Substituting Diagnostic Services

Abstract

To the Editor.— Several of the assumptions by Eisenberg et al 1 cast doubt on the validity of their conclusions that new diagnostic methods do not displace, but rather are performed in addition to, older methods. I question the appropriateness of comparing the radionuclide brain scan to brain computed tomography (CT). A more valid study would compare the incidence of pneumoencephalography and cerebral angiography (the primary neuroimaging studies of the pre-CT era) in 1974 with their incidences in 1980. Pneumoencephalography was performed daily in the preCT era in institutions with active neurology and neurosurgery practices; it was completely discarded within 5 to 6 years of the introduction of CT. The incidence of cerebral angiography has also declined dramatically since the advent of CT. The period studied, 1978 through 1980, was one of rapid change in the capabilities of both CT and ultrasound devices. Many of the devices in use then