Lumbar spine x-rays: a multihospital study

To define the usefulness of the lumbar spine x-ray series in the emergency department and to generate clinical criteria for optimizing its application, we retrospectively studied 552 consecutive emergency department patients for whom lumbar spine x-rays were ordered. Patients were divided into traumatic (47.6%) and nontraumatic (52.4%) groups. Three subgroups were created based on radiological findings: 1) "negative" (55.8%), 2) "possibly significant" (37%), and 3) "positive" (7.2%). The "positive" subgroup was compared with the other two subgroups in an attempt to define physical markers that correlated with positive radiological findings. Four clinical findings were present in significantly different frequencies between the positive group and others: an abnormal physical examination (90% vs 61.5%, respectively) (P less than .0001), tenderness (72.5% vs 41.2%) (P less than .0005), multiple positive findings (42.5% vs 20.7%) (P less than .005), and contusion or abrasion (15% vs 2.7%) (P less than .0005).