Use of lumbar spine films. Statistical evaluation at a university teaching hospital.

Use of lumbar spine films was evaluated in a retrospective review of 1,095 lumbar radiological examinations in 871 patients. Patients' histories and physical examination findings were correlated with four groups of radiological findings: (1) radiologically normal, (2) incidental findings unrelated to patients' complaints, (3) diagnostic films compatible with symptoms, and (4) radiological findings of questionable clinical significance. The high percentage of nondiagnostic and questionable results suggests overuse of lumbar spine films. Forty-six percent of all studies in this investigation were normal or had incidental findings, and an additional 31.8% had radiological findings of questionable clinical significance. Three hundred seventy-five follow-up examinations (64.2%) demonstrated no radiological interval changes. Radiation exposure and medical cost of noncontributory studies may be substantially reduced by judicious consideration of the potential diagnostic yield of the examination and by careful selection of repeated or follow-up studies.