Abstract Body: Introduction: The Pittsburgh Cardiac Arrest Category (PCAC) is an early indicator of illness severity that is predictive of survival after resuscitation from cardiac arrest. Increased PCAC (i.e., higher severity) may bias clinicians towards recommending withdrawal of life-sustaining therapy (WLST), leading to death, even in patients who may have otherwise recovered. In this scenario, increasing PCAC scores may predict decreased survival not only by direct association with worse illness but also by an indirect path of promoting WLST.

Hypothesis: The probability of WLST, and proportion of deaths due to WLST, vary by PCAC.

Methods: In this retrospective cohort study, we identified hospitalized patients resuscitated from cardiac arrest in our prospective registry from January 2010 to April 2025. We extracted demographics, arrest characteristics, PCAC, survival to hospital discharge, and WLST. We used descriptive statistics to summarize cohort characteristics. We compared odds of survival between WLST subjects and non-WLST subjects. Subsequently, we used X² tests to investigate if the distributions of WLST and death due to WLST are different across PCAC scores.

Results: We included 4,617 subjects with median age 62 (IQR 51-71). We had 1,853 (40%) female subjects, 3,001 (66%) out-of-hospital arrests, with 1,422 (31%) shockable rhythms, and median CPR duration 16 (IQR 8-28) minutes. There were 996 (22%) subjects classified as PCAC I, 719 (16%) as PCAC II, 381 (8%) as PCAC III, 1,826 (40%) as PCAC IV, and 694 (15%) as unknown. Subjects who experience WLST have a substantially lower odds of survival compared to those who did not (OR 0.04, 95%CI 0.03 to 0.05). In comparing distributions of WLST and death by WLST with PCAC, we found the proportions of WLST (p < 0.001) and of deaths due to WLST (p < 0.001) are not evenly distributed across PCAC scores (Figure 1).

Conclusions: These data suggest that PCAC influences the probability of WLST. Coupled with the finding that WLST enacts a strong influence on survival after cardiac arrest, there is potential that bias towards WLST may explain some of the association between PCAC and survival, independent of underlying illness severity. Further research is needed to decompose the effects of early illness severity-driven WLST on survival outcomes.