Linear versus volumetric CT analysis in predicting tension-free fascial closure in abdominal wall reconstruction

Background Improved outcomes of abdominal wall reconstruction (AWR) have been shown when tension-free fascial closure (TFFC) is achieved. Our objective was to determine the clinical and radiologic predictors of TFFC in patients undergoing AWR. Study design We conducted a single institution retrospective cohort study of adults who underwent AWR between 2007 and 2018. Demographics, hernia characteristics and operative data were collected. Linear and volumetric variables were obtained from preoperative abdominal CT scans, the latter following 3D reconstruction. Logistic regression was used to evaluate predictors of TFFC. Area under the curve (AUC) >= 0.70 was considered to have acceptable discrimination. Results A total of 108 patients were eligible for analysis. The mean age was 57 +/- 11 years and 53 (49%) were female. 42 (39%) hernias were recurrent, 10 (9%) patients had a stoma and 9 (8%) had a history of open abdomen. The mean defect width was 11 +/- 4 cm and mean defect surface area was 150 +/- 95 cm(2). The most common AWR technique was endoscopic component separation 75 (69%). TFFC was achieved in 90 (83%) patients. No demographics or 3D volumetric measures were predictive of TFFC (all AUC < 0.7). European hernia society (EHS) class M1 was predictive of failure of TFFC [AUC = 0.70; odds ratio 7.0 (referent M3); 95% confidence interval, 2.1-23.8]. Linear variables of rectus muscle separation were the most predictive of TFFC (AUC 0.73-0.77). Conclusion In contrast to clinical characteristics, radiologic characteristics of large incisional hernias requiring AWR are predictive of TFFC. In particular, EHS class M1 and linear variables of rectus muscle separation appear to be better predictors of TFFC than volumetric measurements.