We aimed to compare simple two-dimensional (2D) measurement with comprehensive three-dimensional (3D) volume rendering to determine loss of domain (LOD), a clinically important decision-making feature for incisional hernia repair. In this single-center retrospective study, we analyzed the CT scans of a consecutive cohort of adult patients with a midline incisional hernia. The hernia sac- and abdominal cavity volumes were obtained by two different methods. The 2D method estimated the volumes using the corresponding height, width, and depth. The 3D method comprised of a volume rendering tool. For both methods, LOD was calculated according to the Sabbagh ratio (hernia sac volume / (hernia sac volume + abdominal cavity volume)). Taking the 3D method as the reference standard, the performance of the 2D method was expressed as positive predictive value (PPV) and negative predictive value (NPV) for LOD of more than- and less than 20%. The agreement between both methods was expressed as Cohen's kappa coefficient (kappa). We analyzed 92 CT scans. Agreement between both methods was high (kappa = 0.854, p = 0.0001); all 67 measurements for which the 2D method assessed LOD to be less than 20% were correctly classified (NPV = 100%), and 20 of 25 measurements for which the 2D method assessed LOD to be more than 20% were correctly classified (PPV = 80%). The 2D method can exclude patients fromperioperative actions needed for a more complex hernia. Since this method is easy to use and less time-consuming, it seems useful for the routine radiological assessment of LOD in clinical practice.
Read full abstract