Micro - The Microsurgical Index For Complication Risk And Outcomes
Thomas M. Johnstone, BS1, Daniel Najafali, BS2; Priscila Cevallos, BS3, Clifford Sheckter, MD3, Rahim S. Nazerali, MD MHS3, Gordon K. Lee, MD3.
1Stanford University School of Medicine, Stanford, CA, USA, 2Carle Illinois College of Medicine, Urbana, IL, 3Division of Plastic and Reconstructive Surgery, Stanford University, Stanford, CA, USA.
PURPOSE: Free tissue transfer (FTT) is determined by a multitude of factors. However, no tool exists to quantify risk for complications following FTT. This study developed the Microsurgical Index for Complication Risk and Outcomes (MICRO) to address this.
METHODS: Patients were queried from the 2003-2015 MarketScan Databases with CPT codes for FTT requiring microsurgical anastomosis. ICD9 codes were used to query comorbidity and 90-day postoperative complication data for each patient. The Charlson and Elixhauser Comorbidity Indexes were constructed for each patient. The MICRO was then constructed with forward stepwise selection from Elixhauser comorbidities and domain expert input. Indexes were used as covariates in multivariate logistic regression models with patient age, sex, and flap tissue type to predict complications following FTT. The area under the receiver operating curve (AUC) and five-fold cross validated classification accuracy were determined.
RESULTS: 5,595 patients were included. The final MICRO consists of seven variables (Charlson - nineteen; Elixhauser - thirty). It had the highest AUC (0.60) and accuracy (60.4%) when predicting complications. Table One summarizes cohort characteristics, describes the association of each MICRO component with complications, and compares the performances of included indexes.
CONCLUSION: The MICRO outperforms available comorbidity indexes at predicting complications following FTT with far fewer variables. Future studies could augment the MICRO with surgeon, donor-site, and recipient-site data to create a sharper risk tool for the plastic surgeon. 
Back to 2023 Abstracts