Postoperative lumbar spine MRI: How well does a radiology report that raises suspicion for infection correlate with true clinical infection?

Introduction: Interpretation of a lumbar spine MRI in the immediate postoperative period is challenging, as postoperative tissue enhancement and fluid collections may be mistaken for infection. Radiology reports may use ambiguous language, creating a clinical problem for a surgeon in determining whether a patient needs treatment with antibiotics or revision surgery. Moreover, retrospective criticism of management in instances of a true infection may lead to medicolegal ramifications. Methods: A retrospective review of patients undergoing posterior-approach lumbar decompressive surgery with or without fusion over a 30-month period identified those undergoing postoperative MRI within 10 weeks of surgery. Patients initially operated upon for infection were excluded. The MRI reports were analyzed for language describing findings suspicious for infection and those of these with true infections were identified. Results: Of 487 patients undergoing posterior lumbar spine decompression surgery, 68 (14%) had postoperative MRI within 10 weeks. Of these, the radiology reports raised suspicion for infection in 20 (29%), of which 2 (10%) patients had a true infection. Two patients underwent reoperation for new motor deficit from seroma/hematoma. Of 63 patients who had MRI to evaluate complaints of back and/or leg pain without new motor deficits, the MRI significantly altered management in 3 patients (4.8%). Conclusion: Radiology reports of postoperative lumbar spine MRIs frequently use language that raises suspicion for infection; but it is uncommon, however, that these patients harbor true infections. A radiology report describing possible infectious findings may not be considered significant without corroboration with other laboratory and clinical data.