The Swedish Radiation Safety Authority, SSM, decided in 2011 to investigate the justification of the use of diagnostic computed tomography (CT) for children up to fifteen years of age. The reason is the fact that CT is the main contributor to ionising radiation to the population. During the last decade, the annual number of CT examinations increased by 200%, and during the last 16 years, by 320%.
In the present study, 3,149 CT, magnetic resonance imaging (MRI) and ultrasound (US)examinations in Sweden of children and adolescents aged 0-15 were performed over a 14 day period in 2011, encompassing all health care regions. Evaluation was performed by 18 experienced paediatric clinicians and radiologists. The observers were blinded for the referring practitioners’ choice of method as well as for the performed examination method, but they had access to all other information given in the request forms (referrals). Each examination was judged by at least two observers, in a few cases by three or four observers.
Findings: Radiological examinations were judged as justified or probably justified for 96% of all examinations and as probably unjustified for 4% of the examinations. The information in the request forms was judged to be adequate or almost adequate, while 4% of the requests were considered as not fully adequate.
- The observers agreed on the choice of examination method, between the observers as well as with the requested and performed method, in 88% of requested US, in 68% of MRI and in 51% of CT examinations.
- The observers agreed among themselves on preferring another examination method than the one requested and other than the performed method in 2% of US, in 5% of MRI and in 14% of CT examinations.
- The observers disagreed among themselves and at least one observer agreed on the requested choice of method in 9% of US, in 25% of MRI and in 32% of CT examinations.
Regional variation in the use of CT was demonstrated.
Discussion: There is a high level of national consensus among paediatric clinicians and radiologists concerning the use of US as the adequate method of choice in actual clinical practice. The lower level of agreement regarding CT as the method of choice suggests that clinical guidelines either do not exist, or are suboptimal, or have not yet reached common acceptance in the paediatric healthcare community. During the past decade, parallel to the development of much more optimal techniques leading to steadily increased use of CT, access to MRI has slowly but steadily improved at almost all Swedish county hospitals. Still, there may be clinically unacceptable, long waiting times for MRI (and especially for MRI of children in general anaesthesia), leading to the next best choice of CT for the examination. These circumstances may be important and one reason behind the level of disagreement becoming evident in the present study. The paediatric physician in routine practice may not only be able to choose the most appropriate examination with the lowest radiation dose to the child, but must for example consider choosing between drop-in for CT and very long waiting times for MRI.
In the present study, regional differences were assessed. CT requested by university and county hospitals was considered to be more justified than CT requested by physicians in primary care, though requests from primary care were infrequent.
In a similar investigation in 2004 on the use of MRI, conducted by the same team of experts, the observers considered about one-third of the studied MRI examinations as being unjustified and 20% to 25% of the requests were considered to be inadequate.
It is of material importance to increase efforts to define and establish optimal referral guidelines for paediatric imaging in general and for the use of CT for children in particular.