B104-21 The Epidemiology of Hospital Code Status Designations Across Six US Health Systems

Abstract Rationale Patient elected code status is critical for clinical decision-making and for achieving goal-concordant care. However, we have a limited understanding of how code status designations vary across health systems for similarly ill patients or how these designations evolve day-by-day during acute illness. We therefore aimed to characterize these patterns using electronic health record (EHR) data. Methods We conducted a multi-center retrospective observational study across 6 health systems (39 hospitals) in the Common Longitudinal ICU data Format (CLIF) Consortium. We included adult hospital admissions from 01/01/2016 to12/31/2024 and excluded admissions to psychiatric, labor/delivery, and rehabilitation units as well as inpatient hospice admissions. The primary outcome was the composite of in-hospital death or hospice discharge. We categorized code statuses as (1) Full, (2) Do-Not-Resuscitate and Do-Not-Intubate (DNR/DNI), or (3) limited, partial, or special status (Other). We compared initial code status distributions stratified by admission illness severity (APACHE-II 10, 10-19, 20) and health system. We tracked daily code statuses through hospital day 30 and stratified by illness severity to characterize temporal patterns of code status evolution. Results The study cohort consisted of 2,429,345 hospitalizations, 2,265,217 (93%) of which began as Full Code, 138,030 (5.7%) as DNR/DNI, and 26,098 (1.1%) as any other status. The primary outcome occurred in 3.4% (n = 76,173) of admissions that began as Full Code, 17% (n = 23,943) of admissions that began as DNR/DNI, and 21% (n = 5,536) of admissions that began with other statuses. Among patients admitted with a non-Full-code status, hospice discharges (10.5%, n = 17,184) were more frequent than in-hospital deaths (9.1%, n = 14,995). Within each APACHE II stratum, initial code statuses varied across health systems. Non-Full code status rates ranged from 1.1% to 4.2% among low-acuity admissions (APACHE-II 10, n = 938,137), 2.9% to 14% for moderate-acuity admissions (APACHE-II 10-19, n = 1,175,259), and 5.8% to 21% among high-acuity admissions (APACHE 20, n = 315,949). These patterns persisted through hospital day 30 (Figure). Conclusion Although most hospital admissions began as Full Code, presenting code status varied approximately fourfold across hospitals. This substantial variation—independent of objective measures of illness severity—suggests that code status designations reflect institutional practices and communication norms rather than patient acuity alone. These findings underscore the need to better understand clinical and communication processes driving code status decisions and changes during hospitalization. This abstract is funded by: K08CA270383