Background: Pre-arrest morbidity [PAM] scores could be used as additional tools to help patients and healthcare providers make decisions regarding do not resuscitate [DNR] orders. There are, however, no studies that examine applicability of those scores in Thailand.

Objective: To compare the usefulness of mortality prediction scores at hospital discharge among cardiac arrest patients who received cardiopulmonary resuscitation [CPR] and to identify the optimal cutoff points of the best morbidity scores.

Materials and Methods: A retrospective chart review was conducted of all adult patients who underwent CPR from January 1, 2013 to December 31, 2014 at Srinagarind Hospital, Thailand. Demographic and clinical data to calculate the PAM score, the prognosis after resuscitation [PAR] score, and the modified PAM index [MPI] were collected.

Results: There were enough data available on one-hundred and ninety-two patients to analyze the outcomes. The overall performances of all morbidity scores according to the area under the receiver operating characteristic [ROC] curves were similar; PAM 0.65 (95% confidence interval [CI] 0.56 to 0.74), MPI 0.66 (95% CI 0.57 to 0.75), and PAR 0.6 (95% CI 0.52 to 0.70), p = 0.5. PAM ≥6 and MPI ≥5 were the optimal cutoff points, which provided sensitivities of 49% and 57%, respectively, and specificities of 80.5% and 73.2%, respectively.

Conclusion: PAM, MPI, and PAR scores are not sufficient tools to identify patients who would benefit from resuscitation attempts among Thai patients. Given their high specificities, a combination these tools as part of a shared-decision to identify patients in whom CPR is likely to be unsuccessful recommended.

Keywords: Cardiac arrest, CPR, PAM, PAR, Prognosis, MPI, Survival