Mechanism of sternotomy dehiscence

Abstract

OBJECTIVES Biomechanical modelling of the forces acting on a median sternotomy can explain the mechanism of sternotomy dehiscence, leading to improved closure techniques. METHODS Chest wall forces on 40 kPa coughing were measured using a novel finite element analysis (FEA) ellipsoid chest model, based on average measurements of eight adult male thoracic computerized tomography (CT) scans, with Pearson's correlation coefficient used to assess the anatomical accuracy. Another FEA model was constructed representing the barrel chest of chronic obstructive pulmonary disease (COPD) patients. Six, seven and eight trans-sternal and figure-of-eight closures were tested against both FEA models. RESULTS Comparison between chest wall measurements from CT data and the normal ellipsoid FEA model showed an accurate fit (P < 0.001, correlation coefficients: coronal r = 0.998, sagittal r = 0.991). Coughing caused rotational moments of 92 Nm, pivoting at the suprasternal notch for the normal FEA model, rising to 118 Nm in the COPD model (t-test, P < 0.001). The threshold for dehiscence was 84 Nm with a six-sternal-wire closure, 107 Nm with seven wires, 127 Nm with eight wires and 71 Nm for three figure-of-eights. CONCLUSIONS The normal rib cage closely fits the ellipsoid FEA model. Lateral chest wall forces were significantly higher in the barrel-shaped chest. Rotational moments generated by forces acting on a six-sternal-wire closure at the suprasternal notch were sufficient to cause lateral distraction pivoting at the top of the manubrium. The six-sternal-wire closure may be successfully enhanced by the addition of one or two extra wires at the lower end of the sternotomy, depending on chest wall shape.