Impact of treatment goals on outcomes in critically ill nonagenarians: a retrospective observational study

Abstract

Background: The increasing number of critically ill nonagenarians admitted to intensive care units (ICUs) poses challenges in treatment decision-making. This study examines the short-term impact of treatment goals on ICU and hospital mortality and length of stay (LOS), incorporating APACHE III-J scores for risk adjustment. Methods: This retrospective cohort study included adults aged ≥ 90 years admitted to ICUs in Australia and New Zealand from 2010 to 2023, using data from the ANZICS Adult Patient Database. Patients were categorized by initial goal: full (n = 24,694) vs. limited (n = 10,690). Propensity score matching balanced baseline characteristics, resulting in 10,671 matched pairs. Mortality was analyzed using flexible parametric survival models. ICU and hospital LOS were assessed using log-transformed linear regression models. Competing risks regression was applied to adjust for early mortality bias. APACHE III-J scores were included as a key predictor. Results: Limited treatment was associated with mortality risk that was highest early after admission and declined over time but remained consistently higher than with full treatment. ICU LOS (β = 0.966, 95% CI 0.943–0.990, p = 0.006) and hospital LOS (β = 0.944, 95% CI 0.919–0.969, p < 0.001) were significantly shorter in this group, primarily due to early death. Competing risks analysis confirmed a higher cumulative incidence of ICU mortality (sHR 1.717, 95% CI 1.555–1.900, p < 0.001) and hospital mortality (sHR 1.643, 95% CI 1.541–1.752, p < 0.001). APACHE III-J was an independent predictor of mortality. Conclusions: Limited treatment in critically ill nonagenarians was associated with higher ICU and hospital mortality. The excess risk of ICU mortality was greatest early after ICU admission and gradually attenuated over time, although it remained elevated throughout the ICU stay. These findings support early risk stratification and periodic reassessment of treatment goals as clinical conditions evolve.