Frailty Predicts Morbidity, Complications, and Mortality in Patients Undergoing Complex Abdominal Wall Reconstruction

Introduction: Frailty is becoming an increasingly-established risk factor for adverse postoperative outcomes. Interestingly, the efficacy of frailty indices in predicting morbidity and mortality in the plastic surgery patient population has not yet been examined. As such, given the innate high morbidity involved in complex abdominal wall reconstruction (CAWR) and the propensity for obesity, diabetes, and other comorbidities among this patient population, we sought out to determine the predictive utility of a frailty index in patients undergoing CAWR.

Methods: A retrospective analysis was conducted using the American College of Surgeons (ACS) National Surgical Quality Improvement Project (NSQIP) database from 2005-2013. Patients undergoing CAWR were identified using Current Procedural Terminology (CPT) codes for ventral hernia repair +/- components separation technique, +/- placement of prosthetic or biologic mesh, and complexity of the defect. Pre-operative frailty index was calculating using the Modified Frailty Index (mFI) initially described by Saxton et al. Outcomes included overall morbidity, Clavien-Dindo Grade IV complications, and mortality. Multivariate regression models were used to determine the effect of mFI and each component of the mFI on our outcomes of interest. Recursive partitioning was used to determine a mFI threshold predictive of complications.

Results: Of 70,339 patients identified as above, 9,931 had a complication associated with their procedure. mFI of 0.12 (±0.11) was calculated for these patients and was significantly greater than 0.077 (±0.85) for patients with no complications (p<0.001). When examining mFI correlation with Clavien-Dindo Grade IV complications (n=2,541), mFI once again was significantly greater (0.16 ± 0.12) than those with no Grade IV complications (0.080 ±0.09; p<0.001). Multivariate analyses also showed that all individual factors of the mFI (diabetes mellitus, hx of MI, etc.) were predictive of any complications and Grade IV complications (p<0.001). Calculated odds ratios showed that higher pre-operative mFI also had a 7.77x likelihood of having any complication, 35.71x likelihood of having a Grade IV complication, 3.85x likelihood of having a surgical site complication, and a 62.05x likelihood of death (all p<0.001). Recursive partitioning revealed that a threshold of greater than 3 indicators of mFI conferred a 2.07x likelihood of a Grade IV complication and a 2.33x likelihood of death (both p<0.001).

Discussion: We have shown that frailty as measured by mFI is an accurate predictor of morbidity, complications, and mortality in patients undergoing CAWR. Additionally, we have determined a frailty threshold of complications and mortality rate for plastic surgeons to consider during patient selection.