Importance: Professional society policy statements recommend that clinicians limit (ie, withhold or withdraw) potentially nonbeneficial life-prolonging interventions by (1) achieving a shared decision with patients or surrogates or (2) initiating an institutional process to address disagreement with patients or surrogates. However, in the context of a health care system with a default tendency toward life prolongation, it is unclear whether clinicians rely entirely on these recommended approaches or resort to alternate approaches.
Objective: To characterize the range of decision-making approaches clinicians report using to limit potentially nonbeneficial life-prolonging interventions.
Design, setting, and participants: This qualitative study was conducted at 3 tertiary academic medical centers in Washington and California. Clinicians were sampled from emergency department, medical ward, medical intensive care unit, geriatrics, and palliative care services between February 2018 and June 2022 for in-depth, semistructured interviews. Results were analyzed between August 2023 and May 2025.
Main outcomes and measures: After qualitatively analyzing interviews to identify decision-making approaches, we developed a framework of approaches that categorized each as a recommended or alternate approach.
Results: We conducted 101 interviews (53 attending physicians [52%], 16 trainee physicians [16%], 6 advanced practice clinicians [6%], 21 nurses [21%], 3 chaplains [3%], and 2 social workers [2%]; 59 women [58%], 42 men [42%]; mean age, 42 years [range, 27-74 years]; mean years of experience, 14 [range, 1-52]). We identified 6 decision-making approaches: (1) providing an informed choice regarding interventions, (2) making a recommendation to limit interventions, (3) stating a plan to limit interventions, (4) explicitly not offering interventions, and (5) not mentioning interventions. In rare cases of intractable conflict, clinicians reported using an option of last resort: (6) invoking an institutional process to limit interventions. Respondents reported challenges with limiting interventions via the recommended approaches of shared decision-making (approaches 1-3) and institutional processes (approach 6), which sometimes discouraged the use of these approaches. While respondents recounted successfully using alternate approaches (approaches 3-5), they described interclinician and interhospital practice variation, as well as ethical and practical uncertainties.
Conclusions and relevance: In this qualitative study, clinicians reported substantial challenges using recommended approaches to limit potentially nonbeneficial life-prolonging interventions. Some clinicians reported using alternate approaches that are not supported in professional society policy statements.