anatomical abnormalities, and for teaching purposes. We merely suggested that NICE guidelines for children should be amended until a sufficient evidence base exists.

C. Grebenik

Oxford, UK

References

1 Grebenik CR, Boyce A, Sinclair E, Evans RD, Mason DG, Martin B. NICE guidelines for central venous catheterization in children. Is the evidence base sufficient? Br J Anaesth 2004; 92: 827–30[Abstract/Free Full Text]

2 National Institute for Clinical Excellence. Guidelines on the use of ultrasound locating device for placing central venous catheters. Technology Appraisal Guidance No. 49, 2002

3 Verghese ST, McGill WA, Patel RI, et al. Ultrasound-guided internal jugular vein cannulation in infants. Anesthesiology 1999; 91: 71–7[ISI][Medline]

4 Verghese ST, McGill WA, Patel RI, et al. Comparison of three techniques for internal jugular cannulation in infants. Paediatr Anaesth 2000; 10: 505–11[CrossRef][ISI][Medline]

5 Alderson PJ, Burrows FA, Stemp LI, Holtby HM. Use of ultrasound to evaluate internal jugular vein anatomy and to facilitate central venous cannulation in paediatric patients. Br J Anaesth 1993; 70: 145–8[Abstract]

6 Scott DHT. Editorial II. ‘In the country of the blind, the one-eyed man is king’. Br J Anaesth 1999; 82: 820–1[Free Full Text]

7 Mallinson C, Bennett J, Hodgson P, Petros AJ. Position of the internal jugular vein in children. A study of the anatomy using ultrasonography. Paediatr Anaesth 1999; 9: 111–14[CrossRef][ISI][Medline]





This Article
Full Text (PDF)
E-Letters: Submit a response to the article
Alert me when this article is cited
Alert me when E-letters are posted
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Disclaimer
Request Permissions
Google Scholar
Articles by Verghese, S. T.
Articles by Grebenik, C.
PubMed
PubMed Citation
Articles by Verghese, S. T.
Articles by Grebenik, C.