(217) Selective Nerve Root Injections Can Accurately Predict Level of Nerve Impairment and Outcome for Surgical Decompression: A Retrospective Analysis

Abstract

There remains significant controversy regarding the use of a vertebral selective nerve root injection (SNI) as a diagnostic and therapeutic tool. In addition, the frequency of use of such procedures in patients with radiculopathy has increased dramatically in the last few years. Based on a Medline review there has been no studies combining cervical and lumbar SNI results and comparing preoperative diagnosis to surgical findings and outcome. The purpose of this paper is to retrospectively examine and compare the sensitivity, specificity and predictive value of a good surgical outcome in patients who had an SNI and subsequent surgical intervention. 101 patients from a 1996 thru 1999 database, who were referred to 10 spine surgeons (2 orthopedic and 8 neurosurgeon) for either cervical or lumbar radiculopathy, and had SNI and various imagery studies and subsequent surgery. Patients receive SNIs at our institution if there is a discrepancy between physical exam and radiologic imagery or to confirm a putative pain generator in multilevel pathology. These patients were then retrospectively analyzed with regard to correlation to surgical level and surgical outcome. SNIs were performed by one of three pain specialists in our clinic. Approximation of the appropriate nerve root sleeve was performed using fluoroscopic imagery, a nerve stimulator and contrast. After nerve root stimulation and neurography, 0.5–0.75 cc of lidocaine 2% was injected. Pre- and post-procedural visual analog scale (VAS) pain scores were obtained from the non-sedated patient. A SNI was considered positive or negative if the patient had immediate appendicular pain relief of greater or less then ninety percent respectively. The study was designed to include only those patients that had a SNI, regardless of result, and subsequently had surgical decompression in an attempt to treat the pain that initially prompted the SNI. A statistical analysis was then performed comparing preoperative data to surgical findings and outcome. Overall, 101 patients had SNIs who subsequently underwent surgical decompression. Average duration of symptoms prior to SNI was 1.5–12 months (4 months mean). Fifteen patients presented with cervical and 86 with lumbar radiculopathy. There were a total of 110 procedures performed on these patients. VAS scores of 90% with respect to their pre-procedural appendicular were used to determine if a SNI was positive, negative or indeterminate. All of these patients had an MRI or CT with or without a myelogram and all went to surgery. The results yield that SNIs are able to predict surgical findings with 94% and 90% sensitivity and specificity, respectively. A good surgical outcome was determined if the patient would do the surgery again, if they were satisfied or very satisfied and had a VAS of <3 at 6- and 12-month intervals. Our data revealed that a positive SNI was able to predict a good 6-month outcome with 95% and 64% sensitivity and specificity, respectively. At 12-months, similar results were obtained of 95% and 56%. Preoperative MRI results were also evaluated and revealed a 92% sensitivity in predicting surgical findings. We had 24 false positive MRI results and 0 true negatives. Interestingly we had 8 diabetic (IDDM or NIDDM) patients or nearly 8% of our total. The odds ratio of a diabetic having a bad outcome at 12 months was 5.4 to 1. Diabetics had a 50% likelihood of having a bad 12-month outcome versus 16% for non-diabetics with a p value of 0.066. We also looked at gender, smoking history and presence of cardiovascular disease and found no significant relationship with outcomes. Our data indicate that SNIs, when performed under rigorous method, is a highly valuable tool that can accurately determine level of nerve root impairment and outcome in patients being considered for surgical decompression. With a sensitivity of 94% and a specificity of 90%, SNIs offer a major advantage over other diagnostic modalities in patients with difficult-to-diagnose radiculopathies.