Methods: Three examiners independently identified the spinous processes of L1-L4 on 60 prone volunteers using multiple bony landmarks including the sacral base, L5, Tuffier's line, T12, and the 12th ribs. The spinous processes were marked with radiopaque skin markers. Location of marker placement and presence of anatomical anomalies were determined by posteroanterior lumbar radiographs. Accuracy of marker placement and interobserver reliability were assessed using weighted κ values. Generalized linear mixed models and Cochran-Mantel-Haenszel tests assessed the relationship of accuracy to training level, presence of anatomical anomalies, and participant characteristics.
Results: Examiners identified a spinous process in 91% of vertebral assessments. Correct identification of vertebral level occurred 69% of the time (κ = 0.81; 95% confidence interval, 0.79-0.83). Faculty examiners were significantly more accurate in identifying the correct vertebral level than the resident examiner (67%-78% vs 51%, P ≤ .03). The presence of 12th rib anomalies decreased accuracy for all examiners (P ≤ .05), reducing accuracy from 74% to 55%. Accuracy was higher in male participants than in female participants (P = .01). Obesity significantly decreased accuracy (P = .0003) at L3 (50% vs 73%) and L4 (44% vs 72%).
Conclusions: Identification of lumbar spinous processes using multiple landmarks was more accurate than previously reported values. However, accuracy was dependent on examiner experience, presence of anatomical anomalies, and participant characteristics.
This abstract is reproduced with the permission of the publisher; full text by subscription. Click on the above link and select a publisher from PubMed's LinkOut feature.