Study Design. Post hoc analysis of data from the National Low Back Pai
n Study, a prospective observational multicenter study of patients ref
erred for the evaluation and treatment of persistent low back problems
. Objective. To identify patient characteristics associated with use o
f particular diagnostic imaging examinations in patients with persiste
nt low back problems. Summary of Background Data. The Agency for Healt
h Care Policy and Research clinical practice guidelines on low back pr
oblems suggest that the use of particular diagnostic imaging tests for
a given patient should be based on specific characteristics of that p
atient. Methods. Use of diagnostic imaging examinations in 2,374 patie
nts with persistent low back problems who were enrolled in the Nationa
l Low Back Pain Study from 1986 to 1991 was analyzed. Stepwise logisti
c regression was used to identify patient characteristics that distinq
uish between enrollees who underwent particular imaging studies. Resul
ts. Characteristics that distinguished patients who had undergone magn
etic resonance imaging from those who had received only lumbo-sacral s
pine radiographs included higher socioeconomic status, greater resourc
e use in the preceding 12 months, more functional impairment, presence
of sciatica, and presence of neurologic signs/symptoms suggestive of
nerve root compromise. Suspected soft tissue involvement was character
istic of enrollees who had undergone magnetic resonance imaging, where
as suspected structural involvement characterized patients who receive
d noncontrast computed tomography. Only nonclinical factors, such as h
igher annual household income, disability compensation, and male gende
r distinguished enrollees who had undergone both magnetic resonance im
aging and computed tomography-myelography from those who received only
computed tomography-myelography. Conclusion. Particular patient socio
economic and clinical characteristics are associated with receipt of s
pecific imaging studies in evaluation of persistent low back problems.