A prospective controlled longitudinal study.
To determine whether subjects, asymptomatic for lower back problems, who undergo experimental discography, will develop lower back problems during the medium term to the full term.
Previous work has shown significant pain on discographic injection in approximately 40% of asymptomatic subjects. It has been suggested that those subjects with painful injections would soon develop lower back pain (LBP) syndromes in the near future: that is, the experimental discography was detecting an imminent “pain generator” before clinically symptomatic.
Fifty subjects without low back pain were recruited for clinical and psychometric testing, MRI scanning, and experimental lumbar discography to determine the rate of painful lumbar disc injections in select subjects without LBP history. After determining which subjects had painful injections, all subjects completing the discography protocol were prospectively followed at yearly intervals to determine the occurrence of LBP and LBP disability over time. Statistical methods were then used to determine the correlation, if any, between the asymptomatic subjects’ clinical, MRI, and discography findings, and the subsequent LBP measures. Controls, not participating in the lumbar discography study, were also followed. Controls were matched for clinical features, sex, age, and occupational/recreational exposure. Follow-up examinations were performed at yearly intervals by blinded researchers using a scripted interview and completing standard questionnaires.
A total of 46 of 50 completed the discogram, and all 46 subjects completed the final 4-year follow-up examination. There was a low incidence of LBP episodes in the experimental groups and control. A painful disc injection, independent of psychological profile, did not predict LBP or any other functional outcome measure at follow-up on multivariate analysis. The presence of an anular fissure seen on discography was weakly associated with the cumulative incidence of LBP episodes after discography (P = 0.08). The presence of high intensity zone on MRI in any disc was also weakly associated with the development of LBP episodes (P = 0.09). Psychometric profiles at the start of the study strongly and independently predicted future back pain (P = 0.01), medication usage (P = 0.002), and work loss (P = 0.01) over the 4-year study. Compared with controls not having undergone discography, there was no significant difference in back pain, function, work loss, doctors visits for back pain, or medication intake in any group. A subset in the injection group with somatization disorder had a higher LBP visual analog score compared with somatization disorder controls at 1 year,but this was not significant at 4 years after testing.
Painful disc injections are poor independent predictors of subsequent LBP episodes in subjects initially without active lower back complaints. Anular disruption is a weak predictor of future LBP problems. Psychological distress and preexisting chronic pain processes are stronger predictors of LBP outcomes.
Subjects asymptomatic for lower back pain were tested with experimental discography and followed, along with matched controls, for 4 years thereafter. Positive disc injections were not an independent predictor of subsequent low back pain problems; however, a history of chronic pain problems and psychological distress were strong independent predictors of subsequent lower back pain complaints in both discography and control subjects.
From the Orthopaedic Surgery Department, Stanford University, Stanford, CA.
Acknowledgment date: June 19, 2003. First revision date: December 17, 2003. Acceptance date: January 5, 2004.
The manuscript submitted does not contain information about medical device(s)/drug(s).
Other funds were received in support of this work. No benefits in any form have been or will be received from a commercial party related directly or indirectly to the subject of this manuscript.
Address correspondence to Eugene J. Carragee, MD, Stanford University, Orthopaedic Surgery Division, 300 Pasteur Drive, Room R171, Stanford, CA 94305; E-mail: email@example.com