Association of Care Intensity With Survival and Long-Term Cost-Effectiveness in Sepsis Without Initial Organ Support After Emergency Admission: A Multicenter Cohort Study in Japan

Objectives: To identify subgroups of patients hospitalized for community-acquired sepsis without initial organ support who may derive survival and cost-effectiveness benefits from ICU or high-dependency care unit (HDU) admission. Design: Retrospective cohort study. Setting: A 140-hospital Japanese database (April 2016–March 2024). Patients: Adults admitted via emergency ambulance for community-acquired sepsis not requiring extracorporeal membrane oxygenation, mechanical ventilation, catecholamines, or renal replacement therapy before or on the antimicrobial-initiation day. Interventions: None. Measurements and Main Results: We compared ICU and HDU admissions, each vs. general ward (GW) admission, balancing patient characteristics using overlap propensity-score weighting. In-hospital mortality was estimated as risk ratios (RRs) with 95% CIs using modified Poisson regression. Cost-effectiveness was measured by the incremental cost-effectiveness ratio (ICER) in 3-year costs per quality-adjusted life year (QALY). Among 28,211 patients (GW, 18,349; HDU, 8,706; ICU, 1,156), ICU and HDU admissions were not associated with lower mortality overall (HDU vs. GW; RR, 0.99; 95% CI, 0.92–1.06 and ICU vs. GW; RR, 1.06; 95% CI, 0.92–1.22). However, HDU admission was associated with lower mortality among patients with Sequential Organ Failure Assessment (SOFA) score greater than or equal to 6 (RR, 0.86; 95% CI, 0.75–0.99) or Glasgow Coma Scale (GCS) score less than or equal to 12 (RR, 0.81; 95% CI, 0.71–0.92). Under the Japanese (50,000 USD/QALY) and U.S. (100,000 USD/QALY) thresholds, ICU and HDU admissions were not cost-effective overall (ICERs > 200,000 USD/QALY). However, the HDU subgroup with GCS score less than or equal to 12 (ICER, 47,738 USD/QALY; 95% CI, 45,918–49,558) was cost-effective under both thresholds, and that with SOFA score greater than or equal to 6 (ICER, 55,319 USD/QALY; 95% CI, 52,018–58,624) met the U.S. threshold. Conclusions: ICU or HDU admission was not associated with survival or economic benefits overall, whereas HDU subgroups (SOFA score ≥ 6 or GCS score ≤ 12) were associated with lower mortality and better cost-effectiveness. These findings may help optimize resource allocation.