Abstract 3: Defining the Association of Do-Not-Attempt-Resuscitate Orders with Patients’ Likelihood of Survival after In-Hospital Cardiac Arrest

Background: The long-term prognosis after In-Hospital Cardiac Arrest (IHCA) is variable, but often triggers discussions about further resuscitative efforts. Whether Do-Not-Attempt-Resuscitation (DNAR) orders after IHCA are aligned with patients’ likelihood of meaningful neurological survival is unknown. Methods: Using the Get With the Guidelines-Resuscitation registry, we identified 42,537 patients between 1/2000 and 9/2012 who achieved return of spontaneous circulation (ROSC) after IHCA. Each patient’s likelihood of meaningful survival without severe neurological disability (Cerebral Performance Category score ≤2) was calculated using the previously validated CASPRI prediction model. We examined the proportion of patients made DNAR within each CASPRI score decile to determine whether end-of-life decision-making was correlated with predicted likelihood of favorable neurological survival. Results: Overall, 26,198 patients (61.6%) were made DNAR after ROSC. These patients were older (68.7 ± 15.9 vs. 63.1 ± 17.1) and had higher rates of multiple comorbidities (all P <0.05). Among those with the best prognostic CASPRI scores (Decile 1), 34.6% were made DNAR after ROSC, as compared with 74.2% among those with the worst prognosis (Decile 10; P for trend <0.001). (Figure 1) Only 567 (2.2%) patients made DNAR survived to discharge without severe neurological disability. This rate was consistently low even in the CASPRI deciles with a high predicted likelihood of a good survival outcome (e.g., 6.4% in Decile 1). In contrast, 40.4% of patients not made DNAR survived without severe neurological disability, with a substantially higher rate in the lower CASPRI deciles (e.g., 74.7% in Decile 1). Conclusion: Among patients successfully resuscitated from IHCA, we found reassuring evidence that the decision to adopt DNAR status was generally aligned with a patient’s predicted likelihood of meaningful neurological survival. However, discordance between predicted and actual rates of this outcome among DNAR patients with the highest likelihood of survival suggest either competing, unmeasured interests or the need to inform physicians and patients about prognosis when discussing goals of care. Whether the use of the CASPRI score can further assist in these discussions requires prospective testing.