Objectives: This study examined the clinical- and cost-effectiveness of image-guided Hickman line insertions versus blind Hickman line insertions undertaken by nurses in adult cancer patients.

Design: A cost-effectiveness analysis was carried out alongside a randomized controlled trial.

Setting: The study was undertaken at a large acute cancer center in Manchester, United Kingdom.

Participants: Cancer patients due to have a Hickman line insertion who were older than 18 years of age and were clinically and physically compliant with specified protocols participated.

Interventions: To obtain central venous access for the patient, two interventions were investigated: (i) blind insertion of a Hickman line and (ii) image-guided insertion of a Hickman line. Both interventions involved blind venipuncture of the subclavian vein. In the blind arm, the Hickman line was routinely inserted without the use of image guidance at any point in the procedure. Transfer to the interventional X-ray suite and use of image guidance were options immediately available to the operator during the procedure if required. In the image-guided arm, the position of the guidewire was checked before the Hickman line was introduced and later the Hickman line was positioned with the use of X-ray fluoroscopy.

Main outcome measures: The primary clinical outcome measure was catheter-tip misplacement, and this misplacement was expected to be higher in the blind arm. When comparing the skill level of the trainer and the trainees, pneumothorax was the primary clinical outcome measure. Other outcomes measures included arterial puncture, hematoma, infection, failed insertion, and assistance from other health-care professionals.

Results: No statistically significant difference was found between the mean cost per patient in the two arms of the trial. The only statistically significant difference in clinical outcomes was the frequency of catheter-tip misplacement, which was higher in the blind arm of the trial. At very low costs, the image-guided approach dominates the blind approach as fewer costs and greater benefits are incurred. It is evident that nurses previously inexperienced in the procedure can be trained to insert Hickman lines successfully both at the bedside and under image guidance within a 3-month period.

Conclusions: This report indicates that nurse insertion of Hickman lines in the majority of adult cancer patients is both safe and effective. However, there are a select group of patients for whom image-guided insertion may be preferred. The results reveal that skills and expertise can be transferred from trainer to trainee through a relatively short, but intensive, training course. It is also evident that patients support nurse insertion. Further research is suggested to compare the safety and efficacy of nurse versus doctor insertions in particular subgroups of patients and also to assess the quantity and quality of current service provision to inform National Health Service decision-making in this area.