The correlation between biomechanical experiments and clinical studies in orthopedics: a systematic review.

Purpose: To evaluate the clinical relevance of biomechanical experiments by conducting a systematic review of orthopedic studies that compared treatments with both a biomechanical experiment and a clinical study.

Methods: A systematic search of MEDLINE/PUBMED was performed to identify orthopedic studies that included both a biomechanical experiment and a clinical study to compare treatments. The association between a difference in the outcome of the biomechanical experiment and a difference in the outcome of the clinical study was evaluated.

Results: There were 23 studies that were included. The orthopedic specialty of the studies included trauma (n = 17), pediatric trauma (n = 2), sports (n = 2), and hand (n = 1). The anatomic areas of each study included the hip/femur (n = 11), shoulder/humerus (n = 3), elbow/forearm (n = 2), knee/tibia (n = 3), ankle/foot (n = 2), pelvis/acetabulum (n = 1), and hand (n = 1). The biomechanical experiments involved cadavers (n = 14), synthetic bone (n = 5), finite element analysis (n = 3), and animals (n = 1). The biomechanical experiment from each study, compared to the clinical study, was more likely to report a difference in outcomes (82.6% (19/23) vs. 30.4% (7/23), p = 0.0008). The findings of the biomechanical experiment and the clinical study were in agreement in 43.4% (10/23) of the studies. Studies that reported a difference in the biomechanical outcome, compared to those that did not, were not more likely to report a difference in the clinical outcome (31.6% (6/19) vs. 25.0% (1/3), p = 1.0).

Conclusions: The outcomes of biomechanical experiments did not correlate with clinical study outcomes. Considering these findings, the utility of biomechanical experiments in orthopedics should be reexamined.

Level of evidence: Level 3.