Does lactate clearance prognosticates outcomes in ECMO therapy: a retrospective observational study.

BACKGROUND

ECMO support is a final treatment modality for patients in the refractory cardiogenic arrest and postcardiotomy cardiogenic shock with an utmost importance. Eventhough it is linked to high mortality, its usage gains popularity worldwide. We assessed the fluctuation of lactate levels and the clearance of lactate during the ECMO therapy and its prognostic role on mortality.

METHODS

Data were gathered on all patients receiving ECMO therapy longer than 48 h between January 2015 and December 2017 retrospectively. Blood lactate had been recorded before ECMO implantation and at specific time points during ECMO support as a routine procedure. In this study, the Lactate clearance at specific time points (Lactate clearance-1) and the duration that lactate cleared more than 10% of the initial lactate level (Lactate clearance-2) was measured. Statistical analysis included Mann Whitney U-test and ROC-curves to predict 30-day mortality.

RESULTS

Fourty-eight patients underwent ECMO therapy for refractory cardiogenic shock resulting in 70.8% mortality. The lactate levels before and after ECMO therapy as well as the dynamic changes were significantly correlated with mortality variable. With AUC calculation, LC-2 has a strong discrimination (AUC = 0.97) on 30-day survivors and nonsurvivors. LAE-LBE (AUC = 0.785), L48-LBE (AUC = 0.706) showed moderate predictive power on 30-day mortality.

CONCLUSIONS

Changes in lactate levels after ECMO implantation is an important tool to assess effective circulatory support and it is found superior to single lactate measurements as a prognostic sign of mortality in our study. Based on our results, an early insertion of ECMO before lactate gets high was suggested. Serial changes on lactate levels and calculation of its clearance may be superior to single lactate on both effective circulatory support and as prognostic prediction. LC-2 showed a strong discrimination on 30-day mortality.