Multivariable Diagnostic Prediction Model for Identifying Active Spondylolysis in Young Athletes With Low Back Pain
. 2020 May 30;45:1-6. doi: 10.1016/j.ptsp.2020.05.009. Online ahead of print.
Development of a Preliminary Multivariable Diagnostic Prediction Model for Identifying Active Spondylolysis in Young Athletes With Low Back Pain
- PMID: 32534433
- DOI: 10.1016/j.ptsp.2020.05.009
Aims: The primary aim of this study was to develop a diagnostic cluster of common clinical findings that would assist in ruling out an active spondylolysis in adolescent athletes with low back pain (LBP).
Design: Retrospective case-series.
Setting: Hospital-based sports medicine clinic.
Patients: One thousand and twenty-five adolescent athletes with LBP (age 15.0 ± 1.8 years, 56% female) were reviewed. Active spondylolytic injuries were identified in 22% (n = 228) of these patients.
Main outcome measure: presence or absence of active spondylolysis on advanced imaging.
Results: Through logistic regression analysis, pain with extension (p < 0.001), difference between active and resting pain ≥3/10 (p < 0.001), and male sex (p = 0.002) were identified as significantly associated with active spondylolysis. The clinical cluster had a sensitivity of 88% (95% CI 83%-93%) to help rule out active spondylolysis. The negative likelihood ratio was 0.34 (95% CI 0.23-0.51) and the negative predictive value was 90% (95% CI 86%-93%). Diagnostic accuracy of the cluster was acceptable (area under the curve = 0.72 (95% CI 0.69, 0.76; p < 0.001).
Conclusion: This study found a cluster of three patient characteristics that may assist in ruling out active spondylolysis in adolescent athletes with LBP.
Keywords: Bone stress injury; Low back pain; Lumbar spine; Stress fracture.