Alignment of Do-Not-Resuscitate Status with Patients’ Likelihood of Favorable Neurological Survival After In-Hospital Cardiac Arrest
Metadata[+] Show full item record
After patients survive an in-hospital cardiac arrest, discussions should occur about preferences for future resuscitative efforts. Given the value patients generally place on possessing normal neurological function, these discussions should take into account a patient’s prognosis for survival with good neurocognitive function to ensure autonomy and quality of life. Whether patients’ decisions to become Do-Not-Resuscitate (DNR) after a successful resuscitation are aligned with the likelihood of favorable neurological survival is unknown. Within Get With the Guidelines-Resuscitation, a prospective, observational, multicenter registry of U.S. patients with in-hospital cardiac arrest, we identified 26,327 patients who achieved return of spontaneous circulation (ROSC) after arrest between April 2006 and September 2012. Using the previously validated Cardiac Arrest Survival Post- Resuscitation In-hospital (CASPRI) tool, each patient’s likelihood of meaningful survival without severe neurological disability (Cerebral Performance Category score <2; full recovery, mild or moderate disability) was calculated. We described the proportion of patients made DNR within each CASPRI score decile and calculated measures of association between DNR status adoption and the CASPRI score as a continuous variable using the point-biserial correlation coefficient. A multivariable logistic regression model was constructed using the CASPRI score variables to predict favorable neurological survival within this study cohort. Individual risk estimates were evaluated and the predictive performance of the model was verified using the c-statistic. Finally, we correlated DNR status adoption with actual favorable neurological survival. The 5,944 (22.6%) patients made DNR were older, with higher rates of comorbidities (all P <0.05). The c-statistic for the CASPRI score in this cohort was 0.762. Among those with the best prognostic CASPRI scores (decile 1), 7.1% were made DNR and 64.7% had favorable neurological survival. In contrast, in decile 10 (worst prognosis), 36.0% were made DNR and 4.0% had favorable neurological survival (P for both trends <0.001). While the rate of favorable neurological survival among all non-DNR patients was 30.5%, it was only 1.8% in patients made DNR, and was low (7.1%) even in patients with the best prognosis who were made DNR (decile 1). The point-biserial correlation coefficient for DNR status adoption and continuous CASPRI score was 0.206 (p<0.001), implying low correlation. Decisions to adopt DNR status after in-hospital cardiac arrest were generally aligned with patients’ likelihood of favorable neurological survival. Nevertheless, nearly two-thirds of patients with the worst prognosis were not made DNR, and few of these survived to discharge with a favorable neurological status. Prospective use of the CASPRI tool may better inform patients, families, and clinicians regarding prognosis, and better support shared decision-making about DNR status after in-hospital cardiac arrest.
Table of Contents
Introduction -- Methodology -- Results -- Discussion -- Appendix