Premium
Ultrasound for central venous cannulation: economic evaluation of cost effectiveness
Author(s) -
Scott D. H. T.,
Ho A. M.H.,
Joynt G. M.,
Karmakar M. K.,
Cohen A. M.,
Calvert N.,
Hind D.,
McWilliams R.,
Thomas S.M.
Publication year - 2005
Publication title -
anaesthesia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.839
H-Index - 117
eISSN - 1365-2044
pISSN - 0003-2409
DOI - 10.1111/j.1365-2044.2005.04159.x
Subject(s) - medicine , ultrasound , landmark , central venous catheter , average cost , anatomical landmark , catheter , operations management , surgery , radiology , computer science , artificial intelligence , neoclassical economics , economics
The interesting paper by Calvert et al. [1] contains little new information on ultrasound guided central venous catheter (CVC) insertions since the economic evaluation was published on the internet in 2002 [2]. It has already been dissected in an editorial [3]. In the intervening year, the authors have not addressed many of the criticisms made. Expecting equipment to be replaced every 3 years is unrealistic. The first scanner I used lasted for 13 years. This is a more realistic figure for the length of time we are expected to use equipment in the NHS. Maintenance costs for this type of equipment need not be large. While the authors accept that expensive maintenance contracts are unnecessary, and have reduced their original £5.78 machine cost per insertion [2] to £4.98, they do not describe how this is calculated. Five hours of one-to-one teaching with a Consultant Radiologist is assumed to be necessary for proper training. The cost averages out at £1 per insertion. Anaesthetists used to have to train their trainees to use the landmark technique without ultrasound. Using ultrasound, they can now train them to use both the landmark and ultrasound techniques much more quickly, as they can show the trainees the anatomy and ensure that they get into the right place, first time. Ultrasound lowers the cost of training new operators. The economic model used by Calvert and colleagues takes arterial puncture as its only complication, with a landmark rate of 12%, an ultrasound rate of 3%, and a cost of £40.While they quote pneumothorax as a possible complication, the model did not allow it to be considered [2]. Stroke and death are other possible complications excluded from the model. Fatal stroke has now been described [4] and at least two patients have died from inadvertent carotid puncture in the last 2 years in Scotland (confidential personal communications). These cases are unlikely to be reported in the medical literature. Only two sources were eventually used to provide the input data for the model [5, 6]. Their combined rate of carotid puncture of 12% is much greater than the incidence in most cardiac surgical units [7] and has prompted many clinicians to question the other statements they have made. The paper does have one area where the authors underestimate the cost of using ultrasound guidance. One machine is thought to be sufficient to insert 15 catheters per week, based on the local cardiac surgery unit inserting 30 catheters per week and presumably needing two machines. This is unrealistic. One machine should be located in each place where central lines are regularly inserted, in other words in each theatre performing major surgery and each intensive care area. If scanners are not immediately available for routine CVC insertion, they will not be used. A more realistic figure is seven insertions each per week [3]. Despite the shortcomings of this paper, ultrasound guidance for central venous cannulation, and cannulation of vessels at other sites [8–10], is, I believe, economically justified, but if the nonbelievers are to be convinced they must be given reliable information.