RESUMEN
BACKGROUND: An association between frailty and mortality exists; we hypothesized this is secondary to failure to rescue (F2R). STUDY DESIGN: Data were obtained from the NSQIP (2005 to 2012) for patients undergoing thoracoabdominal operations. Using the Modified Frailty Index, patients were classified as not (0 points), mildly (1 point), moderately (2 points), and severely (≥3) frail. RESULTS: There were 962,913 patients included; a majority were non-frail (52.2%), followed by mildly frail (33.8%). Complications were noted in 15.3%, major complications in 9.5%, mortality in 1.8%, and F2R in 1.3% of patients. On multivariate analysis, increases in frailty were associated with an increase in the risk of major complications (mildly: risk ratio [RR] 1.51; moderately: RR 2.69; and severely frail: RR 5.63 compared with non-frail; p < 0.0001), and death (mildly frail: RR 1.84; moderately frail: RR 4.44; and severely frail: RR 12.4). On univariate analysis, older patients, males, those undergoing small bowel interventions, gastric operations, or other procedures, and the frail were more likely to experience F2R (p < 0.0001). On multivariate analysis, males (RR 1.07), those undergoing small bowel intervention (RR 1.91), gastric operation (RR 1.83), and other procedures (RR 2.43) compared with hernia repair were more likely to experience F2R. As frailty increases F2R increases (mildly frail: RR 1.48; moderately frail: RR 2.41; and severely frail: RR 4.41) (p < 0.0001). Components of the Modified Frailty Index were analyzed separately; measures of impaired functional status were independently associated with increased F2R (RR 2.91; p < 0.0001), and those measuring comorbid medical conditions were not. CONCLUSIONS: Increases in frailty independently predict risk for F2R. The greatest predictors of F2R in the Modified Frailty Index are those associated with disability and not comorbidities.