Identifying Predictors of Cardiopulmonary Resuscitation Failure in Cancer Patients: An Expert Agreement Approach

Aim. Cardiopulmonary resuscitation (CPR) for end-of-Iife and critically ill patients, specifically those with cancer in hospitals, often presents a dilemma between resuscitation and non-resuscitation decisions. This study aims to determine the predictors of failed resuscitation in cancer patients through expert consensus. Materials and Methods. We collected predictor variables by reviewing 20 research journals that reported on cancer patients who had died and were not successfully resuscitated. Among these journals, only 11 articles mentioned variables related to resuscitation failure and death in cancer patients. To identify which predictors were most common in Indonesia, we asked five expert palliative anesthesiologists from a palliative care center in Indonesia to choose the predictors that were most relevant to resuscitation failure. To measure the expert agreement, we used the Fleiss' Kappa statistical method with SPSS software version 20. Results. Based on the experts' agreement (k > 0.8), we identified eight predictors of cardiopulmonary resuscitation failure in cancer patients that are expected to occur in Indonesia. These predictors are: AIDS (acquired immunodeficiency syndrome), CPR for longer than 15 minutes, failure to recover heart rate after CPR for longer than 20 minutes, asystole on electrocardiogram (ECG), early coma, hepatic insufficiency, non-shockable heart rhythm, and pulseless electrical activity (PEA). Conclusion. The identified predictors for cardiopulmonary resuscitation failure in cancer patients, based on expert consensus, are AIDS, CPR for longer than 15 minutes, failure to recover heart rate after CPR for longer than 20 minutes, asystole on electrocardiogram (ECG), early coma, hepatic insufficiency, non-shockable heart rhythm, and pulseless electrical activity (PEA).