Dr Geoffrey Morgan

We looked at our requests for MRI and to try and ascertain whether the result of the MRI (normal and abnormal) influenced a decision to refer.

I searched our clinical system (Torex Health System 5) for all request for MRIís over an 18  month period between January 1998 and June 1999. It became clear that by far the commonest requests were for knee (50 requests), and lumbo-sacral spine (39 requests). Other referrals were for cervical spine (15); brain (2); and wrist (1).

This review looks at referrals for knee and the outcome following the MRI result. There is no statistical breakdown but the results proved helpful in practice discussions re use of investigations.


Over the 18-month period 50 referrals were made. Results show 15 normal scans, 31 abnormal and 4 equivocal

 4 patients with a normal MRI were referred.

 27 of the 31 patients with abnormal findings were referred.

 None of the patients with an equivocal result (not completely normal but only insignificant abnormalities) were referred.

Thus a total of 31 patients were eventually referred. More importantly, based on impressions from the notes and referral letters, the decision whether to refer or not was influenced by the result of the MRI in 41 of the 50 cases.

I think that one can draw two opposing conclusions. It could be argued that the doctors have a high accuracy level of suspicion re serious pathology and use the service appropriately. The counter argument could be that if this is so, referrals could be made on clinical findings alone and MRI is unnecessary.