False-positive findings on lumbar discography. Reliability of subjective concordance assessment during provocative disc injection.

STUDY DESIGN: Experimental disc injections in subjects with no history of low back symptoms.

OBJECTIVE: To determine in an experimental model the reliability of patients' subjective interpretation of pain concordancy during provocative disc injection.

BACKGROUND: Discography in the evaluation of low back pain relies on a patient's subjective assessment of pain magnitude and quality during disc injection. Reproduction of significant pain on disc injection, which is similar to patients' usual pain, is believed to prove that the disc injected is the source of the patient's low back pain. In the current study, this hypothesis was tested in a controlled setting on patients with known nonspinal pain in a common referral area of discogenic pain.

METHODS: Patients with no history of low back pain were recruited to participate in a study of discography. Patients scheduled to undergo posterior iliac crest bone graft harvesting for nonthoracolumbar procedures were evaluated with lumbar radiography, magnetic resonance imaging, and psychometric testing. Two to 4 months after bone graft harvesting, patients underwent lumbar discography by strict blinded protocol. Patients were asked to compare the sensations elicited at discography to their usual back/buttock pain since bone graft harvesting. Pain was rated as 0-5 on a pain thermometer and concordancy was rated as none, dissimilar, similar, or exact.

RESULTS: Eight subjects completed the study, and 24 discs were injected. Of the 14 disc injections causing some pain response, 5 were believed to be "different" (nonconcordant) pains (35.7%); 7 were "similar" (50.0%), and 2 were "exact" pain reproductions (14.3%). The presence of anular disruption predicted concordant pain reproduction (P < 0.05). Of 10 discs with anular tears, injection of 5 elicited pain that was similar to or an exact reproduction of pain at the iliac crest bone graft harvest sites. By the usual criteria for positive discography, 4 of the 8 patients (50%) would have been classified as positive. In these patients, the pain on a single disc injection was very painful, and the pain quality was noted to be exact or similar to the usual discomfort. All subjects had a negative control disc.

CONCLUSIONS: The findings of this study demonstrate that patients with no history of low back pain who had undergone posterior iliac bone graft harvesting for nonlumbar procedures often experienced a concordant painful sensation on lumbar discography with their usual gluteal area pain. Thus, the ability of a patient to separate spinal from nonspinal sources of pain on discography is questioned, and a response of concordant pain on discography may be less meaningful than often assumed.