Dear Editor

We read the article by Reece et al [1] and closely followed the responses to it. We even went ahead to carry out a study looking at identification of the tip of the long lines using inversion of image technique on PACS (picture archiving and communication system).

Background: Positioning of long lines into the heart has serious consequences including death due to cardiac tamponade.[2] The tip of long lines is accurately visible in only 50 % of plain radiographs.[1] Identification of the line using radio opaque contrast media requires caution. The use of an insufficient volume of contrast will falsely identify the tip in an apparently more proximal position, whereas a film taken during active injection may cause the line to appear longer due to a jet of contrast issuing from the tip of the line. Bernard I and Banerjee I, from Glan Clwyd Hospital, wrote in their E-letter on the use of PACS in their hospital to identify the tip of the long lines.[3] Ultrasound may be of value but it requires expertise to perform and interpret.[2]

Methods: At Hinchingbrooke Hospital, Huntingdon, X-rays are taken on a phosphor plate which are later processed through PACS and image is available on computer terminal on the Special care baby unit. The Hospital use Frame wave dicom view version 3.0 software that allows for image inversion, image magnification and image sharpening. Using the technique of image inversion supplemented by image magnification and sharpening of image, the tips of the long line are much better seen than on plain X-rays. Our this observation lead us to carry out a study looking at the tips of the long lines on plain X-ray and then on the same image on PACS with image inversion. Three investigators including an experienced SHO, a consultant paediatrician and a consultant neonatologist participated in this study. The study was retrospective and included the long lines inserted between the periods of January 2000 to July 2001.

Results:

Observer	Tip visible on plain X ray	Tip visible on inverted image	Improvement
A	06/24 (25%)	15/24  (63%)	38%
B	12/23 (52%)	17/23  (74%)	22%
C	40/69 (58%)	65/69  (94%)	36%

Conclusions:
Inversion of image on PACS is better than plain radiograph in identifying the tip of the long lines. Although there is inter observer variation, with experience of using PACS, this may be minimized. It is important to note that there is improvement in recognition of line tip by each observer. Recommendations: We recommend that units having facility of PACS should use them to identify the long line tips and there is a need for prospective randomized study comparing contrast study v/s image inversion technique on PACS, before accepting contrast as the only way of identifying the tips of long lines. Other uses of PACS with regards to neonatal long lines: 1. Accurate line manipulation. Once it is noted that the long line is in right atrium, one can measure the distance on the image by which it should be pulled back to be in acceptable place, rather than guessing the distance by which line is pulled back. 2. Monitoring of long line on subsequent X rays: Most unit that uses contrast to identify the long line tip does it only once to confirm the position and subsequently look at the plain X rays (taken for other clinical indications) to see the long line position. We know that long lines do migrate over a course of time and plain X rays are less sensitive to detect the tip accurately. The use of PACS allows each image to be reviewed with same accuracy as first image. This can detect the potential line migration and allows adjustments in line position to be made.

References

(1)Reece A et al. positioning long lines: contrast versus plain radiography. Arch Dis Child Fetal Neonatal Ed 2001; 84:F129-30.

(2) Review of four neonatal deaths due to cardiac tamponade associated with the presence of a central venous catheter: Recommendations and department of health response. June 2001.

(3) Bernard I, Banerjee I. E-letter. Arch Dis Child 14th May 2001.

Dear Editor, We read the article by Reece et al [1] and followed the subsequent correspondence with interest. In light of the recent review commissioned by The Chief Medical Officer for England, physicians must be aware of potential complications of peripherally inserted central catheters (PICC).[2] While the true incidence of such events will only be known with prospective data collection, retrospective studies suggest a complication rate (pleural/pericardial effusions) of 0.5% per line insertion.[3] The Department of Health (DoH) paper in response to this review recommends placement of central venous lines outwith the cardiac chambers.[4] However, complications related to central lines are not only confined to the thorax. We report 3 cases of delayed detection of peritoneal extravasation related to central venous catheters.

Case 1: An eight-week old malnourished infant was intubated and ventilated for acute onset severe respiratory distress following a brief flu-like illness. His endotracheal secretions were positive for RSV and a previously undetected myopathy was suspected and investigated. He required prolonged ventilation and received blood transfusions through a triple-lumen right femoral venous catheter for anaemia. On the fifth day after admission he developed abdominal distension and a diagnostic tap under ultrasound guidance revealed haemorrhagic ascites. A contrast radiograph of the femoral catheter showed extravascular spillage of dye, in this case in the extraperitoneal space. In retrospect, the infant showed no rise in haemoglobin following 2 packed cell transfusions. Extravascular migration of the catheter tip was diagnosed and the catheter was promptly removed.

Case 2: A 24-week preterm female infant was admitted to the neonatal intensive care unit where she was ventilated and treated with surfactant. Severe respiratory distress of prematurity, patent ductus arteriosus and sepsis complicated the initial course. The PDA did not respond to 2 courses of indomethacin and surgical ligation was scheduled on day 13 of life. A PICC was removed due to suspected catheter sepsis and she was treated with intravenous antibiotics. She had a right femoral venous central single-lumen catheter through which she received parenteral nutrition and blood transfusions on the unit. Her anaemia failed to respond to the three packed cell transfusions and when she developed abdominal distension with bluish discolouration in the groin, the femoral line was radiographically reviewed after injecting contrast material. Extravascular spillage of the dye was noted in the extraperitoneal space and the subcutaneous tissue of the lower abdominal wall.

Case 3: A – week preterm infant with PICC in the leg developed a skin abscess at the xyphisternum. Abdominal radiograph with contrast injected through the catheter, revealed retroperitoneal extravasation of contrast. The PICC was removed and the infant made a full recovery.

In each of these case reports, femoral catheter tip migration was detected following extravascular extravasation of blood or parental nutrition fluid. Haemoperitoneum has been reported in the past as a complication of central catheters but may not be widely recognised. In comparison with PICC, these catheters are shorter and more rigid, hence more likely to perforate through vessel wall. Femoral venous access is readily obtained in infants and is commonly used in intensive care settings for parenteral nutrition, maintenance fluids, blood transfusions and other parenteral therapy. Whilst malpositioning of femoral catheter is readily detected in most instances during placement, a spontaneous extravascular migration of a previously well-placed catheter tip is possible in some cases. As opposed to frank rupture of the blood vessel and haemorrhage into the retroperitoneum, slow extravascular infusion of blood in the low resistance extraperitoneal space may not be promptly detected in the absence of a high index of suspicion. In preterm newborns and malnourished infants the vessel wall integrity may be compromised and migration of a previously normally positioned catheter tip may be more likely. Femoral venous catheter tip positions must be reviewed in all cases of unexplained ascites and abdominal distension. Contrast radiography, digitalized image inversion and ultasonography have a role in determining catheter tip position and diagnosing malpositioned intravascular catheters.[5]

We agree with the DoH recommendation that there should be a prospective National Audit of such cases.

References

(1) Reece A, Ubhi T, Craig AR and Newell SJ. Positioning long lines: contrast versus plain radiography. Arch. Dis. Child 2001; 84: F129-130.

(2) Nadroo AM, Lin J, Green RS, Magid MS, Holzman IR. Death as a complication of peripherally inserted central catheters in neonates. J. Pediatr 2001 Apr; 138(4): 599-601.

(3) Keeney SE, Richardson CJ. Extravascular extravasation of fluid as a complication of central venous lines in the neonate. J Perinatology 1995: 15;No 4, p284-288.

(4) Review of four neonatal deaths due to cardiac tamponade associated with the presence of a Central Venous Catheter. Recommendations and Department of Health response. [http://www.doh.gov.uk/manchesterbabies/manchestersummary/]

(5) Bernard I, Banerjee I. E-letter. Arch. Dis. Child, 14th May 2001.

Authors:

1] Nilesh M Mehta
Specialist Registrar

2] Richard M Nicholl
Consultant Neonatologist, Hon. Senior Clinical Lecturer

Northwick Park Hospital
NorthWest London Hospitals NHS Trust
Harrow HA1 3UJ, UK

Dear Editor,

Long lines are commonplace but putting them in the appropriate place is not so common. The article [1] advocates the use of contrast to position long lines. In a e-letter, Dr Yadav [2] argues against the use of contrast medium until safety is validated. We, at Glan Clwyd Hospital, have found that using the picture archiving and communication system (PACS), line tips are simpler to identify. We use software MedView RV 2.1 in our neonatal unit, which allows for image inversion. Changing a positive image to negative does not offer any additional information but makes it easier to see lines without the use of contrast. This applies for the smaller long lines such as the 27G Medex long line. Filmless radiology is being increasingly implemented [3,4] and visualising long lines adds on to its benefits [4]. We feel that image inversion using a digital system should do away with the debate of contrast versus plain radiography.

Ian Barnard
Consultant Paediatrician
Glan Clwyd Hospital

I Banerjee
Specialist Registrar
Glan Clwyd Hospital

References:

(1) Reece A, Ubhi T, Craig A R, Newell S J. Positioning long lines: contrast versus plain radiography. Arch Dis Child Fetal Neonatal Ed 2001 ; 84:F129-30.
(2) Yadav M. E-letter, Arch Dis Child, May 14, 2001.
(3) Bryan S, Weatherburn G C, Watkins J R, Buxton M J. The benefits of hospital-wide picture acchiving and communication systems: a survey of clinical user of radiology services. Br J Radiol 1999 May;72(857):469-78.
(4) Strickland NH.PACS ( picture archiving and communication systems ): filmless radiology. Arch Dis Child 2000; 83:82-86.

Dear Editor,

I read with interest the recommendation of Reece and colleagues regarding the positioning of long lines in preterm neonates. [1] In their methods the authors state that they aimed to place the tip of the line up to 10 mm into the right atrium (upper limb insertions). Manufacturer and standard text book of neonatology recommend that the line tip should not be sited in the right atrium as there are potential serious complications.[2] I am sure the authors are aware of the recent media attention, and the enquiry into deaths in neonates due to cardiac complications of long lines. [3] The authors commented that almost 50% of the line tips could not be visualised on plain x-ray examination. Some of the currently available lines with guide wires (27 G CV Single lumen catheter Medex Medical Inc. Haslingdon, Lancashire UK ) are particularly suited to the needs of preterm infant. The radio opaque wire enables adequate visualisation on plain x-ray. I would welcome the authors’ views on the use of these lines.

Although the study did not have the power to study the adverse effects of contrast used, unless safety of the material is well established it seems premature to recommend its routine use.

References: (1) Reece A, Ubhi T, Craig AR, Newell SJ. Positioning long lines: contrast versus plain radiography. Arch Dis Child Fetal Neonatal Ed 2001; 84: F129- 30

(2) Brain AJ, Roberton NRC, Rennie JM. Textbook of neonatology. London: Churchill Livingstone 1999:1376

(3)Charter D. Baby heart deaths force inquiry. The Times 2000 Nov 25