External validation of rules for termination of resuscitation in in-hospital cardiac arrest

Importance: Termination of resuscitation rules aim to avoid prolonged, futile resuscitative efforts, reducing patient and provider burden. To date, however, no widely adopted rule exists for termination of in-hospital cardiac arrest. A recent study derived a termination rule (unwitnessed, unmonitored, asystole, and no return of spontaneous circulation within 10 min) in a Scandinavian cohort with an acceptably high positive rate for clinical utility and a very low rate of patients meeting the rule who survived to 30-days. Objective: To externally validate previously derived termination of resuscitation rules for patients suffering in-hospital cardiac arrest. Design: Observational study of a prospectively collected in-hospital cardiac arrest cohort including years 2012–2024. Setting: Participating hospitals from the American Heart Association Get With The Guidelines-Resuscitation registry. Participants: Adult patients (≥18 years) who suffered in-hospital cardiac arrest. Exposure: Meeting a previously described termination of resuscitation rule. Main outcome: Survival to hospital discharge. Positive rates, false positive rates, false discovery rates, and resuscitation time potentially avoided were calculated. Hypothesis formulated prior to analysis. Results: Of 646,794 patients, 359,686 met inclusion criteria across 703 hospitals. Overall survival to discharge was 23.1%. For the primary Scandinavian termination Rule 1, the positive rate (patients meeting Rule criteria) was 1.7%; 32.9% of whom achieved return of spontaneous circulation and 5.1% survived to discharge. Median resuscitation time potentially avoided was 11 min per patient meeting Rule 1 criteria, equating to ∼18 min per 100 events. Across hospitals, positive rates ranged from 0% to 11.9. Other termination rules demonstrated either very low average positive rates (<3%) or unacceptably high survival rates among patients meeting the rule. Findings were similar in sensitivity analyses excluding the COVID-19 era. Conclusions: In this large American cohort, previously derived termination rules for in-hospital cardiac arrest demonstrated either limited applicability or unacceptable error rates. Even the best-performing Scandinavian rule identified few patients for early termination and would have led to premature cessation in ∼5% of survivors. These findings underscore the need for further refinement of termination rules before clinical adoption in the United States.