End-of-Life Trajectories Among Hospitalized Muslim Patients: Evolving from Aggressive to Comfort-Focused Care Across ICU, Post-ICU, and Ward Settings.

BackgroundMuslim patients have historically received aggressive end-of-life care, though palliative acceptance is growing. The factors that influence ICU (aggressive) vs ward (comfort-focused) deaths remain unclear, limiting value-concordant care.PurposeTo examine end-of-life care patterns and predictors of ICU vs ward death among Muslim patients, to guide culturally sensitive care.MethodsRetrospective cohort of adult Muslim decedents (2021-2024) at tertiary hospital in Muslim-majority country, stratified by death location: ward, ICU, and post-ICU ward. Multivariable logistic regression identified predictors of ward deaths.ResultsFrom fifty-thousand admissions, 1859 Muslim decedents were analyzed: ward (53%, n = 981), ICU (39%, n = 722), and post-ICU ward (8%, n = 156). Ward deaths involved older, comorbid patients, often with cancer (63%); ICU deaths were predominantly male with acute illnesses (29% sepsis). Post-ICU deaths showed intermediate profiles but highest tracheostomy rate (22%) and longest stays (48 days). ICU deaths had greatest life-sustaining interventions (82% mechanical ventilation), while ward deaths had highest palliative-care involvement (66%). Do-Not-Attempt-Resuscitation (DNAR) status rose from 20% on admission to 88% at death, with earliest transitions in ward deaths (median 3 days vs ICU: 15; post-ICU: 18). Ward deaths were independently associated with palliative care (aOR 15.11), code status discussion documentation (aOR 4.56), pre-existing terminal disease (aOR 2.98), cancer (aOR 1.61), and comorbidity burden (aOR 1.29), all P < .05.ConclusionMuslim ward deaths were associated with palliative involvement, prior terminal diagnoses, and earlier DNAR; while ICU/post-ICU deaths reflected aggressive care with delayed transitions. DNAR adoption was high (88%) but late, underscoring the need for earlier, culturally aligned goals-of-care discussions.