Evaluating the Association Between Patient Priorities and ICU Care Delivery in Adults Receiving Mechanical Ventilation.

Donovan, Kara A; Polley, Madison; Palmer, Geralyn; Hanlon, Bret M; Stalter, Lily N; Donnelly, Helen K; Wunderink, Richard G; Gao, Catherine A et al. · Chest · 2026

prospective_cohort · Level II

Where this comes from

Abstract

Understanding how ICU treatments align with a patient's priorities is crucial to evaluating and improving critical care. Measuring this alignment is challenging, however, because patient priorities evolve over time and are not systematically documented in electronic health records (EHRs). To what extent is ICU care delivery associated with patient priorities, as reported in real time by their surrogate decision-makers? We conducted a secondary analysis of a prospective observational study of patients who were mechanically ventilated in an adult medical ICU. At study enrollment, surrogates were asked to categorize patients' treatment priorities as: (1) maximize survival; (2) middle ground (try to live as long as possible but with some treatment limitations); or (3) maximize comfort. Demographic characteristics, care delivery, and outcomes were abstracted from the EHR. To evaluate relationships, χ<sup>2</sup> and Kruskal-Wallis tests and linear and ordinal regression models were used. For 393 eligible patients, 217 (55.2%) surrogates responded to the patient priorities questionnaire. The most frequently reported priority was middle ground (n = 105 [48.4%]), followed by maximize survival (n = 75 [34.6%]) and maximize comfort (n = 37 [17.1%]). Hospital mortality was similar across groups (35.1% maximize comfort, 34.3% middle ground, and 28.0% in maximize survival; P = .619), as were frequencies of tracheostomy placement, palliative care consultation, and, among decedents, use of life-sustaining therapies near death. A maximize comfort priority was associated with a shorter hospital length of stay in a linear regression model adjusted for age, race, and Charlson Comorbidity Index score (β = -0.43; 95% CI, -0.78 to -0.09; P < .014). This study did not find strong relationships between surrogate-reported treatment priorities and ICU care delivery. These findings underscore the need for ongoing efforts to achieve individualized, goal-concordant care for patients with critical illness.