Classification Criteria for Neurogenic Claudication Caused by Lumbar Spinal Stenosis

BACKGROUND Classification criteria are recommended for diseases that lack highly specific biomarkers in order to improve homogeneity in clinical research. Since imaging evidence of lumbar spinal stenosis (LSS) or lumbar disc herniations (LDH) may not be associated with symptoms, clinical classification criteria based upon patient symptoms and physical examination findings are required. This study sought to develop a set of criteria identifying patients with neurogenic claudication (NC) caused by LSS and patient with radicular pain (RP) caused by LDH. This abstract reports the results concerning NC caused by LSS.
METHODS Phase 1: Nineteen spine specialists (clinicians and researchers) from 8 countries participated in a Delphi process, using an internet program, to rank symptoms and signs which suggest LSS as the cause of NC or DH as the cause of RP. Phase 2: Nineteen different spine specialists (surgeons and non-surgeons) from 5 countries recruited patients during office visits and classify them with a high degree of confidence as having with either: 1) NC caused by LSS 2) RP caused by LDH or 3) non-specific low back pain (NSLBP) with non-specific leg pain radiation. Patients completed survey items and specialists documented examination signs. Signs and symptoms present in ≥ 10 patients were analyzed by using Generalized Estimating Equations (GEE). Patients with NC caused by LSS or NSLBP served as controls. Items with p<0.1 in univariate analysis were entered in the multivariate analysis. A score to predict NC caused by LSS was developed based on the coefficient of the GEE, and used to obtain a ROC curve and the associated area under the curve (AUC).
RESULTS A list of 46 clinical signs and 28 patient-reported symptoms were selected by the group of spine specialists during the 1st phase. For the 2nd phase, 209 patients with high confidence in the diagnosis were included 63 NC caused by LSS, 89 RP caused by DH, and 57 NSLBP with non-specific leg pain radiation. Items which predicted NC with a p-value <0.1 included age >60, bilateral leg pain, leg pain relieved by sitting, leg pain decreased by leaning or flexing, positive30 seconds extension test, negative straight leg raise test. The score had an AUC of 0.91, and the cutoff to obtain a specificity of 92.1% resulted in a sensitivity of 80.0%.
CONCLUSION An international collaboration of surgeon and non-surgeon spine specialists produced a set of diagnostic criteria with high specificity and sensitivity for identifying patients with NC caused by LSS. Using this set could improve the quality of basic science and clinical research in this field by improving homogeneity within groups of patients.