A prospective audit of the use and costs of myelography in a regional neuroscience unit.

A consecutive series of 397 myelograms performed in 385 patients over a six month period at the Mersey Regional Neurosciences Unit is reported. The reasons for performing the myelogram were to identify the cause of a radicular lesion in 54% of patients, a chronic spinal cord lesion in 30%, an acute cord lesion in 9%, suspected disease at the level of the foramen magnum 6%, and for a variety of other conditions in 8%. For the 385 patients undergoing a myelogram in the study period, the median interval from admission to request, request to myelography and from myelography to discharge was nought, one and three days respectively. The proportion of patients submitted to myelography by individual consultants ranged from 7% to 28%. There was a two-fold variation in the delays in the time to requesting and performing myelograms. There was room for improvement in the clinical information supplied on the myelography request form. The role of ancillary investigations and their effect on myelography was unclear. Only 16 of the patients with suspected cord disease had visual evoked responses performed before myelography. Five of them had myelography after an abnormal result. The estimated annual direct cost of myelography in the unit was at least 486,000 pounds. Reorganisation might have yielded hypothetical "savings" of between 30,000 pounds (6%) and 155,000 pounds (32%), though in practical terms these "savings" represented resources which might have been freed for use in other higher priority clinical problems within the unit, rather than true reductions in monetary cost.