Diagnostic Sacroiliac Joint Injections: Is a Control Block Necessary?

Background: Sacroiliac joint (SIJ) pain presents as a deep and somatic pain, predominantly affecting the lower back and buttock and referring down the leg, sometimes as far as the foot. Given that the features of SIJ pain are non-specific and that this referred pain is similar to lumbar facet joint and lumbar disc pain, diagnostic local anesthetic injections (diagnostic blocks) into the SIJ are used to identify the source of pain. Despite wide use, little is known about the false positive rate of a single diagnostic sacroiliac (SI) block and the requirement for a control block. Objective: To determine whether a control SI block is necessary and to monitor the false positive rate for a single injection. Study Design: A prospective and observational study was conducted as part of a practice audit, with data collected over 3.5 years at the authors’ private practice. Patients & Methods: Under fluoroscopic guidance, 1408 consecutive patients presenting with prominent deep somatic pain over the SIJ region were sterilely injected with anesthetic into the SIJ and/or the deep interosseous ligament (DIL). Pain was measured on the 11-point Numerical Rating Scale (NRS) prior to injection and incrementally over the following 12 weeks. Fully completed and unequivocal data sets were available for 1060 patients. Decreases in pain scores (of >80%) at >2 hours of postinjection were indicative of SIJ pain and recorded as a positive SIJ block. Results: Of 1060 patients receiving a first SIJ diagnostic block, 680 (64.1%) recorded a positive result. Subsequently, 271 positive patients and 22 who were negative for SIJ pain opted to receive a second control block. SIJ pain diagnosis was confirmed in 237/271 (87.5%) of those with an initial positive response, while 18/22 patients (81%) had their initial negative result confirmed. The false positive rate of a single block is therefore calculated at 12.5%, and on a contingency table analysis, a single anesthetic SIJ injection has diagnostic accuracy of 87.03%, with high sensitivity (98.3%), when compared with a second control diagnostic block. Limitations: All injections were performed at one clinical centre. A proportion (348/1408) of initial patients did not return fully completed pain records or had equivocal responses (≥80% pain relief, but transiently, for ≤30 min) and were excluded from further analysis. Conclusion: Given the observed high rates of accuracy in this study, it is reasonable to suggest the use of one diagnostic block as the criterion standard for assessing the SIJ as