Using “LIKE” Score to Predict Prognosis in Postcardiotomy Shock Supported with Extracorporeal Life Support (ECLS)


  • #AC/END 01-O-9
  • Adult Cardiac Surgery/End-stage Heart Failure/Heart Transplantation/VADs/ECMO. SESSION-1
  • Oral

Using “LIKE” Score to Predict Prognosis in Postcardiotomy Shock Supported with Extracorporeal Life Support (ECLS)

Heng-tsan Ho, Feng-Chun Tsai, Kuo-Sheng Liu, Yu-Yun Nan, Shao-Wei Chen, Jao-Ji Chu, Pyng-Jiang Lin

Department of Cardiothoracic and Vascular Surgery, Chang Gung Memorial Hospital, Linkou, Taiwan, Taoyuan City, Taiwan

Date, time and location: 2018.05.27 10:30, Congress Hall, 2F–A

Abstract

Purpose

Refractory postcardiotomy shock supported with extracorporeal life support (ECLS) is a salvage therapy. Though individual risk factors and scoring system have been applied to predict the prognosis, none was specific and conclusive to this population. We try to establish a simple and reproducible scoring system for prognosis.

Materials and Methods

This retrospective study reviewed the adult (> 18y/o) patients of coronary bypass, valvular surgery or combined procedures underwent ECLS support for postcardiotomy shock in a single institute. Demographic, clinical, and laboratory variables post ECLS 48hrs were analyzed and four important parameters, including lactate, INR, kidney failure need renal replacement therapy and ejection fraction by echo were comprised of “LIKE” score. Internal validation by bootstrapping methods was applied to evaluate the power of the scoring system.

Results

Between March 2006 and August 2015, there were 164 patients enrolled to this study, with mean age of 62.6 y/o and male of 63%. Weaned off the ECLS and survival to discharge were 120 (73%) and 78 (48%) respectively and ECLS support time was 126.1 and 183.3 hours for survival and mortality groups. The LIKE score > 6 indicated a good discriminative power of prognosis by using the areas under the receiver operating characteristic curve (AUROC: 0.926 ± 0.034) with sensitivity of 84.6% and specificity of 87.5%.

Conclusions

In patients with postcardiotomy shock supported with ECLS, the LIKE Score after 48hrs demonstrated a good discriminative power of predicting prognosis. Further treatment strategies, either shifting to ventricular assist device or listing to urgent transplant is warranted to improve the final outcome.


To top