Local warming dilates the arterioles and decreases α₂-adrenergic vasoconstriction. The hand or lower arm can be warmed by wrapping it in towels moistened with warm/hot water or immersing it in warm water. Only one randomized controlled trial assessed the effect of this procedure in facilitating insertion of venous cannulae. In adult patients, Lenhard and coworkers 2 showed that local warming facilitates the insertion of venous cannulae, and reduces the time and number of attempts required. A controlled study conducted in children is not available.

Finding an accessible vein in infants is frequently difficult when the skin is coloured, the infant is dehydrated, obese or shocked, or when the commonly accessible veins are exhausted. Transillumination techniques have been used for many years to facilitate arterial puncture 3 and venous access 4. Various devices for transillumination have been described but have not gained popularity. Commonly used are cold-light fibreoptic techniques; complications (burns) are rare but possible 56. A 40% efficacy of palm transillumination using a common otoscope in an emergency department setting was recently reported 7, and venous access could be established in 39 out of 40 patients with just one venipuncture.

Ointments containing nitroglycerine (GTN) have been used to produce local dermal vasodilatation. The first successful use of GTN ointment as an aid to venipuncture was reported in 1983 in a randomized placebo-controlled trial 8. Other investigations showed an increasing diameter of the veins 910, or better success rate for venipuncture 11.

The local application of GTN was efficient in reducing venipuncture failure in children younger than 1 year 12. However, these results could not be reproduced in another series 13. Adverse local and systemic side effects were found in neonates and premature babies, and therefore the use of GTN cannot be recommended in this age group 14.

The combination of GTN with local anaesthetics (EMLA – eutectic mixture of local anesthetics) had a beneficial effect in adults by increasing the ease of venipuncture and decreasing the pain and the dose of local anaesthetic required. These results were confirmed in a double-blind randomized controlled trial in 104 children aged between 1 and 11 years 15. The addition of topical GTN positively affected venous dilatation (P < 0.01), choice of cannulation site (P < 0.001) and ease of cannulation (P < 0.001). Similar findings were achieved when GTN was applied after EMLA removal 16.

The standard procedure for puncturing the femoral vein has a high success rate and a low rate of arterial puncture in paediatric emergency treatment and in the intensive care unit setting 1721, even with unexperienced operators 22. It was successful in about 92% in critically ill patients 23, in about 89–95% in children, and in about 80% in preterm infants weighing less than 1000 g 242526. A higher success rate can be achieved in obese children using ultrasound techniques 27.

Femoral venous catheterization is a safe method 1728 because it does not expose the patient to the potential hazard of intrathoracic complications 29. Stenzel and colleagues 30 demonstrated a 3.7% complication rate for femoral CVC as compared with 7.3% for nonfemoral CVCs. Goldstein reported a catheter-related sepsis rate of 4.9% and mechanical complications of 3.5% in burned children 31. The rate of femoral vein thrombosis varies from 4% to 35%, and this is adversely influenced by age, size and underlying condition of the patient 323334. Heparin-bonded material may significantly reduce femoral catheter-related thrombosis and infection 35. Cannulating the femoral vein is difficult in more than 50% of patients after previous cardiac catheterization, and therefore the contralateral side should be used in the first attempt in such patients 36.

Subclavian vein catheterization has proved to be a rapid alternative to surgical cutdown techniques for venous access in children 2037383940. However, the overall reported complication rate varies significantly from about 3% to 34%, depending on age, indication and side of puncture 192940.

Finck and coworkers 41 reported that subclavian access was successful in 78.8% of patients younger than 6 months (average weight 3.1 kg) and in 96% of those older than 6 months (average weight 7.6 kg); there were no complications. Citak and colleagues 42 found similar results in 148 out of 156 central venous attempts (94.9%) in which subclavian vein catheterization was chosen, with a moderate complication rate (arterial injury 12.8% and pneumothorax in two patients; no mortality occurred) and a high success rate 'in experienced hands'.

However, in a large (2290 central venous catheterizations) prospective multicentre study, Iovino and coworkers43 demonstrated that the overall risk for complications with subclavian puncture was significantly elevated when compared with internal jugular puncture 42. The main complications were pneumothorax and arterial puncture, but a low rate of thrombosis was noted.

Various ultrasound techniques and atrial ECG guidance have been used in an attempt to improve success rates 444546.

The puncture technique and position of the patient are different in children compared with adults. Positioning without a shoulder roll and placing the head in a neutral position (Fig. 1) optimizes vein diameter 47. Jung and coworkers 48 showed that tilting the head toward the catheterization side after successful puncture reduced the incidence of catheter malposition.

The internal jugular vein (IJV) provides a useful and reliable site with a low failure rate, and its cannulation is traditionally performed with the aid of both palpation and anatomical landmarks. Various approaches may be used to reach the IJV 49. Variations between the carotid artery and the IJV, and the depth and size of the IJV may account for failure to locate the vein 505152, and these factors were found to be independent of age and size 53. The IJV was found directly anterior to the carotid artery, at the level between the two heads of the sternocleidomastoid muscle, in about 50% of cases, and anterior or anterolateral at the level of the cricoid cartilage in about 30% of the cases 5455.

Doppler ultrasound guided puncture could reduce the time and the number of attempts for successful cannulation 565758. Direct two-dimensional ultrasound identification proved to be more precise and efficient 4559, especially in small children, and it is now recommended when difficulties are anticipated, complications have been encountered, or when repeated IJV cannulation is required 6061.

Verghese and colleagues 62 reported a 100% success rate using ultrasound guided IJV cannulation in infants, as compared with a 75% success rate using a traditional palpation method, and the incidence of carotid artery punctures was 0% versus 25%. In 2002, these findings were repeated by Asheim and coworkers 63, who found a 100% success rate in 45 consecutive children and a median time to aspirate blood from the IJV of 12 s. Complications include arterial puncture, haematoma formation and catheter malposition, but thrombosis and pneumothorax are rarely reported.

The diameter of the IJV may vary significantly during respiration and after various manoeuvres to increase intravascular filling and pressure 6465, and the Valsalva manoeuvre, the Trendelenburg position and manual compression of the liver increased the size of the IVC 666768.

Percutaneous catheterization techniques for the axillary vein have been described since 1967 for adults, children and neonates 6970. Successful and safe use was demonstrated during resuscitation 71, and the risk for catheter-related infections or thrombosis is similar to that with other CVC sites 7273.

In neonates, axillary vein cannulation was found to be successful in 217 out of 226 patients, with infection occurring in three and shoulder oedema occurring in eight of the 217. Possible but rare other complications are pleural effusion, haematoma and pneumothorax 747576.

In children the axillary approach was found to be an acceptable route for central venous catheterization in about 80%, and the risk of complications was 1.1% per catheter-day 76. Again, the overall success rate can be improved significantly by ultrasound-guided cannulation 77.

Use of the external jugular vein (EJV) for central venous access with a guide-wire technique is associated with a 75–100% success rate in adult patients and a very low complication rate 78.

Applicability in children was first demonstrated by Humphrey and Blitt in 1982 79. The overall success rate reported is lower than that in adults, ranging from 54% to about 92%. When the EJV was visible, there were no serious complications reported 808182. A large series of EJV cannulations was reported by Soong and coworkers 83 in 1995. That group used the EJV in 488 out of 1318 central venous accesses in a paediatric/neonatal intensive care unit. The initial success rate was high (>90%); however, the EJV catheters were used for a shorter period of time.

Over the past two decades, the intraosseous (IO) route into the tibia has become a widely accepted procedure for the resuscitation of critically ill and injured children 8485868788 such as trauma patients 8990 and patients suffering from severe burns 9192. Newer devices, such as the 'bone injection gun', may increase the already widespread use of IO access 9394.

The IO technique is included in standard protocols and training procedures 9596, such as the Advanced Paediatric Life Support textbook, and it is recommended by the American Heart Association, the American Academy of Paediatrics, and the American College of Surgeons 97.

IO lines are not commonly used in newborn infants; however, it is recommended in neonates as an alternative route for medications/volume expansion if umbilical or other direct venous access is not readily available 9899100. Neonates were included in some series of IO therapy 101, but only few studies examined this route in neonates and premature infants 99102. Successful use in an 800 g baby has been reported 103. Additionally, in a model of neonatal emergency vascular access, the IO route was found to be faster and easier than conventional umbilical venous catheterization 104.

Complications from the use of IO access occur rarely and include fractures and osteomyelitis after long-term use of IO access 105106 or when hypertonic solutions have been used. Fat embolism is less likely in children than in adults and has minimal clinical consequences 107. Local extravasation of fluids due to incomplete penetration of the needle into the cortex, IO infusion into a fractured limb, or perforation of the bone may lead to a compartment syndrome 108109. Finally, follow up in neonates ruled out concerns regarding injury to growing bone and the growth plate 110.

The technique of venous cutdown is part of the training programme of Advanced Paediatric Life Support. The most preferred cutdown access site is the saphenous vein above the medial malleolus of the tibia, but antecubital, axillary, cephalic and femoral vessels are also suitable 111 and improved procedures using Seldinger techniques have been reported.

Currently, cutdown procedures are regarded as the methods of last resort. The usual time to achieve access by paediatric surgeons was 6 min in children aged 6–16 years, 8 min in those aged 1 month to 5 years, and 11 min in neonates 112. This time delay makes its use unrealistic for most clinicians, and IO or percutaneous femoral access can be achieved more rapidly 21113.