Evaluation of the LACE plus Index for Short-term Cardiac Surgery Outcomes: A Coarsened Exact Match Study

Background. The potential of length of stay, acuity of admission, Charlson Comorbidity Index score, and emergency department visits in the past 6 months (LACE+) scores in patients undergoing cardiac surgery to predict short-term undesirable outcomes was examined. Methods. Coarsened exact matching was used to assess the predictive capacity of the LACE+ index among all cardiac surgery cases over a 2 -year period (2016-2018) at 1 health system (n = 4001). Study subjects were matched according to characteristics not assessed by LACE+, including duration of surgery, wound class, body mass index, insurance type, median household income, and race. For a comparison of outcomes, LACE score was divided into quartiles and otherwise matched pop- ulations were compared in reference to LACE quartile (Q): Q4 versus Q1, Q4 versus Q2, and Q4 versus Q3. Results. Escalating LACE+ score signi ficantly pre- dicted increased readmission (6.99% versus 25.92% for Q1 versus Q4, 12.79% versus 26.74% for Q2 versus Q4, and 20.52% versus 27.66% for Q3 versus Q4, respectively; P < .001, P < .001, and P = .003), reoperation (2.39% versus 7.73% for Q1 versus Q4, and 4.33% versus 7.67% for Q2 versus Q4, respectively; P < .001 and P = .015, respec- tively), and emergency room visits at 30 days after sur- gery (6.64% versus 13.65% for Q1 versus Q4, and 11.20% versus 14.84% for Q2 versus Q4, respectively; P < .001 and P = .041, respectively) as well as readmission, reopera- tion, and emergency room visits from 30 to 90 days and 0 to 90 days after surgery. Increasing LACE score pre- dicted higher rates of death during follow-up within 30 and 90 postoperative days ( P < .001). Conclusions. The LACE+ index may be suitable as a prediction model for important patient outcomes in a cardiac surgery population. (Ann Thorac Surg 2020;110:173-82) (C) 2020 by The Society of Thoracic Surgeons